Redefining Chip Architecture with Arm CEO Rene Haas

“There's no computing problem that's ever been invented that doesn't utilize and can't utilize the microprocessor. It is the heart of everything.”

Claim: The speaker says that every computing problem can utilize a microprocessor and that the microprocessor is central to everything.

“All roads lead through it, around it, past it. Something has to do the orchestration, arbitration decision around where those tokens go.”

Claim: The speaker says that a system must orchestrate and arbitrate decisions about where tokens are routed.

“All roads lead through it, around it, past it. Something has to do the orchestration, arbitration decision around where those tokens go. That's what CPUs do.”

Claim: The speaker says CPUs perform orchestration and arbitration for routing tokens.

“Chip design can take anywhere from 24 to 36 months depending on the complexity of the chip, etcetera, etcetera. The actual design is not the largest amount of time. The largest amount of time is in the verification, the validation.”

Claim: The speaker says verification and validation take more time than the actual chip design.

“Chip design can take anywhere from 24 to 36 months depending on the complexity of the chip, etcetera, etcetera. The actual design is not the largest amount of time. The largest amount of time is in the verification, the validation.”

Claim: Chip design can take 24–36 months, depending on the chip's complexity.

“stations, the debug. AI is really good at that. And if we were to shut it off, it's like being in the 1990s, you've got internet,”

Claim: The speaker says AI is very effective at chip-development station work and debugging.

“The genie's out of the bottle, and there's no stopping that.”

Claim: The speaker says the spread of the technology under discussion cannot be stopped.

“So we have two positions in the chip supply chain. Primary business is licensing IP, the CPU core that finds its way into smartphones, data centers, automobiles,”

Claim: Arm's primary business is licensing IP, including CPU cores used in smartphones, data centers, and automobiles.

“Our customers are the ones who either build the chips themselves, the Samsung who's got their own fab, or the vast majority of companies that take their chip designs and go to TSMC and get them taped out. So in that world,”

Claim: Arm's customers either manufacture chips themselves or send their chip designs to foundries such as TSMC for fabrication.

“broad in terms of the markets that we serve, we kind of see everything. We have a very good sense of what's going on in automotive, data center, smartphones. So we see the supply chain situation from all angles.”

Claim: Arm says its broad market presence gives it visibility into automotive, data-center, and smartphone markets and the supply chain from multiple angles.

“We also introduced our first product last March, the one you just mentioned at Hot Chips, the ARM A-GICPU.”

Claim: Arm says it introduced its first physical product, the Arm AGI CPU, last March.

“So now we're in that soup ourselves from the standpoint of we're also having to figure out how to buy substrates and buy wafers and buy memory. et cetera, et cetera.”

Claim: Arm says that, as a physical-product supplier, it must arrange purchases of substrates, wafers, and memory.

“it was an evolution from the early days of where we just supplied simply the IP. The IP components, the pieces, the CPU IP, the GPU IP, the system IP, et cetera, et cetera.”

Claim: Arm says its move into physical products evolved from its earlier business of supplying CPU, GPU, and system IP components.

“A few years ago, what we were starting to see was that product cycle times are slowing down, chip manufacturing times are extending,”

Claim: Arm says that a few years ago product cycle times were slowing while chip manufacturing times were getting longer.

“the ability to get solutions out faster was becoming more and more important.”

Claim: Arm says that getting solutions to market faster was becoming increasingly important.

“So we moved from these individual components into what we called compute subsystems. I used, when we did the roadshow a few years ago, I used the Lego analogy where essentially we're providing the blueprint on here's how you stitch it all together.”

Claim: Arm says it moved from supplying individual IP components to compute subsystems that provide a blueprint for integrating them.

“other demand for that was was insane and what we were finding was we were and we initially people thought well people aren't going to want these subsystems because that's what a chip designer does why are you providing that piece but it saved time to market and it saved a whole lot of things in terms of cost speed etc etc the physical product was sort of the next the next leap if you will and there are certain sets of customers that will license ip2 and they've got all the”

Claim: Arm says demand for its compute subsystems was very high and that they reduced time to market and costs.

“all the capability in the world to build chips based on ARM. There's a lot of companies who want to have product based on ARM. customers build products that serve those markets.”

Claim: Arm says some customers can build Arm-based chips, while many companies seeking Arm-based products lack the capability or role to build products for the relevant markets themselves.

“So Meta was that first example. They wanted a general purpose agentic CPU.”

Claim: Arm says Meta was its first example of a customer seeking a general-purpose agentic CPU.

“There wasn't anybody out who could give it to them.”

Claim: Arm says no other supplier could provide Meta with the general-purpose agentic CPU Meta wanted at that time.

“They came to us and said, hey, why don't we do this together? And that's how we got into it.”

Claim: Arm says Meta proposed collaborating on the general-purpose agentic CPU, leading Arm to enter physical-chip development.

“So one of the things that we were very careful about was getting making sure the ecosystem was on board with this because”

Claim: Arm says it took care to secure ecosystem support for its move into physical products.

“We do CPU IP, which is really only as good as the ecosystem, the ecosystem of chip people and the ecosystem of software folks and people who build around that.”

Claim: Arm says the value of its CPU IP depends on the surrounding chip, software, and developer ecosystem.

“And surprisingly, we got a lot less pushback than I thought.”

Claim: Rene Haas says Arm received less customer pushback about its move into physical products than he expected.

“And the reason for that was the more software that's available in the wild, whether it's proprietary and or open source. supports benefits the broader ecosystem and the customers themselves.”

Claim: Rene Haas says greater availability of proprietary and open-source software benefits the broader ecosystem and customers.

“So whether it was NVIDIA, Amazon, Microsoft, Google, all people who build ARM-based server chips, they were all on board.”

Claim: Rene Haas says NVIDIA, Amazon, Microsoft, and Google—all builders of Arm-based server chips—supported Arm's move into physical products.

“But that means you need supply chain operations people. You need to work with, as I said, the TSMCs and Samsungs of the world. You need to work with the Samsungs and the Microns, the SK Hynix to get memory allocation.”

Claim: Arm says selling physical products requires supply-chain operations and work with foundries and memory suppliers.

“And then on the engineering side, you need a lot. A lot more different capabilities. You need back-end people, layout people, implementation people.”

Claim: Arm says physical-chip production requires additional engineering capabilities, including backend, layout, and implementation expertise.

“In AI as a utility, that's going to help productivity for every single industry.”

Claim: Rene Haas believes AI used as a utility will improve productivity across every industry.

“It is going to be the great leveler in terms of companies that get started super quickly.”

Claim: Rene Haas expects AI to level the playing field for companies that can get started very quickly.

“And for industries, whether it's healthcare, infrastructure, robotics, every industry is going to use artificial intelligence as a utility and stop.”

Claim: Rene Haas expects every industry to use artificial intelligence as a utility.

“Since I'm such a believer in this, of course, we use it very heavily inside of ARM.”

Claim: Rene Haas says ARM uses AI extensively within the company.

“On the non-engineering side, we're using it all over the place. But on the engineering side,”

Claim: Rene Haas says ARM uses AI broadly in non-engineering work.

“we're using it all over the place. But on the engineering side, we've seen huge, huge benefit. You mentioned verification.”

Claim: Rene Haas says AI has produced substantial benefits for engineering work at ARM.

“Chip design can take anywhere from 24 to 36 months depending on the complexity of the chip, etc., etc.”

Claim: Rene Haas says chip design typically takes 24–36 months, depending on complexity.

“The actual design of the architecture, the RTL generation, if you will, the mapping of the architecture is not the largest amount of time.”

Claim: Rene Haas says architecture design and RTL generation account for less time in chip design than verification and related work.

“The largest amount of time is in the verification, the validation, the debug, the documentation, etc., etc. AI is really good at that.”

Claim: Rene Haas says AI is well suited to verification and related chip-design tasks.

“The largest amount of time is in the verification, the validation, the debug, the documentation,”

Claim: Rene Haas says verification, validation, debugging, and documentation consume the largest share of chip-design time.

“And I would say we probably have 80 to 90%. of engineers today inside arm who use it on a daily basis and if we were to shut it off uh my analogy i give to people it's like being in the 1990s you've got internet and you're now saying you know only internet between the hours of two and four after that go to the library that we have down the hall that's got all the books that you can go up and look all this information up people it'd be anarchy so the genie's out of the bottle right there's and and there's no there's no stopping that”

Claim: Rene Haas estimates that 80–90% of ARM engineers use AI every day.

“of engineers today inside arm who use it on a daily basis and if we were to shut it off uh my analogy i give to people it's like being in the 1990s you've got internet and you're now saying you know only internet between the hours of two and four after that go to the library that we have down the hall that's got all the books that you can go up and look all this information up people it'd be anarchy so the genie's out of the bottle right there's and and there's no there's no stopping that”

Claim: Rene Haas argues that ARM engineers have become so reliant on AI tools that restricting access would be disruptive, comparing it to limiting internet access to two hours daily and saying the result would be anarchy.

“Now, there's certain things that. The tools are still not that mature of. One of them is really around RTL generation and then physical design and implementation in best of class. And that's simply because the models, you know, they're trained on what's available publicly and a lot of the information is quite proprietary.”

Claim: Rene Haas says AI tools are not yet mature enough to achieve best-in-class RTL generation, physical design, and implementation.

“One of them is really around RTL generation and then physical design and implementation in best of class. And that's simply because the models, you know, they're trained on what's available publicly and a lot of the information is quite proprietary.”

Claim: Rene Haas attributes AI models' limitations in RTL generation and physical design partly to their training on public data while much relevant chip-design information is proprietary.

“That being said, there's massive opportunity between the ecosystems and everyone in the industry to make that better. It's only going to get better.”

Claim: Rene Haas believes industry collaboration can substantially improve AI tools for chip design and expects the tools to improve.

“And I think that's a big, big opportunity.”

Claim: Rene Haas considers collaboration with model makers a major opportunity.

“And one of the things that I'm proud of at ARM is given our business, our core IP business, we probably have the richest IP portfolio, both in terms of not only the IP, and this is the killer, the documentation, the test benches,”

Claim: Rene Haas believes ARM probably has the richest IP portfolio, including its IP, documentation, and test benches.

“you know, how you build the IP. You know, I've worked for chip companies in the past. in the past that have said hey why don't we license this ip that we've got because it's really really valuable and then you get into a wait a minute there's there's no documentation there's no explanation on how you it's unusable it's untestable”

Claim: Rene Haas says he has encountered chip companies that considered licensing valuable IP but lacked the documentation and usage explanations needed to make it usable.

“right and and if it's unusable and untestable it's actually untrainable and if it's untrainable it's not usable for ai i think we have some some built-in advantages based on our business model that will allow us to really be able to take advantage of the tools as they get better”

Claim: Rene Haas believes ARM's business model gives the company built-in advantages in taking advantage of AI tools as they improve.

“right and and if it's unusable and untestable it's actually untrainable and if it's untrainable it's not usable for ai i think we have some some built-in advantages based on our business model that will allow us to really be able to take advantage of the tools as they get better”

Claim: Rene Haas argues that IP that is unusable and untestable cannot be trained on and is therefore unsuitable for AI use.

“I don't know if it's in two to three years away, but five plus years, can you go from idea to a GDS2 file, GDS2 file being the file that you actually send to the fab to go get built for certain designs? Quite possible.”

Claim: Rene Haas considers it quite possible that, in five or more years, AI-enabled processes could take certain chip designs from idea to a GDS2 file.

“I don't know if it's in two to three years away,”

Claim: Rene Haas is unsure whether the idea-to-GDS2 capability will arrive within two to three years.

“So it takes that whole design piece out of the way. It takes that whole piece out of the way relative to the verification. So I think for the more straightforward design, it's quite possible.”

Claim: Rene Haas thinks AI could eliminate much of the design and verification work for more straightforward chip designs.

“Now, if you go into the tool and say, design me something that's 10% faster than Vera Rubin, 20% cheaper and 30% more efficient on this model. you're not going to go press a button and have it happen right away.”

Claim: Rene Haas says current tools cannot instantly produce a chip design that is 10% faster than Vera Rubin, 20% cheaper, and 30% more efficient for a specified model.

“But I think in five to 10 years, you know, our industry as well, we're going to see some amazing differences relative to how chips are designed.”

Claim: Rene Haas expects chip-design methods to change substantially over the next five to ten years.

“I think what's going to matter back to the earlier discussion we had on supply chain is understanding the supply chain impacts, how all that gets built and put into ultimate end products.”

Claim: Rene Haas expects understanding supply-chain impacts and how components are built into end products to become increasingly important.

“selling into an industry where the capital requirements are just massive and the relationships with memory vendors is incredibly critical or the relationship with substrate vendors so companies are going to have to be much”

Claim: AI-chip companies face massive capital requirements and need critical relationships with memory and substrate vendors.

“it yeah all of that and i and i think that is um that's not going to stop in 12 months it's not going to stop in 24 months i think we're going to be in this constrained environment for three to five years at least.”

Claim: Rene Haas expects supply-chain constraints to persist for at least three to five years, not to end within 12 or 24 months.

“So long as the transformer is the unit of energy relative to how you generate AI. training and AI inference by design it is a it is very compute intensive very memory intensive so if you think about that that's going to drive a lot of demand on having supply chain acumen which then goes back to people have got great ideas on chip design they're going to have to need a lot of other things just to be able to get access to capital wafers everything you just talked about”

Claim: AI training and inference are inherently highly compute-intensive and memory-intensive, according to Rene Haas.

“I think building out the data centers is going to be a bottleneck. And when I say building out, if you look at all the projects that are being done today.”

Claim: Rene Haas expects data-center construction to become a bottleneck to AI infrastructure growth.

“Not a lot of them are ahead of schedule and needing less labor than they thought.”

Claim: Few current data-center projects are both ahead of schedule and using less labor than expected, according to Rene Haas.

“And then when you layer on top of that a lot of buzz that's coming from different parts of the country in the United States here relative to slowing down data center development or putting restrictions around it, I think that infrastructure build out could be a headwind just relative to everything going on.”

Claim: Rene Haas thinks efforts in parts of the United States to slow or restrict data-center development could make infrastructure build-out a headwind.

“because if it wasn't, if infrastructure build out was not a headwind, I think capacity for wafers, capacity for memory, that probably would be a headwind.”

Claim: Rene Haas thinks wafer capacity and memory capacity would probably constrain growth if infrastructure build-out did not.

“So I think you're going to see a number of different governors, if you will, not governors of states, but different things that are going to throttle the growth of this, which”

Claim: Rene Haas expects several different constraints to throttle AI growth.

“The bubble in terms of are we over oversupplied to demand? Not even close.”

Claim: Rene Haas says AI infrastructure is not close to being oversupplied relative to demand.

“I think, again, that's because the demand is insatiable, just given the way these models work and infrastructure build out, access to wafers. There is access to memory, all of that's combining.”

Claim: Rene Haas attributes insatiable AI demand to how AI models work, infrastructure build-out, and access to wafers and memory.

“Yeah, so one of the benefits we have at ARM, publicly traded, yes, but a very, very large single shareholder, so I have...”

Claim: ARM is publicly traded and has a very large single shareholder, according to Rene Haas.

“I have lots of informal investor meetings with my chief shareholder all the time about this.”

Claim: Rene Haas says he frequently has informal investor meetings with ARM's chief shareholder.

“We have a big advantage in that there's a lot of things symbiotically we can do together that can help arm advanced initiatives by having SoftBank as our largest shareholder that we look to be very, very innovative around. To your point in terms of young companies,”

Claim: Rene Haas says ARM and SoftBank can pursue symbiotic opportunities to support ARM's advanced initiatives and seek innovative ways to do so.

“I would say strategic partnerships. incredibly early, whether it's with people inside the supply chain, people in private equity, the banks themselves.”

Claim: Rene Haas advises young companies to establish strategic partnerships early with supply-chain participants, private-equity firms, and banks.

“On one hand, semis are kind of back because you now have a wave of semiconductor startups. There was a long time where that was just not happening, investment in the industry.”

Claim: Rene Haas says a wave of semiconductor startups has emerged after a long period with little investment in the industry.

“access to capital is going to be the gate for them in terms of how they get through that.”

Claim: Rene Haas expects access to capital to be a key constraint on semiconductor startups' progress.

“So I think getting much more creative in terms of how they work with the ecosystem is going to be super, super key.”

Claim: Rene Haas says semiconductor startups need to work more creatively with the broader ecosystem.

“And we at SoftBank, that's one of the things we look at very strategically, companies that we can bring in into the portfolio that we can help, that we can provide a combination of either the backstop and or if you think about SoftBank, we just announced we being soft.”

Claim: SoftBank strategically evaluates portfolio companies it could support, potentially by providing a backstop, according to Rene Haas.

“SoftBank, SoftBank Neo, which is our intent to become a neocloud. And in that world, we could become a home for these young companies who have chip technology that in other worlds, they'd have to go up and figure out how to get a design win at Microsoft or Google.”

Claim: SoftBank Neo intends to become a neocloud, according to Rene Haas.

“which is our intent to become a neocloud. And in that world, we could become a home for these young companies who have chip technology that in other worlds, they'd have to go up and figure out how to get a design win at Microsoft or Google. We can provide a lot of interesting avenues for that.”

Claim: Rene Haas says SoftBank Neo could host young chip-technology companies that might otherwise need to win designs at Microsoft or Google.

“So the way we think about it is SoftBank Group, which is headed in Japan, and that is Masa.”

Claim: SoftBank Group is headed in Japan by Masa.

“I have a lot of different operating companies underneath them. One of the largest ones is SoftBank KK, which is essentially SoftBank Mobile.”

Claim: SoftBank Group has multiple operating companies, including SoftBank KK, which is essentially SoftBank Mobile and is one of the largest.

“Inside the U.S., there's a lot of investment activity that's going on with SoftBank Group International.”

Claim: SoftBank Group International conducts substantial investment activity in the United States.

“There's SoftBank Vision Fund.”

Claim: SoftBank has a Vision Fund.

“a lot of the strategies that we're trying to do around SoftBank is helping the strategies that Masa talked about. publicly at his shareholder meeting in Japan, which is around robotics, open AI, infrastructure, and ARM.”

Claim: SoftBank's strategies increasingly support Masa's publicly stated priorities of robotics, open AI, infrastructure, and Arm.

“So I've probably got my eyeballs on a lot of stuff, to be honest with you, in terms of helping Masa really realize the execution of that vision.”

Claim: The speaker is involved in many areas to help Masa realize and execute his vision.

“Yes, I'm leading the direction of Ampere and Graphcore and another company called StackAV that's doing things around autonomous.”

Claim: The speaker leads the direction of Ampere, Graphcore, and StackAV; StackAV works on autonomous technology.

“is that I'm kind of in the room on a lot of discussions that Moss is having and helping him sort of formulate that strategy and more importantly, help execute it.”

Claim: The speaker participates in many of Masa's discussions and helps formulate and execute Masa's strategy.

“Well, one thing it does, it gives us a huge bird's eye view relative to where the broader industry is going, whether it's around infrastructure, whether it's around capital, whether it's around energy.”

Claim: The broader SoftBank ecosystem gives the speaker a broad view of industry direction in infrastructure, capital, and energy.

“But also you can imagine it could provide a home for our products, right? So it doesn't need to be the home, but it certainly can be a home,”

Claim: The SoftBank ecosystem could serve as a home for Arm products, but it need not be their home.

“which is also a big help when we think about the verticals that soft banks evolved with, robotics, energy, data center infrastructure,”

Claim: SoftBank's presence in robotics, energy, and data-center infrastructure could help provide a home for Arm products.

“and then you look at the products that Arm has, the only one we've announced so far is the Arm AGI CPU.”

Claim: As of the conversation, the Arm AGI CPU is the only Arm product announced so far.

“You can start to connect the dots and say, gosh, there could be some. Very interesting opportunities that there could be an opportunity for ARM, which necessarily doesn't mean that we're getting into the broad merchant chip business.”

Claim: Arm may have interesting opportunities that do not require entering the broad merchant-chip business.

“We could be just doing products simply back for South Bank.”

Claim: Arm could make products specifically for SoftBank rather than entering the broad merchant-chip business.

“Oh, broadly speaking, I think whether it's humanoids or dedicated machines to do certain level tasks that can be retrained is going to be enormous, right?”

Claim: The speaker expects the market for humanoid robots and retrainable machines designed for particular tasks to become enormous.

“Robotics 1.0, which is a purpose-built industry, you had a piece of mechanics designed to do a certain task and the software that was optimized for that task.”

Claim: In Robotics 1.0, robots used task-specific mechanics and software optimized for a particular task.

“If you had a brand new automobile line came up or some different piece of equipment, if the robots weren't well suited for that, rip up the line, et cetera, et cetera. So as you can imagine, then the barrier was pretty high.”

Claim: In purpose-built robotics, introducing a new automobile line or equipment that existing robots could not handle could require rebuilding the line, creating a high barrier.

“getting to a world where the robots can learn just based upon either being trained or what they see and then when you then combine that with can you design something mechanically general purpose enough that can take advantage of being reprogrammed and then when you layer on top of that the cost coming down you look at it and say oh my gosh what what will it not be able to do so it's almost like something out of the Jetsons right where”

Claim: The speaker suggests that robots capable of learning from training or observation, mechanically general-purpose and reprogrammable, and lower in cost could perform an extremely broad range of tasks.

“A lot of things will ultimately be done by robots. Construction, infrastructure, service, security.”

Claim: The speaker expects robots eventually to perform many tasks in construction, infrastructure, services, and security.

“Right now you see a lot of stuff on Instagram or TikTok of Olympic races with robots, et cetera, et cetera. I don't think anyone's going to have any interest in watching a sports league of robots. There may be an enthusiast class who might be interested in that,”

Claim: The speaker doubts that a broad audience will want to watch a robot sports league, while allowing that enthusiasts might be interested.

“but the broader utility is going to be around a lot of human labor tasks. tasks that can ultimately easily be replaced by robots.”

Claim: The speaker expects robotics' broader utility to come from human-labor tasks that robots can eventually replace easily.

“I think it's both. Yeah, I guess both. There are a lot of jobs and work tasks that are optimized around a person being six feet tall and having arms of a certain length, etc., etc. But I think it'll be both.”

Claim: The speaker expects robotics to include both humanoid and specialized task-specific form factors.

“Yeah, I guess both. There are a lot of jobs and work tasks that are optimized around a person being six feet tall and having arms of a certain length, etc., etc.”

Claim: Many jobs and work tasks are designed around human dimensions, such as a six-foot height and arms of a particular length.

“And I think the fact that they're going to be smart and can learn.”

Claim: The speaker expects future robots to be intelligent and capable of learning.

“ARM is going to be everywhere.”

Claim: The speaker expects Arm technology to be ubiquitous across robotics.

“We have a tremendous amount of technology from a real-time sensing standpoint around microprocessors that will be out.”

Claim: Arm has substantial real-time sensing technology associated with microprocessors that is expected to become available.

“out at the fingers they can do perception and sensing that's all going to be arm-based today whether it's nvidia or some of the work that qualcomm does most of the brains the brains that you see in the humanoids those are all all running on arm uh today so i think for us going going forward”

Claim: The speaker says that robot-finger perception and sensing systems and most humanoid-robot computing brains currently run on Arm-based technology, mentioning Nvidia and Qualcomm in this context.

“The robotic industry will be powered by ARM.”

Claim: The speaker predicts that Arm will power the robotics industry.

“I think it's still a little bit early because the business models have not. not been actually figured out because the cost of robots are so high right because the cost of robots are so high people buying the robots themselves that's a that's a tough it's a tough model to sort of get people's heads around to does it actually replace so i think costs need to come down and the business model need to be need to be ultimately vetted”

Claim: The speaker considers broad robotics adoption premature because viable business models have not yet been established.

“not been actually figured out because the cost of robots are so high right because the cost of robots are so high people buying the robots themselves that's a that's a tough it's a tough model to sort of get people's heads around to does it actually replace so i think costs need to come down and the business model need to be need to be ultimately vetted”

Claim: The high cost of robots makes a model in which customers buy the robots themselves difficult to justify or understand.

“not been actually figured out because the cost of robots are so high right because the cost of robots are so high people buying the robots themselves that's a that's a tough it's a tough model to sort of get people's heads around to does it actually replace so i think costs need to come down and the business model need to be need to be ultimately vetted”

Claim: For robotics to advance, robot costs need to decrease and the business model needs to be validated.

“But I think it is critically important for the United States to have as much of that technology. inside on U.S. soil.”

Claim: Rene Haas believes it is critically important for the United States to have as much semiconductor technology as possible located on U.S. soil.

“whether it's Micron, I think we need more U.S. fabs.”

Claim: Rene Haas believes the United States needs more semiconductor fabrication plants.

“It's critical for national security.”

Claim: Rene Haas believes U.S. semiconductor fabrication capacity is critical for national security.

“It's also critical for diversification of supply chain.”

Claim: Rene Haas believes U.S. semiconductor fabrication capacity is critical for diversifying the supply chain.

“My personal view is that it's an infinite game, I believe, first in terms of the race. There's not going to be a winner.”

Claim: Rene Haas believes the semiconductor race is an ongoing, effectively infinite competition with no winner.

“But you could get to a situation where a lot of the critical. critical technologies are not U.S. based. And that's not going to be a good thing, right?”

Claim: Rene Haas believes it would be undesirable for many critical technologies to be based outside the United States.

“number one, for both national security reasons and economic, you want to be at the forefront of technology because it drives innovation,”

Claim: Rene Haas believes that being at the technological forefront drives innovation.

“you want to be at the forefront of technology because it drives innovation, but it also drives ecosystems.”

Claim: Rene Haas believes that being at the technological forefront also drives the development of ecosystems.

“I call BS on that because if you think about whether it's around energy, liquid cooling, all of the things that make the data center better, those are all jobs that can be created and done here.”

Claim: Rene Haas believes work on energy, liquid cooling, and other data-center improvements can create jobs locally.

“So I think as a national policy, it's incredibly important for us to be investing A, in the United States, and B, making sure that we stay in the lead.”

Claim: Rene Haas believes U.S. national policy should invest in the United States and ensure it remains a technology leader.

“that there's a fear that AI means job loss. to being an electrician right electrician is either it's not maybe viewed as a highly educated job or you're going to get a phd it's a highly skilled job that requires a lot of training and certification and you need tons of them you know to do this kind of work and that's very very critical to the to the data centers um so i think to your question i think part of the backlash is just from fear i mean there's just a fear that my jobs are going to go away uh the ai boom”

Claim: Rene Haas believes fear of AI-related job loss is one cause of backlash against data centers.

“i think it's i don't think it's well grounded at all i i completely”

Claim: Rene Haas believes the fear that AI will cause job losses is not well grounded.

“to being an electrician right electrician is either it's not maybe viewed as a highly educated job or you're going to get a phd it's a highly skilled job that requires a lot of training and certification and you need tons of them you know to do this kind of work and that's very very critical to the to the data centers um so i think to your question i think part of the backlash is just from fear i mean there's just a fear that my jobs are going to go away uh the ai boom”

Claim: Fear that jobs will disappear is one cause of backlash against the AI boom.

“to being an electrician right electrician is either it's not maybe viewed as a highly educated job or you're going to get a phd it's a highly skilled job that requires a lot of training and certification and you need tons of them you know to do this kind of work and that's very very critical to the to the data centers um so i think to your question i think part of the backlash is just from fear i mean there's just a fear that my jobs are going to go away uh the ai boom”

Claim: Data centers require many electricians, whose highly skilled work requires substantial training and certification and is critical to data-center construction and operation.

“to being an electrician right electrician is either it's not maybe viewed as a highly educated job or you're going to get a phd it's a highly skilled job that requires a lot of training and certification and you need tons of them you know to do this kind of work and that's very very critical to the to the data centers um so i think to your question i think part of the backlash is just from fear i mean there's just a fear that my jobs are going to go away uh the ai boom”

Claim: Rene Haas says electrician work is highly skilled, requires extensive training and certification, and is needed in large numbers for data centers.

“for good or for bad has benefited a lot of people and there's a lot of people have no benefit from it right and there's a lot of america just again on the american political scene who stopped to make the mortgage you know their paychecks haven't gone up and now they've got this ai thing that just look it's going to even harder so i think the the data centers become a bullseye unfortunately for all the things that could be bad about ai which i think is just”

Claim: Some Americans whose paychecks have not increased and who struggle to make mortgage payments fear that AI will make their situation even harder.

“for good or for bad has benefited a lot of people and there's a lot of people have no benefit from it right and there's a lot of america just again on the american political scene who stopped to make the mortgage you know their paychecks haven't gone up and now they've got this ai thing that just look it's going to even harder so i think the the data centers become a bullseye unfortunately for all the things that could be bad about ai which i think is just”

Claim: Data centers have become a target for concerns about AI's potential harms.

“for good or for bad has benefited a lot of people and there's a lot of people have no benefit from it right and there's a lot of america just again on the american political scene who stopped to make the mortgage you know their paychecks haven't gone up and now they've got this ai thing that just look it's going to even harder so i think the the data centers become a bullseye unfortunately for all the things that could be bad about ai which i think is just”

Claim: The AI boom has benefited many people but has provided no benefit to many others.

“it's become the boogeyman for a lot of a lot of things.”

Claim: Data centers have become a scapegoat or symbol of fear for many concerns.

“Yeah, I mean, on first principles, whether it was smartphones, the internet, personal computers, fill in your favorite technology, there is no downside from being the leader. There's just none. There are second and third order effects that you may not like, but to be the laggard.”

Claim: Being the leading country or actor in a technology has no downside, although it may have second- and third-order effects that are undesirable.

“you are having the entire script dictated to you and everything that kind of comes with it. I mean, look at other parts of the world that are just not the leaders in this space. Economically and socially, they're left behind and governments carry the large tax burden of it.”

Claim: Countries that lag in a technology space risk having others dictate the terms, being left behind economically and socially, and imposing a large tax burden on their governments.

“and I would argue to some extent, AI is a little bit of the final frontier. of what can be done with essentially intelligence,”

Claim: Rene Haas argues that AI is, to some extent, a final frontier for what can be done with intelligence.

“of course you want to be in the lead. Of course you want to be driving that because the benefits for society are going to be enormous.”

Claim: Rene Haas expects the societal benefits of leading AI to be enormous.

“I think we are, I feel fortunate every day that we are in the heart of all of this. The fact that we can be a participant in that ecosystem, we can help drive the innovation, we can be involved with leadership companies, develop leadership products,”

Claim: Arm is centrally positioned in the AI ecosystem and can help drive innovation, work with leading companies, and develop leading products.

“all the AI needs some level of compute. That's what ARM does.”

Claim: AI requires compute, and Arm provides compute.

“And that compute needs to be power efficient. That's what we're really, really good at. So those roads all lead through us.”

Claim: AI compute needs power efficiency, an area in which Arm says it is highly capable.

“And now it's like there's no the abundance of opportunity innovation is so great with AI.”

Claim: Rene Haas believes AI creates abundant opportunities for innovation.

“there was so much focus on no matter what the question is the answer the accelerator there's no computing problem that's ever been invented that doesn't utilize and can't utilize the microprocessor it is it is the heart of everything all”

Claim: Rene Haas asserts that every computing problem uses or can use a microprocessor, which he describes as central to computing.

“Well, of course, in a system problem, something has to do the orchestration arbitration decision around where those tokens go,”

Claim: A computing system needs a mechanism to orchestrate and arbitrate the routing of tokens.

“The token factory just generates all these tokens. It's like literally where are the trucks that are going to take the tokens away and give them to the users? That's what CPUs do.”

Claim: The token factory generates tokens, while CPUs transport them to users.

“So until something's invented that says the CPU has gone away and we're now doing it through some other mechanism which has yet been defined or invented. did the cpus can be doing just fine and there's gonna be a lot of demand for it a ton of demand in addition to the accelerators that generate the tokens but the way to think about it is it's a system which again going back to you know memory well of course memory is needed because in a computer von neumann architecture or computing architecture you have a cpu you have some accelerator whether it's a floating point a gpu accelerator and memory”

Claim: Unless a replacement mechanism for CPUs is invented, CPUs will continue to serve their role and face substantial demand.

“did the cpus can be doing just fine and there's gonna be a lot of demand for it a ton of demand in addition to the accelerators that generate the tokens but the way to think about it is it's a system which again going back to you know memory well of course memory is needed because in a computer von neumann architecture or computing architecture you have a cpu you have some accelerator whether it's a floating point a gpu accelerator and memory”

Claim: In the computing architecture described, a system includes a CPU, an accelerator, and memory.

“System design hasn't changed. I think some of the focus kind of moved around,”

Claim: The speaker says system design has not changed, even though some attention has shifted.

“wearables. And in fact, as you get to the smaller footprints where more and more AI is going to take place, That's going to be a sweet spot for ARM because the CPU's table stakes anyway, you have to have it to do all the things that are required in the edge device.”

Claim: Arm sees smaller-footprint devices as a promising area as more AI runs on them, since those edge devices require a CPU.

“But now we have an opportunity with our instructions and architecture to do a lot of things where you just can't put a 50 watt GPU on your head, right?”

Claim: Arm says its instruction set and architecture can support AI tasks in wearable devices where a 50-watt GPU cannot be worn on the head.

“You're going to have to do that AI processing somewhere locally.”

Claim: AI processing in the described edge-device context will need to occur locally.