System Setup

In our testing, we are utilizing the latest official release BIOS with the motherboard’s default settings. We have enabled both PCI-Express Resizable BAR and DOCP in the BIOS to take advantage of features such as NVIDIA GPUs and AMD Smart Access Memory on AMD GPUs. A fresh install of Windows 11 Pro with the latest updates and drivers, including the latest chipset drivers from AMD, is also being used. The “Best Performance” power profile in Windows Settings is enabled and VBS is disabled, while Game Mode is enabled.
Geekbench

 Geekbench 5 is a benchmarking tool used to measure the performance of computer hardware. It has a suite of synthetic tests that test the performance of OpenCL, CUDA, and Vulkan. Results are presented in the form of a total score, with higher scores being better. This benchmark is used to measure how well a computer can handle tasks such as graphics rendering and image manipulation.

 The Geforce RTX 3090 Ti FE has been tested against the Radeon RX 6950 XT in OpenCL compute performance, and the results show that the Geforce RTX 3090 Ti FE is 13% faster in this category. This means that the Geforce RTX 3090 Ti FE has a higher performance rate than the Radeon RX 6950 XT in regards to OpenCL compute performance.

 The GeForce RTX 3090 Ti FE video card is a powerful graphics processor, capable of running the Vulkan compute test with superior results. When compared to the Radeon RX 6950 XT, the RTX 3090 Ti FE is an impressive 88% faster in Vulkan compute performance. Additionally, its Cuda performance is also much higher than the OpenCL score, reaching an impressive 269253. The GeForce RTX 3090 Ti FE is an ideal graphics card choice for anyone who needs superior Vulkan and Cuda compute performance.

Blender Benchmark

We are utilizing the Blender Open Data Blender Benchmark to measure the rendering speed of Blender scenes, in terms of the number of samples per minute. The version of Blender employed in this evaluation is 3.2.0.

Blender Benchmark has been updated to show results in Samples per Minute per scene. The benchmark measures the performance of a GPU by evaluating the number of samples it can generate per minute. The higher the number of samples, the better the performance. In the latest benchmark, the GeForce RTX 3090 Ti FE has been found to be significantly faster than the Radeon RX 6950 XT. It produces 119% more samples per minute in the monster scene.

In the  junkshop, the GeForce RTX 3090 Ti FE outpaces the Radeon RX 6950 XT by a significant margin, boasting a 139% increase in performance.

The GeForce RTX 3090 Ti FE demonstrates a significant increase in performance compared to the Radeon RX 6950 XT, with a 125% faster speed in classroom settings.

LuxMark v4.0


In our testing, we have employed the newest version of LuxMark 4.0, featuring path trace testing. The Hall Bench test utilizes a combination of Path Tracing and Global Illumination cache, while the Food scene utilizes the brute force method of Path Tracing.

The graph illustrates the performance of the Hall Bench scene when utilizing only an OpenCL GPU. The results show that the GeForce RTX 3090 Ti FE is 32% more efficient than the Radeon RX 6950 XT in this particular scenario.

The graph shows the results of testing the performance of the Food scene using an OpenCL GPU. The results indicate that the GeForce RTX 3090 Ti FE is 15% faster than the Radeon RX 6950 XT.

AIDA64 GPGPU Benchmark

The maximum memory data transfer bandwidth was measured by reading memory. The AMD Radeon RX 6950 XT has a 6% higher memory read bandwidth in comparison to the GeForce RTX 3090 Ti FE.

The memory write bandwidth of the Radeon RX 6950 XT is 5% greater than that of the GeForce RTX 3090 Ti FE.

The memory copy performance of the Radeon RX 6950 XT is significantly superior, with a 34% increase in bandwidth compared to the GeForce RTX 3090 Ti FE.

The GeForce RTX 3090 Ti FE exhibits a 57% increase in Single-Precision FLOPS performance compared to the Radeon RX 6950 XT.

The Radeon RX 6950 XT boasts a 154% improvement in Double-Precision FLOPS performance over the GeForce RTX 3090 Ti FE.

The Radeon RX 6950 XT surpasses the GeForce RTX 3090 Ti FE by 17% in 24-bit Integer IOPS performance.

The GeForce RTX 3090 Ti FE outpaces the Radeon RX 6950 XT by 301% in 32-bit Integer IOPS performance.

The GeForce RTX 3090 Ti FE demonstrates a 193% increase in 64-bit Integer IOPS compared to the Radeon RX 6950 XT.

The GeForce RTX 3090 Ti FE outpaces the Radeon RX 6950 XT by 48% in AES-256 performance.

The GeForce RTX 3090 Ti FE has a 38% better performance in SHA-1 Hash compared to the Radeon RX 6950 XT.

The GeForce RTX 3090 Ti FE is 25% faster in single-precision Julia performance than the Radeon RX 6950 XT.

The Radeon RX 6950 XT is significantly more powerful than the GeForce RTX 3090 Ti FE, boasting a 206% faster performance.

PerformanceTEST 10

 The latest graphics card from NVIDIA, the GeForce RTX 3090 Ti FE, has been proven to be 21% faster than the Radeon RX 6950 XT in running the NBodyGravity simulation, making it an ideal choice for those looking to run the simulation with maximum speed and efficiency.

 The Radeon RX 6950 XT is 11% quicker than the GeForce RTX 3090 Ti FE when carrying out a fluid simulation with the MacCormack Method, which is used to simulate a fire reaction in a confined space.

 The GeForce RTX 3090 TI FE graphics card is far more powerful than the Radeon RX 6950 XT, boasting a 71% faster performance. It can generate a four-dimensional Julia fractal set with an epsilon of 0.001, and then render it in three dimensions and animate it for impressive visuals. This extra precision places a heavier load on the GPU, but it is more than capable of performing the task.

The Geforce RTX 3090 Ti FE graphics card is 47% faster than the Radeon RX 6950 XT. This card can be used to view a two-dimensional Mandelbrot fractal set, which is continually zoomed in on to reveal different shapes and patterns.

The Radeon RX 6950 XT GPU is 51% faster than the GeForce RTX 3090 Ti FE in the OpenCL Physics Test.

We have now pitted NVIDIA’s flagship GeForce RTX 3090 Ti FE GPU against AMD’s flagship Radeon RX 6950 XT GPU in compute, content creation, and professional workstation application benchmarks. We have also recently fully reviewed the GeForce RTX 3090 Ti FE, Overclocked it, as well as the Radeon RX 6950 XT, and compared overclock vs overclock in many games at 4K, with Ray Tracing at 1440p and with DLSS and FSR. These compute benchmarks today should provide the full scope, combined with the gaming performance as to how the GeForce RTX 3090 Ti and Radeon RX 6950 XT compete. 

In Geekbench, OpenCL and Vulkan compute performance showed the GeForce RTX 3090 Ti FE to be 13% faster in OpenCL and 88% faster in Vulkan. When benchmarking CUDA, it was even faster. In Blender, the GeForce RTX 3090 Ti FE was 119% faster in monster, 139% faster in junkshop, and 125% faster in classroom. LuxMark’s Hall Bench and Food benchmark scenes also favored the GeForce RTX 3090 Ti FE, with 32% faster and 15% faster performance, respectively. AIDA64’s GPGPU Benchmark showed the GeForce RTX 3090 Ti FE to be 57% faster in Single-Precision, while the Radeon RX 6950 XT was a large 154% faster in Double-Precision. PerformanceTEST and PassMark Advanced GPU Compute testing also favored the GeForce RTX 3090 Ti FE, with 21%, 71%, and 47% faster performance in NBodyGravity, Quaternion Julia Fractals, and Mandelbrot, respectively. Finally, SPECviewperf 2020 showed the GeForce RTX 3090 Ti FE to be 3-63% faster in 3dsmax, catia, maya, and solidworks. However, the Radeon RX 6950 XT did show some strength in energy, medical, and snx, with 8%, 35%, and 267% faster performance, respectively.
 the GeForce RTX 3090 Ti FE is the overall winner for compute workloads, with its faster OpenCL and Vulkan performance, CUDA support, OptiX for 3D rendering, and high encryption and bit-depth Integer performance. It also has 24GB of VRAM, making it great for workstation applications. The Radeon RX 6950 XT, on the other hand, excels in specific workstation applications such as energy, medical, and snx performance, and has 16GB of VRAM.

By FYIPC

Leave a Reply

Your email address will not be published. Required fields are marked *