The US Department of Energy's (DOE) Argonne National Laboratory has deployed hardware from AI-focused chip provider Groq.
Groq this week announced its hardware is now available to researchers through the Argonne Leadership Computing Facility's (ALCF) AI Testbed in Argonne, Illinois.
The company said the deployed GroqRack will allow researchers to work on topics such as fusion energy, material design, imaging sciences, and drug discovery.
"Standing up a GroqRack at Argonne National Laboratory reflects both the hard work of the Groq team and our strong partnership with Argonne, which will provide researchers the flexibility they need to innovate and push the boundaries of what's possible with the ALCF AI Testbed," said Jonathan Ross, CEO and founder of Groq.
The ALCF is a DOE user facility providing high-performance computing (HPC) resources and AI technologies for open science.
Michael E. Papka, ALCF director, and deputy associate lab director for Argonne's computing, environment, and life sciences directorate, said: "Inference is a critical part of leveraging AI for science as it allows researchers to use trained machine learning models to make predictions or discover patterns in complex data. Incorporating Groq inference-based solutions into the ALCF AI Testbed further strengthens our portfolio of AI accelerator offerings to our research community."
The ALCF testbed webpage notes the GroqRack consists of nine nodes with eight accelerators per node, offering a total of 188 teraflops at FP16 and 750 teraflops at INT8.
The testbed also includes hardware from Graphcore, Cerebras, and SambaNova, as well as Intel’s Habana Gaudi tensor processor.
Groq was co-founded in 2016 by Ross, who previously helped lead Google's Tensor Processing Unit development. The company provides dedicated chips and accelerators within its own servers and racks designed for artificial intelligence, machine learning, and high-performance computing.
Aileen Black, president of Groq government & sales public sector, said: "We're thrilled to build upon our successful partnership with Argonne and collaborate with other DOE labs to advance LLM innovation. Our drug discovery work was just the beginning–we're committed to enabling energy-efficient and scalable AI solutions for a sustainable computing future."
