Dihuni OptiReady Supermicro 4028GR-TVRT-8-2 8xNVIDIA Tesla V100 32GB SXM2 NVLINK GPU 2S Xeon 2.4GHz CPU, 256GB Mem, 960GB SSD 10G Deep Learning Server

NVLink Performance

Systems with multiple GPUs and CPUs are becoming common in a variety of industries as developers rely on more parallelism in applications like AI computing. These include 4-GPU and 8-GPU system configurations using PCIe system interconnect to solve very large, complex problems. But PCIe bandwidth is increasingly becoming the bottleneck at the multi-GPU system level, driving the need for a faster and more scalable multiprocessor interconnect. NVIDIA^® NVLink^™ technology addresses this interconnect issue by providing higher bandwidth, more links, and improved scalability for multi-GPU and multi-GPU/CPU system configurations. A single NVIDIA Tesla^® V100 GPU supports up to six NVLink connections and total bandwidth of 300 GB/sec—10X the bandwidth of PCIe Gen 3.

Supermicro Servers Featuring Intel E5-2600 v4 Xeon® Processors

Built on 14 nm process technology, the Intel® Xeon® processor E5-2600 v4 family offers up to 22 cores/44 threads per socket and 55 MB last-level cache (LLC) per socket for increased performance, as well as Intel® Transactional Synchronization Extensions (Intel® TSX) for increased parallel workload performance. You can dynamically manage shared resources efficiently and increase resource utilization with Intel® Resource Director Technology (Intel® RDT) offering cache monitoring and allocation technology, code and data prioritization, and memory bandwidth monitoring. Accelerated cryptographic performance enables encrypted data to move fast over secure connections, plus improved security keys help safeguard network access and deepen platform protection.

Technical Specifications