HPC storage move bolsters Nvidia’s data center position
Nvidia announced Monday it has acquired Tel Aviv, Israel-based Excelero. The move bolsters Nvidia’s position in the data center space: Excelero is a software-defined storage vendor which specializes in NVM Express (NVMe)-based storage for networks and clouds.
“The company’s team of engineers — including its seasoned co-founders with decades of experience in HPC, storage and networking — bring deep expertise in the block storage that large businesses use in storage-area networks,” said Suresh Ollala. Ollala is Senior Director of Engineering for NGC Storage and Data platform services at NVIDIA.
Ollala said that going forward, Excelero will work to expand support for block storage throughout NVidia’s enterprise software stack. Ollala explained that Excelero’s technology will be essential Nvidia’s DOCA software framework which is used in its BlueField line of data processing units (DPUs).
Excelero and Nvidia are long-time collaborators, Ollala noted.
“The company has been an NVIDIA partner since its early days, attracting the former Mellanox, now part of NVIDIA, as an investor. We collaborated on accelerating storage with RDMA, a key technology at the heart of both InfiniBand and RoCE (Ethernet) networks,” said.
Nvidia appears undaunted in its plans for data center dominance following its abandoned attempt to acquire semiconductor maker Arm from SoftBank.
Nvidia said it will honor Excelero contracts, but going forward the technology will be integrated wholly into Nvidia’s enterprise software stack.
Beyond Excelero, Nvidia announced plans in January to acquire Bright Computing, a maker of high-performance computing (HPC) cluster management software. Bright’s software helps automate the creation and management of Linux-based computing clusters in the cloud and in hybrid cloud environments.
The company has seen strong growth in its data center business in recent quarters. Nvidia’s data center-related sales were up 71% year over year, driven by AI uptake. The company’s seen broad use of its processors in data centers, and it expects the trend to continue with new system on a chip (SoC) innovations and other new products headed to market this year and next.
Meta recently took the wraps off a new AI Research SuperCluster (RSC) supercomputer which it says will help it build the metaverse by developing newer and better AI models. Nvidia is playing a central role, as the RSC comprises Nvidia’s DGX A100 systems for its AI data center infrastructure. The companies plan to expand the RSC to much larger proportions as time goes on.
