Under the terms of the partnership the two companies will build a global, Nvidia-powered accelerated computing cloud for deploying, managing, and securing hundreds of petabytes of data for generative AI, high performance computing (HPC) and visual effects (VFX) workloads.
CoreWeave recently secured a $2.3billion funding round and is seeing huge demand for its graphics processing unit (GPU’s) powered infrastructure, that can support the compute intensive workloads required for generative AI and large language models (LLMs).
“CoreWeave’s customers demand the most secure and scalable solutions on top of the industry’s fastest and most flexible infrastructure to keep their data safe,” said Michael Intrator, CEO and co-founder of CoreWeave.
“This partnership is rooted in a deep technical collaboration that will push the boundaries of data-driven accelerated computing to deliver the world’s most optimised AI cloud platform.”
Intrator said CoreWeave were “delighted” to partner with Vast Data to deliver a multi-tenant and zero-trust environment purpose-built for accelerated compute use cases, claiming that the platform would be 35x faster and 80% less expensive than legacy cloud providers.
Although Intrator is delighted now, that wasn’t always CoreWeave’s approach towards Vast.
“When we first spoke to Vast Data in 2019, we flat out told them no. We were wrong” said CoreWeave CTO and co-founder Brian Venturo in a promo video for the partnership reminiscent of a Netflix trailer.
Vast Data say its platform unifies storage, database and virtualised compute engine services.
With the goal of enabling the next step beyond generative AI and LLMs to AI-driven discovery, the platform serves as a one stop shop for disparate public and private cloud services users might require.
“Since our earliest days, Vast Data has had a single vision of building an architecture that could power the needs of the most demanding cloud-scale AI applications,” said Renen Hallak, founder and CEO of Vast Data.
“We could not imagine a better cloud platform to realise this vision than what we’re creating with CoreWeave. We are humbled and honoured to partner with the CoreWeave team to push the boundaries of modern AI computing and to build the infrastructure that will serve as the foundation of tomorrow’s AI-powered discoveries.”
CoreWeave has already announced plans to build out a 14 data centre cloud, with the investment in their most recent Texas facility valued at $1.6B.
The scale of the projects that CoreWeave and now Vast Data will be involved in were summarised in a blog post by Hallak’s co-founder at Vast, Jeff Denworth.
Likening CoreWeave’s to a modern day Fort Knox, Denworth said “Generative AI is expected to create $2-4 trillion dollars of market value to the global economy, per McKinsey. Much, if not all, of this value will be powered by GPUs. If LLMs are the new currency that intelligent businesses trade in, GPUs are the bullion that stands them up.”
Denworth also references the case study of Inflection AI, a $4billion startup with founded by Linkedin co-founder Reid Hoffman and former head of Google AI research lab Deep Mind, Mustafa Suleyman
Inflection recently announced CoreWeave as their cloud to deploy 22,000 H100 Tensor Core GPUs, 3.8x more GPU’s than Nvidia’s Eos, which Nvidia claimed to be the world’s fastest AI supercomputer.
Denworth says the Inflexion project has 2x the AI performance of ORNL’s Frontier system, regarded to be the world’s fastest computer, period.
Finally, “at $25K per GPU (a Raymond James estimate), this system costs more than the total annual Gross Domestic Product of Micronesia or Tonga”.
Considering the scale and ambition of the projects CoreWeave will engage in, Denworth said Vast is the preferred platform of choice for four key reasons: efficiency, uptime, Zero-trust and the level of depth and capability in the Vast Data team.
