The green giant has launched the RTX A4500, a professional graphics card with which NVIDIA has strengthened its high-end range, and which is located just halfway between the RTX A4000 and the RTX A5000, two models that had been available in the market for some time. market, and what are separated by an important space at the level of benefits. By this I mean that the RTX A5000 is not only much more powerful, but also has much more graphics memory than that.
The RTX A4500 perfectly covers that huge gap between the two models, and does so with fairly balanced specifications that, in fact, fit what we expected after seeing the first leaks that had been produced. This graphics card uses Ampere architecture, mounts a GA102 graphics core, has 7,168 shaders, 224 texture units, 96 raster units, 224 tensor cores and 56 ray tracing cores.
From its specifications we can see that the RTX A4500 is prepared to offer high performance in professional applications, in any situation and under any condition, thanks not only to its high power, but also thanks to the presence of hardware specialized in artificial intelligence, its tensor cores, and in ray tracing, its RT cores. To all the above we must add a generous amount of graphics memory, since the RTX A4500 comes equipped with 20 GB of GDDR6, which works on a 320-bit bus to achieve a bandwidth of 320 GB / s.
In FP16 and FP32, the RTX A4500 offers a total power of 23.65 TFLOPs, while in FP64 it is capable of reaching 739.2 GFLOPs. Making a general comparison, its gross power in a broad sense is a bit above the GeForce RTX 3070 Ti.
As we can see in the cover image, the RTX A4500 uses a discreet, cooling system based on a turbine type fan, which means that it can work without problems even in assemblies where the available space is very limited. We do not have details about the sale price, but we are talking about a high-end professional solution, so its price will be according to that range.
