Nvidia's purchase of Mellanox for $6.9 billion will translate into a broader data center play for beyond the graphics and high performance computing markets.
According to Nvidia, the company will buy Mellanox, which makes interconnects, for $125 in cash. Nvidia reportedly outbid Intel the for the company. With one deal, the following unfolds:
Nvidia's rivalry with Intel hits a new stage.
Nvidia gets more revenue diversification and data center sales.
Mellanox gives Nvidia more entries into high performance computing and the data center.
As artificial intelligence workloads become the norm, Nvidia with Mellanox be more an architecture play.
The two companies are also long-time partners. Nvidia's computing platform and Mellanox's interconnects power more than half the world's Top 500 supercomputers. The duo will also have more entries in the hyperscale data center market as well as cloud service providers.
Mellanox is best known for its InfiniBand interconnect technology, which works along with high speed Ethernet products to move data through systems quickly.
The game plan for Nvidia is to use Mellanox to optimize data center workloads across compute, networking and storage. Jensen Huang, CEO of Nvidia, said the acquisition of Mellanox is aimed at addressing "holistic architectures that connect vast numbers of fast computing nodes over intelligent networking fabrics to form a giant datacenter-scale compute engine."
Mellanox will be accretive to Nvidia's non-GAAP earnings and cash flow. Mellanox also gives Nvidia access to the high-speed networking market.
Add it up and Nvidia with Mellanox can better optimize systems. It's a market where Nvidia will own a bit more of high performance systems, which are likely to also include Intel too.
Analysts were upbeat on the purchase. In a research note, Shebly Seyrafi, an analyst at FBN Securities, said:
This deal only increases the competitive rivalry between NVDA and INTC to go beyond GPUs to include Infiniband and Ethernet.
The deal will strengthen NVDA's ability to offer datacenter-scale computing solutions optimized across computing, networking and storage, that can accelerate datacenter-scale workloads. It will also foster the next generation of HPC and AI platforms for hyperscale and enterprise datacenters, as well as supercomputing centers.