![\"koyfin_20260826_122237362.png\"](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/08\/koyfin_20260826_122237362_e5d484807a.png\")
OpenAI Joins The Custom Chips Race<\/p>\nAs the AI boom enters its next phase, investors and analysts have increasingly begun asking if \u2014 and when \u2014 Nvidia\u2019s growth supercycle could slow. <\/p>\n
OpenAI on Tuesday unveiled its first custom chip and said it outperformed Nvidia\u2019s current-generation processors on certain benchmarks.Nvidia\u2019s dominance remains intact for now, analysts say, even though they see OpenAI\u2019s results as a significant development.Nvidia will report its Q2 results after market hours on Wednesday.<\/p>\n
OpenAI\u2019s first custom AI chip is showing an edge over Nvidia Corp.\u2019s Blackwell systems on key inference benchmarks, adding to a growing push by the biggest AI companies to develop their own silicon and potentially reduce their dependence on Nvidia.<\/p>\n
Nvidia continues to post record sales and profits, with investors expecting another strong second-quarter report on Wednesday. But as the AI boom enters its next phase, investors and analysts have increasingly begun asking if \u2014 and when \u2014 the chipmaker\u2019s growth supercycle could slow.\u00a0<\/p>\n
NVDA stock has underperformed the broader chip-sector benchmark, the iShares Semiconductor ETF (SOXX), by a wide margin this year, reflecting growing concerns over its valuation, while critics point out what they describe as the company\u2019s \u201ccircular deals\u201d that artificially bolster revenue growth and its position as the industry\u2019s de facto AI chip supplier.
\u00a0<\/p>\n
OpenAI Joins The Custom Chips Race<\/p>\n
OpenAI on Tuesday\u00a0detailed<\/a> that its Jalape\u00f1o chip delivered 1.5 to 1.9 times more AI work per watt and 1.7 to 3.6 times lower end-to-end latency than Nvidia\u2019s GB200 and GB300 systems across GPT-OSS 120B, DeepSeek R1 and Kimi K2.5 1T.<\/p>\n The chip, developed with Broadcom, is designed specifically for inference \u2014 the stage where trained AI models generate responses \u2014 rather than training models. OpenAI plans to begin deploying Jalape\u00f1o in its own infrastructure by the end of 2026, with second- and third-generation chips already in development.\u00a0<\/p>\n The company also stressed that it will continue deploying Nvidia and other accelerators for both training and inference. That makes Jalape\u00f1o less an immediate Nvidia replacement than a sign of where the AI infrastructure market could be heading: toward increasingly specialized silicon optimized for individual workloads.<\/p>\n The bigger threat comes from custom chips from hyperscalers, who are also Nvidia’s biggest customers. Alphabet has spent years developing its Tensor Processing Units, or TPUs, which are used internally and offered through Google Cloud and are currently on their eighth generation. Amazon has taken a similar approach with Trainium and Inferentia.\u00a0<\/p>\n Amazon\u00a0said<\/a> last month that its custom chips, including Trainium and Graviton, surpassed a $25 billion annual revenue run rate.\u00a0<\/p>\n Morgan Stanley estimates that Google\u2019s TPUs could bring in nearly $200 billion in revenue in the coming years. \u201cWe are raising our TPU 1P sale estimates to $27bn\/GW…selling at a 30% gross margin. In all, we now expect GOOGL to sell 0.3 GW of TPU systems in 2H:26, 3.2 GW in ’27 and 4.2 GW in ’28. This now translates into $84bn\/$108bn of TPU-related Google Cloud revenue in ’27\/’28,\u201d the research firm said in an investor note on Monday.<\/p>\n \u00a0<\/p>\n CompanyLatest AI custom chipLaunchPrimary workloadKey Nvidia riskGoogleTPU 8t \/ 8iApril, 2026Training\/inferenceMore AI workloads can run on Google’s own silicon instead of Nvidia GPUs.Amazon\/ AWSTrainium3December, 2025Training + inferenceAWS can shift more workloads to its own accelerators, lowering infrastructure costs.MetaMTIA 300March, 2026Training accelerator optimized for ranking and recommendation modelsReduces reliance on general-purpose GPUs for Meta-specific workloads and gives Meta more control over performance and costs.<\/p>\n Source: Stocktwits research<\/p>\n Meta has been ramping up its next-generation MTIA custom AI chips,\u00a0detailing<\/a> its MTIA 300 this week. Tesla, meanwhile, has said its upcoming AI chips for the Terafab complex will be built using Intel\u2019s advanced 14A process.<\/p>\n Those chips are not just a threat to Nvidia\u2019s chip sales; they also help Alphabet and Amazon capture more economics from their cloud businesses by lowering the cost of running AI workloads. For Nvidia, the risk is therefore more about the share of future AI compute than an immediate collapse in demand.\u00a0<\/p>\n Amid Custom Chips Race, Nvidia\u2019s Position Appears Strong For Now<\/p>\n Investors and chip watchers were split on whether Jalape\u00f1o changes Nvidia\u2019s outlook.<\/p>\n CNBC analyst Jim Cramer dismissed the threat,\u00a0saying<\/a>: \u201cEvery day I read about some chip that is superior to Nvidia. And every year I see no real competitors.\u201d\u00a0<\/p>\n Noted semiconductor commentator and founder of news site SemiAnalysis Dylan Patel called the result more significant: \u201cUsually first generation chips aren\u2019t competitive, but OpenAI is beating Nvidia Blackwell and even Rubin. This is huge news!\u201d he said in an\u00a0X post<\/a>.<\/p>\n Gavin Baker, managing director at Atreides Management, was similarly impressed but cautious,\u00a0calling<\/a> Jalape\u00f1o the \u201cfirst good ASIC outside of TPU\/Trainium,\u201d while arguing it could underperform more specialized disaggregated accelerator systems.\u00a0\u00a0<\/p>\n Analysts broadly believe that Nvidia\u2019s dominance remains intact, but the economics of AI are increasingly giving its biggest customers a reason to build \u2014 rather than just buy \u2014 the chips powering their models, which create risks for Nvidia.<\/p>\n On Stocktwits, the retail sentiment was \u2018bullish\u2019 for NVDA and META, with all eyes on Nvidia\u2019s quarterly report due after market hours on Wednesday.<\/p>\n Nvidia\u2019s Q2 Report On The Deck<\/p>\n Wall Street expects Nvidia to report roughly $92 billion in quarterly revenue, nearly double the corresponding quarter last year, with analysts looking for continued strength from Blackwell and the upcoming Rubin platforms.\u00a0 The setup is interesting given that NVDA stock declined in seven of the last eight sessions, losing 5.5% cumulatively. Currently, 58 out of 61 analysts have a \u2018Buy\u2019 or higher rating on NVDA stock, with a consensus upside projection of 53.3% over the next 12 months.<\/p>\n NVDA stock is up 14.4% year to date, far below the 71% rise in the iShares Semiconductor ETF (SOXX).<\/p>\n For updates and corrections, email newsroom[at]stocktwits[dot]com.<\/p>\n Read Next:\u00a0Why DOCU, TRI Stocks Slid Over 3% After Google Cloud\u2019s Latest AI Release<\/a><\/p>\n
\u00a0<\/p>\n![\"Data_Center_Chip_Sales_Power_Nvidias_Topline.png\"](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/08\/Data_Center_Chip_Sales_Power_Nvidias_Topline_2668d812dd.png\")
<\/p>\n
![\"Follow](\"https:\/\/chunks-prd.stocktwits-cdn.com\/_next\/static\/media\/google-news-icon.d9c2bfd1.svg\"\/){kind=icon}
<\/a><\/p>\n