Alphabet and Nvidia have joined outstanding risk capital investors to support Superintelligence (SSI), a startup co -founded by former OpenI chief scientist, Ilya Sutskever, who has rapidly increased to the artificial artificial family that is directed at the same time.
Financing illustrates the renewed interest of the great technology and infrastructure suppliers to make strategic investments in the new companies that develop avant -garde that requires massive amounts of computer power. Alphabet, which has its own AI models, at the beginning of the week announced an agreement on its cloud computing arm to sell SSI access to tensioner processing units (TPU), its internal ia chips.
SSI, which according to the sources were recently valued at $ 32 billion in a round directed by Greenoaks, is one of the new highest profile companies that works in the research of the AI model, thanks to the stellar history of Sutkever in the prediction of the next great thing in the development of AI.
Like many of its competitors, it has a great demand for chips.
Reuters could not determine the exact terms of Alphabet’s investment and NVIDIA in SSI. The spokesmen of the three companies declined to comment.
The twin movements of the Corporate Division and in the Alphabet cloud with high -profile AI laboratories, including SSI and Anthrope, show the hardware strategy of the evolution of the technological giant.
Google originally reserved TPU for internal use. The agreement to sell SSI chips in significant quantities to support your border research exemplifies the company’s continuous strategy to expand sales to external customers, Darren Mowry, a managing director in charge of Google associations with the week.
“With these constructors of fundamental models, gravity increases dramatically to us,” he said.
IA developers have historical NVIDIA preferred processing units, which have more than 80% of the AI chips market.
But SSI so far is using mainly TPU instead of GPU for its research and development of AI, two sources said.
Google sacrifices both the GPU Nvidia and their own TPU through their cloud service. Their own chips are destined to excel in specific AI tasks and are more efficient than general Purfosa GPUs. These chips are used in the leg to build large -scale AI models, such as Apple and Anthrope, an OpenI competitor who has received billions of dollars in Google and Amazon funds.
Google and Nvidia also face a challenger on Amazon, who is building their own competitors processors called train and inferentia. Amazon has said since 2023 that Anthrope would develop his technology in those chips. The technological giant announced in December that Anthrope would be the first client to use a mass supercomputer driven by hundreds of thousands or their own chips.
Media, Anthrope continues to use TPUS for its development and has not decreased spending on Google chips, two sources said.
It is increasingly common that the main cloud suppliers invest strongly in new AI companies that not only build fundamental models but also serve as important customers of their infrastructure. For example, Amazon and Google have invested in anthropic, while Microsoft has made substantial bets in OpenAi. Nvidia has also backed Openi, as well as Xai by Elon Musk.
Posted on April 12, 2025