VAST Data Showcases an AI Hall of Fame at SC24
What do you get when you combine football turf, Chick-fil-A, and a bunch of supercomputing operators? Today in Atlanta at the College Football Hall of Fame, that magical combination represented VAST Data’s latest update on its quest to become the operating system for AI with its VAST Data Platform. VAST Data knows the SC crowd well, having heritage in delivery of storage and data solutions for the HPC arena. Their advancement with the VAST Insight Engine and other solutions written about extensively on the TechArena represent a leap forward for VAST, delivering foundational tools to fuel AI training.
Today, VAST CEO Renen Hallak described this transition, highlighting that we’re on the cusp of our next inflection, moving from training to fine tuning and inference. VAST solutions, now integrated with RAG pipelines and vector databases, provide the tools for enterprises to take this bold leap forward with the confidence to keep data secure while delivering the performance offered by the world’s largest models.
This vision has been celebrated by the market with VAST deployments growing among the largest AI training providers and garnering a valuation of over $9 billion dollars based on momentum of its series E funding round. Today’s event was a terrific look into VAST’s traction in the market with leading AI players including XAI, CoreWeave, and Harvard Medical School.
The first up on deck was X-AI, Elon Musk’s AI arm, that just delivered a 100,000 GPU scale cluster called Colossus, in a mere 122 days. Why move so fast? X has notably been less engaged in AI training vs. other cloud behemoths, and Colossus places the firm in a position to accelerate their work quickly with the help of NVIDIA, Supermicro, and yes, VAST Data. The data platform was selected to drive foundational data management for this supercharged work with VAST engineers working closely with X-AI counterparts to get this Memphis based cluster up and running.
One customer that captured attention for their rapid buildout of AI services at scale was CoreWeave. CoreWeave’s Chief Product Officer, Chetan Kapoor, described his company’s mission to drive an accelerated compute platform, custom optimized for AI. Chetan stated “in our environment, you’re not going to find six hundred compute options, we’re specifically designed for AI.” The company has delivered with CoreWeave committing to H100 GPUs a year and a half ago and scaling to levels of performance focused on 100’s of megawatts of accelerated infrastructure and growing rapidly. This places CoreWeave on the path to potentially emerge as on par with the hyperscalers with continued market success.
CoreWeave has been built from the ground up with the VAST data platform. CoreWeave takes this combination of NVIDIA compute and VAST data management, and engages selective customers that are well prepared for AI adoption at scale. They’re targeting the 10,000 GPU scale for their customer base, a statement that highlights the relative demand for GPU cycles in market.
Today’s event wasn’t just about hyperscale. VAST also brought up Spencer Pruitt, CIO of Harvard Medical School, to highlight how scientific discovery is also being propelled by this AI driven innovation. Spencer shared how Harvard is driving genomics research to tackle challenges in disease management like cancer and personalized medicine with use of high-octane technology.
What’s the TechArena take? Data is the gold of this gilded age, and as long as companies are chasing insight from AI, the VAST data platform will continue to earn bold valuations in the marketplace. While much of this infrastructure buildout represents an AI arms race with repercussions beyond the scope of this article, the world will benefit from the accelerated innovation hyperscalers are driving as evidenced by what Harvard Medical School has driven with their data transformation. Color us continually impressed with VAST Data’s ability to execute within this high-pressure environment and waiting to see more innovation from this company in 2025.
What do you get when you combine football turf, Chick-fil-A, and a bunch of supercomputing operators? Today in Atlanta at the College Football Hall of Fame, that magical combination represented VAST Data’s latest update on its quest to become the operating system for AI with its VAST Data Platform. VAST Data knows the SC crowd well, having heritage in delivery of storage and data solutions for the HPC arena. Their advancement with the VAST Insight Engine and other solutions written about extensively on the TechArena represent a leap forward for VAST, delivering foundational tools to fuel AI training.
Today, VAST CEO Renen Hallak described this transition, highlighting that we’re on the cusp of our next inflection, moving from training to fine tuning and inference. VAST solutions, now integrated with RAG pipelines and vector databases, provide the tools for enterprises to take this bold leap forward with the confidence to keep data secure while delivering the performance offered by the world’s largest models.
This vision has been celebrated by the market with VAST deployments growing among the largest AI training providers and garnering a valuation of over $9 billion dollars based on momentum of its series E funding round. Today’s event was a terrific look into VAST’s traction in the market with leading AI players including XAI, CoreWeave, and Harvard Medical School.
The first up on deck was X-AI, Elon Musk’s AI arm, that just delivered a 100,000 GPU scale cluster called Colossus, in a mere 122 days. Why move so fast? X has notably been less engaged in AI training vs. other cloud behemoths, and Colossus places the firm in a position to accelerate their work quickly with the help of NVIDIA, Supermicro, and yes, VAST Data. The data platform was selected to drive foundational data management for this supercharged work with VAST engineers working closely with X-AI counterparts to get this Memphis based cluster up and running.
One customer that captured attention for their rapid buildout of AI services at scale was CoreWeave. CoreWeave’s Chief Product Officer, Chetan Kapoor, described his company’s mission to drive an accelerated compute platform, custom optimized for AI. Chetan stated “in our environment, you’re not going to find six hundred compute options, we’re specifically designed for AI.” The company has delivered with CoreWeave committing to H100 GPUs a year and a half ago and scaling to levels of performance focused on 100’s of megawatts of accelerated infrastructure and growing rapidly. This places CoreWeave on the path to potentially emerge as on par with the hyperscalers with continued market success.
CoreWeave has been built from the ground up with the VAST data platform. CoreWeave takes this combination of NVIDIA compute and VAST data management, and engages selective customers that are well prepared for AI adoption at scale. They’re targeting the 10,000 GPU scale for their customer base, a statement that highlights the relative demand for GPU cycles in market.
Today’s event wasn’t just about hyperscale. VAST also brought up Spencer Pruitt, CIO of Harvard Medical School, to highlight how scientific discovery is also being propelled by this AI driven innovation. Spencer shared how Harvard is driving genomics research to tackle challenges in disease management like cancer and personalized medicine with use of high-octane technology.
What’s the TechArena take? Data is the gold of this gilded age, and as long as companies are chasing insight from AI, the VAST data platform will continue to earn bold valuations in the marketplace. While much of this infrastructure buildout represents an AI arms race with repercussions beyond the scope of this article, the world will benefit from the accelerated innovation hyperscalers are driving as evidenced by what Harvard Medical School has driven with their data transformation. Color us continually impressed with VAST Data’s ability to execute within this high-pressure environment and waiting to see more innovation from this company in 2025.
