Test Drive NVIDIA A100 GPUs with Supermicro and Microway
In partnership with Supermicro, we’re offering an opportunity to test the latest NVIDIA A100 GPUs remotely to qualified customers. Our remote resources include:
NVIDIA A100 Systems
Navion 4U GPU with NVLink Node based upon Supermicro AS-4124GO-NART:
- Up to Eight NVIDIA A100 with NVLink GPUs, fully interconnected
- Up to 320GB of GPU memory
- Nearly 5 AI PFLOPS, 156 FP64 TFLOPS of Tensor Core Performance per Server
- Two 64-core AMD EPYC Rome CPUs
- Up to 1TB DDR4 memory in each node
- NVIDIA Mellanox ConnectX-6 InfiniBand HCAs and switching
Software
Test Drive Software Configuration
- CentOS Linux and SLURM Scheduler
- Ready to run Singularity and NGC (NVIDIA GPU Cloud) container images
- Compatibility with NGC HPC Containers including: NAMD, VMD, GROMACS, LAMMPS, HOOMD-blue, Chroma, MILC, Julia, and more
- Compatibility with NGC AI Training Containers including: Caffe, Caffe2, Chainer, MXNet, Pytorch, TensorFlow, Microsoft Cognitive Toolkit (CNTK) and more
MPI & Compilers
MPI & Compiler Software
- MVAPICH2 and OpenMPI
- NVIDIA HPC SDK
- GNU GCC Compiler Collection (multiple versions, as needed) Provides C, C++ and Fortran compilers.
- AMD Optimizing C/C++ Compiler (AOCC) Provides C, C++, and Fortran compilers with optimizations for the latest AMD EPYC CPUs.