Uncategorized

NVIDIA’s RTX 3080/3090 30-50% Faster in Ray-Tracing Compared to AMD’s Big Navi GPUs

AMD had a ton of Big Navi benchmarks to show off at yesterday’s event, both at both 1440p and 4K, but there wasn’t a single figure indicating the ray-tracing performance of the RDNA 2 GPUs. Digging deep into the press release, it turns out that AMD did provide some initial figures, but they aren’t pretty.

Measured by AMD engineering labs 8/17/2020 on an AMD RDNA 2 based graphics card, using the Procedural Geometry sample application from Microsoft’s DXR SDK, the AMD RDNA 2 based graphics card gets up to 13.8x speedup (471 FPS) using HW based raytracing vs using the Software DXR fallback layer (34 FPS) at the same clocks. Performance may vary.

New to the AMD RDNA 2 compute unit is the implementation of a high-performance ray-tracing acceleration architecture known as the Ray Accelerator.

The Ray Accelerator is specialized hardware that handles the intersection of rays providing an order of magnitude increase in intersection performance compared to a software implementation.

AMD

The company used the Procedural Geometry Sample application from the DXR SDK to show the ray-tracing capabilities of the RDNA 2 hardware. Using the DXR fallback software layer, the GPU scored just 34 FPS, whereas turning on the hardware-level units pushed it to 471 FPS, a boost of 13.8x. Keep in mind that while the same Big Navi GPU was used for the test, we aren’t sure which one.

Although the RDNA 2 GPU is faster than all the RTX 20 series GPUs including the RTX 2080 Ti, it’s just on par with the RTX 3070 and quite a bit slower than the RTX 3080 (-33%) and the RTX 3090 (-50%).

This is likely due to two reasons. While NVIDIA uses its RTCores for both BVH acceleration and ray-triangle intersection, AMD relies on its RT units for only the latter. The shaders are leveraged for the former. Furthermore, NVIDIA’s Ampere architecture can engage its RTCores, shaders and Tensor cores simultaneously, with one’s performance not affecting that of the rest. AMD’s RT accelerators are also linked to the texture units as we saw in the Xbox Series X reveal, meaning they share a common data path and can’t run asynchronously.

However, this doesn’t necessarily mean that AMD’s Big Navi GPUs will be slower in all ray-traced games. Partner titles such as Far Cry 6 and Godfall will likely see the RDNA 2 parts perform on par or even better than the GeForce competition with ray-tracing turned on.

Areej Syed

Processors, PC gaming, and the past. I have been writing about computer hardware for over seven years with more than 5000 published articles. Started off during engineering college and haven't stopped since. Mass Effect, Dragon Age, Divinity, Torment, Baldur's Gate and so much more... Contact: areejs12@hardwaretimes.com.
Back to top button