The $6.9 billion deal further cements Nvidia as a data center player. Credit: Gordon Mah Ung After months of speculation, Mellanox found a suitor — and it was a surprise, to say the least. GPU leader Nvidia snatched up the networking vendor for $6.9 billion, topping a rumored previous offer of $6 billion from Nvidia’s nemesis, Intel. The acquisition ends months of rumors of a suitor for Mellanox. Intel, Microsoft, and Xilinix were all reportedly bidding for the Israeli company, which specializes in high-speed networking. Mellanox Technology was formed in 1999 by a former Intel executive and was a pioneer in the early adoption of InfiniBand interconnect technology, which along with its high-speed Ethernet products is now used in over half of the world’s fastest supercomputers and in many leading hyperscale data centers. It will also be far and away the biggest acquisition in Nvidia’s history. The firm has grown entirely organically, and its few acquisitions have been relatively small, so this is somewhat out of character. “The emergence of AI and data science, as well as billions of simultaneous computer users, is fueling skyrocketing demand on the world’s data centers,” said Nvidia founder and CEO Jen-Hsun Huang in a statement. “Addressing this demand will require holistic architectures that connect vast numbers of fast computing nodes over intelligent networking fabrics to form a giant data center-scale compute engine,” he added. At $6.9 billion, that’s almost half of Nvidia’s $11.7 billion in annual revenue for 2018 and nearly wipes out the $7.4 billion in cash the company was holding. Mellanox’s net profit for fiscal 2018 was $134 million, putting the purchase multiple of 52 times earnings. “I am flabbergasted at what they paid for it. I don’t get those ratios at all,” said Jon Peddie, president of Jon Peddie Research, which follows the graphics market. “Two things we know. One, Jen-Hsun doesn’t do dumb things and two, he doesn’t throw money around. Those two tell me he has something figured out that I don’t understand.” Peddie speculates that this deal is designed to goose Nvidia’s own interconnect effort, NVLink. “It’s not cheap to design high-speed communications networks. It’s as bad as designing a GPU. Mellanox can close the gap with their tech. My speculation is the next-generation GPU will get Mellanox next-gen communications stuck in it that is better than or equal to CXL,” he said, in reference to the new high-speed interconnect announced by Intel and partners. Jim McGregor of Tirias Research also believes it’s a data center play. “Nvidia is trying to shore up its position in the data center because that’s where it’s still growing. It helps build out its strategy from a single box solution to a multi-box solution. So it doesn’t have just the stuff in the box; it has the stuff to connect the boxes,” he said. But had an interesting tidbit to share. He was on a conference call with Huang, who said, “The only reason we acquired them is they called us.” Mellanox said they were getting unsolicited offers and said they were pretty good, and they called Nvidia because the two companies have a long-standing relationship. “I think it was a cultural exchange because they like working with Nvidia and preferred working with them. It sounds like both sides wanted it to happen but it was going to be expensive,” McGregor said. And if it gives Nvidia a chance to thumb its nose at its hated rival Intel, so much the better. The Silicon Valley is a fairly chill place of friendly competition, but you rarely see two companies that hate each other more than Intel and Nvidia. Nvidia and Mellanox expect the deal to close by the end of this year. And Nvidia said “customer sales and support will not change as a result of this transaction.” Related content news High-bandwidth memory nearly sold out until 2026 While it might be tempting to blame Nvidia for the shortage of HBM, it’s not alone in driving high-performance computing and demand for the memory HPC requires. By Andy Patrizio May 13, 2024 3 mins CPUs and Processors High-Performance Computing Data Center news CHIPS Act to fund $285 million for semiconductor digital twins Plans call for building an institute to develop digital twins for semiconductor manufacturing and share resources among chip developers. By Andy Patrizio May 10, 2024 3 mins CPUs and Processors Data Center news HPE launches storage system for HPC and AI clusters The HPE Cray Storage Systems C500 is tuned to avoid I/O bottlenecks and offers a lower entry price than Cray systems designed for top supercomputers. By Andy Patrizio May 07, 2024 3 mins Supercomputers Enterprise Storage Data Center news Lenovo ships all-AMD AI systems New systems are designed to support generative AI and on-prem Azure. By Andy Patrizio Apr 30, 2024 3 mins CPUs and Processors Data Center PODCASTS VIDEOS RESOURCES EVENTS NEWSLETTERS Newsletter Promo Module Test Description for newsletter promo module. Please enter a valid email address Subscribe