Skip to content

Jalapeño

Jalapeño is OpenAI's in-house AI accelerator chip, built with model-generated kernels to rival Nvidia GPUs in power efficiency.

Current stories

buildOne report1 publisher

AI chip-design agents can reproduce patented circuits across thousands of chips, Forte warns

Domenec Forte of the University of Florida says AI chip-design agents can spread an infringed patent across thousands of chips before anyone notices. Teams piloting these tools take on that risk at production volume, while the law on who answers for an agent's accidental copy is unsettled.

Publishers:tomshardware.com

Reality

Evidence40
Adoption35
Hype gap+20
Incentives40
Confidence40
investConfirmed9 publishers

Six customers underwrite Broadcom's $230bn AI chip target for 2028

Custom silicon made up 73% of the $16.7bn AI quarter. Hock Tan says that line doubles twice by fiscal 2028. He has already secured wafers and memory through 2028 to serve promises only six buyers have made.

Perspective Coverage

9 publishers
Builder
Builder 14%
Operator
Operator 21%
Investor
Investor 65%

Reality

Evidence74
Adoption72
Hype gap+22
Incentives62
Confidence70
productOne report1 publisher

JoyIn dates its alien-visitor framing three weeks before OpenAI's essay

JoyIn, a 20-person robotics team in Suzhou with Ant Group backing, says OpenAI distilled its research framing and copied its launch page. The startup is going to court over the two assets any rival could rebuild.

Publishers:thenextweb.com

Reality

Evidence42
Adoption20
Hype gap+40
Incentives80
Confidence45
investOne report1 publisher

OpenAI's cost-per-task argument buys Luna room for ten failed tries before it loses on price

Sarah Friar told Goldman Sachs' technology conference that a pricier model can be cheaper when it needs fewer tries, a claim Artificial Analysis puts at an eleven-to-one price spread against seven index points running the other way.

Reality

Evidence34
Adoption38
Hype gap+30
Incentives80
Confidence45
leadershipOne report1 publisher

Ord's generation-time argument makes runaway AI unlikely, not just slower

The gate on recursive self-improvement is how long one loop takes, and experiments, training runs and new chips all take real time. That makes capability curves fast but bounded, which changes what you sequence first.

Reality

Evidence34
Adoption56
Hype gap+18
Incentives58
Confidence42