"AMPERE" GPUs

Nvidia is Amping Up its fans for its new GPUs (probably the RTX 3000 series with 3080Ti to be the flagship) which are due to release this year based on the new "AMPERE" architecture. The recent GPUs of Nvidia which include the RTX 2000 Series are based on the "TURING" architecture. Ampere architecture is expected to bring a great performance leap as compared to its predecessor with the transition to 7nm node. 

What we Know!


There are high expectations from the AMPERE architecture but all are speculations and rumors till now. There are no official details about performance leap and pricing of these GPUs. What we do know is that the new 7nm node should allow the GPUs to be more efficient and provides room for higher clock speeds and GPU cores. This should also be true for Ray Tracing and Tesnor Cores introduced in the current GPU lineup of Nvidia. Ray Tracing is still not a mandatory and common feature and many games do not support this technology. The improvement in Ray Tracing performance can urge more game developers to go for this amazing technology.

China Times reports that the flagship GPU of Nvidia AMPERE lineup GA100 (3080Ti or whatever they call it we aren't sure yet) could boast up to 8192 GPU Cores, 256 Ray Tracing Cores, 1024 Tensor Cores and a boost clock up to 1.75 GHz with 16 GBPS VRAM Speed. As a comparison, the current flagship RTX 2080Ti has 4352 GPU Cores, 68 Ray Tracing Cores, 544 Tensor Cores and a boost clock up to 1.54 GHz with 14 GBPS GDDR6 VRAM Speed. That's a hell lot of performance bump and with 7nm node, the card is supposed to be much more efficient in handling this serious power.

The specs of potential RTX 3080 (GA103) in the same report looks similar to RTX 2080Ti that we mentioned earlier except for 144 Ray Tracing cores and 16 GBPS of VRAM Speed. Other specs which are leaked are more or less the same.

In conclusion, if these leaks are true (or at least to a greater extent) the flagship RTX 2080Ti can be outmatched by potential RTX 3080 or at least go toe to toe. This is an incredible performance jump but nothing is assured yet!

Is AMD's "Nvidia Killer" in problem?



AMD is battling in two fronts (Microprocessors and GPUs) with Intel and Nvidia. AMD ryzing from its grave with its Ryzen processors do give Intel a lethal damage eating their market share with rapid pace. However, the situation is not quite similar in their battle with Nvidia. Team Green, for many years has remained the leader in GPU technology and production.

With AMD radically revamping its products on their GPU front too with introduction of Radeon RX 5000 series based on TSMC's 7nm node, AMD is still not able to catch up to Nvidia. The performance gap seems a lot wider such that using 7nm node can not overcame it. AMD currently has no offerings to go head to head with RTX 2080Ti, the flagship. The RX 5700 XT can only sightly beat the RTX 2070 which sits at third place in Nvidia's lineup (and we are not counting the RTX Super Series here).

In attempt to beat Nvidia in its own game, RDNA 2 or Navi 2x which was also rumored as "BIG NAVI" earlier seems to bring radical performance jump. It also brings Ray Tracing support to AMD GPUs. We are not sure of its specifications yet but what we do know through leaks is that the flagship could include 5120 GPU cores with 50%  better performance per watt as compared to previous generation GPUs. Its look like it can beat RTX 2080Ti but the problem is that RDNA 2 GPUs will go head to head with AMPERE GPUs not TURING.

If circulating leaks of AMPERE GPUs become a reality, AMD has to up its game if they had to beat Nvidia for sure.

What do you think? Tell us in the comments!