Imbue eschews cloud compute for Voltage Park partnership

After sanctions block new AI chip imports

Creator of the Machine Learning Site Reliability Engineering (ML SRE) organization at Google

Chip engineer Richard Ho leads new division at the generative AI startup to help "influence roadmaps for hardware partners for data center networks, racks, and buildings"

But is now at Microsoft, which has its own AI chips

Azure Cobalt 100 and Azure Maia 100 mark company's entry into custom chip game

Jupiter will also feature SiPearl's CPU

As generative AI compute market heats up

Potentially the first commercial cloud deployment of a RISC-V cluster

After A800 and H800 were also blocked

Chip now generally available in Google's Cloud

Following its earlier acquisition of ITRenew

But Project Silica is still not ready for the cloud

Comes as a cash lump sum or credits

Would value GPU business at $1.5 billion

Project Athena may finally be ready to go head-to-head with Nvidia

Company unable to compete with Nvidia

As it looks to reduce its reliance on Nvidia

That's a lot of particles

Will only support six models at launch

Installed at Datacenter 5, system will be made available through Scaleway

Focuses on automotive and Edge compute

Anthropic will put models on Amazon Bedrock, and use Trainium and Inferentia chips on AWS

Using Zen 4c chiplets