GPU Comparison

The Verdict

The recorded data separates these two GPUs clearly by workload profile. The NVIDIA CMP 30HX holds a decisive advantage in both shared benchmark tests, winning the Geekbench OpenCL test with a score of 65199 against 59589 for the AMD Radeon RX 5600M, a 9.4% lead. The margin widens substantially in the Geekbench Vulkan test, where the CMP 30HX scores 62484 versus 48843, a 27.9% advantage.

The aggregate benchmark average tells a different story, however. The CMP 30HX posts an average benchmark score of 63842, placing it in the 89th percentile of all GPUs in the database. The RX 5600M averages 46601, sitting in the 85th percentile. The nearest rivals for each card reinforce the gap: the CMP 30HX sits within 0.6% of the AMD Radeon Pro WX 9100 (score 64212) and within 0.2% of the AMD Radeon Pro Vega 56 (score 63693). The RX 5600M, by contrast, is nearly tied with the Intel Arc A530M (46614, delta 0%) and the AMD Radeon RX 6550M (46702, delta -0.2%).

For buyers, the choice hinges on whether the target application favors raw compute throughput or balanced mobile integration. The CMP 30HX is a dedicated mining card with no display outputs, so it is only suitable for compute tasks that do not require video output. The RX 5600M is a mobile part, intended for portable devices, with display outputs described as portable device dependent. If the workload is Vulkan or OpenCL compute and the system can accommodate a dual-slot, 125 W card with a 1x 8-pin power connector, the CMP 30HX is the stronger performer. If the requirement is a mobile, integrated form factor with no external power connector and PCIe 4.0 x16 interface, the RX 5600M is the only viable option despite lower raw scores.

Architecture Differences

The two cards come from different architectural generations and process nodes. The NVIDIA CMP 30HX uses the TU116 chip based on the Turing architecture, fabricated by TSMC on a 12 nm process. It integrates 6,600 million transistors on a 284 mm² die, yielding a transistor density of 23.2 million per mm². The AMD Radeon RX 5600M uses the Navi 10 chip based on the RDNA 1.0 architecture, also fabricated by TSMC but on a 7 nm process. It packs 10,300 million transistors on a smaller 251 mm² die, achieving a higher transistor density of 41.0 million per mm².

The RX 5600M belongs to the Radeon RX 5000 series and the Navi Mobile generation, with a predecessor listed as Polaris Mobile. The CMP 30HX belongs to the Mining GPUs generation with no predecessor or successor listed. The AMD card has a higher shading unit count at 2304, compared to 1408 on the NVIDIA card. It also leads in texture mapping units (144 versus 88) and render output units (64 versus 48). The NVIDIA card, however, has a higher base clock at 1530 MHz versus 1035 MHz, and a higher boost clock at 1785 MHz versus 1265 MHz. The AMD card has a game clock of 1190 MHz, which the NVIDIA card does not list.

Memory configurations are similar in capacity and type: both have 6 GB of GDDR6 on a 192 bit bus. The CMP 30HX runs its memory at 1750 MHz (14 Gbps effective), producing 336.0 GB/s of bandwidth. The RX 5600M runs at 1500 MHz (12 Gbps effective), producing 288.0 GB/s. The NVIDIA card leads in pixel rate at 85.68 GPixel/s versus 80.96 GPixel/s, but the AMD card leads in texture rate at 182.2 GTexel/s versus 157.1 GTexel/s. In floating point performance, the RX 5600M achieves 5.829 TFLOPS FP32 and 11.66 TFLOPS FP16 (2:1), while the CMP 30HX delivers 5.027 TFLOPS FP32 and 10.05 TFLOPS FP16.

The power envelopes differ as well: the CMP 30HX is rated at 125 W TDP with a dual-slot cooler and a 1x 8-pin power connector, requiring a 300 W suggested power supply. The RX 5600M is rated at 150 W TDP but is listed as IGP (integrated graphics package) with no power connectors. The bus interface also differs: the NVIDIA card uses PCIe 1.0 x4, while the AMD card uses PCIe 4.0 x16. The NVIDIA card has no display outputs, while the AMD card's outputs are portable device dependent.

FAQ

Q: Which GPU has the higher average benchmark score?

The NVIDIA CMP 30HX has an average benchmark score of 63842, while the AMD Radeon RX 5600M has an average of 46601.

Q: How much faster is the CMP 30HX in the Vulkan test?

The CMP 30HX scores 62484 in Geekbench Vulkan, compared to 48843 for the RX 5600M, a 27.9% advantage.

Q: Does the RX 5600M have more shading units?

Yes, the RX 5600M has 2304 shading units, while the CMP 30HX has 1408.

Q: What is the memory bandwidth difference?

The CMP 30HX provides 336.0 GB/s, while the RX 5600M provides 288.0 GB/s, a difference of 48.0 GB/s in favor of the NVIDIA card.

Q: Which card has a higher boost clock?

The CMP 30HX boosts up to 1785 MHz, whereas the RX 5600M boosts only to 1265 MHz.

Q: Is the RX 5600M usable as a desktop card?

The RX 5600M is an integrated mobile package with no dimensions listed and its display outputs are portable device dependent. The CMP 30HX, being a mining card, has no display outputs at all.

Specification Differences

The two cards differ in nearly every architectural specification except for a few basics. Both use 6 GB of GDDR6 memory on a 192-bit bus, and both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The manufacturing process is different: 12 nm for the CMP 30HX versus 7 nm for the RX 5600M, both from TSMC. The transistor count is 6,600 million versus 10,300 million, and die size is 284 mm² versus 251 mm².

Clock speeds differ markedly: base clock 1530 MHz versus 1035 MHz, boost clock 1785 MHz versus 1265 MHz, and the RX 5600M adds a game clock of 1190 MHz. Memory clock is 1750 MHz (14 Gbps effective) versus 1500 MHz (12 Gbps effective). Bandwidth is 336.0 GB/s versus 288.0 GB/s.

The core configuration differs: shading units 1408 versus 2304, TMUs 88 versus 144, ROPs 48 versus 64. Pixel rate is 85.68 GPixel/s versus 80.96 GPixel/s, and texture rate is 157.1 GTexel/s versus 182.2 GTexel/s. FP32 performance is 5.027 TFLOPS versus 5.829 TFLOPS, and FP16 is 10.05 TFLOPS versus 11.66 TFLOPS.

Power and physical differences are notable: TDP is 125 W versus 150 W, slot width is dual-slot versus IGP, power connectors are 1x 8-pin versus none, and the suggested PSU is 300 W versus none listed. The bus interface is PCIe 1.0 x4 versus PCIe 4.0 x16. The CMP 30HX has dimensions of 229 mm length, 111 mm height, 35 mm width, while the RX 5600M has no listed dimensions. The CMP 30HX has no display outputs; the RX 5600M has portable device dependent outputs.

The release dates differ by about seven months: the CMP 30HX launched on February 24, 2021, while the RX 5600M launched on July 6, 2020. The CMP 30HX has a launch MSRP of 799 USD, while the RX 5600M has no listed launch MSRP. Both are end-of-life products.

Head-to-Head Benchmarks

The database records two direct head-to-head comparisons, both won by the NVIDIA CMP 30HX. In Geekbench OpenCL, the CMP 30HX scores 65199, while the RX 5600M scores 59589. That is a 9.4% advantage for the NVIDIA card. In Geekbench Vulkan, the CMP 30HX scores 62484 against 48843 for the AMD card. That advantage grows to 27.9%.

The margin in Vulkan is nearly three times larger than in OpenCL. This suggests the Turing architecture's Vulkan driver path delivers a substantially stronger performance uplift over RDNA 1.0 in this specific workload. The OpenCL gap, while still in the NVIDIA card's favor, is more moderate at under 10%.

Looking at the broader database, the CMP 30HX's OpenCL score of 65199 is higher than its own average benchmark score of 63842, which indicates that OpenCL is a particularly strong workload for this card. The Vulkan score of 62484 is slightly below the average. The RX 5600M's OpenCL score of 59589 is well above its average of 46601, but its Vulkan score of 48843 is closer to the average, showing that Vulkan is a relative weakness for the AMD card.

The RX 5600M also has a 3DMark Steel Nomad DX12 score of 1320 and a Geekbench Metal score of 76653, but the CMP 30HX has no corresponding scores in those tests, so a direct comparison is not possible.

Where Each One Wins

The NVIDIA CMP 30HX wins all benchmark comparisons that are directly available. It leads by 9.4% in OpenCL and by 27.9% in Vulkan. The data shows that for any workload measured by the database, the CMP 30HX is the faster card. It also has a higher average benchmark score 63842 versus 46601, a difference of 37%. The CMP 30HX achieves higher pixel throughput (85.68 GPixel/s versus 80.96) and higher memory bandwidth (336.0 GB/s versus 288.0), both of which contribute to its benchmark superiority.

The AMD Radeon RX 5600M, however, wins in several architectural categories that may matter for specific uses. It has more shading units (2304 versus 1408), more texture mapping units (144 versus 88), and more render output units (64 versus 48). It also delivers higher FP32 compute at 5.829 TFLOPS versus 5.027 TFLOPS, and higher FP16 compute at 11.66 TFLOPS versus 10.05 TFLOPS. Its texture rate is higher at 182.2 GTexel/s versus 157.1 GTexel/s. These raw compute advantages do not translate into benchmark wins in the recorded tests, but they indicate that the RX 5600M may be more efficient at compute-heavy tasks that are not captured by the OpenCL and Vulkan tests in the database.

The RX 5600M also wins on physical integration. It is an IGP (integrated graphics package) with no power connectors, meaning no separate power cable is required. It uses PCIe 4.0 x16, which provides a wider bus interface than the CMP 30HX's PCIe 1.0 x4. The RX 5600M has portable device dependent display outputs, whereas the CMP 30HX has no outputs at all.

The CMP 30HX wins on power efficiency in terms of performance per watt, based on the data: it delivers its average score of 63842 at 125 W, while the RX 5600M delivers 46601 at 150 W. The CMP 30HX also has a smaller die footprint in terms of process maturity but a larger physical card (dual-slot versus IGP), and it carries a launch MSRP of 799 USD while the RX 5600M has no listed price.