Hear from CIOs, CTOs, and other C-level and senior execs on data and AI strategies at the Future of Work Summit this January 12, 2022. Learn more
The announcement comes after Argonne scientists began using GroqChip, a Tensor Streaming Processor, and GroqWare, a software platform, to launch AI initiatives at the start of 2020.
Throughout 2020, Argonne scientists used GroqChip and GroqWare to power AI initiatives and created machine learning models of the SARS-CoV-2 virus. The scientists used the models to screen a database of billions of candidate drug molecules and identify high-potential lead components for clinical therapy trials.
While the news demonstrates how AI can accelerate the development of medicine and clinical treatments, it also highlights that high-performance computing solutions like GroqChip are necessary to provide faster insights to human users.
High-performance AI chip
GroqChip, the AI chip used by Argonne National Laboratory to power the COVID-19 AI initiatives, is a high-performance semiconductor with 750TOPS @900 MHz and 80 terabytes/sec of memory bandwidth.
The GroqChip also uses TSP architecture to enable it to store more models within its memory to process data with high-computing performance, this makes it ideal for making inferences from large datasets.
“Using the Groq platform, we accelerated our efforts to identify promising COVID-19 drug candidates,” said Tom Brettin, Argonne computational scientist. “The system’s AI capabilities enabled us to achieve significantly more inferences a second, reducing the time needed for each search from days to minutes.”
GroqChip provided the processing power for the AI to analyze mountainous datasets of candidate drug molecules in a way that wouldn’t have been possible with a lower-performance solution.
As Venkat Vishwanath, lead of the Argonne Leadership Computing Facility’s data science team, said, “Data processing and compute performance are key to achieving high performance for inference tasks for scientific applications. Groq technology provides an optimized end-to-end system needed to glean faster insights.”
Vying for the AI-semiconductor top spot
Groq was first launched in 2016 by CEO Jonathan Ross, who left Google to launch a new semiconductor organization. However, it took until 2020 for it to begin making its solutions commercially available.
Earlier this year, the organization announced it had closed its Series C fundraising with $300 million in funding and achieved a valuation of over $1 billion.
Since Groq’s formation, the organization has focused on developing the high-performance and responsive GroqChip, which the organization claims offers speeds of 1000 TOPS, 17 times faster than any competitor’s product.
Despite these claims, Groq has a way to go to unseat NVIDIA, the dominant provider in the AI chip market and a well-known provider of GPU-based chips. In 2020, Nvidia held an 80.6% share of global revenue in artificial intelligence processors used in the cloud and in data centers and had its GPUs used within nearly 70% of the top 500 supercomputers.
However, if Groq continues to achieve more public victories by enabling Argonne National Laboratory to identify potential treatments for COVID-19, then it has a strong chance to steal market shares away from other established competitors.
VentureBeatVentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact. Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:
- up-to-date information on the subjects of interest to you
- our newsletters
- gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
- networking features, and more