The latest NVIDIA technology
As an NVIDIA ELITE partner, we have the latest NVIDIA technologies in the field of computing accelerators and AI available within the NVIDIA Test Drive program. You can try the AI supercomputer DGX station or the most powerful cards A100, V100, A40, T4.
NVIDIA DGX Station A100
AI supercomputer as a workstation.
- 1× AMD Rome 7742, 2.25 GHz (64 cores)
- 512 GB DDR4 RAM
- 4× NVIDIA A100, 40 GB SXM4 NVLink
- 4 GB GPU Memory, 4x Mini DisplayPort
- 1,92 TB M.2 NVME SSD (OS)
- 7,68 TB U.2 NVME SSD (sata)
- dual-Port 10 Gb/s Ethernet
- tower form factor
- internal water cooling CPU a GPU
- NVIDIA SW optimized for HPC an AI
We rent separate cards and a built-in server.
- NVIDIA A100 40 GB PCIe — the most powerful current GP GPU accelerator for HPC (FP64 calculations) and AI (tensor) calculations.
- NVIDIA A40 48 GB PCIe — great performance for FP32, virtualization of the most powerful workstations.
- NVIDIA V100 32 GB PCIe — the predecessor of NVIDIA A100, universal card for HPC, AI and virtualization.
- NVIDIA T4 — the most efficient card for AI inference in terms of price and consumption/power, only 70W.
In cooperation with Intel and VMware, we have prepared a demo environment of the latest software-defined data centre technologies. The latest versions of VMware software (eg VMware Cloud Foundation, VMware Horizon and others) are installed on four Intel DataCenter Building Block servers powered by Intel technology (Intel Platinum processors, Intel Optane Persistent Memory, Intel Optane SSD, Intel networking).
As a standard, we have Intel and AMD processors of the latest generations available for testing; Lenovo, HPE, Intel, or Supermicro servers; Kingston memory, Intel SSDs, Kingston, Western Digital; or IBM, NetApp, and Infortrend disk arrays, and in cooperation with our manufacturers and distributors we can arrange testing of other technologies.