OpenAI deploys Jalapeño ASIC with AMD EPYC Turin hosts
OpenAI is running its first home-built Jalapeño inference ASICs beside AMD EPYC Turin CPUs with 1.5 TB of host memory. Cloud and chip buyers watch whether custom silicon undercuts Nvidia’s grip. Hardware VP Richard Ho called Nvidia’s Vera standalone "a little bit behind" on maturity for the rollout.
OpenAI has deployed its Jalapeño inference ASICs in rack-scale setups hosted by AMD EPYC Turin CPUs, each with about 1.5 TB of memory, Tom’s Hardware reported after interviewing Hardware VP Richard Ho. Ho said pairing Jalapeño with Turin was "pragmatic" de-risking: the team wanted aggressive performance and cost goals without unnecessary platform risk.
Asked about Nvidia’s new Arm-based Vera CPU — which Nvidia markets for agentic workloads and which OpenAI is also exploring — Ho said Vera standalone is "a little bit behind… on that maturity level" versus Turin. Partners already had Turin experience, and socketed x86 hosts are easier to service than board-mounted Vera in early fleets. AI Insiders and SemiAnalysis coverage of the Jalapeño reveal framed the chip as OpenAI’s bid to control inference cost and supply beyond Nvidia GPUs.
The choice matters for every company renting AI compute. If frontier labs can field mature custom ASICs on proven hosts, Nvidia’s pricing power softens even while GPU demand stays huge. OpenAI remains an Nvidia customer and AGI deployment partner; Jalapeño-plus-Turin is a portfolio move, not a divorce.
Backstory: hyperscalers from Google to Amazon already run custom accelerators; OpenAI’s first production ASIC puts a model lab in that club. Engage: Tom’s Hardware interview and SemiAnalysis rack notes.
What we know:
Jalapeño ASICs are deployed with Turin hosts; Ho cited maturity and partner experience; Vera was judged riskier for this program; OpenAI still explores other CPUs.
What we don’t:
Public benchmarks proving Jalapeño beats Nvidia on efficiency at scale; fleet size; when Jalapeño might pair with Vera or Arm AGI instead.
Agree (full): Tom’s Hardware and AI Insiders agree OpenAI’s custom Jalapeño ASIC is live in deployment with AMD hosts.
Differ (full): OpenAI frames a maturity-driven Turin pick; efficiency claims that Jalapeño beats Nvidia remain stronger in trade coverage than in public third-party benchmarks.
Where outlets agree Tom’s Hardware and AI Insiders agree OpenAI has deployed its first custom Jalapeño inference ASIC alongside AMD EPYC Turin hosts.
Where they differ OpenAI calls Turin the mature, low-risk host; efficiency claims that Jalapeño beats Nvidia remain stronger in trade coverage than in public third-party benchmarks.