NVIDIA Turing-Based GPUs
NVIDIA has announced three new products that will be using the new Turing GPU architecture. Unveiled by Jensen Huang, Founder and CEO of NVIDIA at SIGGRAPH, these cards promise to be at the forefront of the global computer graphics industry. Some basic specifications to start with:
GPU | Memory | Memory with NVLink | Ray Tracing | CUDA Cores | Tensor Cores |
---|---|---|---|---|---|
Quadro RTX 8000 | 48 GB | 96 GB | 10 GigaRays/sec | 4608 | 576 |
Quadro RTX 6000 | 24 GB | 48 GB | 10 GigaRays/sec | 4608 | 576 |
Quadro RTX 5000 | 16 GB | 32 GB | 6 GigaRays/sec | 3072 | 384 |
NVIDIA also announced the Quadro RTX Server which combines Quadro RTX GPUs along with new Quadro Infinity software to deliver one of the most powerful servers that cater to the demands of the creative professionals. The new Quadro Infinity software (Q1 2019) will allow multiple users to access a single GPU through virtual workstations and will allow end-users to provision render nodes and workstations according to their specific demands.
Pricing/Part Numbers/Availability
VCQRTX8000-PB | Quadro RTX 8000 with 48 GB memory: $10,000 (Estimated) | Q4 2018
VCQRTX6000-PB | Quadro RTX 6000 with 24 GB memory: $6,300 (Estimated) | Q4 2018
VCQRTX5000-PB | Quadro RTX 5000 with 16 GB memory: $2.300 (Estimated) | Q4 2018
Have any questions? Contact us directly.