One of the biggest cloud hosting companies in the world, OVHcloud, has introduced its second generation of specialized ‘Bare Metal Scale’ servers. This next generation of high-performance dedicated servers, which follows their initial Scale line-up announced in 2021, adds even more computational power to the mix, helping to make innovation more accessible.
The new generation of Scale Bare Metal servers, which would cater to verticals like healthcare, finance, industry, public sectors, and media, would use powerful AMD and NVIDIA processors to meet the most demanding use cases of customers, including virtualization, container orchestration, grid computing, big data, analytics, and confidential computing.
The AMD Secure Encrypted Virtualization (SEV) included in the fourth generation of AMD EPYC processors would help the new OVHcloud Bare Metal second-generation Scale servers by enabling stronger virtual machine isolation.
With single- or dual-socket configurations, OVHcloud’s high-performance and scalable bare metal servers can use the high core count of AMD EPYC CPUs. This would make it possible for the hardware to provide enormous computational power, enabling users to process and evaluate big data sets instantly. Consumers’ data may provide insights that help with decision-making. Customers can quickly process, handle, and use data in addition to storing vast volumes of data with up to 46 TB of NVMe storage.
“Following our announcements at the Very Tech Trip event earlier this year, we are extremely pleased to deliver solid innovations with new x86 CPU generations bringing more performance and capacity,” said Yaniv Fdida, Chief Product Officer at OVHcloud. “This enables our customers to improve their cloud-associated TCO all the while unlocking new use cases with our platforms such as grid computing, VDI and inference.”
NVIDIA L4 Tensor Core GPU for Machine Learning
With fourth-generation EPYC processors, code-named Genoa, the second generation of Scale servers leverages AMD's most sophisticated Zen 4 CPU architecture. In a single-socket setup, the range extends from AMD EPYC 16 cores / 32 threads to 96 cores / 192 threads (SCALE-a1 to SCALE-a6).
Utilizing fourth-generation AMD EPYC processors, code-named Genoa and Bergamo, the latter of which has Zen 4C cores, the second-generation Scale lineup is available in a twin socket configuration with core counts up to 256 cores and 512 threads. With 192 cores and 384 threads (2P AMD EPYC 9654 code called Genoa), OVHcloud’s SCALE-a7 model has the most, while the SCALE-a8 variant has 256 cores and 512 threads (2P AMD EPYC 9754 code named Bergamo). The DDR5 memory configuration, which spans from 128 GB to 1 TB (4800 MHz) and is available for the first time at OVHcloud, offers new Scale references a wide range of storage choices. These range from the standard 2 x 960 GB NVMe to 6 x 7,68 TB SSD NVMe.
NVIDIA L4 GPU-powered SCALE-GPU dedicated servers are being introduced by OVHcloud in addition to new second-generation Scale CPU-based equivalents.
The NVIDIA L4 is a universal GPU designed to efficiently handle tasks including graphics, AI, and video, and it is based on the NVIDIA Ada Lovelace GPU architecture. Use cases including machine learning with image classification, object identification, natural language processing, and VDI would be well-suited for the hardware.
In terms of speed, availability, control, and flexibility, OVHcloud's new second-generation Bare Metal Scale servers would offer all the benefits of dedicated servers, along with guaranteed unmetered public bandwidth of up to 10 Gbps and up to 25 Gbps of unmetered private bandwidth. In addition to ensuring predictable pricing, OVHcloud would offer businesses the greatest support for their workloads, environments, and ad hoc technologies with this solution.
