Cerebras#Technology

Cerebras was founded in 2015 by Andrew Feldman, Gary Lauterbach, Michael James, Sean Lie and Jean-Philippe Fricker.{{Cite web|last=Tilley|first=Aaron|title=AI Chip Boom: This Stealthy AI Hardware Startup Is Worth Almost A Billion|url=https://www.forbes.com/sites/aarontilley/2017/08/31/ai-chip-cerebras-systems-investment/|access-date=2021-04-30|website=Forbes|language=en}} These five founders worked together at SeaMicro, which was started in 2007 by Feldman and Lauterbach and was later sold to AMD in 2012 for $334 million.{{Cite web|last=Hardy|first=Quentin|date=2012-02-29|title=A.M.D. Buying SeaMicro for $334 Million|url=https://bits.blogs.nytimes.com/2012/02/29/a-m-d-buying-seamicro-for-334-million/|access-date=2021-04-30|website=Bits Blog|language=en-US}}{{Cite magazine|title=How Google Spawned The 384-Chip Server|language=en-US|magazine=Wired|url=https://www.wired.com/2012/01/seamicro-and-google/|access-date=2021-04-30|issn=1059-1028}}

In May 2016, Cerebras secured $27 million in series A funding led by Benchmark, Foundation Capital and Eclipse Ventures.{{Cite web|title=A stealthy startup called Cerebras raised around $25 million to build deep learning hardware|url=https://techcrunch.com/2016/12/19/a-stealthy-startup-called-cerebras-raised-around-25-million-to-build-deep-learning-hardware/|access-date=2021-04-30|website=TechCrunch|date=19 December 2016 |language=en-US}}

In December 2016, series B funding was led by Coatue Management, followed in January 2017 with series C funding led by VY Capital.

In November 2018, Cerebras closed its series D round with $88 million, making the company a unicorn. Investors in this round included Altimeter, VY Capital, Coatue, Foundation Capital, Benchmark, and Eclipse.{{Cite web|last=Martin|first=Dylan|date=2019-11-27|title=AI Chip Startup Cerebras Reveals 'World's Fastest AI Supercomputer'|url=https://www.crn.com/news/components-peripherals/ai-chip-startup-cerebras-systems-raises-88-million-series-d-round|access-date=2021-04-30|website=CRN}}{{Cite web|last=Strategy|first=Moor Insights and|title=Cerebras Unveils AI Supercomputer-On-A-Chip|url=https://www.forbes.com/sites/moorinsights/2019/08/19/cerebras-unveils-ai-supercomputer-on-a-chip/|access-date=2021-04-30|website=Forbes|language=en}}

On August 19, 2019, Cerebras announced its first-generation Wafer-Scale Engine (WSE).{{Cite news|last=Metz|first=Cade|date=2019-08-19|title=To Power A.I., Start-Up Creates a Giant Computer Chip|language=en-US|work=The New York Times|url=https://www.nytimes.com/2019/08/19/technology/artificial-intelligence-chip-cerebras.html|access-date=2021-04-30|issn=0362-4331}}{{Cite web|title=The Cerebras CS-1 computes deep learning AI problems by being bigger, bigger, and bigger than any other chip|url=https://techcrunch.com/2019/11/19/the-cerebras-cs-1-computes-deep-learning-ai-problems-by-being-bigger-bigger-and-bigger-than-any-other-chip/|access-date=2021-04-30|website=TechCrunch|date=19 November 2019 |language=en-US}}{{Cite web|title=Full Page Reload|url=https://spectrum.ieee.org/cerebras-giant-ai-chip-now-has-a-trillions-more-transistors|access-date=2021-04-30|website=IEEE Spectrum: Technology, Engineering, and Science News|language=en}}’

In November 2019, Cerebras closed its series E round with over $270 million for a valuation of $2.4 billion.{{Cite news|title=Cerebras Crams More Compute Into Second-Gen 'Dinner Plate Sized' Chip|work=EE Times|url=https://www.eetimes.com/cerebras-crams-more-compute-into-second-gen-dinner-plate-sized-chip/|access-date=2021-05-12}}

In 2020, the company announced an office in Japan and partnership with Tokyo Electron Devices.{{Cite web| title= Cerebras Systems Expands Global Footprint with New Offices in Tokyo, Japan, and Toronto, Canada|url=https://www.yahoo.com/now/cerebras-systems-expands-global-footprint-130000796.html|access-date= August 13, 2021 |work= Press Release |author= Cerebras Systems |language= en-US}}

In April 2021, Cerebras announced the CS-2 based on the company's Wafer Scale Engine Two (WSE-2), which has 850,000 cores.{{Cite web|date=2021-04-20|title=Cerebras launches new AI supercomputing processor with 2.6 trillion transistors|url=https://venturebeat.com/2021/04/20/cerebras-systems-launches-new-ai-supercomputing-processor-with-2-6-trillion-transistors/|access-date=2021-04-30|website=VentureBeat|language=en-US}} In August 2021, the company announced its brain-scale technology that can run a neural network with over 120 trillion connections.{{Cite web|date=2021-08-24|title=Cerebras' Tech Trains "Brain-Scale" AIs|url=https://spectrum.ieee.org/cerebras-ai-computers|access-date=2021-09-22|website=IEEE Spectrum|language=en}}

In November 2021, Cerebras announced that it had raised an additional $250 million in Series F funding, valuing the company at over $4 billion. The Series F financing round was led by Alpha Wave Ventures and Abu Dhabi Growth Fund (ADG).{{Cite web|title=Cerebras Systems Raises $250M in Funding for Over $4B Valuation to Advance the Future of AI Compute|url=https://www.hpcwire.com/off-the-wire/cerebras-systems-raises-250m-in-funding-for-over-4b-valuation/|access-date=2021-11-10|website=HPCwire|language=en-US}} To date, the company has raised $720 million in financing.{{Cite news|date=2021-11-10|title=AI chip startup Cerebras Systems raises $250 million in funding|language=en|work=Reuters|url=https://www.reuters.com/article/cerebras-tech-idUSKBN2HV23G|access-date=2021-11-10}}

In August 2022, Cerebras was honored by the Computer History Museum in Mountain View, California. The museum added to its permanent collection and unveiled a new display featuring the WSE-2—the biggest computer chip made so far—marking an "epochal" achievement in the history of fabricating transistors as an integrated part.{{Cite web |title=AI startup Cerebras celebrated for chip triumph where others tried and failed |url=https://www.zdnet.com/article/ai-startup-cerebras-celebrated-for-chip-triumph-where-others-tried-and-failed/ |access-date=2022-08-04 |website=ZDNet |language=en}}{{Cite web |date=2022-08-03 |title=The Biggest Chip In the World |url=https://computerhistory.org/blog/the-biggest-chip-in-the-world/ |access-date=2022-08-04 |website=CHM |language=en}}

Cerebras filed its prospectus for initial public offering (IPO) in September 2024, with the intention of listing on the Nasdaq exchange under the ticker 'CBRS'. The prospectus indicated that most of its revenue at the time came from Emirati AI holding company G42.{{Cite news |last1=Griffith |first1=Erin |last2=Metz |first2=Cade |date=September 30, 2024 |title=Cerebras, an A.I. Chipmaker Trying to Take On Nvidia, Files for an I.P.O. |url=https://www.nytimes.com/2024/09/30/technology/cerebras-ai-chips-ipo.html |work=The New York Times}} A week after the filing, it was reported that the Committee on Foreign Investment in the United States was reviewing G42's investment into the company, leading to a potential delay in its IPO.{{Cite news |last1=Wang |first1=Echo |last2=Cherney |first2=Max A. |last3=Hu |first3=Krystal |date=October 8, 2024 |title=Cerebras likely to postpone IPO due to CFIUS review delay on G42 deal, sources say |url=https://www.reuters.com/technology/cerebras-likely-postpone-ipo-due-cfius-review-delay-g42-deal-sources-say-2024-10-08/ |publisher=Reuters}} In a May 2025 interview, CEO Andrew Feldman said the company had obtained clearance from a U.S. committee to sell shares to G42 and he hopes that Cerebras will go public in 2025.{{Cite web |last=Novet |first=Jordan |date=2025-05-15 |title=Cerebras CEO says chipmaker's 'aspiration' is to hold IPO in 2025 |url=https://www.cnbc.com/2025/05/15/cerebras-ceo-says-chipmakers-aspiration-is-to-hold-ipo-in-2025.html |access-date=2025-06-17 |website=CNBC |language=en}}

Cerebras was named to the Forbes AI 50 in April 2024{{Cite web |last=CAI" |first="KENRICK |title=Forbes 2024 AI 50 List - Top Artificial Intelligence Startups |url=https://www.forbes.com/lists/ai50/ |access-date=2024-11-01 |website=Forbes |language=en}} and the TIME 100 Most Influential Companies list in May 2024.{{Cite magazine |last=Steinberg |first=Don |date=2024-05-30 |title=TIME100 Most Influential Companies 2024: Cerebras Systems |url=https://time.com/6979980/cerebras-systems/ |access-date=2024-11-01 |magazine=TIME |language=en}}

The Cerebras Wafer Scale Engine (WSE) is a single, wafer-scale integrated processor that includes compute, memory and interconnect fabric. The WSE-1 powers the Cerebras CS-1, Cerebras’ first-generation AI computer.{{Cite web|title=Full Page Reload|url=https://spectrum.ieee.org/cerebrass-giant-chip-will-smash-deep-learnings-speed-barrier|access-date=2021-04-30|website=IEEE Spectrum: Technology, Engineering, and Science News|language=en}} It is a 19-inch rack-mounted appliance designed for AI training and inference workloads in a datacenter. The CS-1 includes a single WSE primary processor with 400,000 processing cores, as well as twelve 100 Gigabit Ethernet connections to move data in and out.{{Cite web|date=2020-06-09|title=Neocortex Will Be First-of-Its-Kind 800,000-Core AI Supercomputer|url=https://www.hpcwire.com/2020/06/09/neocortex-will-be-first-of-its-kind-800000-core-ai-supercomputer/|access-date=2021-04-30|website=HPCwire|language=en-US}} The WSE-1 has 1.2 trillion transistors, 400,000 compute cores and 18 gigabytes of memory.

In April 2021, Cerebras announced the CS-2 AI system based on the 2nd-generation Wafer Scale Engine (WSE-2), manufactured by the 7 nm process of TSMC . It is 26 inches tall and fits in one-third of a standard data center rack.{{Cite web|last=Ray|first=Tiernan|date=April 20, 2021|title=Cerebras continues 'absolute domination' of high-end compute, it says, with world's hugest chip two-dot-oh|url=https://www.zdnet.com/article/cerebras-continues-absolute-domination-of-high-end-compute-it-says-with-worlds-hugest-chip-two-dot-oh/|access-date=August 13, 2021|website=ZDNet|language=en}}

The Cerebras WSE-2 has 850,000 cores and 2.6 trillion transistors.{{Cite magazine|last=Knight|first=Will|date=August 24, 2021|title=A New Chip Cluster Will Make Massive AI Models Possible|language=en-US|magazine=Wired|url=https://www.wired.com/story/cerebras-chip-cluster-neural-networks-ai/|access-date=2021-08-25|issn=1059-1028}}

The WSE-2 expanded on-chip SRAM to 40 gigabytes, memory bandwidth to 20 petabytes per second and total fabric bandwidth to 220 petabits per second.{{Cite news|title=Cerebras Systems Smashes the 2.5 Trillion Transistor Mark with New Second Generation Wafer Scale Engine|work=Bloomberg|url=https://www.bloomberg.com/press-releases/2021-04-20/cerebras-systems-smashes-the-2-5-trillion-transistor-mark-with-new-second-generation-wafer-scale-engine|access-date=2021-06-02}}{{Cite web|last=Cutress|first=Dr Ian|title=Cerebras Unveils Wafer Scale Engine Two (WSE2): 2.6 Trillion Transistors, 100% Yield|url=https://www.anandtech.com/show/16626/cerebras-unveils-wafer-scale-engine-two-wse2-26-trillion-transistors-100-yield|access-date=2021-06-03|website=www.anandtech.com}}

In August 2021, the company announced a system which connects multiple integrated circuits (commonly called "chips") into a neural network with many connections.

It enables a single system to support AI models with more than 120 trillion parameters.{{Cite web|last=August 2021|first=Joel Khalili 25|date=2021-08-25|title=The world's largest chip is creating AI networks larger than the human brain|url=https://www.techradar.com/au/news/the-worlds-largest-chip-is-creating-ai-networks-larger-than-the-human-brain|access-date=2021-09-22|website=TechRadar|language=en}}

In June 2022, Cerebras set a record for the largest AI models ever trained on one device.{{Cite web |author1=Francisco Pires |date=2022-06-22 |title=Cerebras Slays GPUs, Breaks Record for Largest AI Models Trained on a Single Device |url=https://www.tomshardware.com/news/cerebras-slays-gpus-breaks-record-for-largest-ai-models-trained-on-a-single-device |access-date=2022-06-22 |website=Tom's Hardware |language=en}} Cerebras said that for the first time ever, a single CS-2 system with one Cerebras wafer can train models with up to 20 billion parameters.{{Cite web |date=2022-06-22 |title=Cerebras Systems sets record for largest AI models ever trained on one device |url=https://venturebeat.com/2022/06/22/cerebras-systems-sets-record-for-largest-ai-models-ever-trained-on-one-device/ |access-date=2022-06-22 |website=VentureBeat |language=en-US}} The Cerebras CS-2 system can train multibillion-parameter natural language processing (NLP) models including GPT-3XL 1.3 billion models, as well as GPT-J 6B, GPT-3 13B and GPT-NeoX 20B with reduced software complexity and infrastructure.

In September 2022, Cerebras announced that it can patch its chips together to create what would be the largest-ever computing cluster for AI computing.{{Cite web |last=Shah |first=Agam |date=2022-09-14 |title=Cerebras Proposes AI Megacluster with Billions of AI Compute Cores |url=https://www.hpcwire.com/2022/09/14/cerebras-proposes-ai-megacluster-with-billions-of-ai-compute-cores/ |access-date=2022-09-14 |website=HPCwire |language=en-US}} A Wafer-Scale Cluster can connect up to 192 CS-2 AI systems into a cluster, while a cluster of 16 CS-2 AI systems can create a computing system with 13.6 million cores for natural language processing. The key to the new Cerebras Wafer-Scale Cluster is the exclusive use of data parallelism to train, which is the preferred approach for all AI work.{{Cite web |last=Freund |first=Karl |title=New Cerebras Wafer-Scale Cluster Eliminates Months Of Painstaking Work To Build Massive Intelligence |url=https://www.forbes.com/sites/karlfreund/2022/09/14/new-cerebras-wafer-scale-cluster-eliminates-months-of-painstaking-work-to-build-massive-intelligence/ |access-date=2022-09-15 |website=Forbes |language=en}}

In November 2022, Cerebras unveiled the supercomputer, Andromeda, which combines 16 WSE-2 chips into one cluster with 13.5 million AI-optimized cores, delivering up to 1 Exaflop of AI computing horsepower, or at least one quintillion (10 to the power of 18) operations per second.{{Cite web |author1=Paul Alcorn |date=2022-11-14 |title=Cerebras Reveals Andromeda, a 13.5 Million Core AI Supercomputer |url=https://www.tomshardware.com/news/cerebras-reveals-andromeda-a-135-million-core-ai-supercomputer |access-date=2022-11-18 |website=Tom's Hardware |language=en}}{{Cite news |last=Lee |first=Jane Lanhee |date=2022-11-14 |title=Silicon Valley chip startup Cerebras unveils AI supercomputer |language=en |work=Reuters |url=https://www.reuters.com/technology/silicon-valley-chip-startup-cerebras-unveils-ai-supercomputer-2022-11-14/ |access-date=2022-11-18}} The entire system consumes 500 kW, which was a drastically lower amount than somewhat-comparable GPU-accelerated supercomputers.

In November 2022, Cerebras announced its partnership with Cirrascale Cloud Services to provide a flat-rate "pay-per-model" compute time for its Cerebras AI Model Studio. The service is said to reduce the cost—compared to the similar cloud services on the market—by half while increasing speed up to eight times faster.{{Cite web |url=https://www.zdnet.com/article/ai-challenger-cerebras-unveils-pay-per-model-large-model-ai-cloud-service-with-cirrascale-jasper/ |title=AI challenger Cerebras unveils 'pay-per-model' AI cloud service with Cirrascale, Jasper |last=Ray |first=Tiernan |date=2022-11-29 |work=ZDNet}}

In July 2023, Cerebras and UAE-based G42 unveiled the world's largest network of nine interlinked supercomputers, Condor Galaxy, for AI model training. The first supercomputer, named Condor Galaxy 1 (CG-1), boasts 4 exaFLOPs of FP16 performance and 54 million cores.{{Cite web |last=Shilov |first=Anton |title=Cerebras to Enable 'Condor Galaxy' Network of AI Supercomputers: 36 ExaFLOPS for AI |url=https://www.anandtech.com/show/18969/cerebras-to-enable-a-network-of-ai-supercomputers-36-exaflops-for-ai |access-date=2024-11-01 |website=www.anandtech.com}} In November 2023, the Condor Galaxy 2 (CG-2) was announced, also containing 4 exaFLOPs and 54 million cores.

In March 2024, the companies broke ground on the Condor Galaxy 3 (CG-3), which can hit 8 exaFLOPs of performance and contains 58 million AI-optimized cores.{{Cite web |last=Takahashi |first=Dean |date=2024-03-13 |title=Cerebras breaks ground on Condor Galaxy 3, an AI supercomputer that can hit 8 exaFLOPs |url=https://venturebeat.com/ai/cerebras-breaks-ground-on-condor-galaxy-3-an-ai-supercomputer-that-can-hit-8-exaflops/ |access-date=2024-11-01 |website=VentureBeat |language=en-US}}

In March 2024, the company also introduced WSE-3, a 5 nm-based chip hosting 4 trillion transistors and 900,000 AI-optimized cores, the basis of the CS-3 computer. Cerebras also announced a collaboration with Dell Technologies, unveiled in June 2024, for AI compute infrastructure for generative AI.{{Cite web |last1=Kuka |first1=Valeriia |last2=Se |first2=Ksenia |date=June 15, 2024 |title=Cerebras: an Engineering Marvel to Rival NVIDIA |url=https://www.turingpost.com/p/cerebras |access-date=2024-06-15 |website=Turing Post |language=en}}

In August 2024, Cerebras unveiled its AI inference service, claiming to be the fastest in the world and, in many cases, ten to twenty times faster than systems built using the dominant technology, Nvidia's H100 "Hopper" graphics processing unit, or GPU.{{Cite web |title=AI startup Cerebras debuts 'world's fastest inference' service - with a twist |url=https://www.zdnet.com/article/ai-startup-cerebras-debuts-worlds-fastest-inference-with-a-twist/ |access-date=2024-11-01 |website=ZDNET |language=en}}

As of October 2024, Cerebras' performance advantage for inference is even larger when running the latest Llama 3.2 models. The jump in AI inference performance between August and October is a big one, at a factor of 3.5X, and it opens up the gap between Cerebras CS-3 systems running on premises or in clouds operated by Cerebras.{{Cite web |last=Morgan |first=Timothy Prickett |date=2024-10-25 |title=Cerebras Trains Llama Models To Leap Over GPUs |url=https://www.nextplatform.com/2024/10/25/cerebras-trains-llama-models-to-leap-over-gpus/ |access-date=2024-11-01 |website=The Next Platform |language=en-US}}

In March 2025, Cerebras announced six new datacenters across the United States and Europe, increasing the inference capacity twentyfold to over 40 million tokens per second.{{Cite web |last=Nuñez |first=Michael |date=2025-03-11 |title=Cerebras just announced 6 new AI datacenters that process 40M tokens per second — and it could be bad news for Nvidia |url=https://venturebeat.com/ai/cerebras-just-announced-6-new-ai-datacenters-that-process-40m-tokens-per-second-and-it-could-be-bad-news-for-nvidia/ |access-date=2025-06-17 |website=VentureBeat |language=en-US}}

In April 2025, Meta announced a partnership with Cerebras to power the new Llama API, offering developers inference speeds up to 18 times faster than with traditional GPU-based solutions.{{Cite web |last=Nuñez |first=Michael |date=2025-04-29 |title=Meta unleashes Llama API running 18x faster than OpenAI: Cerebras partnership delivers 2,600 tokens per second |url=https://venturebeat.com/ai/meta-unleashes-llama-api-running-18x-faster-than-openai-cerebras-partnership-delivers-2600-tokens-per-second/ |access-date=2025-06-16 |website=VentureBeat |language=en-US}} In May, Cerebras announced that it beats NVIDIA Blackwell in Llama 4 Inference with more than double the performance at more than 2,500 tokens per second/user on the 400B parameter Llama 4 Maverick model.{{Cite web |title=Cerebras beats Nvidia Blackwell in Llama 4 Maverick inference {{!}} Scientific Computing World |url=https://www.scientific-computing.com/article/cerebras-beats-nvidia-blackwell-llama-4-maverick-inference |access-date=2025-06-17 |website=www.scientific-computing.com}}

In May 2025, Cerebras unveiled Qwen3-32B, an open-weight LLM model built for smart, high speed reasoning and performance.{{Cite web |date=2025-05-19 |title=AI performance boost: Cerebras unveils Qwen3-32B model |url=https://siliconangle.com/2025/05/19/cerebras-qwen-ai-performance-reasoning-speed-cerebrassupernova/ |access-date=2025-06-17 |website=SiliconANGLE |language=en-US}}

Customers are reportedly using Cerebras technologies in the hyperscale, pharmaceutical, life sciences, and energy sectors, among others.{{Cite web|date=2020-10-13|title=LLNL, ANL and GSK Provide Early Glimpse into Cerebras AI System Performance|url=https://www.hpcwire.com/2020/10/13/llnl-anl-and-gsk-provide-early-glimpse-into-cerebras-ai-system-performance/|access-date=2021-06-03|website=HPCwire|language=en-US}}{{Cite web |date=2022-03-03 |title=Cerebras Systems Supplies 2nd-Gen AI System to TotalEnergies |url=https://www.enterpriseai.news/2022/03/03/cerebras-systems-supplies-2nd-gen-ai-system-to-totalenergies/ |access-date=2022-03-04 |website=EnterpriseAI |language=en-US}}

In 2020, GlaxoSmithKline (GSK) began using the Cerebras CS-1 AI system in their London AI hub, for neural network models to accelerate genetic and genomic research and reduce the time taken in drug discovery.{{Cite web|last=Ray|first=Tiernan|title=Glaxo's biology research with novel Cerebras machine shows hardware may change how AI is done |url= https://www.zdnet.com/article/glaxos-biology-research-with-novel-cerebras-machine-shows-hardware-may-change-how-ai-is-done/ |date= September 5, 2020 |access-date= August 13, 2021 |website= ZDNet|language=en}} The GSK research team was able to increase the complexity of the encoder models they could generate, while reducing training time.{{Cite web|title=Cerebras debuts new 2.6 trillion transistor wafer scale chip for AI|url=https://www.datacenterdynamics.com/en/news/cerebras-debuts-new-26-trillion-wafer-scale-chip-for-ai/|access-date=2021-06-17|website=www.datacenterdynamics.com|date=21 April 2021 |language=en}} Other pharmaceutical industry customers include AstraZeneca, who was able to reduce training time from two weeks on a cluster of GPUs to two days using the Cerebras CS-1 system.{{Cite web|last=Hansen|first=Lars Lynne|date=2021-04-26|title=Accelerating Drug Discovery Research with New AI Models: a look at the AstraZeneca Cerebras…|url=https://larslynnehansen.medium.com/accelerating-drug-discovery-research-with-new-ai-models-a-look-at-the-astrazeneca-cerebras-b72664d8783|access-date=2021-06-03|website=Medium|language=en}} GSK and Cerebras recently co-published [https://arxiv.org/pdf/2112.07571.pdf research] in December 2021 on epigenomic language models.

Argonne National Laboratory has been using the CS-1 since 2020 in COVID-19 research and cancer tumor research based on the world's largest cancer treatment database.{{Cite news|last=Shah|first=Agam|date=2020-05-06|title=National Lab Taps AI Machine With Massive Chip to Fight Coronavirus|language=en-US|work=Wall Street Journal|url=https://www.wsj.com/articles/national-lab-taps-ai-machine-with-massive-chip-to-fight-coronavirus-11588757403|access-date=2021-06-03|issn=0099-9660}} A series of models running on the CS-1 to predict cancer drug response to tumors achieved speed-ups of many hundreds of times on the CS-1 compared to their GPU baselines.

Cerebras and the National Energy Technology Laboratory (NETL) demonstrated record-breaking performance of Cerebras' CS-1 system on a scientific compute workload in November 2020. The CS-1 was 200 times faster than the Joule Supercomputer on the key workload of Computational Fluid Dynamics.{{Cite web |title=Cerebras Systems and NETL Set New Compute Milestone |url=https://www.hpcwire.com/off-the-wire/cerebras-systems-and-netl-set-new-compute-milestone/ |access-date=2022-03-04 |website=HPCwire |language=en-US}}

The Lawrence Livermore National Lab’s Lassen supercomputer incorporated the CS-1 in both classified and non-classified areas for physics simulations.{{Cite web|date=2020-08-19|title=Cerebras puts 'world's largest computer chip' in Lassen supercomputer|url=https://venturebeat.com/2020/08/19/cerebras-puts-worlds-largest-computer-chip-in-lassen-supercomputer/|access-date=2021-06-03|website=VentureBeat|language=en-US}} The Pittsburgh Supercomputing Center (PSC) has also incorporated the CS-1 in their Neocortex supercomputer for dual HPC and AI workloads.{{Cite web |last=Hemsoth |first=Nicole |date=2021-03-30 |title=Neocortex Supercomputer to Put Cerebras CS-1 to the Test |url=https://www.nextplatform.com/2021/03/30/neocortex-supercomputer-to-put-cerebras-cs-1-to-the-test/ |access-date=2022-03-04 |website=The Next Platform |language=en-US}} EPCC, the supercomputing center of the University of Edinburgh, has also deployed a CS-1 system for AI-based research.{{Cite web |last=Comment |first=Dan Swinhoe |title=EPCC chooses Cerebras' massive chip for new supercomputer |url=https://www.datacenterdynamics.com/en/news/epcc-chooses-cerebras-massive-chip-new-supercomputer/ |access-date=2022-03-04 |website=www.datacenterdynamics.com |date=5 February 2021 |language=en}}

In August 2021, Cerebras announced a partnership with [https://www.peptilogics.com Peptilogics] on the development of AI for peptide therapeutics.{{Cite web|title=Peptilogics and Cerebras Systems Partner on AI Solutions to Advance Peptide Therapeutics|url=https://www.hpcwire.com/off-the-wire/peptilogics-and-cerebras-systems-partner-on-ai-solutions-to-advance-peptide-therapeutics/|access-date=2021-09-22|website=HPCwire |language=en-US}}

In March 2022, Cerebras announced that the Company deployed its CS-2 system in the Houston facilities of TotalEnergies, its first publicly disclosed customer in the energy sector. Cerebras also announced that it has deployed a CS-2 system at [https://nference.com nference], a startup that uses natural language processing to analyze massive amounts of biomedical data. The CS-2 will be used to train transformer models that are designed to process information from piles of unstructured medical data to provide fresh insights to doctors and improve patient recovery and treatment.{{Cite web |title=Cerebras brings CS-2 system to data analysis biz nference |url=https://www.theregister.com/2022/03/14/cerebras_ai_chips/ |access-date=2022-03-15 |website=www.theregister.com |language=en}}

In May 2022, Cerebras announced that the National Center for Supercomputing Applications (NCSA) has deployed the Cerebras CS-2 system in their HOLL-I supercomputer.{{Cite web |title=NCSA Deploys Cerebras CS-2 in New HOLL-I Supercomputer for Large-Scale AI |url=https://www.hpcwire.com/off-the-wire/ncsa-deploys-cerebras-cs-2-in-new-holl-i-supercomputer-for-large-scale-ai/ |access-date=2022-06-03 |website=HPCwire |language=en-US}} They also announced that the Leibniz Supercomputing Centre (LRZ) in Germany plans to deploy a new supercomputer featuring the CS-2 system along with the HPE Superdome Flex server.{{Cite web |last=Comment |first=Sebastian Moss |title=Leibniz Supercomputing Centre to deploy HPE-Cerebras supercomputer |url=https://www.datacenterdynamics.com/en/news/leibniz-supercomputing-centre-to-deploy-hpe-cerebras-supercomputer/ |access-date=2022-06-03 |website=www.datacenterdynamics.com |date=26 May 2022 |language=en}} The new supercomputing system is expected to be delivered to LRZ this summer. This will be the first CS-2 system deployment in Europe.

In October 2022, it was announced that the U.S. National Nuclear Security Administration would sponsor a study to investigate using Cerebras' CS-2 in nuclear stockpile stewardship computing.{{Cite web |date=2022-10-18 |title=NNSA Taps 3 Federal Labs to Research Applications of Cerebras Systems Tech - ExecutiveBiz |url=https://blog.executivebiz.com/2022/10/nnsa-taps-3-federal-labs-to-research-applications-of-cerebras-systems-tech/ |access-date=2022-11-18 |website=blog.executivebiz.com |language=en-US}}{{Cite web |last=Mann |first=Tobias |title=DoE to trial Cerebras AI compute in nuclear weapon sims |url=https://www.theregister.com/2022/10/18/doe_cerebras_waferscale/ |access-date=2022-11-18 |website=www.theregister.com |language=en}} The multi-year contract will be executed through Sandia National Laboratories, Lawrence Livermore National Lab, and Los Alamos National Laboratory.

In November 2022, Cerebras and the National Energy Technology Laboratory (NETL) saw record-breaking performance on the scientific compute workload of forming and solving field equations. Cerebras demonstrated that its CS-2 system was as much as 470 times faster than NETL's Joule Supercomputer in field equation modeling.{{Cite web |title=Cerebras and National Energy Tech Lab Set New Milestones for High-Performance, Energy-Efficient Field Equation Modeling Using Simple Python Interface |url=https://www.hpcwire.com/off-the-wire/cerebras-and-national-energy-tech-lab-set-new-milestones-for-high-performance-energy-efficient-field-equation-modeling-using-simple-python-interface/ |access-date=2022-11-18 |website=HPCwire |language=en-US}}

The 2022 Gordon Bell Special Prize Winner for HPC-Based COVID-19 Research, which honors outstanding research achievement towards the understanding of the COVID-19 pandemic through the use of high-performance computing, used Cerebras' CS-2 system to conduct this award-winning research to transform large language models to analyze COVID-19 variants. The paper was authored by a 34-person team from Argonne National Laboratory, California Institute of Technology, Harvard University, Northern Illinois University, Technical University of Munich, University of Chicago, University of Illinois Chicago, Nvidia, and Cerebras. ANL noted that using the CS-2 Wafer-Scale Engine cluster, the team was able to achieve convergence when training on the full SARS-CoV-2 genomes in less than a day.{{Cite web |last=Peckham |first=Oliver |date=2022-11-17 |title=Gordon Bell Nominee Used LLMs, HPC, Cerebras CS-2 to Predict Covid Variants |url=https://www.hpcwire.com/2022/11/17/gordon-bell-nominee-used-llms-hpc-cerebras-cs-2-to-predict-covid-variants/ |access-date=2022-11-23 |website=HPCwire |language=en-US}}{{Cite web |last=Peckham |first=Oliver |date=2022-11-17 |title=Gordon Bell Special Prize Goes to LLM-Based Covid Variant Prediction |url=https://www.hpcwire.com/2022/11/17/gordon-bell-special-prize-goes-to-llm-based-covid-variant-prediction/ |access-date=2022-11-23 |website=HPCwire |language=en-US}}

Cerebras partnered with Emirati technology group G42 to deploy its AI supercomputers to create chatbots and to analyze genomic and preventive care data. In July 2023, G42 agreed to pay around $100 million to purchase the first of potentially nine supercomputers from Cerebras, each of which capable of 4 exaflops of compute.{{cite news |last1=Nellis |first1=Stephen |last2=Hu |first2=Krystal |date=20 July 2023 |title=Cerebras Systems signs $100 mln AI supercomputer deal with UAE's G42 |url=https://www.reuters.com/technology/cerebras-systems-signs-100-mln-ai-supercomputer-deal-with-uaes-g42-2023-07-20/ |publisher=Reuters}}{{cite news |last=Lu |first=Yiwen |date=20 July 2023 |title=An A.I. Supercomputer Whirs to Life, Powered by Giant Computer Chips |url=https://www.nytimes.com/2023/07/20/technology/an-ai-supercomputer-whirs-to-life-powered-by-giant-computer-chips.html |work=The New York Times}}{{cite web |last=Moore |first=Samuel K. |date=20 July 2023 |title=Cerebras Introduces Its 2-Exaflop AI Supercomputer |url=https://spectrum.ieee.org/ai-supercomputer-2662304872 |work=IEEE Spectrum}} In August 2023, Cerebras, the Mohamed bin Zayed University of Artificial Intelligence and G42 subsidiary Inception launched Jais, a large language model.{{Cite news |last=Cherney |first=Max A. |date=2023-08-30 |title=UAE's G42 launches open source Arabic language AI model |language=en |work=Reuters |url=https://www.reuters.com/technology/uaes-g42-launches-open-source-arabic-language-ai-model-2023-08-30/ |access-date=2023-10-08}}

Mayo Clinic announced a collaboration with Cerebras at the 2024 J.P. Morgan Healthcare Conference, offering details on the first foundation model it will develop with the enablement of Cerebras's generative AI computing capability. The solution will combine genomic data with de-identified data from patient records and medical evidence to explore the ability to predict a patient's response to treatments to manage disease and will initially be applied to rheumatoid arthritis. The model could serve as a prototype for similar solutions to support the diagnosis and treatment of other diseases.

At the January 2025 J.P. Morgan Healthcare Conference, Cerebras and Mayo Clinic announced a new genomic foundation model aimed at harnessing the power of advanced AI and HPC to transform genomics, a field rapidly becoming central to personalized healthcare. The new genomic foundation model is designed to improve diagnostics and personalize treatment selection, with an initial focus on Rheumatoid Arthritis.{{Cite web |last=Eadline |first=Doug |date=2025-01-16 |title=Cerebras and Mayo Clinic Unveil Advanced Genomic AI Model |url=https://www.hpcwire.com/2025/01/16/cerebras-and-mayo-clinic-unveil-advanced-genomic-ai-model/ |access-date=2025-01-31 |website=HPCwire |language=en-US}}

In May 2024, Cerebras in collaboration with researchers from Sandia National Laboratories, Lawrence Livermore National Laboratory, Los Alamos National Laboratory and the National Nuclear Security Administration, for molecular dynamics simulations in which the team simulated 800,000 atoms interacting with each other, calculating the interactions in increments of one femtosecond at a time. Each step took just microseconds to compute on the Cerebras WSE-2. Although that's still 9 orders of magnitude slower than the actual interactions, it was also 179 times as fast as the Frontier supercomputer. The achievement effectively reduced a year's worth of computation to just two days.{{Cite web |title=Giant Chips Give Supercomputers a Run for Their Money - IEEE Spectrum |url=https://spectrum.ieee.org/cerebras-wafer-scale-engine |access-date=2024-11-01 |website=spectrum.ieee.org |language=en}}

In March 2024, Cerebras introduced the CS-3 and third-generation Wafer Scale Engine (WSE-3), which represents the latest development of their technology. It has 2x the performance of CS-2 and hosts 900,000 cores. A CS-3 cluster is capable of training an AI model like Llama2-70B in just one single day.{{Cite web|url=https://cerebras.ai/blog/cerebras-cs3|title=Cerebras CS-3: the world's fastest and most scalable AI accelerator|first=James|last=Wang|date=March 12, 2024}} The WSE-3 was recognized by TIME Magazine as a Best Invention of 2024.{{Cite magazine |last=Booth |first=Harry |date=2024-10-30 |title=Cerebras Systems Wafer-Scale Engine 3: 200 Best Inventions of 2024 |url=https://time.com/7094929/cerebras-systems-wafer-scale-engine-3/ |access-date=2024-11-01 |magazine=TIME |language=en}}

In January 2025, Cerebras announced support for DeepSeek's R1 70B reasoning model, running on the latest Cerebras hardware in US-based datacenters. {{Cite web |last=published |first=Desire Athow |date=2025-01-30 |title=DeepSeek now runs on largest AI chip ever to deliver out-of-this world inference performance and I can't wait to try it. |url=https://www.techradar.com/pro/deepseek-on-steroids-cerebras-embraces-controversial-chinese-chatgpt-rival-and-promises-57x-faster-inference-speeds |access-date=2025-06-17 |website=TechRadar |language=en}}

In April 2025, Cerebras and Canadian photonics company Ranovus announced a contract from DARPA to reduce compute bottleneck challenges.{{Cite web |last=Comment |first=Charlotte Trueman |date=2025-04-02 |title=Cerebras and Ranovus receive $45m DARPA contract to “solve” compute bottleneck |url=https://www.datacenterdynamics.com/en/news/cerebras-and-ranovus-receive-45m-darpa-contract-to-solve-compute-bottleneck/ |access-date=2025-06-16 |website=www.datacenterdynamics.com |language=en}}

Cerebras AI Inference services claims to be the fastest in the world and, in many cases, ten to twenty times faster than systems built using the dominant technology, Nvidia's H100 "Hopper" graphics processing unit, or GPU.

In January 2025, Cerebras announced that it would support DeepSeek's R1 70B reasoning model at 1,600 tokens/second, which the company claims is 57x faster than any R1 provider using GPUs.{{Cite web |author1=Desire Athow |date=2025-01-30 |title=DeepSeek now runs on largest AI chip ever to deliver out-of-this world inference performance and I can't wait to try it. |url=https://www.techradar.com/pro/deepseek-on-steroids-cerebras-embraces-controversial-chinese-chatgpt-rival-and-promises-57x-faster-inference-speeds |access-date=2025-01-31 |website=TechRadar |language=en}}

In February 2025, Cerebras and Mistral AI announced a partnership and helped the French AI player achieve a speed record. Mistral released an app called Le Chat that it said can respond to user questions with 1,000 words per second. Cerebras said it is providing the computer power behind those results.{{Cite web |last=Nellis |first=Stephen |date=6 Feb 2025 |title=AI chip firm Cerebras partners with France's Mistral, claims speed record |url=https://www.reuters.com/technology/artificial-intelligence/ai-chip-firm-cerebras-partners-with-frances-mistral-claims-speed-record-2025-02-07/ |archive-date= |access-date= |website=Reuters}}

Also in February 2025, Cerebras announced a partnership with Perplexity AI that promises to deliver near-instantaneous AI-powered search results at speeds previously thought impossible. The collaboration centers on Perplexity's new Sonar model which runs on Cerebras chips at 1,200 tokens per second making it one of the fastest AI search systems available. Cerebras appears to be leveraging this momentum to establish itself as the go-to provider for high-speed AI inference.{{Cite web |last=Nuñez |first=Michael |date=2025-02-11 |title=Cerebras-Perplexity deal targets $100B search market with ultra-fast AI |url=https://venturebeat.com/ai/cerebras-perplexity-deal-targets-100b-search-market-with-ultra-fast-ai/ |access-date=2025-02-12 |website=VentureBeat |language=en-US}}

• Wafer-scale integration
• Wafer-level packaging
• Semiconductor device fabrication
• Transistor count
• Deep learning processor

{{reflist}}

• {{Official website|https://cerebras.net/}}
• [https://www.servethehome.com/cerebras-wafer-scale-engine-wse-2-and-cs-2-at-hot-chips-34/ Cerebras' presentation at Hot Chips 34 (2022)]

Category:Computer companies of the United States

Category:Companies based in California

Category:Companies based in Sunnyvale, California

Category:Companies based in Silicon Valley

Category:Computer hardware companies

Category:Semiconductor companies of the United States

Category:Fabless semiconductor companies

Category:Computer companies established in 2016

Category:2016 establishments in California