Startup Valued at $2 Billion Integrates New Raptor Chip with Nvidia's Infrastructure, Prepares AI Servers for 2027 Focused on Chatbots, Programming, and Voice Agents - CPG Click Petróleo e Gás
Article excerpt
Highlighted: the sentence this signal was extracted from
Science and Technology d-Matrix will integrate its Raptor processors, specialized in artificial intelligence inference, with Nvidia's server architecture using NVLink Fusion. The partnership paves the way for mixed chip systems within the same "AI factories," as the Microsoft-backed startup aims to compete in a market stage that is growing as models transition from training to daily use. The U.S. startup d-Matrix has decided to leverage one of Nvidia's core technologies to embed its own processors within the chip giant's data center infrastructure. The deal involves NVLink Fusion, an interconnection system that will allow the next generation of Raptor accelerators to be integrated directly into Nvidia's MGX rack architecture. According to Reuters, the first compatible systems are anticipated for 2027 , while d-Matrix expects to finalize the design stage of the Raptor by 2026 . The technology primarily targets a rapidly growing phase of artificial intelligence: inference, when pre-trained models need to respond to users, generate code, engage in real-time conversations, or operate voice agents. Anvisa Approves Import of Experimental Drug for Lito Sousa After Rare Disease Diagnosis, Influencer May Be First to Receive ALN-6457 Without Human Trials Massive Cloud Advances Like a Wall Over SC City, Darkening the Horizon and Turning Day Into Night in Front of Drone, While Region...
Keep reading with a free account
The rest of this article, and every signal for d-Matrix, is in your free account.
