Voltage Park To Deploy $500M Worth Of Nvidia H100s To Milk That AI Hype

AI infrastructure provider Voltage Park revealed Sunday it has acquired 24,000 Nvidia H100 accelerators, which it plans to begin leasing to enterprises, startups and research institutions early next year.

According to Reuters, Voltage Park received $500 million in funding to purchase the chips from a non-profit backed by billionaire Jed McCaleb, who made his fortune in crypto market and has since moved on to building space stations at Vast.

"Not enough people understand how much the compute shortage is affecting AI innovators," CEO Eric Park said in a canned statement. "ML teams and AI founders have to wait months or pay exorbitant sums to access the latest hardware to their models. We hope to redress this imbalance and accelerate cutting-edge work in AI."

Voltage Park claims it's already leasing GPU resources to startups like Imbue, and finalizing clusters for Character.ai, and Atomic AI. The company plans to expand its offering to long and short term leases as well as provide hourly rates for individual GPUs early next year. These resources will be deployed at sites in Texas, Virginia and Washington, the org told Reuters.

As we've discussed in the past, training large language models can be an incredibly expensive prospect. Models like GPT-3 with hundreds of billions of parameters often require thousands of GPUs running for weeks if not months to complete.

According to the Voltage Park website, GPUs can be had for as little as $1.89 an hour per GPU, though it's not clear what kind of commitment is required to get that pricing. Whatever the case, that's still a fraction of the price Amazon Web Services is charging for time on its H100-equipped P5 nodes. Those instances, which our sibling site The Next Platform discussed at length, include eight Nvidia H100s and come in at $98.32 an hour for on-demand pricing and $43.16 an hour when making a three year commitment.

While Voltage Park's pricing remains vague, the company says it will offer between one and eight GPUs for on-demand customers, while those that need more will need more — between eight and 248 GPUs — to sign a short-term lease. Those willing to commit to a year-long lease will be able to access up to as many 4,088 H100s. At $1.89 an hour per GPU, the latter would run you nearly $68 million a year, which is actually quite a bargain.

Voltage is far from the only company taking advantage of demand for AI hardware. AI infrastructure vendor CoreWeave has also built a GPU cluster with more than 22,000 H100 GPUs in collaboration with Nvidia. ®

RECENT NEWS

From Chip War To Cloud War: The Next Frontier In Global Tech Competition

The global chip war, characterized by intense competition among nations and corporations for supremacy in semiconductor ... Read more

The High Stakes Of Tech Regulation: Security Risks And Market Dynamics

The influence of tech giants in the global economy continues to grow, raising crucial questions about how to balance sec... Read more

The Tyranny Of Instagram Interiors: Why It's Time To Break Free From Algorithm-Driven Aesthetics

Instagram has become a dominant force in shaping interior design trends, offering a seemingly endless stream of inspirat... Read more

The Data Crunch In AI: Strategies For Sustainability

Exploring solutions to the imminent exhaustion of internet data for AI training.As the artificial intelligence (AI) indu... Read more

Google Abandons Four-Year Effort To Remove Cookies From Chrome Browser

After four years of dedicated effort, Google has decided to abandon its plan to remove third-party cookies from its Chro... Read more

LinkedIn Embraces AI And Gamification To Drive User Engagement And Revenue

In an effort to tackle slowing revenue growth and enhance user engagement, LinkedIn is turning to artificial intelligenc... Read more