AMD Radeon RX 6600 LE vs NVIDIA GeForce RTX 4090 D Comparison

The Verdict

The NVIDIA GeForce RTX 4090 D is the overwhelming performance leader, winning both recorded head-to-head benchmarks by margins of 240.9% and 302.5%. The AMD Radeon RX 6600 LE, while far slower, occupies a completely different performance tier, sitting at the 91st percentile of all GPUs versus the 98th percentile for the RTX 4090 D. The data shows no scenario where the RX 6600 LE beats the RTX 4090 D in raw compute, but the two cards serve entirely different markets based on their physical and power characteristics.

The RTX 4090 D is for users who need maximum compute throughput, as its average benchmark score of 178,050 places it within 4.9% of the NVIDIA A100 SXM4 40 GB and only 2.2% behind the RTX PRO 5000 Blackwell. The RX 6600 LE, with an average score of 70,829, sits just 1.4% above the AMD Radeon Pro WX 8200 and 1.3% below the RX 6650M, making it a mid-range performer that trades blows with mobile and professional workstation parts. Users choosing between these two should base the decision entirely on required performance level and system constraints, not on any expectation of competitive parity.

Architecture Differences

The two cards share almost nothing architecturally. The RTX 4090 D uses the AD102 chip built on TSMC's 5 nm process, while the RX 6600 LE uses the Navi 23 chip on TSMC's 7 nm node. Transistor counts diverge dramatically: the RTX 4090 D packs 76,300 million transistors on a 609 mm² die, yielding a transistor density of 125.3 million per mm². The RX 6600 LE has 11,060 million transistors on a 237 mm² die, with a density of 46.7 million per mm². The RTX 4090 D achieves roughly 2.7 times the transistor density of the RX 6600 LE, a direct consequence of the newer process node.

The compute architectures differ fundamentally. The RTX 4090 D implements Ada Lovelace with 14,592 shading units, 456 TMUs, 176 ROPs, 114 ray tracing cores, and 456 tensor cores. The RX 6600 LE implements RDNA 2.0 with 1,792 shading units, 112 TMUs, 64 ROPs, and 28 ray tracing cores, with no tensor core equivalent. The RTX 4090 D delivers FP32 performance of 73.54 TFLOPS and FP16 at the same 73.54 TFLOPS with a 1:1 ratio. The RX 6600 LE delivers 8.942 TFLOPS FP32 and 17.88 TFLOPS FP16 at a 2:1 ratio, meaning its FP16 throughput is double its FP32, unlike the 1:1 ratio on the NVIDIA part.

Memory subsystems are equally divergent. The RTX 4090 D runs 24 GB of GDDR6X across a 384 bit bus, producing 1.01 TB/s of bandwidth. The RX 6600 LE runs 8 GB of GDDR6 across a 128 bit bus, producing 224.0 GB/s. Clock behavior also differs: the RTX 4090 D has a 2280 MHz base and 2520 MHz boost, while the RX 6600 LE has a 1626 MHz base, a 2495 MHz boost, and a 2045 MHz game clock. The memory clocks are 1313 MHz with 21 Gbps effective on the RTX 4090 D versus 1750 MHz with 14 Gbps effective on the RX 6600 LE.

Power and physical specifications separate the cards sharply. The RTX 4090 D carries a 425 W TDP, requires an 800 W suggested PSU, uses a triple-slot cooler, and needs a single 16-pin connector. The RX 6600 LE has a 132 W TDP, a 300 W suggested PSU, a dual-slot cooler, and a single 8-pin connector. The RTX 4090 D measures 304 mm long, 137 mm high, and 61 mm wide. The RX 6600 LE measures 190 mm long, 110 mm high, and 40 mm wide. Both use PCIe 4.0, but the RTX 4090 D uses x16 while the RX 6600 LE uses x8. Display outputs are identical at 1x HDMI 2.1 and 3x DisplayPort 1.4a, and both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Where Each One Wins

The RTX 4090 D wins every recorded benchmark, so the more useful question is where each card fits in a system. The RTX 4090 D wins in raw compute, memory bandwidth, and pixel and texture throughput. Its pixel rate of 443.5 GPixel/s is roughly 2.8 times the RX 6600 LE's 159.7 GPixel/s. Its texture rate of 1,149.1 GTexel/s is approximately 4.1 times the RX 6600 LE's 279.4 GTexel/s. The RTX 4090 D also wins decisively in FP32 compute, offering about 8.2 times the RX 6600 LE's throughput.

The RX 6600 LE wins in power efficiency, physical size, and system compatibility. Its 132 W TDP is less than one third of the RTX 4090 D's 425 W, and it requires a 300 W PSU versus 800 W. The dual-slot, 190 mm length design fits in substantially smaller chassis than the triple-slot, 304 mm RTX 4090 D. The RX 6600 LE also uses a single 8-pin connector instead of the 16-pin connector on the RTX 4090 D, which may matter in older or more modest power supplies. Its production status is Active, while the RTX 4090 D is listed as End-of-life, so availability favors the AMD part.

For users who need tensor core acceleration, the RTX 4090 D has 456 tensor cores while the RX 6600 LE has none. For users who need only basic rasterization and moderate compute, the RX 6600 LE's lower power draw and smaller footprint make it the more practical choice despite its large performance deficit.

FAQ

Q: Which GPU is faster in the recorded benchmarks?

A: The NVIDIA GeForce RTX 4090 D wins both head-to-head tests. It scores 278,621 in Geekbench OpenCL versus 69,229 for the RX 6600 LE, a 302.5% advantage, and 246,941 in Geekbench Vulkan versus 72,428, a 240.9% advantage.

Q: How do these GPUs rank against all other GPUs?

A: The RTX 4090 D sits at the 98th percentile of all GPUs, while the RX 6600 LE sits at the 91st percentile. Their average benchmark scores are 178,050 and 70,829 respectively.

Q: What are the closest rivals to each card?

A: The RTX 4090 D's nearest rival is the NVIDIA RTX PRO 5000 Blackwell with an average score of 182,109, which is 2.2% higher. The RX 6600 LE's nearest rival is the NVIDIA RTX A3000 Mobile with an average score of 70,140, which is 1% lower.

Q: How much memory does each card have?

A: The RTX 4090 D has 24 GB of GDDR6X on a 384 bit bus with 1.01 TB/s bandwidth. The RX 6600 LE has 8 GB of GDDR6 on a 128 bit bus with 224.0 GB/s bandwidth.

Q: What power supply do these cards require?

A: The RTX 4090 D has a 425 W TDP and requires an 800 W suggested PSU with a 1x 16-pin connector. The RX 6600 LE has a 132 W TDP and requires a 300 W suggested PSU with a 1x 8-pin connector.

Q: Are these cards still in production?

A: No. The RTX 4090 D is marked as End-of-life. Yes. The RX 6600 LE is marked as Active.

Head-to-Head Benchmarks

The head-to-head data contains only two benchmark tests, and the RTX 4090 D wins both by enormous margins. In Geekbench OpenCL, the RTX 4090 D scores 278,621 against 69,229 for the RX 6600 LE, a 302.5% delta. This means the RTX 4090 D delivers roughly four times the OpenCL compute performance of the RX 6600 LE. In Geekbench Vulkan, the RTX 4090 D scores 246,941 against 72,428, a 240.9% delta, meaning the NVIDIA card is about 3.4 times faster in that API.

These margins are consistent with the underlying hardware differences. The RTX 4090 D has 14,592 shading units versus 1,792, a ratio of approximately 8.1 to 1. Its FP32 throughput of 73.54 TFLOPS versus 8.942 TFLOPS is a ratio of approximately 8.2 to 1. The memory bandwidth gap is a ratio of approximately 4.5 to 1, at 1.01 TB/s versus 224.0 GB/s. The benchmark deltas, while large, are smaller than the raw shading unit and FP32 ratios, which suggests that the RX 6600 LE's smaller memory footprint and narrower bus do not fully constrain its compute in these specific tests, or that the RTX 4090 D is not scaling perfectly to its theoretical peak in these workloads.

The OpenCL margin is larger than the Vulkan margin by about 61.6 percentage points. This could indicate that the RTX 4090 D's advantage is more pronounced in OpenCL workloads, or that the RX 6600 LE's Vulkan implementation is relatively more efficient. Either way, both tests place the two cards in entirely separate performance classes. The RTX 4090 D's nearest rivals are all NVIDIA professional and data center parts (RTX PRO 5000 Blackwell at 182,109, A100 SXM4 80 GB at 183,725, RTX 5000 Ada at 184,664, A100 SXM4 40 GB at 187,147), and it trails all of them by margins between 2.2% and 4.9%. The RX 6600 LE's nearest rivals are a mix of mobile and workstation GPUs (RTX A3000 Mobile at 70,140, Quadro P6000 at 69,986, RX 6650M at 71,768, Radeon Pro WX 8200 at 69,870), and it leads three of them by 1% to 1.4% while trailing the RX 6650M by 1.3%. The data shows the RTX 4090 D competing at the top of the stack while the RX 6600 LE sits in a crowded mid-range cluster.

Specification Differences

The following specifications differ between the two cards:

• Process node: 5 nm (RTX 4090 D) versus 7 nm (RX 6600 LE)
• Transistors: 76,300 million versus 11,060 million
• Die size: 609 mm² versus 237 mm²
• Transistor density: 125.3M / mm² versus 46.7M / mm²
• Base clock: 2280 MHz versus 1626 MHz
• Boost clock: 2520 MHz versus 2495 MHz
• Game clock: not specified versus 2045 MHz
• Memory clock: 1313 MHz, 21 Gbps effective versus 1750 MHz, 14 Gbps effective
• Memory size: 24 GB versus 8 GB
• Memory type: GDDR6X versus GDDR6
• Memory bus: 384 bit versus 128 bit
• Memory bandwidth: 1.01 TB/s versus 224.0 GB/s
• Shading units: 14,592 versus 1,792
• TMUs: 456 versus 112
• ROPs: 176 versus 64
• Ray tracing cores: 114 versus 28
• Tensor cores: 456 versus none
• Pixel rate: 443.5 GPixel/s versus 159.7 GPixel/s
• Texture rate: 1,149.1 GTexel/s versus 279.4 GTexel/s
• FP32 performance: 73.54 TFLOPS versus 8.942 TFLOPS
• FP16 performance: 73.54 TFLOPS (1:1) versus 17.88 TFLOPS (2:1)
• TDP: 425 W versus 132 W
• Slot width: Triple-slot versus Dual-slot
• Power connectors: 1x 16-pin versus 1x 8-pin
• Suggested PSU: 800 W versus 300 W
• Bus interface: PCIe 4.0 x16 versus PCIe 4.0 x8
• Length: 304 mm versus 190 mm
• Height: 137 mm versus 110 mm
• Width: 61 mm versus 40 mm
• Production status: End-of-life versus Active
• Release date: 2023-12-27 versus 2023-12-07
• Launch MSRP: 1,599 USD versus not specified
• Predecessor: GeForce 30 versus Navi
• Successor: GeForce 50 versus Navi III

Identical specifications include the display outputs (1x HDMI 2.1, 3x DisplayPort 1.4a), API support (DirectX 12 Ultimate, OpenGL 4.6, Vulkan 1.4), and manufacturer (TSMC foundry for both). The RTX 4090 D has an average benchmark score of 178,050 with a 98th percentile rank, while the RX 6600 LE has an average of 70,829 with a 91st percentile rank. The RTX 4090 D records two wins in head-to-head benchmarks; the RX 6600 LE records zero.