The Story Behind Cerebras’ $63 Billion IPO with Founder and CEO Andrew Feldman
Netflix used to deliver DVDs and envelopes and when the internet got fast they became a movie studio right it opened up an entirely new business something fundamentally different
Netflix 过去靠邮寄 DVD 起家，等网速变快之后，他们转型成了电影制片公司，对吧？开辟出了一门全然不同的新生意。
that's what happens with speed and I think that's what fast AI does
速度带来的变化就是这样，我认为快速 AI 也会做到同样的事。
right now
就在现在。
we're replacing things that everybody can see like coding design the SAS tools but once we start sort of fundamentally reorganizing around this you're going to see this sort of new business models and fundamental jumps in productivity and I'm eager for that
我们现在替换的是大家都能看得见的东西，比如编程、设计、SaaS 工具，但一旦我们开始围绕 AI 从根本上重组一切，你就会看到全新的商业模式和生产力的跨越式提升。
That's so cool.
太酷了。
Today on No Prize, we have Andrew Feldman, the co-founder and CEO of Sarah Brass.
今天 No Priors 节目邀请到了 Andrew Feldman，Cerebras 的联合创始人兼 CEO。
Sarah Bros was founded in the mid 2010s to focus on new workloads for AI, particularly the machine learning world, and then has made the transition into very fast inference for the foundation model world that we live in today.
Cerebras 创立于 2010 年代中期，专注于 AI 新型工作负载，尤其是推理领域。
Serbust recently went public and is currently worth about $63 billion in the stock market.
Cerebras 近期完成上市，目前市值约为 630 亿美元。
So Andrew, thank you for joining us in No Priors.
那么 Andrew，感谢你来到 No Priors。
Oh, what a pleasure.
哦，非常荣幸。
It's good to see you guys again.
很高兴再次见到你们。
Yeah.
对。
So, first of all, congratulations.
首先，恭喜你们。
So, um, your company, Cerebras, just went public.
嗯，你们公司 Cerebras 刚刚上市了。
Um, as of today, it's a $60 billion market cap, which is pretty amazing.
截至今天，市值 600 亿美元，相当了不起。
Pretty amazing.
确实了不起。
Yeah.
是的。
And you I think you were with us a year or two ago on the show in one of the earlier episodes, and it was a pleasure to talk to you then, and obviously we're very excited to have you on today.
我记得你大概一两年前来过我们节目，是比较早期的一集。
Could you tell us a bit how the business evolved since that time and what you folks just a reminder for our audience what you do, what you're focused on, how you're going forward.
能跟我们说说，从那时候到现在公司是怎么发展的，你们刚刚完成了什么？
We we build AI computers, right?
我们做的是 AI 计算机，对吧？
Computers computers designed to and optimize to accelerate AI workloads.
专为加速 AI 工作负载而设计和优化的计算机。
And right now we're the the fastest at inference, not by little, but by a lot, 15, 18, 20x faster than GPUs.
目前我们的推理速度是最快的，而且不是快一点点，而是快很多，快 15、18、20 倍。
And so what happened was um starting in about 2025 AI models got smart enough to be useful.
大概从 2025 年开始，AI 模型智能到足以真正派上用场了。
People began using them and you know we make AI with training and we we use it with inference.
人们开始使用它们，你知道，我们用训练来构建 AI，用推理来使用它。
So as people began to to use it, it began to to sort of be integrated into their day-to-day work.
随着大家开始使用，它逐渐融入了日常工作之中。
Um speed became fundamentally important and we were just crushed with demand.
速度变得至关重要，我们被需求直接淹没了。
Is it is this faster across the board or is this specific use cases?
这是全面更快，还是只针对特定使用场景？
Faster across the board.
全面更快。
big models, small models, US models, Chinese models, um trillion parameter models, one billion parameter models across the board.
大模型、小模型、美国模型、中国模型，万亿参数的模型、十亿参数的模型。
And then what happened was at the end of the year, we signed a deal with with OpenAI, sort of one of the biggest deals ever in Silicon Valley, sort of north of 20 billion.
然后到了年底，我们和 OpenAI 签了一笔交易，算是某种意义上的里程碑。
And then in March, we signed an agreement with AWS where we will be deployed in their data centers going forward.
然后在三月，我们和 AWS 签了协议，未来我们将部署在他们的数据中心里。
And so it was just a whirlwind year and a half of chasing the chasing supply and trying to trying to sort of meet the demand
过去一年半，我们就是在疯狂追货、追产能，努力跟上需求。
and what what shifted in the year in the last year and a half was it the ramp in manufacturing?
过去一年半里，到底是什么发生了变化？是产能规模提升了，
Was it a new chip design?
还是新的芯片设计？
Was it something else?
还是别的原因？
Could you help educate folks on
能给大家解释一下吗？
what what what happened was
发生的事情是这样的，
we built a really really fast machine and for a long time nobody cared.
我们造出了一台真正快的机器，但很长一段时间没人在意。
Right.
对。
That's because
那是因为，
actually forgive me for saying so, but a lot of people objected and said this is just a weird architecture.
说句实话，那时候很多人都反对，说这不过是个晶圆级方案。
They they called it wrong.
他们都说你们判断错了。
Like Cubas called it wrong.
就像 Cerebras 走错了方向一样。
Yeah.
是。
Yeah.
是的。
They they did.
他们确实这么说。
I I think um to be radically better, right?
我认为，要想真正做到彻底更好，对吧？
You you can't build something that that is a similar architecture, right?
你不能用相似的架构做出来，对吧？
You're not going to get 15 or 20 times better than the GPU wi with a minor modification to their architecture.
靠对 GPU 做小修小改，你不可能比它快 15 倍或 20 倍。
And that's probably true across the board that if you're going to aspire to a radical improvement, your design has to be different.
这条规律大概放之四海皆准：如果你追求的是颠覆性的提升，
And from the beginning, you know, we chose wafer scale, which means we build a 46,000 square millimeter chip, a chip the size of a dinner plate, whereas everybody else is building chips the size of postage stamps.
从一开始我们就选择了晶圆级方案，也就是说我们做的是一块 46,000 平方毫米的芯片，一块餐盘大小的芯片，而其他所有人做的都是邮票大小的芯片。
They told us we were out of our mind.
他们说我们是疯子。
It would never work.
说这根本行不通。
They they listed reasons why it was impossible.
他们列出了一堆不可能成功的理由。
But in 2019, we we proved it was possible.
但在 2019 年，我们证明了它是可行的。
we began delivering it and we improved on it and we improved on it.
我们开始交付，然后不断改进，一次又一次地改进。
Um, but we were fast when AI was a novelty and when it's a novelty, nobody cares that you're fast because it's not being used.
不过，我们在 AI 还是个新奇事物的时候就做到了很快，而当它只是新奇事物的时候，没人在乎速度。
And so from about 2023 to the beginning of 25 sort of people pointed at AI, but nobody used it every day in their work.
所以从 2023 年到 25 年初，大家都在聊 AI，但真正在用的人几乎没有。
And once you use something every day in your work, it can't be slow.
一旦你每天在工作中使用某样东西，它就不能慢。
I mean, how how long will you guys wait for a website to resolve?
你们能忍受网站加载多久？
I'll have no attention,
我会一秒都不耐烦，
right?
对吧？
That's exactly right.
完全正确。
That that's exactly the way it is.
就是这样。
I mean, how big is the market for slow search?
慢搜索的市场有多大？
It's zero.
等于零。
How big is the market for dialup internet?
拨号上网的市场有多大？
It's zero.
等于零。
That's how big the market for slow inference will be.
慢速推理的市场就会是这个大。
But we had to wait until it was smart enough to be useful.
但我们得等到模型足够聪明、真正有用。
And that happened in 2025.
这件事在 2025 年发生了。
And that's why you got this sort of explosion of of demand and companies like Cognition and Cursor and Lovable and and just all these others that began ramping extraordinary.
这也是为什么你看到了需求爆炸，Cognition 这些公司都在快速增长。
Many of the ones you guys have invested in are are ramping like crazy.
你们投资的很多公司都在疯狂扩张。
Open AAI and and and and others and and we were right there with the right product.
OpenAI 和其他几家，我们正好有合适的产品。
I think I first met you back in 2016 or something like that and at the time uh people weren't even like saying AI sounded weird, right?
我记得第一次见你大概是 2016 年前后，那时候大家还在讨论机器学习，
We were talking about machine learning and the models of the time were uh convolutional neuronet networks and RNN's and you know just the emergence of GANs and things like that.
当时我们聊的是卷积神经网络这类模型，
We were trying to tell the difference between a chair and a cat
我们还在努力区分椅子和猫。
right that was qui's great PhD is like a cat
对，那时候某人的博士论文就是研究猫，
or a chair like whoa look how far we've come.
或者说椅子，哇，看看我们走了多远。
It's unbelievable.
难以置信。
Yeah.
对。
Yeah.
是啊。
What do you think um gave you the foresight to build against the market?
你觉得是什么让你有了这份前瞻性，敢于逆着市场去建造？
Because to your point, I think a lot of us believed in that this market would be really important and you more than others, right, since you actually started a company in it.
因为就像你说的，我们很多人当时都相信这个市场会非常重要，
But then it took some time for the market to really expand to the point where uh to your point now it's it's this massive use case.
但市场真正扩大、达到你说的那个程度，还是花了一些时间。
People really care about speed of inference and other things.
大家真的开始重视推理速度和其他东西。
Um what gave you the conviction back then to do this
当初是什么给了你做这件事的信念？
combination of of vision um the right co-founders and a little bit of arrogance, a little bit of luck.
是愿景的结合，加上合适的联合创始人，再加上一点点傲慢，一点点
You know, we we saw AI on the horizon as a new workload and as computer architects, new workloads are opportunity, right?
我们看到 AI 作为新型工作负载出现在地平线上，作为计算机架构师，新的工作负载需要新的架构，
It's very very hard to to to enter in in the x86 world, right?
要在 x86 的世界里立足非常非常难，对吧？
Where there's not nothing new is happening there and nothing has happened for generations.
那个领域早已没有任何新鲜事，几代人以来什么都没发生。
But you know when graphics emerged, you got the discrete GPU and you you you got uh Nvidia and and when uh when the mobile uh compute hit, you you got ARM.
但当图形处理兴起时，你有了独立 GPU，有了 NVIDIA，
And it was interesting that that not Intel, not AMD, not all sorts of people who you would have thought really well positioned to win in that business, they all got no share.
有意思的是，不是 Intel，不是 AMD，不是那些你本以为会胜出的公司赢得了这场比赛。
And so we knew that that that this new workload would eat a lot of compute.
于是我们知道，这种新的工作负载会消耗大量算力。
it would require uh a new architecture, a dedicated architecture, and it ought to be very different.
它需要一种新的架构、一种专用架构，而且应该非常
The architecture could not be a derivative of what's existing.
这个架构不能是现有架构的衍生版。
Those were our big bets and they were 100% Mhm.
这些是我们的核心赌注，而且百分之百。嗯。
And they turned out to be dead, right?
结果证明是完全正确的，对吧？
Were there moments where you just doubted whether this would work given the different we had a period
有没有某些时刻，考虑到我们讨论过的种种不同因素，你会怀疑这能不能成功？
uh you know, we're solving a problem that that had never been solved before.
我们在解决一个从未被解决过的问题。