|
DGX-2, Now Available for Purchase- June 2018 Newsletter |
Microway is pleased to announce its new status as an NVIDIA DGX-2 Advanced Technical Program Partner: we are one of the few partners authorized to sell the DGX-2 solution. Learn about this and other new developments for GPUs in this newsletter. |
|
DGX-2, Now from NVIDIA & Microway
DGX-2 is a groundbreaking new AI appliance to solve the world's toughest AI problems.
- 16 Tesla V100 GPUs, 512GB unified GPU memory for immense datasets!
- NVSwitch interconnect, with full non-blocking BW
- World record DL Training performance
- All in an easy to use DGX appliance
We encourage you to reserve a delivery position for one of these amazing systems before the backorder grows further.
|
|
|
|
GTC 2018: Recap and Key Sessions |
GTC 2018 was an action packed week earlier this spring. The show brought us new hardware, the latest developments in GPU programming, and outstanding technical sessions. Whether you attended or missed this year, don't miss out on the resources. |
|
Missed GTC in March? Recap it with Microway
If you missed the keynote/announcements or the show floor, we have you covered. GTC week is packed, so we built a Photo Journal with the announcements & highlights.
|
|
|
|
GTC Key Session: DGX-2 Intro |
Surprise! NVIDIA has a new DGX appliance to show. This session outlined the basics of DGX-2 and it's advantages for AI. |
|
|
|
GTC Key Session: Inside DGX-2 |
NVIDIA experts take you deep inside the architecture, switching, and hardware design choices of DGX-2. |
|
|
|
GTC Key Session: HPC in Containers |
Containers are making HPC & AI easier for all. See the toolchain to roll custom containers atop NVIDIA's CUDA versions. |
|
|
|
|
|
GTC Announcement: PGI 18.4 w/ OpenACC 2.6, Now Available |
The latest PGI compiler release is available in a free community edition supporting OpenACC 2.6, & OpenMP 4.5. New CPU support for: POWER9, AMD Zen, Xeon Skylake |
|
|
|
GTC Key Session: GPU Porting VASP w/OpenACC |
If you've ever had to get VASP running, you know it can be a very complex code. Learn how the first ever GPU support was added with OpenACC flags, non-destructively. |
|
|
|
GTC Key Session: MPI Programming for Multiple-GPUs |
Walk through common methods of job dispatching, MPI ranks, and key techniques for GPU affinity that deliver the ultimate performance for multi-GPU systems. |
|
|
|
|
|
|
|
New Blog Post: Tesla V100 Price Analysis
NVIDIA has released the Tesla V100 GPU, but how do you know when to select it? Our new blog post dives into the intricacies of Tesla GPU pricing and performance. You'll learn:
- About the pricing and performance gains of 16GB and 32GB Tesla V100 GPUs
- How they compare to Tesla P100 for HPC and Deep Learning/AI
- Recommendations of when to consider (and pick) Tesla P100, V100
- Why price/FLOPS (simple price performance) is an overly simplistic measure of performance
|
|
|
|
|