Taipei, Taiwan, 28 March 2018 – GIGABYTE continues to strengthen its HPC system portfolio by today releasing the G190-G30, a 1U 4 x SXM2 GPU server. Building on the advantages of our current 1U GPU-server range, such as an airflow optimized rack mount, large thermal envelope and uncompromised flexibility, the G190-G30 has been developed to accommodate 4 x NVIDIA SXM2 Tesla® V100 or P100 GPU accelerators with NVLink™ interconnection.
Using NVLink™ technology allows for higher bandwidth, more links, and improved scalability over PCIe for GPU to GPU interconnection. For example, a single NVIDIA Tesla® V100 GPU can support up to six NVLink™ links with a signaling rate of 25 GB/second in each direction, equaling a total bandwidth of 300 GB/sec - 10X the bandwidth of PCIe 3. The higher bandwidth and reduced latency allows the G190-G30 to unlock multi-GPU interconnectivity bottlenecks for better performance in intensive parallel computing applications.
The release of the G190-G30 coincides with GIGABYTE’s participation in this week’s GPU Technology Conference in San Jose, where we will be showcasing our range of 1U, 2U and 4U GPU-server systems fully compatible with NVIDIA’s GPU accelerators for machine learning and AI applications. GIGABYTE is proud to add the G190-G30 to our product family, giving our customers a wider choice of technical and cost options to meet their HPC server requirements.
For more information on GIGABYTE’s GPU server product range, please visit http://b2b.gigabyte.com/GPU-Server
Please follow @GigabyteServer on Twitter for the latest product release news!