Las Vegas
\n \u00a0\u2014\u00a0<\/p>\n
\n Nvidia just provided a closer look at its new computing platform for AI data centers, Vera Rubin, a release that could have major ramifications for the future of AI given the industry\u2019s massive reliance on the company\u2019s tech.\n <\/p>\n
\n Nvidia previously announced some details about Vera Rubin but laid out how the system will work and revealed its launch timing during the CES tech conference in Las Vegas on Monday. Vera Rubin is currently in production and the first products running on it will arrive in the second half of 2026, the company said.\n <\/p>\n
\n Nvidia<\/a> has become the poster child for the AI boom, with the pervasiveness of its AI chips and platforms propelling it to briefly become the world\u2019s first $5 trillion company last year. But the company is also combatting fears of an AI bubble <\/a>amid growing competition and a push by tech companies to make their own AI chips to decrease reliance on Nvidia.\n <\/p>\n \n Nvidia CEO Jensen Huang, clad in his signature leather jacket, addressed the question of where AI funding is coming from \u2013 a point central to the bubble debate \u2013 in his opening remarks on stage at the theater inside the Fontainebleau Las Vegas. He said companies are shifting budgets in research and development in classical computing methods to artificial intelligence.\n <\/p>\n \n \u201cPeople ask, where is the money coming from? That\u2019s where the money is coming from,\u201d he said.\n <\/p>\n \n The Vera Rubin platform is an attempt by Nvidia to position itself as the answer to the computing challenges posed by increasingly demanding AI models \u2013 such as whether existing infrastructure <\/a>can handle increasingly complicated AI queries. The company claims in a press release that its upcoming AI server rack, called Vera Rubin NVL72, \u201cprovides more bandwidth than the entire internet.\u201d\n <\/p>\n \n With Vera Rubin, Nvidia says it\u2019s developed a new type of storage system to help AI models process more complex, context-heavy requests more quickly and capably. Existing types of storage and memory used by traditional computers and even the graphics processing units powering data centers won\u2019t be enough as companies like Google, OpenAI and Anthropic shift from offering simple chatbots to full-fledged AI helpers.\n <\/p>\n \n Huang walked through what the transition from chatbots to agents on Monday. In a video demonstration, a person built a personal assistant by connecting a friendly-looking tabletop robot to multiple AI models running on Nvidia\u2019s DGX Spark desktop computer. The robot was able to do things like recount the user\u2019s to-do list and even tell the dog to get off the couch.\n <\/p>\n \n Huang said creating such an assistant would have been unimaginable several years ago but is \u201cutterly trivial\u201d now that developers can rely on large language models rather than traditional programming tools to build apps and services.\n <\/p>\n \n In other words, the old way simply won\u2019t cut it as AI grows more sophisticated and \u201creasons\u201d on tasks that take multiple steps like these, Nvidia claims.\n <\/p>\n \n \u201cThe bottleneck is shifting from compute to context management,\u201d Dion Harris, Nvidia\u2019s senior director of high-performance computing and AI hyperscale solutions, said on a call with reporters ahead of the press conference.\n <\/p>\n \n \u201cStorage can no longer be an afterthought,\u201d he added.\n <\/p>\n \n Nvidia also just entered into a licensing agreement with a company called Groq<\/a> that specializes in inference ahead of CES, another sign that it\u2019s investing heavily in that branch of AI.\n <\/p>\n \n \u201cInstead of a one-shot answer, inference is now a thinking process,\u201d Huang said, referring to the process AI models go through to \u201cthink\u201d and \u201creason\u201d through answers and accomplish tasks.\n <\/p>\n \n All of the major cloud providers \u2013 Microsoft, Amazon Web Services, Google Cloud and CoreWeave \u2013 will be among the first to deploy Vera Rubin, Nvidia said in its press release. Computing companies like Dell and Cisco are expected to incorporate the new chips into their data centers, and AI labs such as OpenAI, Anthropic, Meta and xAI are likely to embrace the new tech for training and to provide more sophisticated answers to queries.\n <\/p>\n \n Nvidia also deepened its push<\/a> into autonomous vehicles with new models called Alpamayo and \u201cphysical AI,\u201d the type of AI that powers robots and other real-world machinery, building on the vision it laid out during its GTC conference in October.\n <\/p>\n \n But Nvidia\u2019s progress and prevalence also means it shoulders the burden of consistently surpassing Wall Street\u2019s high expectations and assuaging concerns that spending on AI infrastructure is far outpacing tangible demand.\n <\/p>\n \n Meta, Microsoft and Amazon, among others, have spent tens of billions in capital expenditures<\/a> this year alone, and McKinsey & Company<\/a> expects companies to invest nearly $7 trillion in data center infrastructure globally by 2030. And much of the support being poured into AI seemingly involves a relatively small group of companies trading money and technology back and forth in what\u2019s known as \u201ccircular funding.\u201d\n <\/p>\n \n Google and OpenAI<\/a> have also been leaning into developing their own chips, allowing them to better tailor hardware to the specific needs of their models. Nvidia has also been facing growing competition from AMD<\/a>, and chipmaker Qualcomm also recently announced it\u2019s getting into the data center business.\n <\/p>\n