AMD Radeon PRO V620 vs NVIDIA GeForce RTX 4090 D Comparison

# NVIDIA GeForce RTX 4090 D vs AMD Radeon PRO V620

The data presents a stark contrast between two professional-grade GPUs from different architectural generations. The NVIDIA GeForce RTX 4090 D, built on the Ada Lovelace architecture with a 5 nm process, faces the AMD Radeon PRO V620, which uses the older RDNA 2.0 architecture on a 7 nm node. While both target high-end workstation workloads, their benchmark scores reveal a substantial performance gap, with the RTX 4090 D winning both head-to-head tests by wide margins. The average benchmark score for the RTX 4090 D sits at 178,050, placing it in the 98th percentile of all GPUs, while the Radeon PRO V620 averages 136,472, which still places it in the 96th percentile. This positioning suggests both are elite performers, but the NVIDIA card operates in a different performance tier entirely.

Head-to-Head Benchmarks

The two GPUs were tested in two shared benchmarks: Geekbench OpenCL and Geekbench Vulkan. In both cases, the NVIDIA GeForce RTX 4090 D emerges as the clear winner, and the margins are substantial. In the Geekbench OpenCL test, the RTX 4090 D scores 278,621 points against the Radeon PRO V620's 128,580 points. That represents a 116.7% delta, meaning the NVIDIA card more than doubles the AMD card's performance in this compute-oriented benchmark. This is not a marginal victory; it is a decisive generational leap. The OpenCL result is particularly telling because it exercises raw compute throughput, where the RTX 4090 D's 73.54 TFLOPS FP32 performance and 14,592 shading units vastly outnumber the Radeon PRO V620's 20.28 TFLOPS and 4,608 shading units.

In the Geekbench Vulkan test, the gap narrows slightly but remains overwhelming. The RTX 4090 D scores 246,941, while the Radeon PRO V620 manages 144,364. The delta here is 71.1%, still a commanding lead for NVIDIA. Vulkan results often reflect driver optimization and architectural efficiency in graphics workloads, and the data indicates the Ada Lovelace architecture handles these tasks with significantly more headroom. The RTX 4090 D's 114 ray tracing cores and 456 tensor cores provide hardware acceleration that the Radeon PRO V620, with its 72 ray accelerators and no tensor core equivalent, cannot match in these tests.

The wins tally is 2-0 in favor of the RTX 4090 D, and every available data point reinforces this conclusion. It is importantly the Radeon PRO V620 has no corresponding benchmark in the 3DMark Steel Nomad DX12 test, which the RTX 4090 D completes with a score of 8,587. This absence limits direct comparison in that specific workload, but the existing data provides no evidence that the AMD card could close the gap in DX12 scenarios.

Where Each One Wins

Looking strictly at the benchmark data, the NVIDIA GeForce RTX 4090 D wins every category in which both cards were tested. However, the Radeon PRO V620 is not without its own strengths when considering the broader specification landscape. The AMD card offers 32 GB of GDDR6 memory on a 256-bit bus, compared to the RTX 4090 D's 24 GB of GDDR6X on a 384-bit bus. While the NVIDIA card delivers higher memory bandwidth at 1.01 TB/s versus 512.0 GB/s, the AMD card's larger capacity could be advantageous for workloads that require massive datasets resident in VRAM, such as certain scientific simulations or large language model inference. The data does not include a benchmark that isolates memory capacity benefits, but the specification difference is real and could matter in niche scenarios.

The Radeon PRO V620 also draws less power, with a 300 W TDP against the RTX 4090 D's 425 W, and it uses a dual-slot cooler versus the RTX 4090 D's triple-slot design. For dense multi-GPU server configurations where power density and physical space are constrained, the AMD card's lower profile and power requirement could be decisive. The card also features no display outputs, indicating it is designed for headless compute or virtualization environments, whereas the RTX 4090 D includes HDMI 2.1 and DisplayPort 1.4a outputs for direct display attachment.

For the RTX 4090 D, the wins are not just in raw performance but also in architectural features. Its 5 nm process node from TSMC allows for 76,300 million transistors packed into a 609 mm² die, yielding a transistor density of 125.3M per mm². The Radeon PRO V620, by contrast, uses 26,800 million transistors on a 520 mm² die at 7 nm, achieving only 51.5M transistors per mm². This density advantage translates directly into the performance margins observed in the benchmarks.

The Verdict

The data is unambiguous: the NVIDIA GeForce RTX 4090 D is the superior performer in every benchmark where both cards were tested. Its 116.7% lead in Geekbench OpenCL and 71.1% lead in Geekbench Vulkan are massive, and its average benchmark score of 178,050 places it 30.5% above the Radeon PRO V620's 136,472 average. The RTX 4090 D also sits in a higher performance tier relative to all GPUs, at the 98th percentile versus the 96th percentile for the AMD card.

For users who prioritize raw compute performance, ray tracing capability, or the highest possible frame rates in graphics-intensive professional applications, the RTX 4090 D is the clear choice based on the data. Its 73.54 TFLOPS FP32 performance and 1,149.1 GTexel/s texture rate dwarf the Radeon PRO V620's 20.28 TFLOPS and 633.6 GTexel/s. The NVIDIA card also supports a broader range of API features, including DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, matching the AMD card's API support but with significantly more horsepower behind it.

The Radeon PRO V620, however, is not without a rationale for selection. Its 32 GB memory capacity exceeds the RTX 4090 D's 24 GB, and its lower 300 W TDP and dual-slot design make it more suitable for power-constrained or space-constrained environments. The AMD card's 96th percentile ranking still places it among the top GPUs worldwide, and its nearest rivals — the AMD Radeon Pro W6800X Duo at 135,774, the AMD Radeon PRO W6800 at 135,396, the NVIDIA A10M at 135,230, and the NVIDIA RTX 4000 Ada Generation at 135,218 — all cluster within a 0.9% delta, indicating a highly competitive mid-tier segment.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce RTX 4090 D has an average benchmark score of 178,050, which is 30.5% higher than the AMD Radeon PRO V620's 136,472.

Q: How large is the performance gap in Geekbench OpenCL?

A: The RTX 4090 D scores 278,621 compared to the Radeon PRO V620's 128,580, representing a 116.7% delta in favor of NVIDIA.

Q: Does the AMD Radeon PRO V620 have any benchmark win over the RTX 4090 D?

A: No. In the two head-to-head benchmarks available — Geekbench OpenCL and Geekbench Vulkan — the RTX 4090 D wins both, with a 2-0 wins tally.

Q: What memory capacity does each card offer?

A: The AMD Radeon PRO V620 offers 32 GB of GDDR6 memory on a 256-bit bus, while the NVIDIA GeForce RTX 4090 D offers 24 GB of GDDR6X on a 384-bit bus.

Q: Which GPU has higher power consumption?

A: The NVIDIA GeForce RTX 4090 D has a 425 W TDP, while the AMD Radeon PRO V620 has a 300 W TDP.

Q: How do the GPUs compare in terms of process node and transistor density?

A: The RTX 4090 D uses a 5 nm process with 76,300 million transistors at a density of 125.3M per mm², whereas the Radeon PRO V620 uses a 7 nm process with 26,800 million transistors at 51.5M per mm².

Architecture Differences

The architectural divide between these two GPUs is profound. The NVIDIA GeForce RTX 4090 D is built on the Ada Lovelace architecture, which represents NVIDIA's latest generation of GPU design. It uses TSMC's 5 nm process node, enabling a massive 76,300 million transistors to be packed into a 609 mm² die. The chip, codenamed AD102, features 14,592 shading units, 456 texture mapping units, 176 raster operation units, 114 ray tracing cores, and 456 tensor cores. This configuration yields a pixel rate of 443.5 GPixel/s and a texture rate of 1,149.1 GTexel/s. The FP32 compute performance is 73.54 TFLOPS, with FP16 performance also at 73.54 TFLOPS due to a 1:1 ratio.

The AMD Radeon PRO V620, conversely, is built on the RDNA 2.0 architecture, which is a previous-generation design. It uses TSMC's 7 nm process, with 26,800 million transistors on a 520 mm² die. The Navi 21 chip contains 4,608 shading units, 288 texture mapping units, 128 raster operation units, and 72 ray accelerators. Notably, it has no tensor cores, as AMD does not have a direct equivalent in this architecture. The pixel rate is 281.6 GPixel/s, and the texture rate is 633.6 GTexel/s. FP32 performance is 20.28 TFLOPS, while FP16 performance reaches 40.55 TFLOPS due to a 2:1 ratio, indicating that the AMD card can accelerate half-precision workloads at a higher rate than full precision.

The memory subsystems also differ significantly. The RTX 4090 D uses 24 GB of GDDR6X memory with a 384-bit bus, achieving 1.01 TB/s bandwidth. The Radeon PRO V620 uses 32 GB of GDDR6 memory on a 256-bit bus, with 512.0 GB/s bandwidth. The NVIDIA card's memory clock is 1313 MHz (21 Gbps effective), while the AMD card's memory clock is 2000 MHz (16 Gbps effective). These differences explain the bandwidth gap, though the AMD card's larger capacity is a countervailing advantage.

Specification Differences

The two cards differ across nearly every specification field. The RTX 4090 D has a base clock of 2280 MHz and a boost clock of 2520 MHz, while the Radeon PRO V620 has a base clock of 1825 MHz and a boost clock of 2200 MHz. In terms of physical dimensions, the RTX 4090 D measures 304 mm in length, 137 mm in height, and 61 mm in width, while the Radeon PRO V620 measures 267 mm, 120 mm, and 50 mm respectively. The NVIDIA card is triple-slot and requires a 1x 16-pin power connector with an 800 W suggested PSU, whereas the AMD card is dual-slot with 2x 8-pin connectors and a 700 W suggested PSU.

Display outputs also differ: the RTX 4090 D offers 1x HDMI 2.1 and 3x DisplayPort 1.4a, while the Radeon PRO V620 has no display outputs at all, making it a compute-only card. Both use PCIe 4.0 x16 interfaces and support DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4. The release dates are notably different, with the Radeon PRO V620 launching on 2023-12-27 and the RTX 4090 D on 2021-11-03, meaning the NVIDIA card is actually the newer product despite the AMD card's later release in the data. The RTX 4090 D has a launch MSRP of 1,599 USD, while the AMD card has no listed MSRP. Production status for both is end-of-life, with the RTX 4090 D having a successor in the GeForce 50 series and the Radeon PRO V620 having no listed successor.