Η Nvidia has announced its new Titan RTX graphics card, Turing architecture, the company's most powerful desktop graphics card, suitable for research in artificial intelligence, science and a host of business applications. The computing power of the card reaches 130 teraflops, while in raytracing it achieves 11 GigaRays. However, these performances do not come in handy, with the card costing by Nvidia at $ 2499.
The new Titan uses the TU102 GPU found on the RTX 2080 Ti graphics card as well, but with the GPU on the RTX 2080 Ti being a slightly more streamlined version than the Titan. So Titan has more CUDA and Tensor cores, while the memory bus is now 384bit. The Turbo frequency of the GPU is also higher, while the memory of the card reaches 24GB, a quantity more than double that of the RTX 2080 Ti. Another significant difference is found in the performance of the Tensor cores, which are also more than double what the RTX 2080 Ti achieves. This performance brings the Titan to the level of the Quadro RTX 6000, making it the ideal choice for researchers and programmers working on artificial intelligence, with neural networks as well as large data sets.
Below you can see a comparison of the technical features of the new Titan RTX, with the Titan V, RTX 2080 Ti Founders Edition and Tesla V100 (PCIe).
| NVIDIA Compute Accelerator Specification Comparison | ||||||
| Titan RTX | Titan V | RTX 2080 Ti Founders Edition |
Tesla V100 (PCIe) |
|||
| CUDA Cores | 4608 | 5120 | 4352 | 5120 | ||
| Tensor Cores | 576 | 640 | 544 | 640 | ||
| Core Clock | 1350MHz | 1200MHz | 1350MHz | ? | ||
| Boost Clock | 1770MHz | 1455MHz | 1635MHz | 1370MHz | ||
| Memory Clock | 14Gbps GDDR6 | 1.7Gbps HBM2 | 14Gbps GDDR6 | 1.75Gbps HBM2 | ||
| Memory Bus Width | 384-bit | 3072-bit | 352-bit | 4096-bit | ||
| Memory Bandwidth | 672GB / sec | 653GB / sec | 616GB / sec | 900GB / sec | ||
| VRAM | 24GB | 12GB | 11GB | 16GB | ||
| L2 Cache | 6MB | 4.5MB | 5.5MB | 6MB | ||
| Single Precision | 16.3 TFLOPS | 13.8 TFLOPS | 14.2 TFLOPS | 14 TFLOPS | ||
| Double Precision | 0.51 TFLOPS | 6.9 TFLOPS | 0.44 TFLOPS | 7 TFLOPS | ||
| Performance Tensor (FP16 w / FP32 Acc) |
130 TFLOPS | 110 TFLOPS | 57 TFLOPS | 112 TFLOPS | ||
| GPU | TU102 (754mm2) |
GV100 (815mm2) |
TU102 (754mm2) |
GV100 (815mm2) |
||
| Transistor Count | 18.6B | 21.1B | 18.6B | 21.1B | ||
| TDP | 280w | 250w | 260w | 250w | ||
| Form Factor | PCIe | PCIe | PCIe | PCIe | ||
| cooling | Active | Active | Active | passive | ||
| Manufacturing Process | TSMC 12nm FFN | TSMC 12nm FFN | TSMC 12nm FFN | TSMC 12nm FFN | ||
| Brand | Turing | Volta | Turing | Volta | ||
| Launch Date | 12/2018 | 12/07/2017 | 09/20/2018 | Q3'17 | ||
| Price | $2499 | $2999 | $1199 | ~ $ 10000 | ||
In terms of card design, it follows the design we saw with the new RTX cards, with a two-fan cooling system, but in gold color to stand out from the GeForce series cards. It has four video outputs, three DisplayPorts 1.4 and one HDMI 2.0b, as well as another USB C, which supports both DisplayPort signal and VirtualLink for virtual reality headsets.
The card will be available through the Nvidia website within the month.
[the_ad_group id = ”966 ″]
