Global compute, storage and networking technologies vendor, Supermicro, has introduced several new SuperServer solutions optimized for NVIDIA Tesla P100 accelerators with the new Pascal GPU architecture at the GPU Technology Conference (GTC) Washington, DC 2016, October 26-27.
Supermicro’s SuperServer 4028GR-TXR(T) is optimized to support eight of the most advanced NVIDIA Tesla P100 SXM2 accelerators for maximum high bandwidth for important HPC clusters and hyperscale workloads. This server solution incorporates the latest NVIDIA NVLink GPU interconnect technology “with over 5x the bandwidth of PCI-E 3.0.”
This SuperServer solution would provide a good option for parallel computing – with eight Tesla P100 SXM2 accelerators with an independent GPU thermal zone design, which would ensure uncompromised performance and stability under the most demanding workloads with up to 170 TFLOPS FP16 of peak performance.
Density Optimized Server
Supermicro has also debuted the density optimized 4U SuperServer 4028GR-TR(T)2 system supporting up to 10 PCI-E Tesla P100 accelerators for up to 187 TFLOPS FP16 peak performance with GPU Direct RDMA support.
Supermicro’s “innovative” and GPU optimized single root complex PCI-E design would be proven to dramatically improve GPU peer-to-peer communication efficiency over QPI, “with up to 21% higher QPI throughput and 60% lower latency compared to previous generation products.”
Both 1U and 4U SuperServers support dual Intel Xeon processor E5-2600 v4/v3 product families, up to 3TB DDR4-2400MHz memory, optional dual onboard 10GBase-T ports, and redundant Titanium level (96%) digital power supplies.
“Our high-performance computing solutions enable deep learning, engineering, and scientific fields to scale out their compute clusters to accelerate their most demanding workloads and achieve fastest time-to-results with maximum performance per watt, per square foot, and per dollar,” said Charles Liang, President and CEO of Supermicro. “With our latest innovations incorporating the new NVIDIA P100 processors in a performance and density optimized 1U and 4U architectures with NVLink, our customers can accelerate their applications and innovations to address the most complex real world problems.”
With the convergence of Big Data Analytics, the latest GPU architectures, and improved Machine Learning algorithms, Deep Learning applications would require processing power of multiple GPUs that must communicate efficiently and effectively to expand the GPU network. Supermicro’s single-root GPU system would allow multiple GPUs to communicate efficiently to minimize latency and maximize throughput as measured by the NCCL P2PBandwidthTest.
“Supermicro‘s new high-density servers are optimized to fully leverage the new NVIDIA Tesla P100 accelerators to provide enterprise and HPC customers with an entirely new level of computing horsepower,” said Ian Buck, General Manager of the Accelerated Computing Group at NVIDIA. “The new SuperServers deliver superior energy-efficient performance for compute-intensive data analytics, deep learning and scientific applications while minimizing power consumption.”