Matt Murphy, CEO of Marvell Technologies, finds himself at the center of the AI revolution. As Joubin notes, Marvell is truly "in the heart of the action... kind of under the radar."

"I think in a lot of ways... for sure," Matt acknowledges about being in the eye of the hurricane. "From a broad public perspective, for sure. Friends of mine, family... nobody understands this stuff."

Despite the relatively low public profile, Marvell's position in the AI and data center space is significant:

"We're 80% of our revenue next year is going to be like data center and AI," Matt explains. "That concentration from a semiconductor perspective, or actually any supplier if you think about it into that sort of ecosystem, it's us and NVIDIA in terms of the biggest exposure leverage to it."

Matt notes this concentrated exposure is "good and bad," creating more volatility but also tremendous opportunity. What makes Marvell's journey even more remarkable is how they've positioned themselves as a neutral supplier:

"We're Switzerland... we supply to everybody."

Timestamp: [01:41-03:03]

Joubin breaks down Marvell's extraordinary growth trajectory under Matt Murphy's leadership, highlighting the remarkable stock performance despite recent volatility:

"$9 a share was Marvell's stock price when you joined right... and that was 2016... In January of this year it was at $124 a share, right? It was a hundred billion dollar company like 3 months ago."

Matt provides even more context about the company's transformation:

"When I started in July of 2016, even more remarkable is the actual firm enterprise value was only 3.5 billion because the company had like a billion six of cash. So the market cap was like five billion, no debt, so really firm value was three and a half. And then... just before all this stuff happened it was 100 billion plus."

Joubin notes the dramatic stock appreciation in 2024:

"In 2024 the stock appreciated 83%, right? In a year, right?"

Yet despite Marvell having its "best quarter ever as a business" and "guiding the best quarter in the history of the company," the stock recently dropped dramatically from $100 billion to around $50 billion market cap.

Matt responds with calm perspective:

"You got to take the long view. We've had periods over the last nine years of a lot of volatility, and we've always just powered through. I mean, I've been doing this a long time."

Timestamp: [03:03-05:11]

Matt Murphy shares his perspective on the cyclical nature of the semiconductor industry, drawing from his extensive experience in the field.

"Semiconductor business... you have to look way through any kind of cycle in terms of your investment profile. The stuff we're working on today, decisions we make right now, it's really not going to generate meaningful revenue until three, four, five, six years from now."

He explains why despite periodic claims to the contrary, semiconductors remain fundamentally cyclical:

"Semiconductors is inherently a cyclical business, it always has been. If anyone ever tells you at some point that 'hey, the good news is it's so structurally so much better now and it won't be as cyclical' - it's not. I mean, everybody says that in periods between huge cyclicality, but I'm like 31 years doing this."

Matt recounts historical cycles that many have forgotten:

"Since 1995 was the first big semi up cycle, followed by a massive crash in 1996, actually huge reset. But nobody remembers that. You all you remember is like the point from like early 90s to '99, 2000, and it was just all up and to the right, but in between you had a '96 reset, there was a 1998... there was this Asian contagion, the tiger countries had a structural crisis in their economies, big downturn in the industry. Nobody remembers this stuff."

Despite these cycles, Matt emphasizes the importance of maintaining conviction:

"You got to look at where your total opportunity is. You have to stay very committed to your R&D into your programs. You have to have a lot of conviction."

Timestamp: [05:11-07:04]

Matt explains why extraordinary persistence is essential in the semiconductor industry, where feedback loops are exceptionally long.

"You got to be super gritty to be in this industry because you have to have very strong conviction of what you're doing with little to no positive feedback or reinforcement along the way other than like customer input."

He elaborates on why these long development cycles require such resilience:

"The external world, they can't see it right until the stuff... the things you conceive about what you can do in terms of products go through their development cycle, come out the other side."

When Joubin asks if the lifecycle from conception to market is about 5 years, Matt breaks down the timeframe in detail:

"Let's say we engage in a new concept right now, some new type of chip... you might spend like six months vetting the idea, coming up with a product definition, getting some customer sponsorship. Some point you do your financial analysis... and then you commit this project and typical IC development cycle times are 18 to 36 months in terms of just a whole team working on a chip."

The complexity continues to compound after development:

"Once the product is out, especially now on these advanced technology nodes... the wafer fab cycle times are like 6 months just from when you put it in the fab, get it out. And then you got to package it, test it... even from when you think you're done and you've taped out the chip to TSMC or your foundry of choice, you're waiting like six months to nine months to even know if it works."

The process doesn't end there:

"And then assuming it works, you got to go get it in your customer hands, then they have to go through a qualification cycle. Even if they're jamming, that could take like in a hyperscale cloud account, that could take like a year sometimes."

Timestamp: [07:04-09:05]

Matt Murphy draws parallels between semiconductor leadership and venture capital, emphasizing his core philosophy on the CEO's most critical responsibility.

Joubin notes the similarity with venture capital's delayed feedback loops:

"It reminds me of the feedback loops of early stage venture capital where you make a bet, you make a bunch of bets, you have no idea for five years right, at least, if the bets are good or if you're good."

Matt agrees while noting you can still track progress along the way:

"You can checkpoint them, you can status them, you can milestone it, markup..."

He then reveals his primary focus when he took the CEO role at Marvell:

"When I took the job at Marvell, my concept really coming in, and I had sort of gotten religion on this even when I was at Maxim in my later years... I told the board when I was interviewing, I said, 'You know my view is my number one job is capital allocation. That's my number one job.'"

Matt explains why this capital allocation mindset is so crucial in semiconductors:

"Getting these bets correct on where you deploy your most precious resource, which is in a semiconductor company is your R&D expense, your new product development expense... given that you don't have a lot of flexibility once you commit to these, you got to get it right."

He emphasizes the consequences of poor capital allocation decisions:

"You steer it the wrong way and once it's steered the wrong way, it's really hard to turn the ship."

Timestamp: [09:05-10:26]

• Marvell Technologies is positioned at the center of the AI revolution with 80% of next year's revenue coming from data center and AI – giving them exposure similar only to NVIDIA
• Despite having their best quarter ever, Marvell experienced dramatic stock volatility – dropping from $100B to $50B market cap, demonstrating the importance of long-term thinking
• Under Matt Murphy's leadership, Marvell transformed from a $3.5B enterprise value company in 2016 to a $100B+ company at its peak
• The semiconductor industry is inherently cyclical, with Matt emphasizing that claims of reduced cyclicality are historically inaccurate
• Semiconductor leadership requires exceptional grit due to 5+ year development cycles with minimal feedback along the way
• The entire chip development process involves multiple stages: 6 months of planning, 18-36 months of development, 6-9 months in fabrication, and up to a year of customer qualification
• Matt considers capital allocation his #1 job as CEO, as directing R&D resources effectively determines the company's future path
• Once semiconductor development paths are set, they're extremely difficult to redirect, making initial strategic decisions critical

Timestamp: [01:41-10:26]

Companies:

• Marvell Technologies - The company led by Matt Murphy, positioned at the center of AI and data infrastructure
• NVIDIA - Mentioned as the only other company with comparable exposure to the AI and data center market
• TSMC - Referenced as a semiconductor foundry that manufactures chips for companies like Marvell
• Maxim - Matt Murphy's previous company where he spent two decades before joining Marvell

People:

• Matt Murphy - Chairman and CEO of Marvell Technologies since 2016
• Jim Cramer - Financial television host mentioned as giving Marvell recognition

Industry Concepts:

• Semiconductor cycles - Historical patterns of boom and bust in the chip industry, including specific downturns in 1996 and 1998
• Asian contagion/Tiger countries crisis - Economic crisis mentioned that affected the semiconductor industry in 1998
• Advanced process technologies - Manufacturing improvements like "finfet" and "gate all around" that increase chip complexity
• Tape out - The point when chip design is completed and sent to fabrication
• Hyperscale cloud - Large cloud service providers that are Marvell customers
• Capital allocation - The strategic deployment of resources, particularly R&D funding, which Matt considers his primary responsibility

Timestamp: [01:41-10:26]

Matt Murphy describes how his strategic vision for Marvell was shaped by recognizing the major shift in market value toward large technology platforms.

"When I joined the company... it's really important semiconductor companies what your end market exposure is," Matt explains, highlighting the critical importance of understanding where your revenue concentration lies. "The biggest sort of thing you got to look at is your revenue concentration and your opportunity by end market because the end market ultimately is what's going to determine your success."

When Matt became CEO, he found Marvell heavily exposed to consumer electronics markets with limited growth potential:

"By the time I became CEO, consumer electronics in general had become ex-unit growth. Like there's a kind of a fixed amount of smartphones that are selling every year, fixed number of PCs that hadn't changed in forever."

This realization that consumer markets offered minimal growth was particularly concerning given Marvell's business mix:

"At Marvell we were like 60-70% consumer when I joined," including phones, smart TVs, and Wi-Fi chips for various consumer devices. Matt remembers, "One of the employees was really proud and posted on LinkedIn that Marvell had been selected as a supplier for the Wi-Fi chips for the latest Mattel Barbie smart house, and I'm thinking to myself, 'I don't think this is where the future is here.'"

Matt's strategic insight centered on where the market was heading:

"My strong view in joining Marvell was that from the industry perspective, basically all the market cap in the world was going to shift to the big tech platform companies. It was already underway back then, but I don't think we understood how profound that shift was going to be."

Timestamp: [10:32-14:46]

Matt details how he rapidly developed and executed a bold transformation plan for Marvell, presenting it to the board just months after taking the CEO role.

"I joined in July of 2016, and I told the board in my December 2016 board meeting, so it was 5 months later... I told them that was my view, that I wanted—I believe Marvell with the right M&A strategy, because we couldn't do it on our own, could transform itself to be a leading provider of what I termed at that time, I called it the data infrastructure opportunity."

This strategy was clear and direct:

"So shift out of the consumer, harvest it, and go after the other one."

Matt came prepared with specific acquisition targets in mind:

"I listed four companies that I thought we should acquire. I still have the slide from like that board meeting. And those four companies were Cavium, Inphi, Aquantia, and Mellanox. And we actually ended up acquiring three of the four, and NVIDIA bought the fourth."

These acquisitions were ambitious and carried significant risk, particularly the Inphi deal:

"Inphi was the actually the highest multiple ever paid for a public asset in semiconductors. I mean, I took a huge risk on that. It was a $10 billion buy. It was supposed to do $840 million in the year ahead of us. It had done $680 or something in terms of revenues. So 13 times kind of sales multiple."

The gamble paid off tremendously:

"What a home run. I mean under Marvell, it's totally thrived. I mean, big round numbers, it's over tripled since we closed in 2021."

The first major acquisition, Cavium, was equally bold:

"That was like $6 billion and we paid 85% of our enterprise value to buy it. And I was CEO for like 18 months, so this was like a huge bet."

Timestamp: [14:46-17:00]

Matt elaborates on the strategic rationale behind the Cavium acquisition and how it fundamentally reshaped Marvell's business mix and prospects.

"Cavium was a pure play. They had no consumer," Matt explains. "They were like an $800 million revenue company that was going to grow, and they were in enterprise, they were in carrier, so they had a very good position in 4G and they were going to have a stronger position in 5G in terms of the infrastructure chips."

The acquisition brought critical new capabilities:

"They had a number of emerging very exciting technologies for cloud computing, for compute products, ARM-based microprocessors, there's an inference chip, there's a few different things they had."

This first major deal served multiple strategic objectives:

"That was a way to get leverage to the cloud, leverage to the carrier and the 5G opportunity, and then add some additional enterprise heft. And then by adding that revenue, which was 100% infrastructure and zero consumer, and I merge it with my consumer-heavy business, it shifted the mix."

Matt also reveals a shocking reality about Marvell's legacy business:

"Half of the company revenue, so most of that consumer business, was actually in storage. We were the leading company for storage controllers for hard disc drives in notebooks. So how many notebooks today have a hard disc drive? None. Back then it was still transitioning. And that was actually like a big chunk of my revenue and it was a huge portion of my operating income."

This dependence on a declining technology made transformation imperative:

"I knew I needed to do something right, 'cuz that was not going to be sustainable. I could manage it, but it certainly wasn't going to be a growth opportunity. It was going to probably shrink over time."

Timestamp: [17:00-18:41]

Matt describes the market's surprising response to Marvell's acquisition strategy and how these deals created substantial shareholder value despite their significant cost and risk.

When Marvell announced the Cavium acquisition, something unexpected happened:

"There was a multiple mismatch. So Cavium was trading at, say, I don't know, 18, 19, 20 times forward earnings. We were trading at 12 or 13. We literally when we announced it—it leaked first, and our stock appreciated quite—our stock appreciated when that happened. And then when we finally announced it, our stock ended up appreciating almost as much as the Cavium stock did."

This created a remarkable financial outcome:

"What happened is our—we got basically our multiple on the announcement rerated up to their multiple. So all that revenue rerated, and then we announced some pretty compelling cost synergies because there was overlap and things we were doing, and we got credit for all the synergies up front."

The result was dramatic for Marvell's share price:

"By the time we did the Cavium thing, the stock was like in the low 20s where it had been like nine bucks."

Despite the boldness of these moves, Matt maintained financial discipline:

"I've always maintained investment grade, like that's a kind of a line in the sand I drew from the very beginning. If I'm going to do an M&A strategy, I want to never get—never get over my skis."

This disciplined approach to transformation has radically changed Marvell's business mix:

"Now where 95% of our revenue is infrastructure and five is consumer."

Timestamp: [18:41-20:28]

• Matt Murphy's strategic vision recognized early that market capitalization would shift dramatically toward big tech platform companies, driving semiconductor demand toward cloud computing and infrastructure
• Within 5 months of becoming CEO, Matt presented a bold M&A strategy to transform Marvell from a consumer-focused company to a data infrastructure leader
• Marvell was 60-70% consumer-focused when Matt joined, with significant revenue dependent on hard disk drive controllers for notebooks—a technology rapidly becoming obsolete
• The Cavium acquisition ($6B) was a massive bet representing 85% of Marvell's enterprise value at the time, but dramatically shifted the company's market exposure
• The Inphi acquisition ($10B) set a record for the highest multiple ever paid for a public semiconductor company (13x sales) but tripled in value under Marvell's ownership
• When the Cavium acquisition was announced, the market "rerated" Marvell's stock multiple higher to match Cavium's, creating immediate shareholder value
• Matt maintained investment-grade debt levels throughout the transformation, setting it as a "line in the sand" for financial discipline
• Through strategic acquisitions, Marvell transformed from 60-70% consumer exposure to 95% infrastructure and only 5% consumer

Timestamp: [10:32-20:28]

Companies:

• Marvell - The company Matt Murphy transformed from consumer-focused to infrastructure-focused
• Cavium - An $800M revenue company acquired by Marvell for $6B, representing 85% of Marvell's enterprise value
• Inphi - Acquired by Marvell for $10B (13x sales multiple), has since tripled in value
• Aquantia - One of the four companies on Matt's initial acquisition target list
• Mellanox - The fourth company on Matt's acquisition list, ultimately purchased by NVIDIA
• Mattel - Mentioned in context of Marvell supplying Wi-Fi chips for a Barbie smart house
• Cisco - Mentioned as an enterprise customer of Marvell
• Juniper - Mentioned as an enterprise customer of Marvell
• Meta - Referenced as example of big tech companies gaining market cap
• Amazon - Referenced as example of big tech companies gaining market cap

Industry Concepts:

• End market exposure - A critical factor in determining a semiconductor company's success
• Consumer electronics - Market with limited unit growth that Marvell was heavily exposed to
• Data infrastructure - The term Matt used for the strategic opportunity he wanted Marvell to pursue
• Hard disk drive controllers - A major portion of Marvell's revenue that was at risk as notebooks transitioned to SSDs
• 5G infrastructure - Growth opportunity targeted through the Cavium acquisition
• ARM-based microprocessors - One of the strategic technologies acquired through Cavium
• Investment grade - The debt rating level Matt insisted on maintaining throughout acquisitions

Timestamp: [10:32-20:28]

Matt Murphy explains how his strategic vision for Marvell aligned perfectly with the rise of AI, creating an enormous market opportunity beyond his initial expectations.

Matt recalls his original strategic vision for Marvell from 2016:

"I told the board my vision for the company and I basically said look, we want to be the leading semiconductor company and provider of solutions that move data, store data, process data and secure data, do it faster, more reliably, better than anybody else."

This data-focused vision positioned Marvell ideally for what was to come:

"We got to 2020, 2021, and then the advent of the sort of acceleration in AI, and it's like AI literally became the ultimate killer data infrastructure app for sure, because you need all those things—moving data, storing data, processing data, securing it. It's what the whole thing is run on in terms of the complexity of those systems."

Matt describes how their acquisition of Inphi turned out to be even more strategically valuable than they initially projected:

"When we acquired it, most of our DCF on that company was based on their position in the traditional cloud networks which were all transitioning to a new type of optical interface... and by the way there's this AI kind of kicker if it ever really hit. That's where I was in 2020. But it wasn't an accident in terms of how we developed. So Inphi and then Marvell, we stayed very engaged, and then boom, when this all changed, our business just rocketed, but because we were in those designs already."

Timestamp: [20:36-24:00]

Matt provides a historical perspective on why semiconductors have been critical to economic growth for decades, but explains how the AI boom differs fundamentally from previous technology cycles.

"Fundamentally, at least for the last 30-40 years, semiconductor technology has been one of the key growth drivers of global economic growth. And it's the foundation and basis for almost everything we do."

He explains how previous technology cycles benefited semiconductor companies more broadly:

"Having been through all these different products cycles, there's different benefits it's brought to consumers and human beings. But in every one of those cycles, there was a huge sort of value attribution if you will to a certain set of companies."

Matt provides examples from his career:

"Notebook PCs, when I started in the chip industry, were a luxury item, they were very expensive, they were heavy. That trend happened, hundreds of millions of notebooks started shipping. Intel benefited big time, but so did the rest of the ecosystem. The memory vendors—I was at an analog company, we were doing all the power management back then for notebooks, helped us tremendously."

This broad industry benefit occurred in multiple waves:

"The networking wave, or optical communications, the cloud computing trend, smartphones—there's always this sort of big boon that comes from that into the chip industry."

However, Matt highlights how the AI revolution differs critically:

"The interesting thing in those prior cycles is it was very broad-based, like almost every semiconductor company had their own angle in one of these types of applications. So the whole industry gets lifted. But the AI opportunity is so unique that it's a lot more value being created because the chip technology has a bigger lever on the end result, which is ultimately productivity—I'm talking about like global productivity—and there's fewer companies that possess that unique technology."

This has created a significant division within the semiconductor industry:

"In the AI wave, it's a big disconnect between the top three, four, five companies that are really in there and then everybody else who has some exposure to it, but it's not enough to move their needle because they're in other end markets too."

Timestamp: [24:00-26:15]

Matt describes the complex, tightly integrated technology stack that powers AI systems and why only a handful of companies can participate meaningfully in this ecosystem.

"If you think about the value that GenAI is creating and where that's coming from, it is coming from the fundamental, at the most basic layer, the coupling of breakthroughs in high performance computing, which NVIDIA has really driven. But now other people are also coming in with their own homegrown solutions, the large cloud providers. But NVIDIA pioneered all this in terms of the compute layer."

Matt also highlights the critical role of networking and connectivity:

"Coupling it and going back to Mellanox which they did very well, coupling it with the networking, the connectivity, the overall system design. And we participate in that too, within some within their ecosystem, but also within all the other cloud providers."

What makes this technology so valuable and difficult to replicate is its integrated nature:

"It's just this very complex, very kind of tightly-knit closed system that only a few companies have the basic technology to provide it all."

Joubin draws a parallel with Cisco's role in the internet boom, asking if the current AI infrastructure providers might face a similar fate after their period of dominance. Matt acknowledges the parallel while highlighting differences:

"There are parallels in terms of, has there been a buildout cycle and capex and money being spent and then is there digestion and then there's economic factors that happen in the sort of networking boom, which I referred to and you kind of brought back with the Cisco story. And now we're in an AI capex boom right now, and at some point there'll be a digestion, there'll be some moderation like you would normally expect."

However, Matt sees a critical difference in the scale of opportunity:

"The interesting one is that on the AI side, I mean there is—and you pick your number, you guys have number in your firm and others—but there's, I don't know, four or five trillion dollars of productivity out there that AI could really go after in the world. It really could create a massive, for those that win in it, it could create massive, massive economic opportunity."

Timestamp: [26:15-30:03]

Matt describes a fundamental shift in what drives semiconductor advancement, moving from consumer electronics to AI and high-performance computing.

"For the last 20-plus years, what drove Moore's Law was consumer electronics. In the 90s and early 2000s, it drove Intel's road map in terms of personal computing, right? That's what drove Moore's Law, bleeding edge process technology."

This driver shifted with the rise of mobile devices:

"With the advent of smartphones and sort of that whole—and just higher-end feature phones and TSMC's rise—their leading edge technology was all mobile focused."

Now, Matt points to another fundamental shift in what's driving semiconductor advancement:

"If you look now at TSMC, the majority of their revenue is now coming from what they call HPC—high performance computing—which is this stuff, data infrastructure. And so now the bleeding edge process nodes and packaging, it's all driven by AI."

The barrier to entry in this sector is substantial:

"I've been doing this a long time, we're in this business and we are at the bleeding edge, and this is not for the faint of heart. The moat that's created around the companies that have the right team, the right IP, the right skill set, the right scale to invest in this technology and drive the breakthroughs—there's no best practice to follow right now."

Timestamp: [30:03-31:31]

• Matt's original vision for Marvell to be the leader in moving, storing, processing, and securing data perfectly positioned the company for AI, which emerged as "the ultimate killer data infrastructure app"
• The acquisition of Inphi, initially valued mainly for its traditional cloud networking technology, delivered unexpected additional value when AI accelerated
• Unlike previous technology cycles that benefited the semiconductor industry broadly, the AI wave concentrates value among a small number of companies with specialized capabilities
• Historical technology cycles (PCs, networking, smartphones) created opportunities across the semiconductor ecosystem, while AI requires more specialized, complex technology from fewer providers
• The economic opportunity from AI productivity gains could reach $4-5 trillion globally, creating massive potential for companies that can enable these capabilities
• A fundamental shift has occurred in what drives semiconductor advancement: from consumer electronics (PCs in the 90s, smartphones in the 2000s) to high-performance computing for AI today
• TSMC's revenue now comes primarily from high-performance computing rather than mobile, reflecting this shift in technology drivers
• Companies in the AI semiconductor space have created significant competitive moats through specialized expertise, with "no best practice to follow" in this rapidly evolving field

Timestamp: [20:36-31:31]

Companies:

• Marvell - Matt Murphy's company that positioned itself strategically for the AI revolution
• Inphi - Company acquired by Marvell whose value increased dramatically with AI acceleration
• NVIDIA - Pioneer in high-performance computing that drove breakthroughs enabling AI
• Mellanox - Company acquired by NVIDIA that provided critical networking technology for AI systems
• Cisco - Former CEO John Chambers mentioned; company's role in the internet boom compared to NVIDIA/Marvell in AI
• Intel - Referenced as driving Moore's Law through PC chips in the 1990s and early 2000s
• TSMC - Taiwan Semiconductor Manufacturing Company whose revenue now predominantly comes from high-performance computing rather than mobile
• OpenAI - Mentioned as one of the few AI companies capturing value beyond the semiconductor layer
• Anthropic - Mentioned alongside OpenAI as capturing value in the AI revolution

Industry Concepts:

• Moore's Law - The observation that the number of transistors on a chip doubles approximately every two years
• High Performance Computing (HPC) - The category of computing now driving TSMC's revenue, focused on AI and data infrastructure
• PAM4 modulation technology - Advanced optical interface technology that Inphi had led in developing
• GenAI - Generative AI, the technology creating massive economic opportunity
• Bleeding edge process nodes - The most advanced semiconductor manufacturing techniques
• Productivity gains - The potential $4-5 trillion economic impact of AI technologies
• Best practice - Matt notes there is "no best practice to follow" in the rapidly evolving AI semiconductor space

Timestamp: [20:36-31:31]

Matt Murphy draws parallels between the semiconductor industry's position in the tech ecosystem and Cisco's role during the internet boom, highlighting how infrastructure providers remain valuable regardless of which specific applications dominate.

"Cisco didn't care if Myspace or Facebook was going to become the social network player... Cisco didn't care if Napster or Spotify was going to win... Cisco didn't care if Airbnb existed... or if it was Google or the 19 search engines before. At some point like you're abstracted from the winners and losers."

Matt confirms this applies to Marvell's position in the AI ecosystem as well:

"We are very agnostic to those types of things you mentioned. We supply our fundamental technology in in some different forms, right? Whether it's sort of merchant products that we develop, or we do a lot of custom silicon products."

This technology-agnostic positioning means Marvell benefits regardless of which specific AI applications ultimately succeed:

"Whatever model wants to run, whatever killer app wants to run, whatever the new thing in the future on the app store that's going to be the number one app that's using generative AI and the latest models to go make something magical happen—our view is that'll all at some point, all those little hops that go through the data center and all the proc—that'll run on Marvell technology or a competitor."

Joubin notes the current imbalance between supply and demand: "Right now it's Marvell and all of your competitors and there's still not enough chips."

Timestamp: [31:38-33:15]

The conversation shifts to TSMC's pivotal role in the global semiconductor ecosystem and the growing geopolitical concerns around chip manufacturing concentration, with Matt discussing his involvement in industry policy.

Matt begins by sharing his long-held position on the semiconductor industry's importance:

"I had this belief that the US semiconductor industry was critical for US and I'd say Western nations at large national and economic security."

He highlights how semiconductor technology has been a driver of economic growth globally:

"What has driven significant growth in economies, which has created economic opportunity for everybody, is countries that adopted and invested in leading edge technology. And you saw this, by the way, in Southeast Asia—look at the rise of Korea, but it's on the back of semiconductor technology and consumer electronics and advanced manufacturing. The rise of Taiwan, same thing."

Matt explains when these concerns gained higher-level government attention:

"It did become a very critical issue which really started, I think the recognition at the national level right at the end of the Obama administration, his second term."

Matt shares his extensive involvement in these industry policy discussions:

"I've been on the board as an alternate, then a board member of the US Semiconductor Industry Association since 2011. I'm actually, I think, the longest serving current member now if you can believe that. A lot of guys have come and gone and retired."

Discussing the geopolitical shift, Matt notes:

"In 2016 when I became CEO, the world was wide open for business. We had a huge team at Marvell in China, big R&D team, big salesforce. I was going there once a quarter, the world was open, free trade was everywhere. And you fast forward a couple years and all of a sudden, tension, right? And the rivalry emerged and two different sort of points of view on how US and China were going to go forward, and they were no longer completely compatible."

Timestamp: [33:15-36:46]

Matt provides a rich historical perspective on the semiconductor industry's policy challenges, drawing parallels between previous U.S.-Japan tensions and current U.S.-China relations, while explaining the origins of industry advocacy.

Matt emphasizes that industry concerns predated broader political awareness:

"These are things we were worried about as a semiconductor industry long before 2016, 2017, and not in a protectionist way. I mean, this is like historically been like a free trader—global markets drive growth, every time a big market opens up for semiconductors everybody wins. But it was to the point where a lot of things were happening that just looked very anti-competitive."

He explains his leadership role during a critical period:

"I was the chairman of the US SIA in 2018." He defines the organization: "That's the Semiconductor Industry Association, and it's a pretty cool story. It was formed in the 1980s by the founders—by Robert Noyce who was founder of Intel, Jerry Sanders from AMD, a bunch of industry legends and icons that were very competitive with each other, big rivalries."

Matt draws a compelling historical parallel:

"In the 1980s if you study it, there was a similar set of issues between the US and Japan in terms of semiconductor technology and misappropriation of IP and patent violations, and ultimately accusations of dumping, which means a company would come and sell something below product cost in a region to gain share and starve the competitor, and using non-market things like subsidies."

The situation became serious enough to warrant presidential intervention:

"It got so serious in the US that actually Ronald Reagan got involved and there was tariffs put on Japanese automobiles and there's a whole kind of defensive strategy that formed. And the US SIA was the one that sort of blew the whistle and said if we don't do something, this anti-competitive behavior is going to put us out of business, and we can't afford that from a national security standpoint."

Matt describes the SIA's ongoing approach to industry challenges:

"The industry came together and formed this organization called the Semiconductor Industry Association. It's comprised of CEOs of the largest companies, and you've always got a full board and really amazing people. And we all sort of put down our differences and we're looking out for the industry."

Timestamp: [36:46-39:31]

Matt details the development of the CHIPS Act as a strategic response to growing concerns about semiconductor manufacturing concentration and international competitiveness.

Matt explains the dual approach needed to address industry challenges:

"What's your defensive strategy and then what's your offensive strategy? You got to do both right, you can't just sort of complain, you've got to also advance your own agenda."

He outlines the origins of the CHIPS Act:

"The CHIPS Act was part of that. That was actually conceived during the first [Trump] administration. The concept was very well understood, which was we were going to need a way to incentivize companies to, if they've got a decision to make about where to put factory capacity and so forth, put it in the US. And then what would that take?"

Matt notes the bipartisan nature of the legislation:

"It didn't get enacted and passed by Congress until the Biden administration, but they really built on the work of the first term. And by the way, I suspect if it gets—it's going to probably get renamed and maybe recast a little bit, but I think even in the current administration, I think there's very much an understanding that this is still really important."

He clarifies the core purpose of these initiatives:

"An incentive structure to make sure that we are less dependent on foreign manufacturing, and in particular the hot point going back to Taiwan is that Taiwan being in the—less dependent on TSMC in Taiwan."

Timestamp: [39:31-40:45]

• Semiconductor companies like Marvell are "abstracted from the winners and losers" in AI applications, similar to how Cisco benefited regardless of which specific internet applications succeeded
• Marvell supplies fundamental technology that powers AI infrastructure regardless of which specific models or applications become dominant
• The geopolitical landscape for semiconductors has transformed dramatically, from a globally open market in 2016 to one of increasing tensions and trade restrictions
• Matt has been involved in semiconductor industry policy as a board member of the Semiconductor Industry Association (SIA) since 2011
• The SIA was formed in the 1980s by industry legends like Robert Noyce and Jerry Sanders to address similar competitive issues with Japan that the industry now faces with China
• The CHIPS Act originated during the Trump administration as a way to incentivize domestic semiconductor manufacturing but wasn't passed until the Biden administration
• There appears to be bipartisan recognition of the importance of reducing dependence on foreign manufacturing, particularly from TSMC in Taiwan
• Countries that have invested in leading edge semiconductor technology (Taiwan, Korea, Singapore) have seen significant economic growth and prosperity

Timestamp: [31:38-40:45]

Companies:

• Marvell - Matt Murphy's company that provides semiconductor technology regardless of which specific AI applications succeed
• TSMC - Taiwan Semiconductor Manufacturing Company, critical for global chip production and a focus of supply chain security concerns
• Cisco - Used as an analogy for how infrastructure providers can thrive regardless of which applications win
• Facebook - Mentioned as an example of a social network winner that Cisco would have supported regardless
• Myspace - Mentioned alongside Facebook as a platform Cisco was agnostic about
• Spotify - Mentioned as a music streaming service winner
• Napster - Early music sharing service mentioned alongside Spotify
• Airbnb - Mentioned as another application layer company
• Google - Referenced as a search engine winner
• Huawei - Mentioned in context of trade tensions ("peak Huawei")
• Intel - Co-founder Robert Noyce helped establish the Semiconductor Industry Association
• AMD - Founder Jerry Sanders helped establish the Semiconductor Industry Association

People:

• Robert Noyce - Intel co-founder who helped establish the SIA in the 1980s
• Jerry Sanders - AMD founder who helped establish the SIA
• Ronald Reagan - President who imposed tariffs on Japanese automobiles during 1980s semiconductor trade tensions
• Barack Obama - Recognition of semiconductor industry issues began toward the end of his administration

Industry Concepts:

• Semiconductor Industry Association (SIA) - Industry trade group formed in the 1980s where Matt served as chairman in 2018
• CHIPS Act - Legislation to incentivize domestic semiconductor manufacturing
• Dumping - Anti-competitive practice of selling products below cost to gain market share
• Custom silicon - Specialized chips designed for specific customers, an area where Marvell operates
• National and economic security - Key drivers for semiconductor policy
• IP misappropriation - Intellectual property concerns that have driven trade tensions

Timestamp: [31:38-40:45]

Matt Murphy explains the critical role of TSMC in advancing semiconductor technology and the strategic importance of Moore's Law in the context of AI computing.

"TSMC [is the] undisputed leader at the moment in driving Moore's Law and process technology leadership."

When asked to define Moore's Law, Matt provides a comprehensive explanation:

"Moore's Law was a concept that was coined by the co-founder Robert Noyce's co-founder of Intel, named Gordon Moore, who posited concepts around by driving improvements in process technology—wafer process technology—and improving the transistor density of those processes, which would enable more computing power on every successive generation, more power at a lower cost if you will, because you get more density, more die-per-wafer."

The result of these improvements created a compounding effect:

"You could basically have sort of a doubling effect every few years in terms of your power performance in area. And that's continued today, now it's slowed."

Matt highlights a significant competitive shift in the industry:

"TSMC took the lead over Intel in this maybe seven or eight years ago something like that, and Intel has not been able to catch up."

He then explains how Moore's Law applies differently to AI applications today:

"What's happened with these AI applications is even though the technology is very expensive and you're not getting the kind of cost per transistor improvement anymore, the power per millimeter square is still improving like 20-30% per generation."

This power efficiency has become the new driving force for semiconductor advancement:

"Maybe you're not making a chip any cheaper when you go from one process node to the other, but if you can save 20-30% power consumption, and that's direct opex for a data center provider—there's a rule of thumb that says basically every watt you can save of power is about a dollar of opex."

This creates a compelling economic case even without cost reduction:

"You might not be able to benefit by getting like a lower cost solution anymore, it's probably going to get more expensive actually, these chips, but the power savings is so compelling, the economics pencil out."

Timestamp: [40:50-42:59]

Matt discusses the recent dramatic stock decline of Marvell, explaining the context of semiconductor company valuations and market expectations.

Joubin raises the challenging question:

"Why is the stock getting hammered? Like why are you down 50% in a few months?"

Matt responds thoughtfully:

"It's always fun to speculate on why you're up and why you're down. I think in general, if you look—I mean, I'll just give you some empirical data. For like the last 5 years if you look at all semiconductor companies, there have been three companies that always traded in a very similar PE multiple range that were different than everyone else, and that was Marvell, AMD, and NVIDIA."

He details the valuation premium these companies have traditionally enjoyed:

"The three of us have historically for like the last 5-6 years commanded something in the 30 to 35 times range, when the average for the overall semiconductor index is maybe 20 or low 20s, which is a little elevated of where it usually was. Usually that index trades at like 17-18 times."

Matt explains what happened during Marvell's peak:

"When we announced our December quarter, our PE... probably went to like 45 times when you got to a hundred billion dollar market cap."

The subsequent market correction has been dramatic:

"Now with all the global economic concerns, there's been an AI trade-off that's happened, the tariffs—you name it—multiples have come down. So now what was normally 30-35, shot to 45, is like 18. And by the way, that's kind of where those other companies—they're trading at too, NVIDIA and AMD."

Beyond multiple compression, Matt acknowledges broader market concerns:

"I think Mr. Market is also saying—and this is a broad statement for semiconductor companies—'Hey, there's a lot of risk of recession and what's going to happen in the global economy.' So probably Mr. Market's got the 'E' part of it lower in the future. It doesn't quite match the sell-side numbers."

Despite these market fluctuations, Matt emphasizes his team's focus on execution:

"Our view is we've got to just execute. We got to deliver on the revenue we committed, we got to meet these critical programs for our customers, and we got to make sure we stay three steps ahead so that we stay well positioned for this—I think at least kind of 10-year cycle we're on relative to the AI spend and where the technology is going. And we're just going to power through it."

Timestamp: [42:59-47:02]

Matt discusses Intel's importance to the American technology ecosystem and reflects on his personal connection to Silicon Valley's semiconductor heritage.

Joubin mentions an article stating that retail investors view Marvell as having the second-best AI growth story behind NVIDIA, and also referenced rumors about Matt being considered for the Intel CEO position.

Matt sidesteps commenting directly on the CEO rumors but offers profound respect for both Intel and its current CEO:

"Pat's a good friend. I'm very happy for him. I think Intel is really blessed to have someone of Pat's capability and character."

He emphasizes Intel's critical importance to American technology:

"We need this company. We absolutely—it's arguably the most important U.S. manufacturing and technology company we've had."

When Joubin suggests it must be flattering to be associated with such an institution, Matt expresses deep appreciation for Intel's legacy:

"Oh my God, absolutely. We talk about people like Robert Noyce as an example, or Andy Grove, who was always sort of one of my role models."

Matt shares a personal story about meeting Andy Grove:

"I got a chance when I attended the Stanford Executive Program in 2010, Andy was still alive and came and spoke to us, group of like 80 people or something. It's one of the highlights of my life to be able to—and I got to meet him afterwards and read all the books."

He acknowledges the unexpected honor of being mentioned as a potential Intel CEO:

"It's a legendary company. I never thought I would wake up one day and see a Bloomberg article with my name in it associated with Intel, and it's flattering to even have that sort of out there."

However, Matt reaffirms his commitment to Marvell:

"The reality is—and I've kind of given you some of this background now on the Marvell story—it's really a special place to me. I came in, there was a lot of issues going on, I left a great job I had, I took a lot of risk, and I'm just so excited with where we are."

He recalls his public response to these rumors:

"Somebody asked me about it right around when that thing came out on an earnings call, and I just said 'Look, I'm not commenting on anything, but I'm all in.'"

The conversation concludes with reflections on Silicon Valley's semiconductor heritage:

"We forget... it's called Silicon Valley. It's the silicon, and Silicon Valley. I mean, it's such an amazing place to have worked for the last 30 years. I take a lot of pride and I'm like so lucky to be in this industry. I think it's really a kind of a golden era for semis."

Timestamp: [47:02-51:12]

• TSMC has overtaken Intel as the "undisputed leader" in advancing Moore's Law and semiconductor process technology for approximately the last 7-8 years
• In AI applications, the economic value of chips is increasingly driven by power efficiency (20-30% improvement per generation) rather than cost reduction, with every watt saved worth approximately one dollar in operational expenses
• Marvell, AMD, and NVIDIA have historically traded at premium valuations (30-35x PE) compared to the broader semiconductor index (17-20x PE), reflecting their growth trajectories
• During its peak, Marvell's PE ratio expanded to approximately 45x, contributing to its $100 billion market cap, before contracting to around 18x amid broader market concerns
• The current stock decline for semiconductor companies reflects both multiple compression and market concerns about potential recession risks
• Intel remains "arguably the most important U.S. manufacturing and technology company," with Matt expressing respect for its current CEO Pat Gelsinger
• Despite market volatility, Matt remains focused on execution, meeting customer commitments, and staying "three steps ahead" to position Marvell for what he sees as a "10-year cycle" in AI
• Silicon Valley's name directly reflects its semiconductor heritage, with Matt viewing the current period as a "golden era for semis" despite previous challenging cycles

Timestamp: [40:50-51:12]

Companies:

• TSMC - Taiwan Semiconductor Manufacturing Company, described as the "undisputed leader" in driving Moore's Law and process technology
• Intel - American semiconductor company that lost process leadership to TSMC around 7-8 years ago; described as "arguably the most important U.S. manufacturing and technology company"
• Marvell - Matt Murphy's company, identified as having "the second best AI growth story" behind NVIDIA according to retail investors
• AMD - Semiconductor company mentioned as having similar valuation patterns to Marvell and NVIDIA
• NVIDIA - Leading semiconductor company in AI, mentioned as sharing similar valuation patterns with Marvell and AMD
• Kleiner Perkins - Venture capital firm mentioned as having "a long history with Andy [Grove]" and origins connected to the semiconductor industry
• Bloomberg - News outlet that published an article connecting Matt Murphy's name with the Intel CEO position

People:

• Gordon Moore - Co-founder of Intel who formulated Moore's Law, the observation that transistor density doubles approximately every two years
• Robert Noyce - Co-founder of Intel mentioned alongside Gordon Moore
• Pat Gelsinger - Current Intel CEO, described by Matt as "a good friend" with "capability and character"
• Andy Grove - Former Intel CEO described as one of Matt's "role models"; Matt recalls meeting him during the Stanford Executive Program in 2010
• "Mr. Market" - Personification of market sentiment used by Matt to explain valuation changes

Industry Concepts:

• Moore's Law - The observation that transistor density doubles approximately every two years, driving improvements in computing power
• Process technology - Manufacturing techniques that enable semiconductor advancement
• PE (Price-to-Earnings) ratio - Valuation metric discussed extensively; Marvell, AMD and NVIDIA historically traded at 30-35x, peaked at 45x, now at around 18x
• Power per millimeter square - Key metric for AI chips improving 20-30% per generation
• Opex (Operating Expense) - For data centers, power consumption is a direct operating expense
• Die-per-wafer - Manufacturing efficiency measure related to how many chips can be produced from a single silicon wafer
• Sell-side numbers - Analyst estimates that may not match current market expectations

Timestamp: [40:50-51:12]

I'll create a new set of cards for this segment, following the specified format requirements, and include a references card with only the information from this current segment.

Matt Murphy shares the challenging circumstances he faced when taking over as CEO of Marvell, highlighting the significant risks and problems that needed to be addressed.

"Marvell was founded in 1995, actually it's our 30-year anniversary this year," Matt begins, noting a fitting coincidence: "Our little theme of the whole 30-year anniversary we came up with was 'all-in.'"

While Marvell had a strong engineering foundation, Matt describes the serious problems that emerged before his arrival:

"Marvell always had a very strong reputation as a really capable engineering company, products were good, very innovative culture. But they ran into some trouble and in 2015 there was a whistleblower complaint about somebody in the finance team that there was revenue ship-aheads happening."

This triggered a cascade of governance issues:

"That was kind of declared in front of the company auditor at the time, which was Price Waterhouse, and basically a whole saga ensued. The CFO left, the auditor resigned, at some point the board fired the two founders, which was a husband and wife—Sehat and Weili. They remained on the board."

The situation grew more complex with activist investor involvement:

"An activist investor came in, which was Starboard Value. They bought a chunk of the company, they said basically, 'Deal with us or we're going to run a proxy contest.' The company settled with Starboard to add new directors."

Matt initially hesitated when approached about the CEO role:

"I got a call to interview to be the CEO, and I thought, 'I'm not sure I want to touch this.' But something intrigued me."

He sought counsel from several people, including the incoming lead director:

"I remember interviewing with him and he was a former senior partner at a big audit firm, and I said, 'Hey, how bad is it?' And he said, 'I'm not sure I've ever seen any more hair on anything than this one, but it's fixable.'"

Despite these severe challenges, Matt saw potential:

"I looked at it and I thought, 'Okay, there's potential here.' I think I could take—it's big enough in terms of the scale, there's enough sort of operating cash flow once I clean it up, and I could probably use this as a platform to reshape what I thought. And I did it, played out, but it wasn't clear."

The magnitude of these challenges was extraordinary:

"Everybody was terrified. Can you imagine we had no financials on file for like three quarters, we had a delisting notice from NASDAQ, and I had no CFO."

Timestamp: [51:18-55:36]

Matt shares how his father's influence as a technology executive shaped his career path, while explaining that becoming a CEO wasn't necessarily his explicit goal.

When Joubin asks if Matt's father was a CEO, Matt confirms:

"My dad was a longtime tech executive. He had worked at Apple in their heyday when Steve was there, and then he became the CEO of a company. At one point he was head of Apple USA in kind of the end of his career, and then he left and he became CEO of a modem company."

Joubin wonders if Matt always aspired to follow in his father's footsteps:

"When you were growing up then, like you always aspired to be kind of like a business person like your dad? Like, did you have the CEO checkbox in your mind?"

Matt responds thoughtfully:

"Maybe in the back of it. I think I never—it's fair by the way to have that, people are shy—but no, I think I never thought I would be capable of it necessarily. But I always admired my father."

His relationship with his father deepened as Matt entered the business world:

"He was a huge influence on me. In particular, I think we grew the closest actually after I started working because he was so experienced, and so I would like—well, first of all, I lived at home for the first year I was working."

This arrangement facilitated daily conversations with his father:

"I'd see him every night and talk to him every day. I'd say for—he got very sick later, he got Parkinson's in like the 2010 time frame, maybe a little bit earlier—but there was a period of time where I talked to him almost every day."

Matt describes how his father's mentorship evolved as his own career progressed:

"It was kind of interesting as somebody who had no experience and all of a sudden I started like managing people, and then I kind of became like a director level person, and it's almost like along the way there was a new layer to unfold. He didn't expose me to everything he knew, you know what I mean? So as I progressed until I became a CEO, he really helped me."

Timestamp: [55:36-57:45]

Matt shares the invaluable career advice and leadership lessons he received from his father, which helped shape his approach to his professional journey.

Matt recounts how his father helped temper his early career impatience:

"He really helped me, and he helped me in a lot of ways with my impatience and understanding that when you're young and you're ambitious and your career is going like this, a lot of times it takes a while for the management to figure out like actually how good a job you're doing."

His father offered a particularly memorable perspective on being underpaid early in a role:

"He says, 'You should be—it's like an honor if you're the lowest paid guy in any given position you have, because that means you got there early and you'll eventually catch—they'll all catch up. Don't worry.' He couldn't have been more right."

Matt has tried to pass this wisdom along:

"That's a really good [lesson]—and I try to tell that to people that have worked for me, because I was that guy. 'I can't believe I didn't get promoted, I can't believe I didn't get this or that.'"

The breadth of his father's mentorship was comprehensive:

"Great lessons in leadership, great lessons in management, great lessons in patience and perspective, and how to conduct yourself. It really helped me to have a mentor like that."

Timestamp: [57:45-58:45]

Matt shares an emotional story about showing his father the Marvell headquarters after becoming CEO, creating a powerful moment of connection despite his father's declining health.

When asked if his father was still alive when Matt became CEO of Marvell, Matt explains:

"He was. He passed away in 2019, and he was in pretty significant decline when I became CEO in 2016, but he was still with it enough so I could have a conversation with him, but it was a little slow and stilted. I think his cognitive function was still pretty good, but it was just hard to communicate."

Despite these challenges, Matt arranged a special visit:

"I started in July, and I think within a few weeks I said, 'Hey, can I—I'd love to show you the place.' So he had a great—we had a couple of people that were helpers for him to get around. A lot of mobility issues. So he came over to Marvell on a Saturday, and I met him."

Matt describes the impressive Marvell campus:

"If you remember a company called 3Com, they had built a huge, beautiful facility right off 237 in Great America. And basically Marvell bought it in 2005. We own this facility. It's like a million square feet, acres and acres campus. And it's just physically a huge building. When people know it, they're like, 'Oh my God, that's like a huge company.' It almost helps you punch above your weight in a way."

This visit led to an emotional moment:

"He came over and he was just like, 'You're the CEO of this place?' And then I took him around and showed him the whole tour of the facility. I took him through my building, showed him the boardroom, my office. And we were standing there kind of looking at—it's right on 237, so you look out over the bay—and we were standing in the boardroom. I remember just him and I, and he was like tearing up and he's like, 'I just can't—I can't believe what you've accomplished. I'm so proud of you.'"

Matt reflects on the significance of this moment:

"It was very cool. So he did get to see it. And yeah, I think—and it was great 'cause all that advice and impact he made on me, it really—it made a difference."

Timestamp: [58:45-1:01:18]

• Marvell was in serious crisis when Matt became CEO, with accounting irregularities, the departure of founders, activist investor involvement, no filed financials for three quarters, and a NASDAQ delisting notice
• Despite the tremendous risks, Matt saw Marvell's strong engineering reputation and operational cash flow as a potential platform to execute his vision for transformation
• Matt's father was a significant influence on his career, serving as a mentor and confidant as Matt progressed through increasingly responsible roles
• One of the most valuable lessons Matt learned from his father was patience with career progression—being "the lowest paid guy" in a position means you got there early and will eventually catch up
• Matt's father had been a technology executive who worked at Apple during Steve Jobs' era and later became CEO of a modem company
• Matt didn't explicitly aspire to be a CEO but was deeply influenced by his father's example and guidance
• Despite his father's declining health due to Parkinson's disease, Matt was able to show him the Marvell headquarters after becoming CEO, creating an emotional moment when his father expressed pride in his accomplishments
• Marvell's impressive campus, a former 3Com facility with approximately one million square feet, helped the company "punch above its weight" in terms of perception

Timestamp: [51:18-1:01:18]

Companies:

• Marvell - The company Matt Murphy leads as CEO, founded in 1995 and celebrating its 30th anniversary with the theme "all-in"
• Price Waterhouse - The auditor for Marvell that resigned following the whistleblower complaint
• Starboard Value - Activist investor that bought a stake in Marvell and pressured for board changes
• NASDAQ - Stock exchange that issued Marvell a delisting notice prior to Matt joining
• Apple - Company where Matt's father worked "in their heyday when Steve was there" and later as "head of Apple USA"
• 3Com - Former technology company whose facility Marvell purchased in 2005, described as "a million square feet, acres and acres campus"

People:

• Sehat and Weili - The husband and wife founders of Marvell who were fired by the board but remained on the board
• Rick Hill - Former chairman of Marvell who had been chairman and CEO of Novellus Systems, described as "very successful" and "well known in the valley"
• Mike Strachan - Matt's lead director, a former senior partner at a big audit firm who told Matt about Marvell's problems: "I'm not sure I've ever seen any more hair on anything than this one, but it's fixable"
• Steve [Jobs] - Referenced as being at Apple when Matt's father worked there
• Matt's father - A technology executive who worked at Apple, became CEO of companies, provided valuable mentorship to Matt, and passed away in 2019 after suffering from Parkinson's disease

Industry Concepts:

• Revenue ship-aheads - Accounting issue identified by a whistleblower at Marvell where revenue was recognized before it should have been
• Proxy contest - Threatened action by Starboard Value to force board changes at Marvell
• Delisting notice - Formal warning from NASDAQ about potential removal from the exchange due to non-compliance with listing requirements
• Career progression patience - Matt's father's advice that being "the lowest paid guy in any given position" is an honor because "it means you got there early"

Timestamp: [51:18-1:01:18]

Matt Murphy discusses his practice of sending regular emails to his company, including a particularly meaningful message about grit that resonated with Joubin.

"I send one every week," Matt explains about his email communications to the company. "They're not all this detailed, but some of them are."

He traces the origins of this communication practice:

"I started this on a regular basis during COVID. During the pandemic, I always found even when I was at my prior company, at Marvell, my written communication like via email is high impact. It's not the same as a blog, it's not the same as something on the Slack channel or a little audio recording."

Matt has observed a unique impact from his personal emails:

"For whatever reason, for me, with people that have ever worked with me or for me, email from Matt makes an impact, and there's something about just getting an email."

While he had sent occasional emails before, the pandemic prompted a more consistent approach:

"When we hit the pandemic, I just decided, and I told my staff and Chris Koopmans, who's been with me from the beginning, 'Hey, I need your help because I got to—I'm going to have to communicate, like, overcommunicate.' And so the first email I sent was like March 20th, 2020, subject: Coronavirus. And [I've] kept going ever since."

Joubin notes that Matt referenced Angela Duckworth's research on grit in one particular email, quoting:

"She found that even beyond a person's natural talent and IQ, grit was the differentiating factor in those who would succeed, more than the school you went to, standardized test scores, or other metrics commonly associated with long-term success."

Timestamp: [1:01:25-1:04:02]

Matt shares how his background as a competitive runner helped shape his resilience and determination, drawing parallels between athletic endurance and professional perseverance.

Joubin highlights a personal story from Matt's email:

"You shared your own story where as a college senior, you would train twice a day and ran up to 90 miles a week for your—you were in track, obviously."

Matt confirms:

"Yeah, I ran cross country and track. 90 miles a week."

Joubin expresses that the email felt like a rallying cry during difficult times, and Matt explains his intentional timing:

"I timed it around that. I can't remember exactly what was happening at that time. I don't know if there was a—I think there was a down quarter right around when I sent it."

He clarifies that this was different from their recent quarter:

"Well, no, different—this one I guided up, right? That one I guided down. And it was to be expected. It was, I think, it was a quarter right after every company went through their reset after the pandemic. Semiconductors are volatile, people built up too much, so we had this sort of guide down, stock got hit, everyone's a little bummed out."

Matt reveals that he keeps a collection of personal messages ready for the right moment:

"I always have a queue of these, like, very personal 'Call: Matt' notes, and so I was saving the grit one. So I tried to time it with sort of like, 'Hey, we got to power through. You can't just get all mopey-dopey because your stock went down and like the world's coming to an end.'"

He felt confident about their trajectory despite the temporary setback:

"This whole AI thing was taking off for us. I actually felt very good about, once we got through that quarter, what it was going to look like."

Matt emphasizes the importance of sharing his own experiences with persistence:

"I think it's helpful to share experiences that it's okay that you have to grind sometimes, and you do get beaten down. It's kind of your defining thing though—like, do you decide to pack it up and give up, or do you just keep going?"

Timestamp: [1:04:02-1:06:34]

Matt delves deeper into Angela Duckworth's concept of grit, explaining how the ability to persist without external validation has been crucial to his success.

"The grit thing was really interesting to me because—I don't know if I put in the note—but the other factor that Duckworth talks about is being able to do that, basically long periods of sustained sort of effort and pain potentially with little to no positive reinforcement."

Matt reflects on how this quality has been instrumental in his career:

"I think that is something that either I had innately or I developed over time, but definitely has helped me a lot to get through these big ups and downs that you deal with in your professional life."

He draws a direct connection to his athletic background:

"Certainly as an athlete, both running in college and then I ended up doing triathlons and racing, it's kind of invaluable."

Matt observes a broader pattern in successful business leaders:

"It has a huge correlation—people that do endurance sports or played sports in college or were competitive athletes and their ability to excel and succeed in the business world."

When Joubin asks if Matt still maintains his intense running regimen, he admits:

"When times get bad, do you still run like that? No, not as much."

Timestamp: [1:06:34-1:07:33]

• Matt sends weekly emails to his company, a practice he formalized during the pandemic beginning March 20, 2020, with the first subject line "Coronavirus"
• These personal emails from Matt have proven to have a unique impact on his team compared to other communication channels like blogs or Slack
• Matt strategically times certain messages to coincide with company challenges, such as sending his "grit" email during a down quarter to rally the team
• He maintains a queue of personal notes ready to deploy at the right moment, saving the grit message for when the company needed resilience during a downturn
• Matt ran cross-country and track in college, training twice daily and logging up to 90 miles per week, later transitioning to triathlons
• Angela Duckworth's research on grit resonated with Matt, particularly her finding that grit is a stronger predictor of success than natural talent, IQ, or educational pedigree
• Matt emphasizes that a crucial aspect of grit is the ability to sustain effort and endure pain with "little to no positive reinforcement"—a quality he developed through athletics
• He observes a strong correlation between people with backgrounds in endurance sports or competitive athletics and their ability to excel in business, particularly in handling adversity

Timestamp: [1:01:25-1:07:33]

People:

• Matt Murphy - CEO of Marvell who sends weekly emails to his company, including a notable one about grit
• Angela Duckworth - Researcher whose work on grit was referenced in Matt's email; found that grit predicts success more than talent, IQ, or educational background
• Chris Koopmans - Marvell executive mentioned as having been with Matt "from the beginning" and helping with communication strategy

Companies and Organizations:

• Marvell - The company where Matt serves as CEO and implemented his email communication strategy
• Kleiner Perkins - Venture capital firm where Joubin works and realized the importance of grit in successful founders

Concepts:

• Grit - Angela Duckworth's concept referring to perseverance and passion for long-term goals; described as "the differentiating factor in those who would succeed"
• Sustained effort with little reinforcement - A key component of grit that Matt highlights as crucial for business success
• Email communication - Matt's preferred method for high-impact communication with his team, which he finds more effective than blogs or Slack
• Overcommunication during crisis - Strategy Matt adopted during the pandemic, beginning March 20, 2020
• Athletic mindset in business - The correlation Matt observes between competitive athletes and business success, particularly in handling adversity
• Endurance training - Matt's background running 90 miles per week in college and later competing in triathlons

Events:

• COVID-19 pandemic - Catalyst for Matt's formalized email communication strategy, beginning March 2020
• Post-pandemic reset - Business cycle referenced when semiconductor companies experienced volatility after initial pandemic demand

Timestamp: [1:01:25-1:07:33]

Matt Murphy shares how his father's illness prompted a renewed focus on his own health, leading to a significant period of athletic achievement through triathlons, and his subsequent struggles to maintain that commitment as CEO.

"When my father got sick, which was in 2010, 2011, I was turning 40—turned 40 in 2012—and I had gone through a period where, you know, I was traveling so much and my kids were younger and it was just everything was busy and I really let my health kind of go."

His father's battle with Parkinson's became a powerful motivator:

"Especially seeing what my dad was going through—not that he was unhealthy and that caused it—but you know, it gets your wakeup call, and how much he was battling with physical therapy as the Parkinson's got worse, to try to stop the entropy... It was just really hard to watch. So I'm like, this is something I can still get control of."

Matt describes his journey back to fitness:

"I did like a little sprint triathlon in 2011 in the summer. It was so painful, I was so out of shape. And I signed up for a half Ironman in March of 2012 in Oceanside, California, and I trained for it. And it was just hellacious."

He persisted and achieved remarkable progress:

"By the end of the summer, I was actually like pretty fit. I'd lost a bunch of weight, I was in really good shape. I started placing pretty high, and I raced the 2012 season, 2013, 2014, 2015."

His last serious race coincided with a pivotal career moment:

"My last kind of semi-serious race was Escape from Alcatraz in 2016. And I remember that right after the race, I literally had to go to a hotel I had booked and I was like signing my Marvell offer letter. It's crazy. It was like this final negotiation over what my severance agreement was or something."

Despite being in shape for an Iron Man at that time, Matt made a practical decision:

"I was actually in shape to do an Iron Man, so I had the Vineman on the calendar. That was like end of July. And I thought, 'There's no way I'm doing this.' I was in such bad shape over the two Iron Mans I did. I'm like, 'I can't take this CEO job and like two weeks later I'm trying to do some Iron Man in 90° weather. I'm going to be a mess.'"

Since becoming CEO, maintaining his athletic regimen has proven challenging:

"Since I've been CEO, I got to tell you, I've had some fits and starts, and I've tried my best, but the job really, really makes it difficult."

Timestamp: [1:07:39-1:10:10]

Matt reflects on how leadership roles can become convenient excuses for neglecting health priorities, and his refusal to grant himself that latitude despite the demands of being CEO.

When asked if he still works out every day and if the job gets too demanding, Matt offers a thoughtful perspective:

"Here's my take on this. This is where I hold myself accountable. It's an easy excuse to use your job as an excuse. And by the way, the higher you go, the more you get away with it. 'Well, you're so busy, you're a CEO.' I understand that everyone's going to give you sympathy on that."

He refuses to make excuses for himself:

"I actually have no sympathy for myself. It's 100%—well, not 100% because I'm also getting older, I'm like 10 years older now than I was when I was doing this before—but it's mostly, in my view, a discipline issue and a prioritization issue."

Matt points out that he maintained his fitness regimen during other demanding roles:

"I was super busy when I was doing all this training. It wasn't like I didn't have an easy job. I was running sales for this company, for Maxim when I was there, I was SVP of a division, I was number two executive."

He emphasizes his long-term experience in demanding positions:

"I've been a Section 16 officer of a public company for 19 years. So I've always had stuff, but I found periods of my life where it was easier to get momentum and get consistent, and I'm still striving to do that."

Timestamp: [1:10:10-1:11:38]

Matt shares some remarkable athletic accomplishments, including swimming between continents, while offering advice to Joubin who is preparing for his first triathlon.

When Joubin mentions he's doing his first triathlon, the Escape from Alcatraz, Matt expresses concern about his preparation:

"Oh, you are? First one? Good luck. Have you ever—have you swam in the bay? Have you acclimatized yourself to that?"

Joubin admits his limited swimming experience:

"I went into a pool like a few months ago, and I was like, 'Oh, I barely know how to swim.' I got goggles gifted to me from my fiancée's family... I get to one end of the pool and I'm turning back and I'm like, 'I'm exhausted.'"

Matt reveals his impressive open water swimming background:

"I've done a bunch of open water swims. There's a race in LA every year I like to do, which is the Hermosa Beach Pier to Manhattan Beach Pier. It's called the Pier-to-Pier."

He then shares a remarkable achievement:

"I did once in 2013, I swam from Asia to Europe—the Cross-Continental Swim on the Bosphorus Strait in Istanbul."

When asked about the distance, Matt explains:

"7 and a half kilometers, with the current. It's like 4 miles and change. It's like 8 km is like 5 miles, so just a little bit less than that."

He describes the unique experience:

"They literally start on the Asia side, get in the water, and then when you finish, you finish on the European side. It's pretty cool. I have a whole certificate and all that stuff. It was crazy."

Joubin shares his recent attempt to swim in San Francisco's cold waters:

"A few weeks ago I went for a long run and then I was like, 'All right, I'll just go swim in Aquatic Park just in my running shorts.' Jumped in, swam for, I don't know, 1 minute, and I was like, 'No, I'm hurting, I'm so cold.' I'm like, 'All right, I'm out of here.'"

Matt offers practical advice:

"Don't do that. And don't swim after you run either. That's really hard."

Timestamp: [1:11:38-1:14:09]

In closing the interview, Matt shares his optimism about Marvell's position in the technology ecosystem despite market volatility, emphasizing the company's innovative capabilities and strong team.

Matt expresses excitement about the company's trajectory:

"It's a really interesting time to be doing this. There's tremendous opportunity out there, I think, for the types of technologies. I think the long term is really good."

He acknowledges current market uncertainties:

"There's a lot of uncertainty in the business community right now on the impact of trade and tariffs and things like that, and what's it going to do."

Despite these challenges, customer demand remains exceptionally strong:

"We look at our customers and what they're asking us to do and what their ambitions are, and quite frankly, we can't go fast enough. I mean, it's a really, really amazing time."

Matt separates the company's fundamental value from stock market fluctuations:

"If I just didn't have to look at the equity markets, I'd be saying we're in great shape. And so that'll take care of itself over time. I mean, three months ago it was where it was, and now it's where it is, and I'm the same person and the same guy and the same company."

He expresses genuine enthusiasm for his role:

"I love what I do. I think I've got this company to a point where we're really in fighting shape, and we're not a technology laggard. We're not trying to go—we're pioneering stuff."

Matt concludes by praising his team at Marvell:

"I've got such a capable, incredible team of people at Marvell. I mean, very, very dedicated individuals—no politics, no BS-ing, great teamwork. We are still hiring."

Timestamp: [1:14:09-1:15:40]

• Matt's father's battle with Parkinson's disease in 2010-2011 served as a wake-up call that prompted him to prioritize his own health after years of neglect
• Starting with a painful sprint triathlon in 2011, Matt worked his way up to completing half Ironman competitions and became quite competitive by 2012-2015
• His last serious race was the Escape from Alcatraz in 2016, which coincided with signing his offer letter to become Marvell's CEO
• Matt acknowledges that maintaining fitness as CEO has been challenging but refuses to use his position as an excuse, viewing it primarily as a matter of discipline and prioritization
• He emphasizes that he managed to maintain rigorous training while in other demanding executive roles, having been a Section 16 officer of public companies for 19 years
• Among Matt's athletic accomplishments is swimming between continents in the 2013 Cross-Continental Swim across the Bosphorus Strait from Asia to Europe (7.5 kilometers)
• Despite market volatility affecting Marvell's stock price, Matt maintains strong optimism about the company's future, noting they "can't go fast enough" to meet customer demands
• Matt takes pride in having transformed Marvell into a company that's "in fighting shape" and "pioneering stuff" rather than being a technology laggard

Timestamp: [1:07:39-1:15:40]

People:

• Matt Murphy - CEO of Marvell who revitalized his fitness through triathlons after his father's illness and later struggled to maintain the same regimen as CEO
• Matt's father - Battled Parkinson's disease starting around 2010-2011, which inspired Matt to focus on his own health

Companies:

• Marvell - The company Matt leads as CEO, which he describes as being in "fighting shape" with a capable team and pioneering technology
• Maxim - Previous company where Matt worked as a sales executive and SVP of a division while maintaining his fitness regimen

Events and Places:

• Oceanside Half Ironman - Matt's first half Ironman competition in March 2012
• Escape from Alcatraz - Triathlon Matt completed in 2016 just before becoming Marvell CEO; also the event Joubin is preparing for
• Vineman - Iron Man competition in Napa that Matt was training for but skipped due to taking the CEO position
• Cross-Continental Swim - 7.5-kilometer swim across the Bosphorus Strait from Asia to Europe that Matt completed in 2013
• Pier-to-Pier - Annual swim race from Hermosa Beach Pier to Manhattan Beach Pier that Matt has participated in
• Aquatic Park - Location in San Francisco where Joubin attempted to swim after a run

Industry Concepts:

• Section 16 officer - Corporate executives required to report their securities transactions to the SEC; Matt has been one for 19 years
• Trade and tariffs - Current sources of uncertainty in the business community that Matt mentions
• Equity markets - Matt notes the volatility of stock prices despite the company's strong fundamentals
• "Pioneering stuff" - Matt's description of Marvell's current innovation position rather than being a "technology laggard"

Timestamp: [1:07:39-1:15:40]

Matt Murphy discusses the specialized talent areas where Marvell is actively recruiting, highlighting the critical importance of certain engineering specialties and the company's commitment to bringing in new talent.

Matt explains the often-overlooked aspects that determine a chip's success:

"You think a lot about, 'Hey, like how much compute power is on it,' or 'What are the features?' But it turns out, this is something that we're really good at—what ends up defining, in a lot of ways, the success or failure of your chip is actually what's known as the IO or the interconnect."

This area requires specialized expertise:

"There's a lot of specialization that occurs, specialized engineering capability that's in what's called the analog and mixed signal domain versus the digital domain. You have a bunch of electrical signals on one of these digital logic circuits like a big AI processor complex or something, but you've got to get the data on and off the chip at an incredibly high rate."

Matt notes this expertise is particularly valuable now:

"How you kind of hook up all these chips together now and how you get all the data in between them and on and off, including between the chips—think about all the memory now that gets stacked around these high-performance AI processor chips, how do you optimize that throughput, how do you do it in the smallest possible size so you can have more compute—that whole area is like ripe for innovation."

Beyond these technical specialties, Matt shares a decision he made on his first day as CEO that has become a cornerstone of Marvell's talent strategy:

"It was my first major decision as CEO. When I first day, I was doing like my orientation... and I said, 'Hey, what's going on here with the intern program? Like how many interns do we have every year? Do we have a program?' And she [said], 'I don't know. We had one once, but budget got cut and no funds for it.'"

This struck Matt as shortsighted:

"I'm thinking this company's spending like a billion dollars a year on opex on people, and we can't afford some interns? Are you serious? The place was leaking oil, there was so much cost to go fix, and yet we couldn't find the interns."

He recognized the long-term implications:

"I'm going to have to create the flywheel here because over time, there's this issue in any industry, but especially semiconductors—it's a pretty mature industry. So if you're not bringing that infusion of talent, you get older and you age out, people retire, and you haven't passed on the knowledge."

The program has since grown dramatically:

"We started with humble beginnings. We've now got a program of like 500 interns that come in every year. If they're going to be a senior, if they've done a good job, we offer them a job by the end of the summer. I think we convert 70-80%. They leave Marvell the summer before their senior year, they got a job the next year."

This initiative has significantly impacted Marvell's hiring:

"You add that sort of flywheel we've built, plus just new college hiring—23% of my new hires last year were out of school: bachelors, masters, or PhDs directly in. It's a huge infusion of young talent, and it's exciting."

Matt emphasizes his unwavering commitment to this program:

"That one I'm going to kind of protect at all costs. No matter how bad the year, I never cut that intern program. I always make sure it's world class."

Timestamp: [1:15:47-1:20:10]

In the final segment of the interview, Matt offers a profound and personal definition of grit, drawing from his experiences in both business and endurance sports.

When asked what the word "grit" means to him, Matt responds:

"It's a combination of things, right? It's certainly not giving up—that's kind of an obvious one, but it's easy to give up or lose motivation."

He emphasizes the importance of acknowledging difficulties without dwelling on them:

"It's also about not wallowing in it. Take your time, process it—you had a bad race, you had a bad quarter, your chip didn't work, you had some bad thing happen. Look, at some point, like you're a CEO like me, and I got bad stuff happening all the time. But process it first, like just acknowledge it like, 'Hey, that sucked, that wasn't good.'"

Matt offers a perspective that helps maintain resilience:

"Realize that for every bad day you have, it could be the worst day you're having—guess what, probably within the next seven days ahead of you, you're going to have a good day if you're good at what you do and you're competent and capable and you put in the effort."

He identifies what often causes people to falter:

"Sometimes it feels like the end of the world, and that's where I see people lose their confidence or spin out of control. And that's why it's important to have mentors."

Matt shares a personal story about a discouraging triathlon experience:

"I did the Wildflower Triathlon, the half—it's really hard, it's kind of like doing a three-quarter Iron Man, also a horrible race. I blew up on the run, walked a ton of it. I finished, I was almost in tears, and not because my time was slow. I was like, 'I don't think I can do this. How am I going to do a whole one, and I can't even get through this thing?'"

He recounts how a friend's support proved crucial:

"My good buddy and training partner was there at the finish. He finished way ahead of me and he was waiting for me, and I was just kind of getting emotional, and he's like, 'Look at me. You're a badass. You've put in the time. I've been riding with you. You're smoking me. You're killing it in the pool. Like, bad race, get over it.'"

This encouragement helped Matt gain perspective:

"Having people around you at those moments... he was right, actually. I got stronger after that race. I ate it and it was terrible, but guess what—I was so freaking strong after doing that, suffering for 6 and a half hours or whatever that one took."

The difficult experience ultimately made him stronger:

"By the time I did this first Iron Man, which was Iron Man Lake Tahoe in 2013, the first year they had it, I was just rearing to go. And by the way, in that race, the last four miles of the run were my fastest four miles of the whole marathon—finished strong."

Matt explains how grit involves rational analysis amid emotional challenges:

"It's always coming back to remembering like, 'Did I do something wrong to have this bad thing happen to me, or just did it happen? Do I need to course correct or not?'"

The confidence from thorough preparation became his foundation:

"I remember thinking to myself, I don't actually feel any different physically, but I know I did the work. I know I put every—I know I dialed that training in, and that actually carried me. Every time you kind of get low, it's like, 'No, no, I've done the fundamentals. I put in the time.'"

Matt concludes with a psychological insight:

"It's just interesting to decouple kind of your rational brain from your emotional brain, and I think grit is the ability to fundamentally do that and then reset yourself and power through, keep going."

Timestamp: [1:20:10-1:24:38]

Joubin concludes the episode with brief closing remarks.

"That's it for now. If you like the episode, please leave us a review or go back into the archives where we've done more than 200 episodes with some fantastic folks. This podcast is a Kleiner Perkins production, and I'm Joubin. Thanks for listening."

Timestamp: [1:24:46-1:25:01]

• The success of advanced semiconductor chips often hinges not just on compute power but on the critical "IO" or interconnect capabilities—the specialized engineering needed to get data on and off chips at high speeds
• One of Matt's first decisions as CEO was revitalizing Marvell's internship program, which has grown from nearly non-existent to approximately 500 interns annually with a 70-80% conversion rate
• 23% of Marvell's new hires last year came directly from schools (bachelors, masters, or PhDs), creating a crucial "flywheel" of young talent in an aging industry
• Matt protects the internship program "at all costs," recognizing it as essential for knowledge transfer and talent pipeline in a mature industry like semiconductors
• Grit involves not just perseverance but also the ability to process setbacks without "wallowing" in them or losing confidence
• Matt emphasizes the importance of mentors and supportive peers during challenging times, sharing how a friend's encouragement helped him overcome a discouraging triathlon experience
• True grit includes the ability to "decouple your rational brain from your emotional brain"—recognizing your preparation and capabilities even when emotions suggest otherwise
• Matt's experience in the Iron Man Lake Tahoe 2013 exemplifies grit—despite harsh conditions (27°F at race start), his preparation and mental strength allowed him to finish with his fastest miles at the end
• The "fundamentals" and preparation provide the foundation for resilience—knowing you've "put in the time" creates confidence that carries through difficult moments

Timestamp: [1:15:47-1:25:01]

People:

• Matt Murphy - CEO of Marvell who shares his perspectives on hiring strategies and the meaning of grit
• Joubin - Host of the podcast (mentioned in the outro)
• Matt's training partner - Friend who encouraged Matt after his disappointing performance at Wildflower Triathlon

Companies:

• Marvell - Semiconductor company led by Matt Murphy that's actively hiring in specialized areas and investing in their internship program
• Kleiner Perkins - Venture capital firm that produces the podcast (mentioned in the outro)

Technical Concepts:

• IO (Input/Output) - Critical component of chip design that moves data on and off the chip
• Interconnect - The connections between chips that are crucial for performance
• Analog and mixed signal domain - Specialized engineering area focused on non-digital signal processing
• Digital domain - Area of chip engineering focused on digital logic circuits
• Package technology - How multiple chips are integrated together, becoming "as important as the chip"
• CMOS - Complementary metal-oxide-semiconductor technology mentioned in relation to process technologies

Events:

• Wildflower Triathlon - Challenging half-triathlon where Matt "blew up on the run" but ultimately grew stronger from the experience
• Iron Man Lake Tahoe 2013 - Matt's first full Iron Man competition, which started in 27°F weather and where he finished strongest in the final miles

Business Concepts:

• The talent flywheel - Matt's concept of creating a self-reinforcing cycle of talent acquisition through internships and college hiring
• Opex - Operating expenses; Matt mentions Marvell was spending "a billion dollars a year on opex"
• Knowledge transfer - Critical process of passing expertise from experienced to newer employees

Psychological Concepts:

• Grit - Defined by Matt as the ability to acknowledge setbacks without dwelling on them, maintain confidence in your preparation, and decouple rational thinking from emotional responses
• Decoupling rational from emotional brain - Matt's description of the mental process that enables perseverance through challenges

Timestamp: [1:15:47-1:25:01]