Cerebras CEO Andrew Feldman Talks Earnings

Summary

---

Bloomberg Talks: Cerebras CEO Andrew Feldman Talks Earnings

Date: June 24, 2026
Host: Bloomberg Tech Host
Guest: Andrew Feldman, CEO of Cerebras Systems

---

Episode Overview

This episode features an in-depth conversation with Andrew Feldman, CEO of Cerebras, following the company’s first quarterly earnings report since going public. The discussion centers on Cerebras' financial performance, the challenges and opportunities of scaling in the AI hardware industry, unique aspects of the company’s technology, and how Cerebras differentiates itself from competitors facing supply chain and capacity constraints. Feldman provides both technical and business context, directly addressing concerns regarding margin contraction, market potential, and future capital strategies.

---

Key Discussion Points & Insights

1. Earnings Recap & Market Reaction

• Cerebras posted record revenues of $191 million, a 92% increase year over year.
• Cloud business grew dramatically—up 167% YoY.
• Sales guidance beat Wall Street estimates, but shares dropped 17.5% due to concerns over margin contraction.

“We put forward a plan at the start of ‘26...we delivered record revenues of 191 million, up 92% year over year...we beat margin consensus substantially and then we guided for full year the gross margins would be 10% better than plan.”
— Andrew Feldman (01:20)

• Margin decline resulted from buying back gear to meet unexpected demand; this was deliberately done to keep key customers supplied, despite compression of gross margins (by about 10-15 points).

2. Ramp-Up & Capacity Constraints

• The company used IPO proceeds primarily to expand capacity.
• Physical data center space, not compute hardware, is the main constraint.
• Data center deals are now a global priority, including a 120-megawatt partnership with Bell Canada (to be delivered in 2027).

“Capacity is the largest constraint right now for everyone...We announced a data center partnership with Bell Canada for 120 megawatts...We are pursuing data centers across the US, Canada, Europe, and the Middle East.”
— Andrew Feldman (02:37)

• The AI market moves rapidly, but real estate and construction for data centers lag—posing a bottleneck for all players.

“The AI market is moving at blistering speed and we are being constrained by data centers which move with the speed of real estate.”
— Andrew Feldman (03:45)

3. Technology Differentiation: Memory & Supply Chain Resilience

• Cerebras does not rely on traditional off-chip HBM (High Bandwidth Memory), avoiding major industry bottlenecks (HBM supply is dominated by just three companies, with expensive, long lead times).

“Because of our innovative architecture, because of our wafer scale approach, we don't use HBM...There is a global shortage, it's extremely expensive. Lead times are long and we don't use it. So we have a tremendous advantage there.”
— Andrew Feldman (04:28)

• Avoids other major constraints like TSMC’s 3nm node and co-loss processes.

“Our architecture has allowed us to deliver the fastest inference in the world by an order of magnitude while avoiding the main supply chain constraints faced by others in the field.”
— Andrew Feldman (05:05)

4. Customer Delivery & Speed to Deploy

• Cerebras claims to have unmatched speed in bringing compute online.
• Example: Signed with OpenAI on December 24, full production by February 1.

“We signed our contract with OpenAI on December 24th and had them in full production on February 1st. That's unheard of.”
— Andrew Feldman (05:49)

5. AI Compute Market Structure & Competition

• AI compute is a heterogeneous (diverse) market; will not be “consolidated around GPUs.”
• Market will include ASICs from various players, and novel architectures from companies like Cerebras.
• The overall AI compute market is exceptionally large and growing.

“This market will not be consolidated around GPUs. There will be ASICs from hyperscalers, there’ll be ASICs from labs, and then there’ll be companies like Cerebras with pioneering architectures.”
— Andrew Feldman (06:22)

6. Financial Position & Capital Plans

• Cerebras has more than $9 billion on its balance sheet.
• Open to additional fundraising if it accelerates growth, but comfortable for now.

“We have more than 9 billion on the balance sheet...of course we are always scanning both the capital markets, both equity and debt for ways to accelerate our growth.”
— Andrew Feldman (07:22)

7. Personal Reflections & Success Metrics

• Execution against aggressive plans is Feldman’s measure of success.
• Emphasis on setting and beating high targets, supported by strong Q2 and Q3 guidance.

“I think my mother's proud of me. I think that's the biggest thing you can ask for...When you lay out a plan that is aggressive and you crush it, that's how you feel good about your ability to execute.”
— Andrew Feldman (07:48)

---

Memorable Quotes & Moments (w/ Timestamps)

• On Margin Contraction and Customer Management:
  “We would go back to some of our customers and we would rent back gear that we’d sold them to try and keep up with demand and that would have a margin impact on the order of 10 or 15 points...to keep our customers close.”
  (01:20)

• On the Real Bottleneck:
  “The AI market is moving at blistering speed and we are being constrained by data centers which move with the speed of real estate.”
  (03:45)

• On Wafer-Scale Architecture’s Supply Chain Resilience:
  “We have a tremendous advantage there...our architecture has allowed us to deliver the fastest inference in the world by an order of magnitude while avoiding the main supply chain constraints faced by by others.”
  (04:28–05:05)

• On Industry Diversity:
  “This market will not be consolidated around GPUs.”
  (06:22)

• On Speed of Execution:
  “We signed our contract with OpenAI on December 24th and had them in full production on February 1st. That's unheard of.”
  (05:49)

• On Personal Satisfaction:
  “I think my mother's proud of me. I think that's the biggest thing you can ask for.”
  (07:48)

---

Important Segments & Timestamps

• [01:20] Feldman outlines the quarter’s record results, margin strategy, and demand from key customers.
• [02:37] IPO proceeds and global data center expansion.
• [03:45] Real estate challenges for the AI market.
• [04:28] Explanation of technical differentiators and supply chain resilience.
• [05:49] OpenAI case study—unmatched delivery speed.
• [06:22] Discussion on market structure and the future of AI compute.
• [07:22] Company’s robust financial position and openness to future capital markets activity.
• [07:48] Defining success and personal reflections.

---

Closing Notes

In this concise but wide-ranging post-earnings conversation, Andrew Feldman emphasizes Cerebras' rapid growth and ability to avoid major supply chain disruptions thanks to its bespoke chip architecture. While the stock faced a sharp drop over margin concerns, Feldman articulates that these were strategic moves to sustain customer relationships and respond to persistent, extraordinary demand. The episode highlights the shifting competitive landscape for AI hardware, the practical limits imposed by real estate, and a CEO intent on both execution and long-term industry leadership.

Transcript

[Indiana University Narrator]

Indiana University strengthens tomorrow's workforce with practical real world experience. IU grads make a difference in your community, serving as teachers, nurses and engineers who rise to tomorrow's challenges and meet them. Learn more at iu. Edu Impact

[Andrew Feldman]

Bloomberg Audio Studios Podcasts, Radio

[Bloomberg Tech Host]

news Cerebras reported quarterly earnings for the first time since going public last month. Its sales outlook beat Wall street estimates but still disappointed investors. Hoping to see the company carve out a bigger slice of the identity center market. Right now, shares down 17.5%, its biggest drop in its quite short history as a public company. What's behind that? CEO Andrew Feldman's with us. Welcome back to Bloomberg Tech. Andrew. You know there was a time where we would talk about the merits of top to tail serv ownership, how owning all of the content of. Now we're going to talk about margin contraction and we're going to talk about the stock being down 17%. That, that to me is kind of the mismatch. The outlook on the sales side beat Wall street estimates. I think a lot of people are trying to understand the sequential margin decline and for me, that this is about ramping output for two big customers. Is that true?

[Andrew Feldman]

Yeah, I think what we did is we put forward a plan in, in the start of 26. We shared it with investors as we went public and we're ahead of plan. You know, we delivered record revenues of 191 million, up 92% year over year. And for our cloud business, you know, it was up 167% year over year. We, we beat margin consensus substantially and then we guided for full year the gross margins would be 10% better than plan. We also shared that in Q2 and Q3 we would go back to some of our customers and we would rent back gear that we'd sold them to try and keep up with demand and that would have a margin impact on the order of 10 or 15 points. We did that to keep our customers close to be sure we could keep up with their extraordinary demand for our product for fast inference. And so that was the story on every metric we put out. We're, we're ahead of plan.

[Bloomberg Tech Host]

Have the proceeds from the IPO actually allowed you to move more quickly in ramping up capacity?

[Andrew Feldman]

Yeah, I think capacity is the largest constraint right now for everyone. Data centers are and we've significantly increased our, our, our ability and our pipeline for data centers, which is now very large. You know, we announced a datacenter partnership with Bell Canada for 120megawatts that will be delivered in 2027 we are pursuing data centers across the US in Canada, in Europe, in the Middle East. There are the vast resources that we have now at our disposal give us tremendous advantage in the pursuit of of this. The limiting factor data centers

[Bloomberg Tech Host]

you're talking about like matter of factly is buildings, not necessarily the compute. Right? It's not what you guys are offering. How difficult right now is it to to get moving in America or other markets to to get planning approval, get the concrete, get the labor, get the thing built.

[Andrew Feldman]

Now that's the irony of this market that the, the AI market is moving at blistering speed and we are being constrained by data centers which move with the speed of real estate. And so that is a problem that is being confronted by by everybody in the category, by the hyperscalers, by the NEO clouds, by the new generation clouds. Everybody is confron similar problem.

[Bloomberg Tech Host]

Andrew, Cerebras does not rely on traditional off chip hbm. Would you just explain that the basics of the technology. But how insulated are you from the memory bottleneck that others are experiencing?

[Andrew Feldman]

Yeah, that's a really good point. Because of our innovative architecture, because of our wafer scale approach, we don't use hbo. HBM is a type of DRAM and it's made by three companies, one of whom is reporting shortly. Right, that's Micron, HYN and Samsung. There is a global shortage, it's extremely expensive. Lead times are long and we don't use it. So we have a tremendous advantage there. The other constraints in the supply chain for for many are co loss which is a process inside of TSMC. Again we don't use it.

[Bloomberg Tech Host]

Right.

[Andrew Feldman]

And the third is capacity at the 3 nanometer node that space in TSMC factory that makes 3 nanometer chips. Again we don't use it, we're at the nanometer node. So our architecture has allowed us to deliver the fastest inference in the world by an order of magnitude while avoiding the main supply supply chain constraints faced by by others in the field.

[Bloomberg Tech Host]

Can you say hand on heart, not just winning business but if you actually been able to go to a customer and say we can get this compute online faster than others for those reasons you just outlined and then actually gone and done it?

[Andrew Feldman]

Oh for sure.

[Bloomberg Tech Host]

I mean any case studies we signed,

[Andrew Feldman]

we for example we signed our contract with OpenAI on December 24th and had them in full production on February 1st. That's unheard of.

[Bloomberg Tech Host]

That's quick, that's really quickly Andrew, this morning OpenAI is out with Jalapeno Intelligence processor But everyone is compute constrained, right? How do you interpret the frontier labs going to custom Asics alongside their other compute options?

[Andrew Feldman]

And one of the things we've been saying all along is that this market is enormous and is going to be met with a heterogeneous collection of, of architectures for hardware. This market will not be consolidated around GPUs. There will be ASICs, there will be Asics from hyperscalers, there'll be Asics from, from labs, and then there'll be companies like Cerebras with pioneering architectures who all of us will take big bites of this enormous market. I think one of the things that's difficult to get your head around is just how big the compute market is right now. That one of the things AI does is it it makes tractable for computer much of the world around us. And that really wasn't the case prior to I

[Bloomberg Tech Host]

Andrew, if it's, if the world is simple and it's money that greases the wheels to get your industry going, can we see you come back to capital markets in some form just so you can move quicker?

[Andrew Feldman]

We have more than 9 billion on the balance sheet, I think so we are, we're really pleased with our, our position there. But of course we are always scanning both the capital markets, both equity and debt for, for ways to accelerate our growth.

[Bloomberg Tech Host]

Finally, how are you judging success yourself? What is the milestone that you'd pitch to the market to keep closest attention to?

[Andrew Feldman]

Well, I think, I think my mother's proud of me. I think that's the biggest thing you can, you can ask for. I think for markets. I think when you lay out a plan that is aggressive and you crush it, that's how you feel good about both your ability to execute and your ability to predict your own execution. And so you know, we, we announced yesterday that we would beat our, our full year. We gave full year guidance that was 10 gross margin points above consensus. You ought to be proud of that. And we got to continue to execute and continue to set extraordinarily high bars and then continue to, to beat them.

[Bloomberg Tech Host]

Cerebral CEO Andrew Feldman, thank you very much indeed.

[4imprint Advertiser]

Whatever your goal, trade show giveaways, client Gifts or Team Gear 4 imprint has the promo products to match with thousands of options from apparel and drinkware to tech and totes, it's easy to find the right fit for your brand and budget. With standout choices at every price point and with their 360 degree guarantee, you can be four imprint certain your order will show up just right, right on time. Explore more at 4imprint.com 4imprint for certain.