One Stop Systems Shipping Gen 5 4U Pro Accelerator System Doubling Performance Over Previous-Gen
Includes upgraded power and cooling to support multiple Nvidia H100 Tensor Core GPUs with PCIe Gen 5.
This is a Press Release edited by StorageNewsletter.com on February 16, 2024 at 2:02 pmOne Stop Systems, Inc. (OSS) has begun shipping its latest Gen 5 4U Pro Accelerator system to a large composable infrastructure provider.
It expects shipments of this compute accelerator to the customer to total between $4 million and $6 million over the next 3 years.
For AI workflows at the edge, this latest Gen 5 4U Pro Accelerator delivers 2x the interconnect bandwidth performance over Gen 4. The accelerator also includes upgraded power and cooling to support multiple Nvidia H100 Tensor Core GPUs with PCIe Gen 5, resulting in 4.8x the AI inference performance using FP8 precision compared to the previous-gen.
“Our customer’s adoption of this product for their leading-edge composable infrastructure platform represents a strong validation of our superior capabilities in PCI Express expansion and AI accelerator technology,” stated Mike Knowles, president and CEO. “Our customers are able to stay at the forefront of the latest high-performance computing technology with our highly scalable, interoperable solutions.”
This customer has used the previous-gen the firm’s 4U Pro to rapidly scale GPU power and NVMe storage. Now with this latest version, the company anticipates its flexibility, advanced features and pure acceleration will unleash the capability of the customer’s software-defined composable infrastructure solutions for both its commercial and defense end customers.
The compute accelerator is expected to be deployed at the customer’s locations worldwide, which include commercial cloud data centers and military command posts that demand ‘performance without compromise.’
HPC and resource-intensive AI workloads are on the rise in the edge computing market. Demand for edge solutions is projected to grow at a 37.9% CAGR to reach $155.9 billion by 2030, with the mobile edge segment expected to grow even faster at a 49.7% CAGR to reach $58.4 billion by 2032.
This growth is driving increased demand for expansion platforms like the 4U Professional Accelerator, which is capable of processing, analyzing and acting upon vast amounts of sensor and other data in real time using the fastest and most rugged interconnect technology that OSS provides.
About OSS 4U Professional Accelerator system
It combines the power of the latest PCIe Gen 5 high-speed connectivity with an optimized, feature-rich and rugged design for the most demanding high-performance edge applications.
The system supports up to 8 Nvidia H100 GPUs that deliver 2.6x the FP64 performance compared to the previous-gen. It also supports 4xPCIe Gen 5 x16 HIB/NIC slots for up to 256GB/s of sustained data throughput.
The 4U Pro can be configured to provide 16 single-width PCIe Gen 5 x8 slots for FPGA data ingest or the latest storage add-in cards. It also includes proprietary firm’s Unified Baseboard Management Controller (U-BMC) software that enables dynamic fan speed control and an Intelligent Platform Management Interface (IPMI) for system monitoring that integrates with the host server system management.
U-BMC manages, monitors and controls the telemetry of the company’s products with features that set firm’s systems apart. It securely enables the firm’s customers to unify their systems management and enjoy unique benefits when using several OSS products simultaneously, such as for data acquisition, data storage, AI acceleration and autonomous vehicle compute. The rugged chassis design makes it suitable for use outside the datacenter environment. All together, these features make the Gen 5 4UP the accelerator expansion platform for the entire AI workflow at the edge.
OSS Gen 5 4U Pro and other AI Transportable solutions can be ordered directly from the company.