PC Games Benchmarks
Comparing AMD’s RDNA 2 architecture to Nvidia’s Ampere is like comparing — as much as I bemoan the clichė — apples and oranges. The two are based on different technologies, use different manufacturing processes, and use different metrics. Looking at which has the larger number proves little.
RTX 3070 buy From Amazon:====================================
ZOTAC Gaming GeForce RTX 3070 Twin Edge OC 8GB:
ZOTAC Gaming GeForce RTX 3070 Twin Edge 8GB: https://amzn.to/36ITENZ
MSI Gaming GeForce RTX 3070 8GB OC: https://amzn.to/3f1TLYL
MSI Gaming GeForce RTX 3070 8GB: https://amzn.to/32Oculr
Gigabyte GeForce RTX 3070 Gaming OC 8G: https://amzn.to/2IJLoVn
EVGA 08G-P5-3751-KR GeForce RTX https://amzn.to/35AsdGB
Buy RTX 3080 from Amazon:====================================
MSI RTX 3080: https://amzn.to/32M12Hi
Geforce RTX 3080: https://amzn.to/32M12Hi
Zotac RTX 3080: https://amzn.to/32M12Hi
Gigabyte RTX 3080: https://amzn.to/32M12Hi
ASUS RTX 3080: https://amzn.to/32M12Hi
Buy RTX 3090 from Amazon:====================================
MSI RTX 3090: https://amzn.to/3nuoLUk
Geforce RTX 3090: https://amzn.to/3nuoLUk
Zotac RTX 3090: https://amzn.to/3nuoLUk
Gigabyte RTX 3090: https://amzn.to/3nuoLUk
ASUS RTX 3090: https://amzn.to/3nuoLUk
===========================================================
For this generation, AMD has added Ray Tracing hardware on its cards, with research and development collaboration with Microsoft. The goal for both companies has been to optimize DirectX 12 Ultimate, as evidenced by Microsoft’s full hardware support on Xbox.
RDNA 2, like the original RDNA, is a 7 nanometer chipset, but with 50% performance per-watt improvement. Even though RDNA 2 is using the same manufacturing processes as last generation, AMD says its optimized data paths on the back end to eek out that increased performance per watt. There’s also a 30% clock increase over last gen and it sticks to a 300 watt TDP.
Ampere is Nvidia’s architecture with the 3000 series of GPUs. It’s made some major changes over last generation’s Turing. Ampere chips, manufactured by Samsung, sit on an 8 nanometer process, a huge improvement over Turing’s 12 nm. It does require a higher TDP at 350 watts. But over Turing, Ampere brings second generation Ray Tracing cores and third generation Tensor Cores, delivering two times the throughput. While Nvidia claimed in its presentations that 3080 would bring double the performance over 2080 Ti, it was more like a 30-70% increase, depending on the title. Even then, those are dramatic gains.