AMD Radeon RX 6600 LE vs NVIDIA GeForce RTX 5090 D Comparison

NVIDIA GeForce RTX 5090 D and AMD Radeon RX 6600 LE occupy opposite ends of the performance spectrum, yet both hold near-identical percentile rankings among all GPUs. The RTX 5090 D sits at the 92nd percentile with an average benchmark score of 77,712, while the RX 6600 LE ranks at the 91st percentile with a score of 70,829. This odd statistical proximity masks a massive absolute gap; the data indicates the NVIDIA card is in a different performance class entirely, though the AMD card's efficiency and compact design may serve specific use cases better.

Where Each One Wins

The RTX 5090 D wins every benchmark in which both cards were tested, making its case as a pure compute and graphics powerhouse. In Geekbench OpenCL, it scores 310,674 against the RX 6600 LE's 69,229—a 348.8% advantage. In Geekbench Vulkan, the margin widens to 420.4%, with scores of 376,915 versus 72,428. These are not incremental gains; the NVIDIA card delivers quadruple the performance in API-level compute workloads. The data shows no scenario in the head-to-head results where the AMD card takes a win.

However, the RX 6600 LE wins in ways that raw scores do not capture. Its 132 W TDP and 300 W suggested PSU mean it can operate in systems where the RTX 5090 D's 575 W TDP and 950 W PSU requirement would be impractical. The AMD card's 190 mm length and 40 mm width make it far more adaptable to small-form-factor builds. For users prioritizing power draw, thermal footprint, or physical size over absolute throughput, the RX 6600 LE is the only viable choice between these two. The RTX 5090 D, with its dual-slot width and 304 mm length, demands a full-size chassis and substantial cooling infrastructure.

FAQ

Q: Which card is faster in Geekbench Vulkan?

A: The NVIDIA GeForce RTX 5090 D scores 376,915 points, which is 420.4% higher than the AMD Radeon RX 6600 LE's 72,428 points. This is the largest performance gap between the two cards in any shared test.

Q: How do their average benchmark scores compare?

A: The RTX 5090 D averages 77,712 points, while the RX 6600 LE averages 70,829 points. The NVIDIA card leads by roughly 9.7% in average score, despite being far ahead in individual tests, because the AMD card's percentile ranking is only one point lower.

Q: What is the power consumption difference?

A: The RTX 5090 D has a TDP of 575 W and requires a 950 W suggested PSU, whereas the RX 6600 LE has a TDP of 132 W and a 300 W suggested PSU. The AMD card consumes roughly one-quarter the power of the NVIDIA flagship.

Q: Which card has more memory bandwidth?

A: The RTX 5090 D provides 1.79 TB/s of bandwidth via a 512-bit bus and 32 GB of GDDR7 memory. The RX 6600 LE offers 224.0 GB/s through a 128-bit bus and 8 GB of GDDR6 memory.

Q: Do both cards support DirectX 12 Ultimate?

A: Yes, both the RTX 5090 D and RX 6600 LE support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. API compatibility is identical, despite the vast difference in hardware capabilities.

Q: How does the RTX 5090 D compare to its nearest rivals?

A: The RTX 5090 D is 1.1% faster than the AMD Radeon RX 6650M XT, 1.6% slower than the AMD Radeon RX 6850M XT, and 2.1-2.4% slower than the NVIDIA Tesla P100 PCIe variants. Its average score of 77,712 places it just above the 76,904 of the RX 6650M XT.

Head-to-Head Benchmarks

The two shared benchmark results tell a story of overwhelming NVIDIA dominance. In Geekbench OpenCL, the RTX 5090 D posts 310,674 points, dwarfing the RX 6600 LE's 69,229. The 348.8% delta means the NVIDIA card processes nearly 4.5 times more OpenCL work per unit time. This level of separation suggests the RX 6600 LE is not competing in the same segment for compute-heavy tasks like rendering, simulation, or machine learning inference.

Geekbench Vulkan results are even more lopsided. The RTX 5090 D scores 376,915, while the RX 6600 LE manages only 72,428—a 420.4% difference. Vulkan is often used in modern game engines and professional visualization tools, so this gap indicates the NVIDIA card will deliver dramatically higher frame rates and smoother performance in Vulkan-based titles. The RX 6600 LE's Vulkan score is roughly 19.2% of the NVIDIA card's output, which aligns with its lower shader count and memory bandwidth.

The RTX 5090 D's additional benchmarks, which the RX 6600 LE lacks, reinforce its positioning. In 3DMark Steel Nomad DX12, it scores 14,326. Passmark DirectX 10, 11, and 12 results are 231, 371, and 219 respectively, while DirectX 9 reaches 434. Its Passmark G3D score of 44,065 and GPU compute score of 28,396 further demonstrate multi-API strength. The RX 6600 LE has no comparable data in the pack, so direct comparison beyond OpenCL and Vulkan is impossible; the available evidence shows the NVIDIA card winning all shared tests by margins exceeding 348%.

Specification Differences

The two cards differ in nearly every measurable specification. Memory configuration is starkly different: the RTX 5090 D uses 32 GB of GDDR7 on a 512-bit bus, delivering 1.79 TB/s, while the RX 6600 LE uses 8 GB of GDDR6 on a 128-bit bus, yielding 224.0 GB/s. The NVIDIA card has eight times the memory capacity and roughly eight times the bandwidth.

Compute resources show similar disparity. The RTX 5090 D features 21,760 shading units, 680 TMUs, and 176 ROPs, versus 1,792 shading units, 112 TMUs, and 64 ROPs on the RX 6600 LE. The NVIDIA card has 170 RT cores and 680 tensor cores, while the AMD card has 28 RT cores and no tensor cores. Pixel rate is 423.6 GPixel/s versus 159.7 GPixel/s, and texture rate is 1,636.8 GTexel/s versus 279.4 GTexel/s. FP32 throughput is 104.8 TFLOPS versus 8.942 TFLOPS.

Clock speeds and power profiles diverge as well. The RTX 5090 D runs at 2017 MHz base and 2407 MHz boost, while the RX 6600 LE runs at 1626 MHz base, 2045 MHz game, and 2495 MHz boost. The AMD card boosts higher but has far fewer cores. TDP is 575 W versus 132 W, and the suggested PSU is 950 W versus 300 W. The NVIDIA card uses a 16-pin connector, the AMD card an 8-pin. Physical dimensions differ: 304 mm versus 190 mm in length, 137 mm versus 110 mm in height, and 48 mm versus 40 mm in width. Both are dual-slot, but the RTX 5090 D is substantially larger.

Architecture Differences

The architectural split is fundamental. The RTX 5090 D uses NVIDIA's GB202 chip built on Blackwell 2.0 architecture, manufactured on a 5 nm process at TSMC with 92,200 million transistors on a 750 mm² die. Transistor density is 122.9 million per mm². The RX 6600 LE uses AMD's Navi 23 chip based on RDNA 2.0, built on a 7 nm process with 11,060 million transistors on a 237 mm² die, achieving a density of 46.7 million per mm². The NVIDIA chip has over eight times the transistor count and roughly three times the die area.

Memory technology differs by generation: GDDR7 on the RTX 5090 D versus GDDR6 on the RX 6600 LE. The NVIDIA card's memory clock is listed as 1750 MHz with 28 Gbps effective, while the AMD card runs at 1750 MHz with 14 Gbps effective. FP16 processing differs in ratio: the RTX 5090 D achieves 104.8 TFLOPS at 1:1 ratio with FP32, whereas the RX 6600 LE reaches 17.88 TFLOPS at 2:1 ratio, indicating half-rate FP16 on the AMD side.

Bus interface and display outputs also diverge. The RTX 5090 D uses PCIe 5.0 x16, while the RX 6600 LE uses PCIe 4.0 x8. Display connectivity favors the NVIDIA card with 1x HDMI 2.1b and 3x DisplayPort 2.1b, versus 1x HDMI 2.1 and 3x DisplayPort 1.4a on the AMD card. The NVIDIA card's launch MSRP is 2,299 USD; the AMD card has no listed launch MSRP. Both share the same DirectX, OpenGL, and Vulkan API support, but the NVIDIA architecture includes tensor cores, which the AMD RDNA 2.0 design lacks entirely.