HPC

AMD EPYC Servers with GPU Acceleration for AI, HPC, and Datacenter Workloads

June 10, 2020
4 min read
275432-EPYC-7002-background-1920x450.png

AMD EPYC Servers Set the Standard for Performance

The AMD EPYC 7002 Series Processors have taken the compute world by storm. The 7nm techology, PCIE 4.0, encrypted memory, and other innovations allow researchers to tackle demanding HPC workloads with superior performance, I/O, bandwidth, and industry-leading interconnect speeds. Adding GPU acceleration with an AMD EPYC Processor allows for even more groundbreaking research. With this in mind Exxact has developed a couple GPU Server options featuring AMD EPYC Processors for modern HPC and AI workloads.

AMD EPYC SERVER SALE

Exxact TensorEX 2U AMD EPYC 4x GPU Server

This AMD EPYC based GPU server is a well rounded choice for many different HPC and Artificial Intelligence development tasks. Featuring 2x AMD EPYC 7002 Processors and up to 4x GPUs, this is a highly scalable solution built for continuous use in many popular applications.

  • Great for applications needing a balance of CPU/GPU compute such as molecular dynamics.
  • Standard 3 year warranty included.

AMD EPYC GPU SERVER

Exxact TensorEX 4U AMD EPYC 8x GPU Server

This GPU Server is made to handle heavy-duty heterogeneous workloads. Accommodating up to 8x GPUs this is a popular choice for AI developers needing GPU Acceleration for neural network training.

  • Dense GPU compute capability suitable for deep learning and HPC tasks.
  • Large stand-alone storage capacity for big datasets.
  • Standard 3 year warranty included.

AMD EPYC GPU SERVERS

Let’s check out HPC Performance for AMD EPYC CPUs

AMD EPYC Benchmarks
2) Cloud and Virtualization Performance is measured with VMmark® 3.1, Integer Performance is measured with SPECrate®2017_int_rate_base, Floating-Point Performance is measured with SPECrate®2017_fp_rate_base, JAVA® Performance is measured with SPECjbb®2015-MultiJVM max-jOPS, HPC Performance is measured with ANSYS FLUENT®

Not sure what GPU is right for your system?

Luckily, we’ve got you covered. Check this blog for details on selecting the right GPU for the job.

Contact us - AMD EPYC SERVER

Topics

275432-EPYC-7002-background-1920x450.png
HPC

AMD EPYC Servers with GPU Acceleration for AI, HPC, and Datacenter Workloads

June 10, 20204 min read

AMD EPYC Servers Set the Standard for Performance

The AMD EPYC 7002 Series Processors have taken the compute world by storm. The 7nm techology, PCIE 4.0, encrypted memory, and other innovations allow researchers to tackle demanding HPC workloads with superior performance, I/O, bandwidth, and industry-leading interconnect speeds. Adding GPU acceleration with an AMD EPYC Processor allows for even more groundbreaking research. With this in mind Exxact has developed a couple GPU Server options featuring AMD EPYC Processors for modern HPC and AI workloads.

AMD EPYC SERVER SALE

Exxact TensorEX 2U AMD EPYC 4x GPU Server

This AMD EPYC based GPU server is a well rounded choice for many different HPC and Artificial Intelligence development tasks. Featuring 2x AMD EPYC 7002 Processors and up to 4x GPUs, this is a highly scalable solution built for continuous use in many popular applications.

  • Great for applications needing a balance of CPU/GPU compute such as molecular dynamics.
  • Standard 3 year warranty included.

AMD EPYC GPU SERVER

Exxact TensorEX 4U AMD EPYC 8x GPU Server

This GPU Server is made to handle heavy-duty heterogeneous workloads. Accommodating up to 8x GPUs this is a popular choice for AI developers needing GPU Acceleration for neural network training.

  • Dense GPU compute capability suitable for deep learning and HPC tasks.
  • Large stand-alone storage capacity for big datasets.
  • Standard 3 year warranty included.

AMD EPYC GPU SERVERS

Let’s check out HPC Performance for AMD EPYC CPUs

AMD EPYC Benchmarks
2) Cloud and Virtualization Performance is measured with VMmark® 3.1, Integer Performance is measured with SPECrate®2017_int_rate_base, Floating-Point Performance is measured with SPECrate®2017_fp_rate_base, JAVA® Performance is measured with SPECjbb®2015-MultiJVM max-jOPS, HPC Performance is measured with ANSYS FLUENT®

Not sure what GPU is right for your system?

Luckily, we’ve got you covered. Check this blog for details on selecting the right GPU for the job.

Contact us - AMD EPYC SERVER

Topics