{"id":597969,"date":"2026-08-27T21:22:13","date_gmt":"2026-08-27T21:22:13","guid":{"rendered":"https:\/\/www.newsbeep.com\/il\/597969\/"},"modified":"2026-08-27T21:22:13","modified_gmt":"2026-08-27T21:22:13","slug":"openai-built-an-nvidia-beating-inference-chip-in-nine-months-broadcom-avgo-helped-make-it-happen","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/il\/597969\/","title":{"rendered":"OpenAI Built an Nvidia-Beating Inference Chip in Nine Months. Broadcom (AVGO) Helped Make It Happen"},"content":{"rendered":"\n

OpenAI says its first homegrown inference chip can do up to 1.9 times more AI work per watt than the comparison systems, and it went from initial design to tape-out in nine months. The more investable surprise is who helped turn that sketch into silicon. <\/p>\n

Broadcom Inc. (NASDAQ:AVGO<\/a>) handled Jalape\u00f1o’s physical implementation and supplied the Tomahawk networking. Celestica is responsible for board, rack and system integration, while TSMC manufactured the chip. OpenAI designed the architecture and said its own models helped it move from initial design to tape-out in nine months, much faster than usual. <\/p>\n

OpenAI’s Nvidia Win Is Narrower Than It Sounds <\/p>\n

In OpenAI’s published InferenceX results, Jalape\u00f1o delivered 1.5 to 1.9 times more work per watt and 1.7 to 3.6 times lower end-to-end latency across three large open models. The comparisons included Nvidia GB200 and GB300 systems. <\/p>\n

\"OpenAI<\/a> OpenAI Built an Nvidia-Beating AI Chip in Nine Months. Broadcom (AVGO) Helped Make It Happen <\/p>\n

Those are impressive results, but bear in mind that Jalape\u00f1o is an inference chip, not a training accelerator. It was not compared with Vera Rubin, and some tests emphasized single-token prediction, while Nvidia systems can use multi-token techniques. Initial deployment is expected by year-end, so production economics remain unproven. <\/p>\n

OpenAI rated Jalape\u00f1o at 700 watts but said measured sustained power stayed at or below 550 watts. <\/p>\n

The Bull and Bear Case for Broadcom and Nvidia <\/p>\n

Broadcom Inc. (NASDAQ:AVGO) bulls can connect Jalape\u00f1o to a much larger prize: the companies plan to deploy 10 gigawatts of OpenAI-designed accelerators through 2029. Broadcom’s AI semiconductor revenue already jumped 143% to $10.8 billion last quarter. Bears should watch customer concentration, costly project financing and margin pressure from expensive HBM memory. <\/p>\n

For NVIDIA Corporation (NASDAQ:NVDA<\/a>), the bull case is that OpenAI still says it needs plenty of Nvidia hardware, and Jalape\u00f1o does not replace the CUDA training ecosystem. The bear case is strategic, in that every efficient custom inference chip shifts another recurring workload away from premium GPUs. <\/p>\n

Hedge Funds Remains Heavily Invested in Both <\/p>\n

Nvidia’s hedge-fund holder count rose from 277 to 286 in Q2, a 3.2% increase. Insider Monkey’s latest Broadcom count was 171 funds in Q2, down from 174 in Q1. As of August 14, short interest represented only 1.20% of Broadcom Inc. (NASDAQ: AVGO)’s float and 1.23% of NVIDIA Corporation (NASDAQ:NVDA)’s. Short sellers are barely leaning against either stock. <\/p>\n

Jalape\u00f1o is not yet an Nvidia killer. It is proof that Broadcom can help Nvidia’s biggest customers become chip companies and get paid even when Nvidia loses the socket. <\/p>\n","protected":false},"excerpt":{"rendered":"OpenAI says its first homegrown inference chip can do up to 1.9 times more AI work per watt…\n","protected":false},"author":2,"featured_media":597970,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[345,343,344,938,85,270455,46,270456,4136,125],"class_list":["post-597969","post","type-post","status-publish","format-standard","has-post-thumbnail","category-artificial-intelligence","tag-ai","tag-artificial-intelligence","tag-artificialintelligence","tag-broadcom-inc","tag-il","tag-initial-design","tag-israel","tag-jalapeno","tag-nvidia","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/posts\/597969","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/comments?post=597969"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/posts\/597969\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/media\/597970"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/media?parent=597969"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/categories?post=597969"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/tags?post=597969"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}