Cerebras Systems
| Private | |
| ISIN | 🆔 |
| Industry | Semiconductor industry Artificial Intelligence |
| Founded 📆 | 2016 |
| Founder 👔 | |
| Headquarters 🏙️ | , , USA |
Area served 🗺️ | |
Key people | |
| Products 📟 | |
| Owner | Privately funded |
| Members | |
Number of employees | |
| 🌐 Website | www |
| 📇 Address | |
| 📞 telephone | |
Cerebras Systems is an American company with offices in Silicon Valley, San Diego, Toronto, and Tokyo.[1] Cerebras is building a new class of computer system for complex artificial intelligence and deep learning applications.[2]
History
Cerebras was founded in 2015 by Andrew Feldman, Gary Lauterbach, Michael James, Sean Lie and Jean-Philippe Fricker.[3] 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.[4][5]
In May of 2016, Cerebras secured its first official funding round for $27 million led by Benchmark Foundation Capital and Eclipse Ventures.[6][3]
By December 2016, Cerebras secured series B funding, led by Coatue, which was followed in January 2017 with series C funding led by VY Capital, for a total raise of $112 million.[3]
In November 2018, Cerebras closed its series D with $88 million, making the company a unicorn. Investors included Altimeter, VY Capital, Coatue, Foundation Capital, Benchmark and Eclipse.[7][8]
On August 19, 2019, Cerebras announced its Wafer-Scale Engine (WSE).[9][10][11]
In November 2019, Cerebras closed its series E with over $270 million for a valuation of $2.4 billion. Investors included Altimeter, Coatue, Eclipse, FalconEdge and Moore Capital.[12]
Core Products and Technology
The Cerebras Wafer Scale Engine (WSE) is a single, wafer-scale integrated processor that includes compute, memory and interconnect fabric.[13] [14][8]A single algorithm can use all the cores and memory. The aggregate memory and interconnect bandwidth are achieved at low latency, resulting in high performance.[13]
The WSE, Cerebras’ first-generation processor, has 1.2 trillion transistors, 400,000 compute cores and 18 gigabytes of memory. As of August 2019, it was the largest chip ever made, only to be surpassed by the Cerebras WSE-2 announced in April 2021.[9][10][11] The Cerebras WSE-2 more than doubles the core count of the first-generation processor, with 850,000 AI optimized cores and 2.6 trillion transistors.[15] Built on TSMC’s 7nm node, the Cerebras WSE-2 also expands on-chip SRAM to 40 gigabytes, memory bandwidth to 20 petabytes per second and total fabric bandwidth to 220 petabits per second. Each WSE is cut from a single 300mm wafer and is fabricated by TSMC.[16][17]
The WSE powers the Cerebras CS-1, which is Cerebras’ first-generation AI computer. It is a rack-mounted appliance designed for AI training and inference workloads in the datacenter. The CS-1 includes a single WSE primary processor with 400,000 processing cores, 18 Gigabytes of on-chip memory, 9 Petabytes per second of on-die memory bandwidth; the system has 1.2 Terabits per second I/O via 12 - 100 Gigabit Ethernet connections to move data in and out of the CS-1 system, and uses 20 kilowatts of power.[18][10]
In April 2021, Cerebras announced the Cerebras CS-2. Powered by the 7-nanometer WSE-2, the CS-2 is Cerebras’s second generation AI computing solution. It is 26 inches tall and fits in one-third of a standard data center rack.[15][1]
Early Customers and Artificial Intelligence Deployments
Cerebras products are built to accelerate deep learning in the data center.[2] Customers are using Cerebras solutions in the pharmaceutical and life sciences sectors, in supercomputing and in government applications.[19]
In 2020, GlaxoSmithKline began using the CS-1 in their new London AI hub, using new kinds of neural network models to accelerate genetic and genomic research and reduce the time taken in drug discovery.[20] The GSK research team was able to increase the complexity of the encoder models they could generate, while reducing training time by 80x over their prior system.[21] Other pharmaceutical industry customers include AstraZeneca, who was able to reduce training time on a large cluster of GPUs from two weeks to two days using the Cerebras CS-1.[22]
Tokyo Electron Devices is a customer and partner to Cerebras as the company increases its global reach by offering local sales and support in the Japan region.[23]
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. They have been able to accomplish in a few months what would previously have taken years to do. A series of models running on the CS-1 to predict tumor response to drugs achieved speed-ups of many hundreds of times on the CS-1 compared to their GPU baselines. Argonne also noted large speed-ups in generative molecular design for COVID-19 research.[24][19]
The Lawrence Livermore National Lab’s Lassen supercomputer has incorporated the CS-1 in both classified and non-classified areas for physics-based HPC simulations.[25]
Recognition
2020 IEEE Spectrum’s Emerging Technology Awards[26]
2020 HPCWire’s Readers’ and Editors’ Choice Awards[27]
2020 Global Semiconductor Alliance “Startup to Watch”[28]
2019 CBInsights AI 100[29]
2021 Fast Company’s Most Innovative Companies[30]
2020 Fast Company’s Best World Changing Ideas Awards[31]
2020 Forbes AI 50[32]
See Also
References
- ↑ 1.0 1.1 "Cerebras launches new AI supercomputing processor with 2.6 trillion transistors". VentureBeat. 2021-04-20. Retrieved 2021-04-30.
- ↑ 2.0 2.1 "Cerebras Systems deploys the 'world's fastest AI computer' at Argonne National Lab". VentureBeat. 2019-11-19. Retrieved 2021-04-30.
- ↑ 3.0 3.1 3.2 Tilley, Aaron. "AI Chip Boom: This Stealthy AI Hardware Startup Is Worth Almost A Billion". Forbes. Retrieved 2021-04-30.
- ↑ Hardy, Quentin (2012-02-29). "A.M.D. Buying SeaMicro for $334 Million". Bits Blog. Retrieved 2021-04-30.
- ↑ "How Google Spawned The 384-Chip Server". Wired. ISSN 1059-1028. Retrieved 2021-04-30.
- ↑ "A stealthy startup called Cerebras raised around $25 million to build deep learning hardware". TechCrunch. Retrieved 2021-04-30.[permanent dead link]
- ↑ Martin, Dylan (2019-11-27). "AI Chip Startup Cerebras Reveals 'World's Fastest AI Supercomputer'". CRN. Retrieved 2021-04-30.
- ↑ 8.0 8.1 Strategy, Moor Insights and. "Cerebras Unveils AI Supercomputer-On-A-Chip". Forbes. Retrieved 2021-04-30.
- ↑ 9.0 9.1 Metz, Cade (2019-08-19). "To Power A.I., Start-Up Creates a Giant Computer Chip". The New York Times. ISSN 0362-4331. Retrieved 2021-04-30.
- ↑ 10.0 10.1 10.2 "The Cerebras CS-1 computes deep learning AI problems by being bigger, bigger, and bigger than any other chip". TechCrunch. Retrieved 2021-04-30.[permanent dead link]
- ↑ 11.0 11.1 "Full Page Reload". IEEE Spectrum: Technology, Engineering, and Science News. Retrieved 2021-04-30.
- ↑ "SharesPost". app.sharespost.com. Retrieved 2021-04-30.
- ↑ 13.0 13.1 "Full Page Reload". IEEE Spectrum: Technology, Engineering, and Science News. Retrieved 2021-04-30.
- ↑ "Semiconductor Startup Shows Off the World's Biggest Processor". BloombergQuint. Retrieved 2021-04-30.
- ↑ 15.0 15.1 Ray, Tiernan. "Cerebras continues 'absolute domination' of high-end compute, it says, with world's hugest chip two-dot-oh". ZDNet. Retrieved 2021-04-30.
- ↑ "Cerebras Systems Smashes the 2.5 Trillion Transistor Mark with New Second Generation Wafer Scale Engine". Bloomberg.com. 2021-04-20. Retrieved 2021-04-30.
- ↑ Cutress, Dr Ian. "Cerebras Unveils Wafer Scale Engine Two (WSE2): 2.6 Trillion Transistors, 100% Yield". www.anandtech.com. Archived from the original on 2021-06-26. Retrieved 2021-04-30.
- ↑ "Neocortex Will Be First-of-Its-Kind 800,000-Core AI Supercomputer". HPCwire. 2020-06-09. Retrieved 2021-04-30.
- ↑ 19.0 19.1 "LLNL, ANL and GSK Provide Early Glimpse into Cerebras AI System Performance". HPCwire. 2020-10-13. Retrieved 2021-04-30.
- ↑ Ray, Tiernan. "Glaxo's biology research with novel Cerebras machine shows hardware may change how AI is done". ZDNet. Retrieved 2021-04-30.
- ↑ "Cerebras debuts new 2.6 trillion transistor wafer scale chip for AI". www.datacenterdynamics.com. Retrieved 2021-04-30.
- ↑ Hansen, Lars Lynne (2021-04-26). "Accelerating Drug Discovery Research with New AI Models: a look at the AstraZeneca Cerebras…". Medium. Retrieved 2021-04-30.
- ↑ "Cerebras Systems Expands Global Footprint with New Offices in Tokyo, Japan, and Toronto, Canada". www.yahoo.com. Retrieved 2021-04-30.
- ↑ Shah, Agam (2020-05-06). "National Lab Taps AI Machine With Massive Chip to Fight Coronavirus". Wall Street Journal. ISSN 0099-9660. Retrieved 2021-04-30.
- ↑ "Cerebras puts 'world's largest computer chip' in Lassen supercomputer". VentureBeat. 2020-08-19. Retrieved 2021-04-30.
- ↑ "Honors 2020: Cerebras Systems Wins the IEEE Spectrum Emerging Technology Award". IEEETV. Retrieved 2021-04-30.
- ↑ "2020 HPCwire Awards - Readers' & Editors' Choice". HPCwire. Retrieved 2021-04-30.
- ↑ "Cerebras Systems Wins Global Semiconductor Alliance 2020 "Start-Up to Watch" Award". Bloomberg.com. 2020-12-04. Retrieved 2021-04-30.
- ↑ "The Top 100 AI Startups Of 2019: Where Are They Now?". CB Insights Research. 2019-12-10. Retrieved 2021-04-30.
- ↑ Staff, Fast Company (2021-03-09). "The 10 most innovative companies in artificial intelligence". Fast Company. Retrieved 2021-04-30.
- ↑ Clendaniel, Morgan (2020-04-28). "World Changing Ideas Awards 2020: North America Finalists and Honorable Mentions". Fast Company. Retrieved 2021-04-30.
- ↑ Ohnsman, Alan. "AI 50: America's Most Promising Artificial Intelligence Companies". Forbes. Retrieved 2021-04-30.
External Links
Category:Computer companies of the United States Category:Companies based in California Category:Electronics companies established in 2016 Category:Electronics companies of the United States
This article "Cerebras" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Cerebras. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.
|
This page exists already on Wikipedia. |
