Artificial intelligence has undoubtedly turned the tables in the technology industry. When it comes to big players, such as Google LLC and Amazon Web Services Inc., the race is on when it comes to the cloud.

Over the past 10 years, the two companies have made inroads to various levels of success. Two or three different cloud paradigms have a kind of edge cloud, such as Cloudflare Inc., according to Jerry Chen (pictured), general partner at Greylock Partners.

“You have classic clouds like Amazon did in the early days, and these AI clouds now are all about GPUs, InfiniBand, low power, whereas edge clouds are close to the user and the classic clouds,” Chen said. “Clearly, Amazon, Azure and Google also want to own the AI clouds. So, I think there’s a race now [as to] which company can add the AI cloud portfolio and win in that market.”

Chen spoke with theCUBE industry analysts John Furrier and Dave Vellante at the “Supercloud 5: The Battle for AI Supremacy” event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed data centers and how cloud still remains an inevitable wave.

The opportunities at play

Recent trends have seen purpose-built, large-scale GPU clusters emerging. When it comes to data centers, there may not necessarily be a pendulum swing back, according to Chen.

“But [one may see] more inertia to private clouds and private data centers for security reasons and privacy reasons. So, there’s this inevitable shift towards public cloud,” he said. “But now, large enterprises want to train their data. They want private data.”

Because of that, more organizations want to run open-source models in its data centers or in its private cloud, according to Chen. For that reason, cloud remains the inevitable wave.

“You’re going to see this new trend of new stacks of GPUs and supercomputers inside your data centers,” he said. 

That begs the question: Is there an opportunity for startups to go after that, or will it be Dell Technologies Inc. or Hewlett Packard Enterprise Co. that does that? In fact, Nvidia Corp. is a stock all “should have bought,” according to Chen.

“I think there’ll be an opportunity for startups. Not necessarily build systems, because systems are hard to make margin, but software,” Chen said. “Think about all the software you need to train, fine-tune, run the models and build these applications.”

How might one replicate the stack that Bedrock, Azure or Google has in one’s data center? It’s all about retrieval augmented generation, according to Chen. That’s the process of providing additional context from outside data sources to enable more accurate query responses from large language models, improving the performance of generative AI applications.

“How to build these data pipelines of your enterprise data with these new models inside a public cloud or private cloud,” he said. “I think you’re going to see new software stacks be created.”

What about when a company, such as OpenAI Inc., does large language models or retrieval-augmented generation out of the box? If it’s just one giant model, all OpenAI or all Bard, then the startup ecosystem probably is decimated, according to Chen.

“But, there’s a long fat tail of open-source models, other foundation models, not just open source,” he said.

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the “Supercloud 5: The Battle for AI Supremacy” event:

Photo: SiliconANGLE