Although it is still too early to talk about the RDNA 3 architecture. On which the RX 7000 series of graphics cards will be based, as well as the integrated GPU of several of its future Ryzen processors, little by little new information and details . And taking advantage of an interview that AMD itself gave to a media outlet, we decided to collect and order this information on how the Radeon Technology Group intends to outperform NVIDIA’s RTX 40 with its new architecture. Between them the possibility that the TDP of the AMD RX 7000 is 600 W.
Officially, about RDNA 3, we know quite a bit. First of all, it will be the first GPU architecture broken down into several chips. In a move very similar to what AMD itself has already done with its Ryzen desktop CPUs. On the other hand, the use of the TSMC 5nm node for the manufacture of part of the chips that will be part of the GPU and finally, as in previous generations of the RDNA architecture, we will see a model that will have a 50% additional performance under the same consumption compared to an existing graphics card of the current RX 6000. However, an interview with Tom’s Hardware by Sam Naffzigerone of the architects of the Radeon graphics card, allowed us to discover new details hitherto unknown.
RX 7000 will hit 600W TDP with new PCIe slot
In an almost hour-long interview, Naffizger answered various questions posed to him that helped us gain some new details and get an idea of AMD’s plans for its RX 7000. key issues has been support for PCIe Gen 5 12+4 pin connectors, which we know will be used by NVIDIA, but we haven’t had official confirmation from its rival. Well, everything indicates that AMD could also join this trend.among others. Which means we could see patterns of Consumption 600 W RX 7000or at some point. Although this is something they have yet to decide, so there would be no specific product on the roadmap at this time.
The other point is power consumption, the fact that you can get more performance for less power implies that in a similar configuration in terms of number of cores, you can achieve the same speed with lower voltage. In the step from RDNA to RDNA 2, the additional 50% efficiency was between the RX 5700 XT and the RX 6700 XT. At the moment, we do not know which model of the RX 7000 will have 50% more efficiency compared to the RX 6000.
Anyway, the performance of a GPU is hard to measure by all these fixed function units out there, however, one way to compare architectures in terms of performance per watt is by floating point compute rate. In other words, the blessed TFLOPS. In the end, therefore, it all comes down to the fact that the number of operations per unit of consumption increases gradually over time.
AMD will continue without betting on Tensor Cores in the RX 7000
In addition to the unidades analogs to the Tensor Cores of the RTX of NVIDIA or the Matrix Cores of its own architectures CDNA for high performance computing, the respuesta of AMD has sido clara: el soporte para FP16 a través de Rapid Packed Math es sufficient. Therefore, these units will continue to be exclusive to the AMD Intinct for the time being and, therefore, the RX 7000 will not have Tensor Cores, at least in principle. Which means we won’t see any future versions of FidelityFX Super Resolution based on deep learning.
