Cerebras (CBRS) Q3 2026 Earnings Call Transcript

Operator00:00

Good day and thank you for standing by welcome to the Cerebras System third quarter 2026 earnings conference call at this time all participants are in a listen only mode after the speaker's presentation there'll be a question and answer session to ask a question during the session you'll need to press star 11 on your telephone you will then hear an automated message advising your hand is raised to withdraw your question please press star 1 again please be advised that today's conference is being recorded I would now like to turn the conference over to Sean Dorsey head of investor relations please go ahead Sean

Sean Dorsey01:11

Thank you operator good afternoon everyone and welcome to Cerebras Systems Q3 2026 earnings call earlier today we issued our press release and posted our supplemental earnings presentation to the investor relations section of our website a replay of this webcast will also be available on our investor relations website following the call joining me today are Andrew Feldman our co-founder chief executive officer and president and Bob Komin our chief financial officer before we begin I would like to remind everyone that today's discussion will include forward-looking statements under the safe harbor of the Private Securities Litigation Reform Act of 1995. These statements include but are not limited to statements regarding our future financial performance business strategy market opportunity customer demand product roadmap technology leadership supply chain operating model and outlook for Q4 and full year 2026. forward looking statements are based on our current expectations and assumptions and are subject to risks and uncertainties that could cause actual results to differ materially from those expressed or implied.

Sean Dorsey01:26

These risks are described in our SEC filings including our final prospectus related to our IPO and our future periodic filings with the SEC. We undertake no obligation to update these forward looking statements except as required by law.

Sean Dorsey01:40

During today's call we will also discuss certain non-GAAP financial measures. Reconciliations between GAAP and non-GAAP results are included in today's press release and supplemental materials available on our investor relations website.

Sean Dorsey01:43

With that, I'll turn the call over to Andrew.

Prepared remarks

Andrew Feldman02:14

Thank you, Sean, and thank you all today for joining us. As we look to the months ahead, we see two certainties, rapid change and compute shortages. In the last 90 days, the AI ecosystem has seen the tidal of leading frontier model change hands multiple times It's seen dozens of interesting open models released major security incidents occurred and powerful new consumer agents emerged and many other developments that none of us could even remember.

Andrew Feldman02:21

It's impossible for anyone to predict AI capabilities 90 days for now let alone a year from now.

Andrew Feldman02:34

Yet through all that change intelligence and speed remain the two essential features for every use case from coding to security to personal agents

Andrew Feldman02:58

At Cerebras our job in this rapidly changing world is to make sure that no matter how AI grows Cerebras grows a foundation models grow we grow as open models grow we grow as personal agents grow we grow whether enterprises or startups grow we grow and whether customers want hardware or cloud capacity we grow.

Andrew Feldman03:27

Our goal is to position ourselves so that whatever drives the industry Cerebras wins. And in Q3 we made remarkable progress towards that goal. First I'd like to report that we exceeded all guided business metrics and while I'm pleased with these results I believe even greater importance is the fact that we closed a number of deals with key AI industry leaders since our last earnings call.

Andrew Feldman03:48

With the latest being our announcement that today we signed an important cloud deal with Microsoft which along with the AWS and Oracle announcements previously made bring us to three hyperscalers who have chosen to leverage the Cerebras platform as part of their inference strategy.

Andrew Feldman03:59

These new deals alongside continued growth and execution with our existing customers enable us to raise Q4 core guidance across the board.

Andrew Feldman04:13

And continue to see revenue tripling in 2027 with expanding gross margins. Bob will share all of the details regarding our Q3 performance and forward looking guidance in his section.

Andrew Feldman04:25

Let me take a minute to clarify the shape of our business as an AI accelerator company our pioneering innovations are in the chips systems and software that make AI compute fast.

Andrew Feldman04:41

We are an AI hardware company with two channels to market. First we sell systems installing racks of Cerebras hardware in our customers data centers. This is a great business revenue is recognized quickly and cash is received quickly.

Andrew Feldman04:58

It is also lumpy and depends on customers having data center capacity ready to receive the hardware. Our second route to market is through our cloud there's a very rapidly growing category of AI natives who want to consume tokens rather than deploy hardware.

Andrew Feldman05:32

And given the fact that data centers and compute are in very short supply the fact that we have both the system and the data center is a strategic advantage. The AI cloud business is a particularly good business if you make your own hardware our cost per megawatt is much lower and as such we can bring more capital efficiency than Neo clouds The mix of these two channels will vary from quarter to quarter but we believe both will grow significantly over time.

Andrew Feldman05:44

Let me also address speculation about our OpenAI partnership. OpenAI is a very large customer and a close partner and I believe they will continue to be so going forward.

Andrew Feldman05:56

OpenAI grew in Q3 as compared to Q2 and we expect more growth in the future. The success of OpenAI's ultra fast 056 preview based on Cerebras was overwhelming.

Andrew Feldman06:01

Demonstrating unequivocally that the demand for fast tokens is enormous.

Andrew Feldman06:25

It is also true that Cerebras didn't have sufficient capacity to support ultra fast in a broad rollout including for Astra. As others have stated we believe that capacity is key driver for business decisions in our industry in fact even with the enormous supply of GPUs available to OpenAI usage based restrictions still needed to be implemented.

Andrew Feldman06:39

After signing the agreement with OpenAI took us longer than we would have liked to find contract four and bring online the substantial increase in data center capacity required to meet the aggressive requirements in the agreement.

Andrew Feldman06:45

And as such we find ourselves running behind schedule on the initial tranches.

Andrew Feldman06:57

Given the importance of this partnership to our business we understand the close scrutiny it gets from investors. But details of the agreement or news about it are unlikely to come from any source other than Cerebras or OpenAI.

Andrew Feldman07:04

For example there were rumors that Cerebras couldn't support Astra for technical reasons. These are false.

Andrew Feldman07:13

CS3 CS4 and CS5 are more than capable of supporting Astra scale models and larger models as well.

Andrew Feldman07:33

At the same time I expect and you should expect OpenAI to build capacity using a range of hardware solutions. OpenAI will retire older models deploy new ones across those compute options for internal and external use their compute needs are vast with plenty of opportunities for different solutions to play a role as they scale.

Andrew Feldman07:53

Before turning to our business framework let me make one more general point despite the experimentation low batch GPU startup announcement mega kernels and hyperscalers internal custom silicon we continue to be the fastest inference solution bar none.

Andrew Feldman08:00

We believe our roadmap will enable us to maintain that advantage for years to come.

Andrew Feldman08:12

Our framework for evaluating progress remains unchanged. We assess our business across three dimensions, customers, capacity, and capabilities.

Andrew Feldman08:24

Let's begin with customers. We are rapidly expanding our customer base to broaden our reach and diversify our revenue stream. Q3 was a successful quarter to that end.

Andrew Feldman08:41

We had an important win with two new hyperscalers. We signed an agreement to bring Cerebras hardware into Oracle Cloud where Cerebras will be deployed in Oracle data centers serving AI labs, enterprises, and developers building interactive and intelligent applications.

Andrew Feldman08:53

We expect this Cerebras powered Oracle Cloud inference service to launch in the summer of 2027. We also signed an exciting deal with Microsoft about which we will share more next year.

Andrew Feldman09:22

That brings us to three hyperscaler wins in the last six months this is an important achievement and it positions us well for 2027 and 2028. We also expanded our reach through a variety of Neo clouds with off-takers of high caliber like fireworks diversifying our routes to market and giving customers more ways to access Cerebras high speed inference.

Andrew Feldman09:35

We also benefited from open weight models which continue to enable new applications and expand what AI can do in everyday life and across industries.

Andrew Feldman09:47

For example just this week we announced that Cerebras Fast Inference is powering the built neighborhood OS platform and personal assistant bringing blisteringly fast inference to millions of people.

Andrew Feldman10:07

With consumer apps like built we're putting the power of Cerebras directly into people's pockets. That is why fast tokens command a premium and our cloud customers tell us that fast tokens generate exceptional gross margins and that Cerebras fast tokens are the most profitable tokens they sell.

Andrew Feldman10:19

Let's now turn to capacity. Capacity has two components data center capacity and manufacturing capacity. Both are critical to growth.

Andrew Feldman10:34

Data center availability remains a constant constraint across the industry and we are no exception. Moreover we believe the evolving US policy landscape could create greater challenges for data center development in the US.

Andrew Feldman11:04

In 2027 more than 75% of our data center capacity comes online outside of the US in fact in the next several months new facilities in Canada in France and in Norway come online and during the quarter we signed contracts for additional capacity in Norway and Finland further expanding our European footprint and we're in the final stages of contracting for a data center in Sao Paulo our first in South America.

Andrew Feldman11:23

So far we've secured approximately 700 gross megawatts of data center capacity for delivery by the end of 2027 our first party cloud is already among the larger non hyperscaler AI clouds and we're far down the path to negotiations for gigawatts of capacity.

Andrew Feldman11:33

Finally the best way to increase data center capacity is by increasing the token generating capability of the hardware in the data center.

Andrew Feldman11:46

But simply to generate more tokens per megawatt. To this end this aggregated inference plays a role in expanding our capacity by increasing by up to 5x the number of tokens that a data center can deliver.

Andrew Feldman11:53

More super fast tokens per megawatt means each data center produces markedly more revenue.

Andrew Feldman12:10

Turning now to the second element of capacity manufacturing and supply. We're scaling manufacturing with Flex and Sanmina we expect to exit 2026 with more than 10 times the manufacturing capacity we had at the start of the year.

Andrew Feldman12:22

In 2027 we plan to add factories and further increase capacity to support the substantial growth we anticipate. our supply chain is well positioned to support this growth as well.

Andrew Feldman12:34

Most importantly we have sufficient wafer commitments from TSMC which continues to distinguish itself as a world class partner. It is also important what our supply chain does not need.

Andrew Feldman13:01

That is also key to us. Unlike other competitors we do not use HBM memory. we believe memory shortages and the price increases associated with them will continue through 2028 raising competitors costs and limiting supply industry-wide memory challenges create an opportunity for Cerebras Factories and supply are necessary but not sufficient to scale capacity.

Andrew Feldman13:19

We must also have high yields Cerebras has invented a fundamentally new computing architecture applying metrics and mental models designed for other architectures can be like examining a Tesla and asking where's the exhaust pipe or trying to measure its tailpipe emissions.

Andrew Feldman13:34

Electric cars require a different frame of reference and so does wafer scale computing. We welcome questions about every aspect of our design and we'll work to share architectural details more clearly.

Andrew Feldman14:06

But there are also a set of persistent falsehoods proselytized to the financial community namely that our systems can't support large models obviously false as they currently do and that our manufacturing yields are low both are false I already addressed model size earlier in my comments and in previous discussions Since manufacturing yields are harder to observe I'll take some time here to address the topic directly.

Andrew Feldman14:25

There are two types of yield fab yield and parametric yield. Our yields from TSMC which are the fab yields are high. So are our parametric yields. While our technical blog covers this subject in detail I'll provide a brief overview

Andrew Feldman14:35

All wafers contain defects. For decades those defects were a central reason that industry considered wafer scale processors impossible.

Andrew Feldman14:55

We solved that problem through redundancy. Our design contains roughly a million identical tiles with a small portion held in reserve. When a defect disables a tile the hardware routes around it and activates a spare.

Andrew Feldman15:09

This mechanism works both at the fab and in the field. delivering high wafer yields and strong infield reliability. The second challenge is parametric variation.

Andrew Feldman15:21

Even when silicon functions its electrical characteristics vary across the wafer. Some regions run faster or slower than other regions at the same voltage.

Andrew Feldman15:32

This is a basic fact of chip design and one of the obstacles we have to solve to make wafer scale computing possible. We address parametric variation in three ways.

Andrew Feldman15:49

First, we remove the wafer's outer edges. The edges are the weakest part of a wafer and account for much of the parametric variation. We use the largest square that fits into a 300 millimeter wafer excluding much of the weakest silicon.

Andrew Feldman16:10

Second, we design and test for a broad range of conditions. Our cores tolerate a wider range of variation during time enclosure the process of verifying that circuits meet timing requirements we validate the design across operating conditions much broader than those used by others in the industry.

Andrew Feldman16:21

This enables expected performance despite variations across the wafer. Finally we tune voltage independently across the wafer.

Andrew Feldman16:37

We invented technology to adjust core voltage independently across 84 wafer sections. We can raise voltage in slower sections or lower it in faster sections. giving us precise control over performance.

Andrew Feldman16:52

We know of no other chip with this capability. The result is exceptionally high yield at both the fab and the factory. We achieve this by engineering solutions to the problems others said made wafer scale computing impossible.

Andrew Feldman17:01

As a final note there are those who have pointed to our relatively low gross margins on hardware shipments and suggested that this is evidence of a yield problem.

Andrew Feldman17:14

When in fact it simply reflects the ongoing ramp of our manufacturing capabilities and the final shipments of hardware covered under special pricing programs offered to one of our early customers.

Andrew Feldman17:23

We expect our hardware margins to improve markedly over the next several quarters eventually exceeding our long-term target of 60%.

Andrew Feldman18:03

Turning now to capabilities. we are advancing our capabilities on every front new systems and processors improved software advanced manufacturing techniques and partnerships that deliver best in industry solutions despite the rapid pace of change across the AI industry the importance of inference speed continues to grow and Cerebras remains the undisputed leader we believe this performance advantage will endure over time and to that end our product roadmap continues to advance rapidly CS4 will soon be in full production doubling both performance and throughput.

Andrew Feldman18:15

Our next generation processor the wafer scale engine five which will be deployed in our fifth generation CS5 system is back from the fab ahead of schedule.

Andrew Feldman18:35

The WSC five taped out on June 15th the chip is back has been brought up and is already running models and is undergoing in chassis testing. Moving from tape out to system test in less than five months is exceptional execution.

Andrew Feldman18:51

The CS5 will more than double the speed of the CS4 it will enter production in early summer of 27 it is built on TSMC's five nanometer node the competition for wafers is less intense and wafer costs are much lower.

Andrew Feldman19:14

Our innovations also extend to complete solutions to our knowledge Cerebras is the first to deliver production scale heterogeneous disaggregated inference today our cloud serves customers using AMD GPUs for prefill and Cerebras CS3 systems for decode.

Andrew Feldman19:48

To enable the production launch we built a disaggregated inference software platform it is built to work with AMD as well as other GPUs the same solution will be deployed across the Cerebras inference cloud and to select data center customers Disaggregated inference delivers substantially more high speed tokens for the same power that means more profit per megawatt for cloud owners and lower cost per token for AI users.

Andrew Feldman19:57

We believe disagg will become one of the default offerings in major cloud deployments.

Andrew Feldman20:09

Our software stack continues to mature our cloud services deliver impressive uptime our systems keep getting faster and the time needed to bring up new models has fallen to just a few days.

Andrew Feldman20:44

Cerebras was a day zero launch partner for Kimi K3.1 Moonshot AI's 2.8 trillion parameter frontier class open weight model perhaps the premier open weight model in the market Kimi 3.1 on Cerebras is not only the fastest in the industry but it runs at up to 2000 tokens per second Nearly 11 years into our journey we continue to innovate at a tremendous pace our team has stayed together through good times and tough times and there is more energy and excitement in our building than ever.

Andrew Feldman20:51

We take pride in tackling and solving some of the industry's hardest problems.

Andrew Feldman21:32

By way of summary we beat on all metrics and we're raising guidance for Q4 across the board and expect triple revenue in 2027. we made important progress in each of our key domains including adding two hyperscale customers and have now won three hyperscalers in the last six months we began shipping our CS4 our fifth generation processor the WSE5 is back and in CS5 system test in record time production disagg is now live with AMD a day zero support for yet another 2.8 trillion parameter model was successfully achieved at industry best performance.

Andrew Feldman21:36

With that let me turn things over to Bob.

Bob Komin22:00

Thank you Andrew and good afternoon everyone. Before I get into the specifics on our Q3 results and Q4 outlook I'd like to reflect on how we've scaled the business. We chose to ramp a cloud offering in early 2025 because in this rapidly advancing AI landscape AI natives want to scale compute quickly and don't want to own data centers.

Bob Komin22:10

And because hyperscalers plan their data center compute allocations years in advance, they often don't have near term availability in their own facilities for new hardware.

Bob Komin22:23

In order to win the large deployments the foundation labs and hyperscalers require, Cerebras needed to demonstrate that our systems could deliver our industry leading inference performance at scale.

Bob Komin22:33

After having the opportunity to thoroughly evaluate our solution in our private cloud these customers can then choose to buy our hardware or purchase our cloud offering.

Bob Komin22:46

To be able to provide enough cloud capacity quickly to meet these demands we chose to begin to lease data center capacity and grow our private cloud as quickly as possible towards gigawatt scale.

Bob Komin22:56

This general strategy appears to have been sound as data center capacity is now very scarce and we have a lot of it under contract.

Bob Komin23:40

Even more importantly once we had our private cloud we were able to demonstrate the value of fast inference to the entire market and drive more demand for it but large new data center capacity can take 15 to 18 months to bring online instead of waiting for these data centers and to accelerate our ability to meet the overwhelming demand we already had we rented some of our own equipment back from one of our customers as we've previously described this applied downward pressure to gross margin in the short term but it also allowed us to grow existing customers and win significant new ones because we had some capacity when they were ready to buy.

Bob Komin23:59

We have now won Frontier labs like OpenAI and hyperscalers like AWS and recently added two new hyperscalers to our customer base. We have other leading customers like Cognition, Figma, Block, and Built among others on our cloud.

Bob Komin24:17

We meet these customers where they are in some cases selling them hardware and in others delivering them our cloud services. Soon new Neo cloud customers we have signed will begin distributing fast inference services running on Cerebras hardware in their clouds.

Bob Komin24:35

In our prior calls we also shared that 2026 is going to be a year of building out capacity capabilities and adding customers to set us up to more than triple our revenue in 2027. With our strong progress so far this year we are well positioned to achieve that target.

Bob Komin24:49

Now turning to Q3 results. I'd like to talk about our execution in the third quarter and I'm pleased to say that Q4 will be an important inflection point for our growth and core financial trends.

Bob Komin25:05

We've exceeded our core financial guidance across the board for the third quarter in a row. We also delivered positive non-GAAP net income for the first time as a public company and ahead of schedule.

Bob Komin25:16

As a reminder the definition of our core metrics and the associated reconciliations to GAAP are included in today's earnings release and on our website.

Bob Komin25:33

We delivered record core revenues of $231 million up 10% sequentially and 70% year over year exceeding the $215 million midpoint of our guidance by 16 million or 7%.

Bob Komin25:44

Revenue from our core cloud and other services was also a record increasing by 272% year over year to 145 million.

Bob Komin25:56

Sequential cloud revenue growth in Q3 was 13%. We currently expect flattish sequential cloud revenue in Q4.

Bob Komin26:07

This trend is not driven by a pricing or demand issue as the pricing environment is healthy and demand for our solutions continues to outpace supply by a wide margin.

Bob Komin26:21

Our ramping cloud revenue is limited entirely by our data center capacity which we will begin to ramp in December as we build momentum going into next year but unfortunately that's too late to affect Q4.

Bob Komin26:37

We expect much more substantial increases in data center capacity across 2027. The bulk of this starts in the middle of next year as we deliver larger portions of the 700 gross megawatts of data center capacity currently under contract.

Bob Komin26:58

Core hardware revenue was 86 million in the quarter increasing 5% sequentially. We are seeing increasing customer demand for fast inference delivered not just through our private cloud but also through new hardware deals that we've closed with several new customers.

Bob Komin27:16

With those hardware shipments scheduled to begin in Q4 we will see a significant step up in total core revenue. Sequential core revenue growth from Q3 to Q4 will approach 40% and year over year growth will be about 85%.

Bob Komin27:23

These hardware orders including Oracle will continue to ship through at least the first half of 2027.

Bob Komin27:51

As Andrew mentioned earlier today we also announced Microsoft as a new customer. Over the last few months our pipeline has grown by much more than 10x driven by a diverse set of opportunities from both new prospects and existing customers including hyperscalers Neo clouds sovereigns enterprises and AI natives

Bob Komin28:05

Turning now to gross margin Q3 core gross margin was 41% a sequential improvement of 40 basis points as compared to Q2 and 200 basis points above our guidance for the quarter.

Bob Komin28:17

In addition to the gross margin impact from renting back our systems we also face some headwinds in the quarter as we were burning off the last shipments from a hardware purchase agreement signed years ago at lower pricing.

Bob Komin28:28

We guided that all of this would create temporary gross margin declines and the low point would be reached in Q3.

Bob Komin28:34

Well we did better than that and the low point was actually last quarter in Q2

Bob Komin28:49

Now a break total Q3 core gross margin into its components. Core hardware gross margin declined sequentially to 31% driven by the final shipments under the year old customer agreement.

Bob Komin29:01

We are now shipping hardware orders at significantly higher current prices which also means significantly better margins starting in Q4.

Bob Komin29:11

Gross margin from core cloud and other services improved significantly to 46.9% 510 basis points higher than in Q2.

Bob Komin29:17

The primary driver of the increase was new business delivered at higher pricing and margin.

Bob Komin29:35

This was partially offset by our program of renting back data center capacity and our hardware at higher cost from a customer. We expect this to continue through at least the middle of 2027 until we have enough of our own data center capacity to fulfill our high levels of cloud demand.

Bob Komin29:49

With all of that as a backdrop in Q4 we expect total core gross margin to improve by a step function of 700 basis points to 48% at the midpoint.

Bob Komin29:59

We expect total core margin to exceed 50% in the near term and to continue to improve towards our 60% long term target.

Bob Komin30:20

We also made strong progress towards profitability in Q3 and year to date. Core operating margin improved sequentially by 170 basis points to a loss of 14.3% which exceeded our guidance midpoint by nearly 10 full margin points.

Bob Komin30:42

Core operating loss was flat in Q3 versus Q2 at approximately 33 million core net income turned positive and was 30 million adjusted core EBITDA year to date was also positive at 2 million dollars

Bob Komin30:56

With strong increases in revenue and gross margin forecast for Q4 we expect that all of our core key profit measures including operating margin will be positive next quarter.

Bob Komin31:21

Now turning to RPO and the OpenAI contract. Remaining performance obligations increased to 26.1 billion at September 30th we update RPO for new business that was signed by quarter end and adjust existing contract values for any changes to estimates, of what remains to be recognized over the remaining committed term.

Bob Komin31:33

Q3 RPO did not include the impact of new deals signed after September 30th including Oracle which is large.

Bob Komin31:53

There have been a lot of questions about the status of our OpenAI contract since it represents the majority of our reported RPO to date. As a reminder OpenAI contract which was filed with our S1 before we went public is a take or pay contract for 750 megawatts of capacity.

Bob Komin32:09

Among many contract requirements that apply to both parties it includes a term that if we experience delays in delivering capacity they are entitled to credits during that time to use against future committed purchases.

Bob Komin32:19

We have calculated the amount for potential delay credits and deducted them from RPO at September 30th.

Bob Komin32:35

These credits would have minimal impact on revenue and gross margin because as a percentage of RPO they are estimated to be relatively small and then they would get spread ratably as the majority of the revenue gets recognized over the next few years.

Bob Komin32:47

Future changes to these estimates will be updated each quarter and reflected in both RPO and guidance if needed.

Bob Komin33:11

If a delay in delivering capacity continues for 90 days or more both OpenAI and us have the option to cancel that portion of capacity only the amount we estimate could be subject to cancellation this year would be less than 5% of the total committed megawatts

Bob Komin33:26

Given the relentless demand for data center capacity around the world it has taken us some additional time against a very aggressive schedule to find and bring up the initial data center capacity required to meet the delivery schedule.

Bob Komin33:43

Based on our team executing frequent in-person inspections of each site we now project we will be back on schedule by the second half of 2027 and this should reduce future delay credit risk.

Bob Komin33:57

We've taken a conservative approach and have not assumed that any of these delays may get excused due to dependencies that are outside our control. If that happens it would reduce the amount of the delay credits we have assumed.

Bob Komin34:11

Also if OpenAI or Cerebras were to elect an option to cancel a delayed capacity tranche we believe there's sufficient demand and we would find other customers to absorb it.

Bob Komin34:25

OpenAI also retains an option to purchase an additional 1.25 gigawatts beyond their initial 750 megawatt commitment that could come online starting in 2029.

Bob Komin34:36

Since that's still three years away and the world is changing as quickly as it is in AI we've not even begun to discuss this yet.

Bob Komin34:43

Now turning to our balance sheet and capital efficiency.

Bob Komin34:52

We continue to have strong liquidity and in Q3 with 7.8 billion in cash equivalents restricted cash and marketable securities.

Bob Komin35:24

We also have a revolving credit facility of up to 850 million dollars that remains unused to date. The upfront 1 billion working capital loan we received in January and continuing payments we receive for non-recurring pass-through expenses to fit out new data center capacity to be delivered as part of the OpenAI contract have funded the majority of our net working capital requirements and capital expenses so far in 2026.

Bob Komin35:45

We are not a Neo cloud we are vertically integrated inventor of the fastest inference solution in the industry our business comes with a level of durability in technology differentiation that we believe will drive much higher long term margin potential and cash generation over the long run.

Bob Komin36:35

Our model also requires lower capital investment for the mix of compute equipment we've deploy. For a significant portion of the equipment used in our data centers that we manufacture similar to others who've chosen to make and deploy their own AI chips we do not pay a third party 70% gross margins nor do we use high price high priced HBM we only pay our bomb cost which reduces the relative capital expense per megawatt while we also benefit from producing the fastest and most valuable tokens this shows up in our capital structure to date where the only debt on our balance sheet despite our rapid growth and private cloud build out is the OpenAI working capital loan.

Bob Komin37:02

We've also received and continue to get upfront payments from customers which is another lever for capital efficiency. Our capital structure today includes significant cash and investments on hand and little leverage which provides us with more flexibility to have multiple ways to continue to fund our rapid growth which is an advantage especially in the current capital market environment.

Bob Komin37:26

Now turning to our outlook. For Q4 2026 we expect to raise core revenue to the range of 310 to 330 million core gross margin to the range of 47 to 49% and core operating margin to the range of minus 1% to 1%.

Bob Komin37:46

For the full year 2026 we're raising core revenue to the range of 940 million to 960 million dollars we're raising core gross margin to the range of 42 to 44% and we're raising core operating margin to the range of minus 10% to minus 8%.

Bob Komin38:04

In closing Q3 was another record quarter for Cerebras. And Q4 will be too. Our flexible model of selling both hardware and cloud is driving strong growth in revenue and improvements in gross margin that began a quarter earlier than expected.

Bob Komin38:22

We've made significant progress scaling our manufacturing and data center capacity which will enable us to triple revenue with core gross margin progressing towards our 60% long term target in 2027.

Bob Komin38:33

We've demonstrated strong operating leverage and are projecting core operating profit to be approximately break even in Q4

Bob Komin39:06

Then we expect to make significant progress towards our long term 40% operating margin target during 2027. this mix of scale growth and profitability positions us as a unique and leading AI systems provider also as a final note I'd like to remind everyone that we've now been a public company for more than 180 days and as such our IPO stock lockup releases were fully complete as of November 9th Thank you for listening to our prepared remarks.

Bob Komin39:12

I'd like to ask the operator to please open the line for questions.

Questions and answers

Operator39:26

Thank you. As a reminder, if you would like to ask a question, please press star 11 on your telephone. You will then hear the automated message advising your hand is raised. If you would like to remove yourself, please press star 11 again.

Operator39:38

We also ask that you limit yourself to one question and one follow-up. As well, wait for your name and company to be announced before proceeding with your question. One moment while we compile the Q&A roster.

Operator39:46

First question will be coming from Tom O'Malley of CraigHallum. Please go ahead.

Thomas O'Malley40:07

Hey Andrew. Hey Bob. Uh thanks for taking the questions and congrats on the really good results. Uh first question is in regards to the new deals OpenAI and Microsoft. If you could just give us any metrics perhaps on the relative sizes of those new deals and then Bob you mentioned that some of OpenAI is already in the RPO perhaps giving us how much is already in the RPO?

Andrew Feldman40:20

Hey Tom, this is Andrew. Good to hear from you.

Andrew Feldman40:40

we'll be announcing more details about the Microsoft opportunity in the deal we struck with them we are really obviously very excited about that we'll be announcing that in more detail in the first half of next year and that's when you can expect revenue to start from that transaction Bob

Bob Komin41:25

Hey, Tom, in the RPO, what we've included for OpenAI is our estimate of the committed amount of revenue under the contract under the take or pay contract no extensions no no additional options adjusted by the delay credits that we think would be due could be due and it also includes for data centers that we've signed for them it includes the amounts that they pay for non-recurring costs set them up and also for monthly recurring costs over the term of the contract.

Bob Komin41:32

And those are the pass-through revenues and pass-through expenses that we recognize in the P&L.

Thomas O'Malley41:40

Gotcha, Bob. I I said OpenAI, I meant Oracle, but we can circle to that. But I guess the other one is just we've seen the ability from frontier.

Bob Komin41:41

Yeah.

Bob Komin41:46

We fight really hard to win hyperscalers we can't

Bob Komin41:54

The answer for Oracle is even easier so far no Oracle in the RPO we signed it after

Thomas O'Malley41:56

Gotcha. Thank you.

Bob Komin41:56

Yeah.

Thomas O'Malley42:17

Alright and then the follow-up is so we've seen the ability from Frontier AI labs to just not only to build the hardware quickly like Holograph but to also accelerate model deployment across different compute hardware and increase fungibility. So how do you guys benefit from this deployment and what do you expect in terms of the future of heterogeneous infrastructure to look like? How does that play into your hands?

Andrew Feldman42:24

Sure, that's a really good question.

Andrew Feldman42:38

And in their I think in their S1 or on a tweet talked about how it took them about three days to bring up AMD GPUs

Andrew Feldman42:45

And this is a very interesting and important development

Andrew Feldman43:16

Particularly those who have a long term focus I don't think people have really grokked what this means to our ecosystem compilers and low level software have historically been a fundamental impediment to new architectures they have favored the incumbent they were the software moat that you heard so much about

Andrew Feldman43:33

If Infropic can in fact do what they say they do it means that the value of this low level software shrinks and over time disappears. And it means that the moats from that low level software also shrink and disappear.

Andrew Feldman43:44

And if the low level software has little value the thing that is left to provide value is innovative hardware

Andrew Feldman44:03

And it means that sort of differentiation will be more dependent on hardware differentiation. I think this is a tremendous advantage to innovative hardware companies like Cerebras and others who build pioneering hardware and use new and different architectures.

Andrew Feldman44:18

And I I think this will remove an advantage that the incumbents have long had. But that's a really good question and I I think something that the industry needs to keep an eye on. Thank you.

Operator44:30

Thank you for your question. One moment for the next question. That question will be coming from the line of Joe Moore of Morgan Stanley. Please go ahead.

Joseph Moore44:48

Great. Thank you. And congratulations on the strong hardware numbers. Can you talk about the RPO issue with the scheduling on the cloud? You know is there anything specific that we can focus on and you're sort of ability your confidence in your ability to resolve that before it turns into additional penalties?

Bob Komin45:01

I think what we put in the RPO was a very conservative number of

Bob Komin45:45

What in the absolute worst case we could be hit with and so I I think we are comfortable today that we will be better than that these are commonplace in big data center deals we have similar agreements with our data center builders there are penalties when they're late and I I think it varies over time whether you choose to implement them and you make a business decision at that time I I think that's the way these work and that's the way we expect these to work going forward

Joseph Moore45:55

Okay. And and what was there a cause of the delay like I think we all recognize issues with land power shell and things like that was it something like that or was it something else that that sort of caused it

Bob Komin45:59

Our delays right now are exclusively land power show

Joseph Moore46:00

Got it.

Bob Komin46:10

We have capacity we have all materials we are waiting on the completion of land power shell

Joseph Moore46:12

Great. Thank you.

Joseph Moore46:14

Thanks, y'all.

Operator46:23

Thank you. The next question will be coming from the line of Atif Malik of Citi please go ahead.

Atif Malik46:53

Hi thank you for taking my questions Andrew I have a question the product roadmap when are you expecting crossover of CS4 versus CS3 is it in 2027 and then on just your product roadmap in CS4 you improved throughput significantly and if you stay on that path into CS5 and beyond can at some point your chip deliver enough fast speed and throughput that your system cannot be used its system can be used alone and no GPUs are needed.

Andrew Feldman46:59

Well, so hey, if I were you

Andrew Feldman47:23

It's good to talk to you. Um first until a few weeks ago all our deployments were CS3 alone it has only recently that we've begun deploying disaggregated solutions and we saw a meaningful bump in throughput which is really exciting

Andrew Feldman47:41

Um so that's observation number one so obviously they work really well in an aggregated form but disaggregation which we are now deploying and it sort of GA with

Andrew Feldman47:57

Produces even more high speed tokens when high speed tokens are constrained. I would say the crossover for CS3 versus CS4 is in Q1 27

Andrew Feldman48:10

The roadmap is looking really good right now we are ahead of schedule on the CS5 And while I mentioned it in my prepared remarks,

Andrew Feldman48:16

Getting from tape out to chip and system

Andrew Feldman48:28

In five months is extraordinary. And we expect that to begin shipping in early summer 27

Andrew Feldman48:40

And we are that doubles performance over the four the four double performance over three

Atif Malik49:00

Great. And then for Bob your guiding to flat cloud sales for Q4 so within that tripling of sales in 2027 can you just help us understand the mix of cloud versus hardware

Bob Komin49:57

I think the best way to think about the mix of hardware is the size of those deals can be very very large what we would like to do is provide some color on the shipment over the next quarter or two it it's very reliant on our customers having their data centers ready for us so what I can say is we think that the hardware mix in Q4 and into Q1 will be closer to 50% than where it's been over the last few quarters as we look forward into the rest of 2027 you know we'll have more information for the Q4 call and I think we'll give a more complete view of our outlook on sort of everything for 2027 then

Atif Malik50:00

Thank you.

Operator50:11

Thank you. One moment for the next question. Our next question is coming from the line of Quinn Bolton of Needham and Company. Please go ahead.

Quinn Bolton50:30

Hey, guys, let me add my congratulations on the progress. Andrew, just wanted to make sure I heard you right. In the RPO and your assumptions, you're making assumptions on potential delay credits, but you have not received any cancellations for delayed capacity at this point. Is that correct?

Bob Komin50:33

That is correct.

Quinn Bolton50:57

Excellent. Okay, good. Um a second question is uh you know there's you know fears in the market about um you know fast inference from low batch GPUs um that sounds like an expensive way to serve tokens. So wondering if you can address sort of the cost per token uh for Cerebras versus uh deployment on the low batch uh GPU configuration.

Andrew Feldman51:26

Sure. The the GPU allows you to reduce throughput by trading off for additional speed. And you do that by bringing down batch. So you have a lot of throughput at low speed, and as you try and get higher speed, your throughput hemorrhages.

Andrew Feldman51:31

To the point where you're in the low batch domain

Andrew Feldman51:38

Now that that's the characteristic of the architecture.

Andrew Feldman51:57

The the way to think about that is they can get to some modest performance by going to low batch. But when they do that, the cost per token, the cost to generate a token jumps.

Andrew Feldman52:14

They have fewer simultaneous customers so the total number of tokens generated is very small. As you continue to go faster, we become stronger and stronger.

Andrew Feldman52:33

And we get to a point where no matter what you do with the GPU, they can't deliver the performance we can at low batch, at batch size one, they still can't do it. And so the question is, where do you cross over?

Andrew Feldman52:49

I think uh around 250, 300 tokens per second, we are more economical by the time you get to 500, 800, 1000 tokens we are vastly more economical.

Andrew Feldman53:02

Somewhere between 800 and 2000 tokens the GPU can't play. And that's the way it is right now. that's sort of the lay of the land. That's a good question on a complicated topic.

Quinn Bolton53:07

Thank you, Andrew.

Operator53:12

Thank you for your question. One moment for the next question.

Operator53:20

Our next question is coming from the line of Joshua Bachalter of TD Cowen. Please go ahead.

Joshua Buchalter54:03

Hey guys. Thank you for taking my question and congrats on the Microsoft deal. Um maybe to start can you speak to us about how you know your allocating capacity in light of the land power shell issues and you know there's a couple layers to that you know like it doesn't sound like silicon or rack availability is an issue so I guess how much do you how much outlet do you see is there for incremental hardware sales um to customers now and then I guess given the challenges with getting land power and shell for cloud customers has there been any discussions with OpenAI about them potentially taking hardware sales and putting them in their own or their partners data centers.

Bob Komin54:04

Thank you.

Andrew Feldman54:31

Sure. The the question really had two parts, right, Josh? The first is is it really the case that that land power shell and the construction thereof is the limiting factor? The answer is absolutely. TSMC has committed sufficient capacity wafers to us for 2027 we're very comfortable with that.

Andrew Feldman54:43

The fact that we don't use HBM obviously hugely beneficial as that's in very very tight supply.

Andrew Feldman55:02

The you are correct in the second part of your question which is doesn't that create opportunity for you to sell hardware while your data centers are being built. I think the only caveat to that is that to the best of our knowledge everybody's data centers are late.

Andrew Feldman55:21

That includes OpenAI, that includes all the data centers we know of being built for hyperscalers, data centers are late across the board and across the world. So the constraint of data centers isn't unique to us, it is also manifest in our customers.

Andrew Feldman55:39

And so the yes we have supply for them yes we are shipping to our customers with data centers yes it's the fact that the case that they too are very late make sense?

Joshua Buchalter55:52

Yeah, no, totally. But so the last part of the annoying three part question was, you know, is there any potential for the OpenAI contract to switch to hardware sales in light of the

Joshua Buchalter55:53

So

Joshua Buchalter55:54

Land power shell?

Andrew Feldman56:06

Sorry. It it is in the contract they have that that right. But as I'm saying our understanding from them is their data centers are also behind schedule.

Joshua Buchalter56:13

Okay, got it. Thank you. And then for my follow-up, can you just help us?

Andrew Feldman56:15

Seven parts, Josh?

Joshua Buchalter56:18

It wasn't my best. I'm sorry.

Joshua Buchalter56:20

All right. Check.

Joshua Buchalter56:38

I acknowledge that. Can you maybe talk about how Oracle and Microsoft are positioning you know your offering with their customers? Is this a case of it'll be hardware sales and then it'll just basically track based on how the demand for you know ultra fast inferences for their customers? Thank you.

Andrew Feldman57:04

Sure. We've announced more information about the Oracle engagement and partnership and we will set aside for 2027 more details on Microsoft. In Oracle we are they are purchasing hardware it is being deployed in their data centers those data centers are in the US we are collaborating and they are building pipeline for that now there's a ton of

Andrew Feldman57:23

And so uh we do not think that is uh by any stretch of the imagination the last deal we do together. it is a start of a large partnership in which our equipment will be deployed in Oracle data centers and part of the OCI cloud.

Joshua Buchalter57:29

Thank you Andrew and I'll give up a question next quarter.

Andrew Feldman57:33

You you trading me a hamburger today for a question next quarter?

Joshua Buchalter57:38

Fair deal.

Operator57:48

Thank you so much. That does conclude today's Q&A session. I would like to turn it back to Sean for closing remarks go ahead Sean

Sean Dorsey58:04

In closing, please note we are attending the Tomorrow X Summit in Austin next week on November 17th as well as the Barclays Global Technology Conference in San Francisco on December 9th. Thank you for joining us today. Operator please go ahead and close the call.

Operator58:08

Thank you. This concludes today's programming. You may all disconnect.