Yesterday, thanks to the latest Nvidia graphics drivers, it was announced that the Nvidia CMP 30HX and Nvidia CMP 40HX mining graphics were using the Turing architecture at a 12nm manufacturing process from TSMC.
Specifically, the CMP 30HX makes use of a custom TU106 graphics chip (used in the GeForce RTX 2060 / SUPER and GeForce RTX 2070 Mobile) to a 12nm manufacturing process from TMSC, while the Nvidia CMP 30HX uses TU116 silicon. also custom (GeForce GTX 1650 SUPER / GTX 1660 / GTX 1660 SUPER).
In summary, these mining cards do not imply a decrease in terms of the production of graphics cards for next-generation gamers.
Today the surprise comes with the Nvidia CMP 50 HX, which is also a graphics card based on the Turing architecture. Specifically, this model makes use of the TU102 graphics chip, the same used by the Nvidia GeForce RTX 2080 Ti and some previous generation Nvidia Quadro graphics.
In essence, the only Ampere chip manufactured by Samsung Foundry that we see in this product line is the Nvidia CMP 90 HX, and it is not surprising, since this model offers almost double the hash rate mining Ethereum compared to the previous model.
Now we just have to wait to know the graphics chip that it integrates, although we always talk about custom versions, which may involve isolating a certain number of CUDA Cores, Tensor Cores, RT Cores, TMUs, ROPs, etc., which are not functional, that is, isolate a problem in the graphics chip to get a product usable for another use since it would not work as a gaming graphics.
| 30HX | 40HX | 50HX | 90HX | |
| Ethereum Hash Rate | 26 MH / s | 36 MH / s | 45 MH / s | 86 MH / s |
| Energy consumption | 125 W | 185 W | 250 W | 320 W |
| PCIe connectors | 1x 8 pin | 1x 8 pin | 2x 8 pins | 2x 8 pins |
| Memory | 6GB | 8GB | 10GB | 10GB |
| Availability | Q1 | Q1 | Q2 | Q2 |
0 Comments