Cerebras SUPERNOVA 2025 | Andrew Feldman, Cerebras | Cerebras SUPERNOVA 2025

Clips
More from Cerebras SUPERNOVA 2025

Andrew Feldman

Founder & CEO

Cerebras Systems

play_circle_outline Cerebras Unveils Qwen 32B: Revolutionizing AI with Unmatched Speed and Performance

play_circle_outline Collaboration with Meta and IBM for fast inference

play_circle_outline Strategy for targeting big cap expenses and enterprise market

play_circle_outline Offering developers a million free tokens per day for AI applications

Info
Transcript

Andrew Feldman, Cerebras | Cerebras SUPERNOVA 2025

Andrew Feldman

Founder & CEO Cerebras Systems

Andrew Feldman, founder and chief executive officer of Cerebras Systems Inc., joins theCUBE’s John Furrier at Cerebras SUPERNOVA 2025 to discuss the company’s latest advancements in AI performance and model deployment. Set against the backdrop of Fort Mason in San Francisco, the conversation dives into Cerebras' release of the Qwen 32B model, which is drawing attention for its unmatched speed and agentic reasoning capabilities.

Feldman breaks down how Cerebras is redefining compute benchmarks by delivering performance 20 to 30 times faster than tradi... Read more

explore Keep Exploring

What are some key requirements that people should pay attention to when considering AI models? add

What companies announced they would be providing API space services for fast inference to their clients? add

What strategy are you optimizing for as you prioritize your business growth in the market? add

What announcement was made regarding developers receiving tokens and what is the message to developers wanting to get in on the action? add

bolt Powered by CUBE AI

Andrew Feldman, Cerebras | Cerebras SUPERNOVA 2025

search

>> Hello, I'm John Furrier with theCUBE. We are here in San Francisco at Fort Mason. We are here for the Supernova event. Cerebras is having a big news launch. It's turning into almost like a conference. A lot of people coming, oversubscribed event, a lot of buzz. And of course Andrew Feldman is the star of the show. The founder and CEO of Cerebras. Andrew, great to see you again.

Andrew Feldman

>> It's good to see you again.

>> It's our pop-up cube. But we will do whatever it takes to get the stories that matter.

Andrew Feldman

>> We appreciate that. We appreciate that.

>> So obviously we've had a couple conversations before. You guys really on a tear recently. Meta announcement, I saw the IBM news, just continuing to thunder away at pushing the envelope on the speed.

Andrew Feldman

>> Yeah.

>> And this is not new for people who know what you've been doing and saying speed matters. The market wants more.

Andrew Feldman

>> It does.

>> They want more faster. Slow is bad, fast is good. Simple message, but hard to do. What's the core news that you guys are launching here today?

Andrew Feldman

>> Well, we're putting out a new model, Qwen 32B and as with all our models, it is the fastest in the industry, bar none. Somewhere between 20 and 30 times faster than the fastest GPU offering. Qwen's really interesting model. It's a reasoning model and it's one of the smartest models. It behaves in line with the leading closed source models, while it's an open source model. So fast model, a smart model. I think it'll give some open AIs models a real run for their money.

>> You've obviously been at this and seeing all the, I call it the AI Olympics. Say, "I got the faster thing." A lot has changed. Reasoning is putting a lot of pressure on the tokens and so there are some key requirements that people are paying attention to. Could you share in your mind the key requirement that people should pay attention to? We had Russ on, one of your early partners here, from LiveKit. He saw my first token out. How fast you get those tokens. Speed of course matters, but what are some of the things that people should pay attention to? Because it seems like the goal posts are shifting a lot. People are taking advantage of the stats, saying certain things. What do you advise people to look at as you guys push the envelope on being the fastest?

Andrew Feldman

>> Reasoning models behave much more like humans, right? If you set out to write a paper, you don't just disgorge the paper. You don't just think for a second and then write the whole paper. Likely you'd write a draft, you'd edit it, you'd think about it again, you'd review it and you would iterate. Reasoning models perform that workflow. They develop an answer, they review it, they improve it. And this requires more work.

>> More tokens.

Andrew Feldman

>> And more tokens. That's right.

>> Yeah.

Andrew Feldman

>> And tokens are just a measure of computational work. And so nobody really wants to wait 30 minutes for their research report to come out, right?

>> Mm-hmm.

Andrew Feldman

>> You'd like it out extremely quickly. And so for this type of work, for agentic workflows, for the reasoning for the smartest models, performance is particularly important because you have so many more tokens to work through. So if you are 20 or 30 times faster, you get something in a second or two instead of several minutes.

>> In your talk, you've got some stats and then always the bar charts. Break that down for me. Simplify what the message is. Is it first token out? What are some of the key bullet points and stats for the new model?

Andrew Feldman

>> So you spoke to some of our customers and partners. I think it varies based on your workload. For somebody who does voice, time to first token is a crucial measure. How fast can you get the first part of the response? I think for most other applications, output speed is the best measure. How fast can you finish the work? And that's measured by output speed. I think the best place to look is you can go every day and you can look on Artificial Analysis. They're a thoughtful benchmarking firm. They go to everybody's API, they ping it without you knowing it, and they report the results every single day. That's a good way to measure us.

>> So I got to ask you about the application. I love talking to your partners because you had a new AI entrepreneurial class. I was just making a comment on a keynote I was moderating this morning across town in San Francisco and I said when I was breaking, my first startup was an internet startup, they called it. I was an internet entrepreneur and then web entrepreneur, whatever you want to call it. But it was categorical. Now AI entrepreneurs are coming to the scene. You're attracting a lot of them. You also got the big partners. So it's like this big old and new. Or big and small. So what is an AI entrepreneur and what are the use cases that you're seeing that you guys are enabling? Because now that the horsepower is coming out, speed-

Andrew Feldman

>> Right....

>> what use cases are coming out? What kinds of use cases and what kind of startups are emerging in your mind?

Andrew Feldman

>> Let me share with you several different examples. One from the largest and one from the smallest. We'll take Kim Branson at GlaxoSmithKline. They're a 300-year-old company. They're a leader in the design and development of pharmaceuticals. They've built an agentic flow that can ask the entire corpus of literature a question. The hundreds of thousands of papers or millions of papers that have been written around the world in any language, they can distill that down. They can then develop a hypothesis. They can then have pros and cons of the hypothesis. It can then search data in support of the hypothesis and eventually it will be able to link to automated labs, which could then run experiments in support of the hypothesis.

>> On robotics?

Andrew Feldman

>> That's right. On robots.

>> Robots that have precision.

Andrew Feldman

>> That's righ.t.

>> The physical AI meets the human.

Andrew Feldman

>> That's right. And it asked a question, it scoured the existing knowledge. It played the role of a professor, of a researcher. It scoured what was known, it thought about, it delivered some questions to the community. It then went out and it answered them. And it then criticized itself. It critiqued itself, whether it was a good answer, good hypothesis or not. You still-

>> You bring up a good point that comes up a lot. The robotics is hot right now, but it's also phase one was get the machines right, get the computer vision. Now that's a lot of machine learning, a lot of tech. Now the open source software world is now plugging in to these robots as a software interface. So smaller, faster, cheaper hardware. That's like the Moore's Law world, you live and make it faster, make next version better.

Andrew Feldman

>> Yup.

>> But the software has to get smarter, more intelligent and smaller code.

Andrew Feldman

>> Sure. For those robots, understanding the world around them is the fundamental challenge.

>> And computer vision is going to be a huge inference use case.

Andrew Feldman

>> It is a big eater of inference compute for sure.

>> You've seen that too.

Andrew Feldman

>> For sure.

>> Okay, so what's the relationship with Meta? I just want to review a little bit of some past couple of weeks for Cerebras. You had some good run-

Andrew Feldman

>> It's good news.

>> You got a good run in. Not a bad week for you guys, or a couple of weeks. Take us through what's happened because a lot of breakthroughs happened. Just walk us through storyboard. You had the Meta LlamaCon, you had that event and then IBM, a bunch of other things. Take us through the news.

Andrew Feldman

>> Sure. So Meta announced that they would be providing an API space service. So you could go to their API and it would serve Llama 4. And that behind their API was Cerebras. So they would be our fast tokens that they were being served through their offering. So this was the first hyperscaler win for Cerebras. Immediately following that, IBM announced that they would be doing approximately the same thing. We would be providing fast inference to their clients. So if you think of Meta as at one end, developer friendly, mobile first, developer friendly and huge.

>> And huge by the way. The numbers are up. Huge. The numbers' up. Huge.

Andrew Feldman

>> Huge. Huge. And then IBM does business with the largest 100 customers in the world, largest accounts. We have expanded our reach from the developer all the way to the largest financial institutions in the world through IBM.

>> It's interesting, the IBM is not obvious, but I cover IBM as well pretty deeply. And they have a lot of, I call, hyper scale enterprises, like the big banks. They look like hyperscalers.

Andrew Feldman

>> Yeah. They're enormous.

>> They are huge. I think JPMorganChase is a $10 billion annual IT budget. They don't even call it IT. Just say tech budget.

Andrew Feldman

>> Right.

>> That's 10 billion a year. Lori Beer who's the CIO and I talked at Reinvent last year and it's off the charts. They are huge and that's just one.

Andrew Feldman

>> Huge opportunities.

>> But there's still other big ones too. They need inference.

Andrew Feldman

>> They do. They need inference and they need training. And I think those financial institutions, the JPMorganChases, the Citibanks, the Deutsche Banks, these are some of the largest institutions in the world.

>> You're powering basically the model deployments. Make sure I get this right. If I can read the tea leaves to put this together, you're basically going to Meta, say, "Hey, whatever models you want, we're serving it out fast." So if I've got Mistral over here, I got whatever, you can just say pick your model. You're powering it. Is that oversimplifying it?

Andrew Feldman

>> No. We bring that message to our customers around the world. We brought it to Perplexity, to Mistral, to Meta, to IBM, to our strategic partner in the Middle East, G42. We're delivering that message around the world to our partners who are here. AlphaSense doing really interesting things that you pick the model and we'll make it blazing fast for you.

>> So the startup ecosystem is here. We're in the little startup room. It's going to get populated in a few hours. Love that growth. That means there's entrepreneurship involved, so I love to see those applications. Are there other things you could share around how you guys win? I mean obviously these startups love horsepower and they want to go it.

Andrew Feldman

>> Sure.

>> But when you do a bake-off ... That's an old term, but in the old days you do a bake-off. Two people would gets a ... I'm sure there's a lot of bake-offs going on because you're claiming the fastest.

Andrew Feldman

>> Sure.

>> When you go into a bake-off, take us through what happens. Do they line up Cerebras against the other guys or-

Andrew Feldman

>> I think sometimes they do. I don't think there's any doubt we're the fastest. In fact, the others have modified their claims faced with the truth. I think what they do is they jump up on your cloud and go through your API. They run through a set of inference tasks and they compare. And I think we win those with extraordinary regularity.

>> I know you don't like to talk about competition in a direct way, but the strategy, always talk about your benefits. But I get this question a lot so I'll ask you, is people want to know whose benchmarks are real. There's a lot of benchmark Olympics going on.

Andrew Feldman

>> Sure.

>> "Hey, I got the fastest."

Andrew Feldman

>> Sure.

>> What should people look at, to pay attention? You gave that site as good reference.

Andrew Feldman

>> Artificial analysis is-

>> If I just use the power, I could say the tokens could pop out faster or is that my oversimplifying it?

Andrew Feldman

>> No, that's not quite true. There is no amount of power you can throw at a GPU to get it to perform inference at the rate we can and that's because the problem is one of memory bandwidth. They just can't move information from memory to compute fast enough to compete with us. And so I think my strategy with benchmarking has always been how do we make it the most realistic? And you don't want to run a drag race and then say, well look therefore it's better to go to the grocery store. You don't want people with months to tune and to try and cheat and to try and do all these things that you don't get when you buy the product. I think many historic benchmarks were that way. What I like about some of the modern benchmarks; artificial analysis, I think is the leader; is they go to your service when you don't know.

>> Yeah. .

Andrew Feldman

>> They do a little bit of work.

>> Yeah. They're smart enough.

Andrew Feldman

>> That's right. They do a little bit of work, they collate the results, they publish the results every single day. This is exactly what your customers get. It's not a special tuned version, maybe like MLPerf. We used to do that in the compute world in SPEC and SPECint where people would tune and tune for months for a test. That's not what you want because that's not what customers-

>> But not all workloads are considered equal too. It's all situational.

Andrew Feldman

>> That's right. And so what you want is if your customers are going to use it, to try and evaluate what their performance will be. It should mirror what their performance should be. And I think a blind test of your service is an excellent test.

>> Well Andrew, great to have you on theCUBE. Love to get the perspective. On a business question, as you guys hit your growth strategies, hit escape velocity, got the best product that's cool, that's a win. Doesn't hurt. That's great.

Andrew Feldman

>> Doesn't hurt. Product matters.

>> Being ahead of the pack is a good one.

Andrew Feldman

>> Marketing is good. The product really matters.

>> If you're the lead dog, the view is still the same, right?

Andrew Feldman

>> That's right.

>> Okay, get that. Now as you prioritize your business, obviously you want to win the big cap expense.

Andrew Feldman

>> Sure.

>> That's obviously the Metas of the world. The enterprise is a huge market and that's wide open. The data is on-prem, startups are coming online. You're starting to see the signs of the maturization of Cerebras.

Andrew Feldman

>> True.

>> What are you optimizing for? Do you have a sequenced approach? Is there a plan? Can you share as much as you can publicly to the strategy that you guys are taking?

Andrew Feldman

>> Sure. We pursued at the start those who had tremendous data assets for training and for inference. And so who has huge assets? People in life sciences like Mayo Clinic, people like GlaxoSmithKline or Nova Nordisk. And we won across the board there. As inference came to the fore, they were able to move into inference and we were able to move very quickly into AI-first companies, companies like Mistral and Perplexity and AlphaSense, and deliver to their customers value through them. To deliver Mistral blazing fast, to deliver AlphaSense's product blazing fast. And I think that helps them separate themselves.

>> Yeah.

Andrew Feldman

>> And I think in our cadre of developers, of startups, you've got the next writer, you've got the next cursor, you've got the next ... all the leaders. .

>> There's always new brands that come out of the woodwork.

Andrew Feldman

>> That's exactly right. And today they're building on blazing fast inference from Cerebras.

>> We're doing a special focus starting in the second half of the year with on theCUBE, both Silicon Valley and New York, and then going to 2026 around the AI developer. It's a little bit categorical, loosely defined, but I want to share this with you. I want to get your thoughts. You guys are announcing today that all developers receive a million free tokens per day, no wait list. That's awesome. So developers want to know where's the action? What do I do? What do I code? Am I spinning up models? What's your message to developers? Because they want to get in on the action because I hate to say picks and shovels, but there's a market of real good software that can bring into a long tail I was talking to someone who was just doing one unique thing in the enterprise and that startup is going to run the table in the category. It wouldn't have been funded 10 years ago by any VC, but they'll do a 100 million.

Andrew Feldman

>> I think you're quite right. I think maybe through the end of '23 or middle of '24, it was AI for AI's sake. And now the question is which problem do you solve with AI and how much better are you at solving it than other techniques? Now we've flipped the script and we're using AI to solve particular problems. Can you do better marketing? Can you use fewer people in HR? Can you find drugs more quickly? Can you find-

>> These white spaces are actually hugely profitable.

Andrew Feldman

>> That's right.

>> And if I can tap into the best tokens per second engine-

Andrew Feldman

>> That's right....

>> I could create a niche domain-specific based on my expertise or product.

Andrew Feldman

>> That's exactly right.

>> It's a SaaS app.

Andrew Feldman

>> It is. It becomes a SaaS app. It is the engine underneath, I think, not just a SaaS app, but an ecosystem.

>> All right, so I'm that developer. Tell me what do I do with Cerebas? I'm that developer. What do I do? I log in, sign up-

Andrew Feldman

>> Log in, you get an API, and you think about how to use inference to solve the problems that are top of mind for your customers.

>> And it's free for me for the first million tokens per day?

Andrew Feldman

>> Free for the first million tokens. You can buy in little buckets so it's friendly to the little guy.

>> All right, You're open for business with developers?

Andrew Feldman

>> We're open for business with developers

>> All right. Andrew, thanks for coming on theCUBE. I know you're busy, you got to go up over the main stage. Great to see you again.

Andrew Feldman

>> Thank you so much.

>> Congratulations for breaking the speed record. Need for speed here-

Andrew Feldman

>> Need for speed....

>> at Supernova. Love the name. Implies things are going Supernova here at Cerebras. I'm John Furrier, theCUBE. Thanks for watching.