Company: Cerebras Systems
Company Description: Cerebras Systems is a team of computer architects, scientists, deep learning researchers and engineers who have accelerated generative AI by building a new class of computer.
Nomination Category: Company / Organization Categories
Nomination Sub Category: Most Innovative Tech Company of the Year - Up to 2,500 Employees
 Nomination Title: Cerebras Systems
- Which will you submit for your nomination in this category, a video of up to five (5) minutes in length about the achievements of the nominated organization since January 1 2022, OR written answers to the questions for this category? (Choose one):
Written answers to the questions
- If you are submitting a video of up to five (5) minutes in length, provide the URL of the nominated video here, OR attach it to your entry via the "Add Attachments, Videos, or Links to This Entry" link above, through which you may also upload a copy of your video.
N/A
- If you are providing written answers for your submission, you must provide an answer to this first question: Briefly describe the nominated organization: its history and past performance (up to 200 words):
Total 200 words used.
Cerebras Systems is a team of computer architects, scientists, deep learning researchers and engineers who have accelerated generative AI by building a new class of computer. We enable enterprises, government, and academic researchers to speed up time to AI insights, powering the next wave of scientific breakthroughs in drug discovery, computational fluid dynamics, cancer research, energy research, astronomy, physics, material science, medicine, and more.
Cerebras’ customers run massive AI models and need a faster, more capable system to speed time to insights. The Cerebras team addresses this need with their CS-3 AI system, powered by the third-generation Wafer Scale Engine (WSE-3), the world’s largest and fastest AI processor, which simplifies training large models by avoiding the complexity of distributed computing. Cerebras CS-3s are clustered together to make the largest AI supercomputers in the world, including Condor Galaxy.
Cerebras solutions are available through the cloud and on premise. Cerebras AI Model Studio cloud access offers customers a flat rate “pay-per-model” to reduce the cost of running AI models while increasing speed up to 8x faster.
Cerebras has 77 paying customers such as Mayo Clinic, AbbVie, G42, TotalEnergies, Bayer, GlaxoSmithKline, Jasper, AstraZeneca, Argonne National Laboratory. Cerebras is valued at over $4 billion.
- If you are providing written answers for your submission, you must provide an answer to this second question: Outline the organization's achievements since the beginning of 2022 that you wish to bring to the judges' attention (up to 250 words):
Total 250 words used.
Confidential: On March 13, 2024, Cerebras will announce the latest version of our AI system, the CS-3, powered by the third-generation Wafer-Scale Engine (WSE-3). Built at the 5nm node, WSE-3 will more than double the performance of WSE-2. WSE-3 will be the world’s fastest AI processor – an order of magnitude faster and more capable than the latest Nvidia GPU.
In addition to building the world’s most powerful AI computers, Cerebras has trained and contributed to top open-source large language models (LLMs). In March 2023, Cerebras released Cerebras-GPT, the first family of open-source models as an alternative to ChatGPT. In July, Cerebras released BTLM, a state-of-the-art 3B parameter model, downloaded over one million times.
Last summer, Cerebras announced a partnership with UAE-based G42 and unveiled Condor Galaxy 1, the first of nine interconnected AI supercomputers totaling 36 exaFLOPs of AI compute. Each of the supercomputers is among the largest in the world at 4 exaFLOPs of AI compute and containing 64 connected Cerebras CS-3 systems. The full constellation of nine Condor Galaxy AI supercomputers, offered via Supercompute-as-a-service, will provide enough power to fundamentally impact the worldwide inventory of AI compute.
In January 2024, Cerebras and Mayo Clinic announced a partnership on generative AI for development of LLMs for medical applications. The collaboration is producing pioneering LLMs to improve patient outcomes and diagnoses leveraging Mayo Clinic’s robust longitudinal data repository and Cerebras’ generative AI compute. Mayo Clinic selected Cerebras for its experience designing and training large scale, domain-specific generative AI models.
- If you are providing written answers for your submission, you must provide an answer to this third question: Explain why the achievements you have highlighted are unique or significant. If possible compare the achievements to the performance of other players in your industry and/or to the organization's past performance (up to 250 words):
Total 240 words used.
Cerebras is addressing some of the biggest issues facing AI today – across AI compute, datasets and model development – by building and deploying AI accelerators that are orders of magnitude faster, use a fraction of the power and lead to AI breakthroughs benefitting society. The CS-3, powered by the WSE-3, will more than double the performance of WSE-2 and will be the world’s fastest AI processor – an order of magnitude faster and more capable than the latest Nvidia GPU.
In November 2023, Cerebras announced a 130X performance improvement over Nvidia’s A100 GPUs on a key nuclear energy HPC simulation kernel developed by researchers at Argonne National Laboratory, showcasing the significant benefit of using Cerebras’ dedicated AI supercomputers over traditional GPUs.
In September, Cerebras’ work with King Abdullah University of Science and Technology in Saudi Arabia on multi-dimensional seismic processing was selected as a finalist for the prestigious Gordon Bell Prize.
According to Cerebras customer GlaxoSmithKline (GSK), who generates extremely large datasets in its genomic and genetic research, “the Cerebras CS-2 is a critical component that allows GSK to train language models using biological datasets at a scale and size previously unattainable. These foundational models form the basis of many of our AI systems and play a vital role in the discovery of transformational medicines.”
Cerebras continues to innovate with newer and faster versions of our AI supercomputers, as well as to develop new partnerships, models and research.
- You have the option to answer this final question: Reference any attachments of supporting materials throughout this nomination and how they provide evidence of the claims you have made in this nomination (up to 250 words):
Total 178 words used.
| Attachments/Videos/Links: |
|---|
|
