Nvidia Corp and Cisco Systems Inc are teaming up to make it easier for corporations to build their own AI computing infrastructure, an attempt to push the technology beyond the big data-center providers.
Under an arrangement announced yesterday, Cisco is to offer Nvidia-based equipment — popular for developing artificial intelligence models — along with its networking gear.
For Nvidia, the partnership opens up another channel for its technology. Cisco, meanwhile, gets to capitalize on an AI spending boom that has helped turn Nvidia into the world’s most valuable chipmaker.
Photo: AFP
“Working closely with Cisco, we’re making it easier than ever for enterprises to obtain the infrastructure they need to benefit from AI, the most powerful technology force of our lifetime,” Nvidia chief executive officer Jensen Huang (黃仁勳) said in a statement.
Cisco is to offer M7 server computers that use Nvidia’s latest processors, as well as the chipmaker’s software, AI models and development tools. It is also offering its own cloud-based software and services for monitoring and managing such systems.
“Strengthening our great partnership with Nvidia is going to arm enterprises with the technology and the expertise they need to build, deploy, manage and secure AI solutions at scale,” Cisco CEO Chuck Robbins said.
The two companies made the announcement at the Cisco Live event in Amsterdam early yesterday. It is part of an effort to spread the deployment of AI hardware beyond its current concentration in data centers owned by Microsoft Corp, Amazon.com Inc, Alphabet Inc’s Google and Meta Platforms Inc.
That group of so-called hyperscalers represents a potentially risky concentration of revenue for Nvidia. Though demand for the company’s chips remains insatiable, those customers could ultimately use more of their own technology. They are all either working on their own chips or have already begun adopting them in some way.
Cisco’s largest sales team is to sell Nvidia-based servers combined with its Ethernet networking gear. Nvidia offers an alternative called InfiniBand in its own products. The closer tie-up with Cisco gives Nvidia better access to customers who prefer to stick with the more widely used standard for transferring data between computers.
Taiwan Semiconductor Manufacturing Co (TSMC, 台積電) would not produce its most advanced technologies in the US next year, Minister of Economic Affairs J.W. Kuo (郭智輝) said yesterday. Kuo made the comment during an appearance at the legislature, hours after the chipmaker announced that it would invest an additional US$100 billion to expand its manufacturing operations in the US. Asked by Taiwan People’s Party Legislator-at-large Chang Chi-kai (張啟楷) if TSMC would allow its most advanced technologies, the yet-to-be-released 2-nanometer and 1.6-nanometer processes, to go to the US in the near term, Kuo denied it. TSMC recently opened its first US factory, which produces 4-nanometer
GREAT SUCCESS: Republican Senator Todd Young expressed surprise at Trump’s comments and said he expects the administration to keep the program running US lawmakers who helped secure billions of dollars in subsidies for domestic semiconductor manufacturing rejected US President Donald Trump’s call to revoke the 2022 CHIPS and Science Act, signaling that any repeal effort in the US Congress would fall short. US Senate Minority Leader Chuck Schumer, who negotiated the law, on Wednesday said that Trump’s demand would fail, while a top Republican proponent, US Senator Todd Young, expressed surprise at the president’s comments and said he expects the administration to keep the program running. The CHIPS Act is “essential for America leading the world in tech, leading the world in AI [artificial
REACTIONS: While most analysts were positive about TSMC’s investment, one said the US expansion could disrupt the company’s supply-demand balance Taiwan Semiconductor Manufacturing Co’s (TSMC, 台積電) new US$100 billion investment in the US would exert a positive effect on the chipmaker’s revenue in the medium term on the back of booming artificial intelligence (AI) chip demand from US chip designers, an International Data Corp (IDC) analyst said yesterday. “This is good for TSMC in terms of business expansion, as its major clients for advanced chips are US chip designers,” IDC senior semiconductor research manager Galen Zeng (曾冠瑋) said by telephone yesterday. “Besides, those US companies all consider supply chain resilience a business imperative,” Zeng said. That meant local supply would
BIG INVESTMENT: Hon Hai is building the world’s largest assembly plant for servers based on Nvidia Corp’s state-of-the-art AI chips, Jalisco Governor Pablo Lemus said The construction of Hon Hai Precision Industry Co’s (鴻海精密) massive artificial intelligence (AI) server plant near Guadalajara, Mexico, would be completed in a year despite the threat of new tariffs from US President Donald Trump, Jalisco Governor Pablo Lemus said. Hon Hai, also known as Foxconn Technology Group (富士康科技集團), is investing about US$900 million in what would become the world’s largest assembly plant for servers based on Nvidia Corp’s state-of-the-art GB200 AI chips, Lemus said. The project consists of two phases: the expansion of an existing Hon Hai facility in the municipality of El Salto, and the construction of a