OpenAI’s Jalapeño chip is built for fast inference at scale, benchmarks show | TechCrunch · AuraTracer

SummaryAI generated

According to TechCrunch, the Jalipeño chip launched by OpenAI is designed for large-scale fast inference. Related benchmark results have been published, confirming the chip’s performance in high-performance computing scenarios.

Related eventsRELATED EVENTS

2026-09-08 10:00“The starting price is over 10,000! The Xiaomi 18 Fold has been released. Lei Jun talks about chip manufacturing: As soon as we start to catch up, Xiaomi will be on the road to victory.”
2026-09-07 13:02“China launched the International Earth-Moon Cube Constellation Program: 30 satellites will be deployed by 2030 to build a global continuous observation network in Earth-Moon space”
2026-09-07 20:42Xiaomi 18 Fold is the first to be equipped with LPDDR6, and Changxin has officially announced that LPDDR6 will be mass-produced for flagship smartphones for the first time

Key entitiesKEY ENTITIES

BroadcomNvidiaOpenAIRichard Ho

Event frameEVENT FRAME

Launch

OpenAI Jalapeño chip 发布 Jalapeño 芯片，性能优于当前 SOTA

Coverage · reports per dayLANGUAGE SPLIT

Entity relations

SignalsSIGNALS

Keyword heat

OpenAI1
Broadcom1
Nvidia1
Richard Ho1

All reports (1)SOURCES

↗

T TechCrunch 归档 8月 p05 en 2026-08-25 08:00

OpenAI’s Jalapeño chip is built for fast inference at scale, benchmarks show | TechCrunch

OpenAI 在周二举行的 Hot Chips 会议上展示了其 Jalapeño 芯片的首批基准测试结果，显示其在 SemiAnalysis 的 InferenceX 测试中比现有的最先进的推理处理器拥有更高的每用户令牌数和每千瓦吞吐量。OpenAI 硬件负责人 Richard Ho 表示，Jalapeño 能更高效地处理更多 AI 工作并降低延迟，其对比对象为英伟达 Blackwell 系统。该芯片由 OpenAI 与博通合作开发，旨在通过全栈方法优化推理过程中的预填充和通信阶段，减少数据移动和通信延迟。Jalapeño 计划于 2026 年底以极小规模部署，主要大规模部署预计将在 2027 年进行。