AMD Radeon HD 8730M vs NVIDIA Quadro K620 Comparison

The NVIDIA Quadro K620 and AMD Radeon HD 8730M are both end-of-life mobile or small-form-factor graphics solutions built on a 28 nm TSMC process, but they target different segments and deliver measurably different performance. The data shows a clear, if narrow, overall winner in the single available head-to-head benchmark, with architectural and feature differences that matter for specific workloads.

Head-to-Head Benchmarks

The only direct benchmark comparison available is Geekbench OpenCL, and it decisively favors the NVIDIA Quadro K620. The Quadro K620 scores 6693 points, while the AMD Radeon HD 8730M scores 5955 points. This represents a 12.4% advantage for the NVIDIA part. That is not a marginal gap; it is a solid, consistent lead that shows up in compute-oriented tasks. The Quadro K620’s average benchmark score across all tests is 6282, while the HD 8730M averages 5955, reinforcing the same trend.

Looking at the broader competitive landscape from the nearestRivals data, the Quadro K620 sits in a tight cluster. Its average score of 6282 is within 0.2% of the NVIDIA GeForce RTX 5070 Ti SUPER and the NVIDIA GeForce RTX 4070 Ti SUPER AD102, both of which average 6270. It is also 0.7% behind the AMD Radeon R7 M350 (6327) and 0.8% behind the AMD Radeon Pro WX 4100 (6330). This means the Quadro K620 is not an outlier; it is right in the middle of a competitive pack of modern and older parts.

The AMD Radeon HD 8730M, meanwhile, sits in a slightly lower performance tier. Its average score of 5955 is essentially tied with the NVIDIA Quadro K620M (5957, deltaPct 0%), 0.3% behind the AMD Radeon HD 8750M (5970), 0.4% ahead of the Intel UHD Graphics 730 (5929), and 0.5% behind the NVIDIA Quadro K4000 (5982). The HD 8730M’s percentile rank of 34 versus the Quadro K620’s 36 confirms that the NVIDIA part outperforms a larger share of the GPU population.

The wins tally is straightforward: the Quadro K620 has 1 win in the head-to-head benchmarks, and the HD 8730M has 0. There is no benchmark where the AMD part comes out ahead. The 12.4% delta in OpenCL is the headline number, but the consistency across the average scores and the rival comparisons tells the same story.

The Verdict

From the data, the NVIDIA Quadro K620 is the superior performer for compute-heavy tasks. Its 12.4% lead in Geekbench OpenCL over the HD 8730M is the single most important metric here, and it is backed by a higher average benchmark score (6282 vs 5955) and a better percentile ranking (36 vs 34). Anyone running OpenCL workloads, which include many professional and scientific applications, should choose the Quadro K620 without hesitation.

The HD 8730M is not without merit, but its strengths lie elsewhere. It is a smaller chip (77 mm² vs 148 mm²) with fewer transistors (950 million vs 1,870 million), which suggests lower power consumption and heat output, though the FACT PACK does not list a TDP for the AMD part. It is also a mobile-oriented part, as indicated by the "M" suffix in its name, so it may be the only option in certain laptops or compact systems. However, the benchmark results do not show any performance advantage for the AMD part.

Who should pick which? Strictly from the data, a user needing maximum OpenCL compute performance should pick the Quadro K620. A user constrained by a mobile form factor or a smaller physical footprint may have no choice but to use the HD 8730M, and the data shows it is a capable part that trades blows with the Quadro K620M and the Quadro K4000. But if performance is the only criterion, the verdict is unambiguous: the Quadro K620 wins.

Architecture Differences

The two GPUs come from different architectural generations and design philosophies. The NVIDIA Quadro K620 is built on the Maxwell architecture, specifically the GM107 chip. The AMD Radeon HD 8730M uses the GCN 1.0 architecture, with the Mars chip. Both are fabricated by TSMC on a 28 nm process, but the similarities end there.

The transistor counts and die sizes differ substantially. The Quadro K620 packs 1,870 million transistors into a 148 mm² die, yielding a transistor density of 12.6M per mm². The HD 8730M has 950 million transistors on a 77 mm² die, with a density of 12.3M per mm². The NVIDIA chip has roughly twice the transistor budget, which explains its higher compute throughput.

Clock speeds strongly favor the NVIDIA part. The Quadro K620 runs at a base clock of 1058 MHz and a boost clock of 1124 MHz. The HD 8730M is significantly slower, with a base clock of 650 MHz and a boost of 700 MHz. This nearly 400 MHz difference in base clock is a major factor in the performance gap. Memory clocks are identical on paper: both use 900 MHz memory with 1800 Mbps effective data rate, and both have a 128-bit bus width, resulting in the same 28.80 GB/s bandwidth.

The compute resources are partially matched. Both GPUs have 384 shading units and 24 texture mapping units (TMUs). However, the ROP count differs: the Quadro K620 has 16 ROPs, while the HD 8730M has only 8. This directly impacts pixel fill rate. The Quadro K620 achieves 17.98 GPixel/s, while the HD 8730M manages just 5.600 GPixel/s. The texture rate also favors NVIDIA: 26.98 GTexel/s versus 16.80 GTexel/s. The FP32 performance gap is even larger: the Quadro K620 delivers 863.2 GFLOPS, while the HD 8730M delivers 537.6 GFLOPS.

The memory subsystems are equally matched in bandwidth, but the rest of the architecture is not. The Quadro K620 supports a single-slot design with no power connectors, a 200 W suggested PSU, and a PCIe 2.0 x16 interface. The HD 8730M uses a PCIe 3.0 x8 interface, which offers higher per-lane bandwidth but fewer lanes. Display outputs also differ: the Quadro K620 has 1x DVI and 1x DisplayPort 1.2, while the HD 8730M has no listed display outputs.

API support shows a few notable differences. Both support DirectX 12, but the Quadro K620 is listed as "12 (11_0)" while the HD 8730M is "12 (11_1)". OpenGL support is identical at 4.6. Vulkan support favors NVIDIA: the Quadro K620 supports Vulkan 1.4, while the HD 8730M supports Vulkan 1.2.170. The Quadro K620 also has a Geekbench Vulkan score of 5870, while the HD 8730M has no Vulkan benchmark listed. The Quadro K620’s generation is listed as "Quadro Kepler (Kx200)" despite being Maxwell-based, while the HD 8730M is from the "Solar System (HD 8700M)" generation.

FAQ

Q: Which GPU has a higher Geekbench OpenCL score?

A: The NVIDIA Quadro K620 scores 6693, which is 12.4% higher than the AMD Radeon HD 8730M’s 5955.

Q: How do the FP32 compute performances compare?

A: The Quadro K620 delivers 863.2 GFLOPS, while the HD 8730M delivers 537.6 GFLOPS, a substantial advantage for NVIDIA.

Q: Are the memory bandwidths the same?

A: Yes, both GPUs use 2 GB of DDR3 memory on a 128-bit bus, resulting in an identical 28.80 GB/s bandwidth.

Q: What is the pixel fill rate difference?

A: The Quadro K620 achieves 17.98 GPixel/s, which is over three times the 5.600 GPixel/s of the HD 8730M, due to having 16 ROPs versus 8.

Q: Which GPU has better Vulkan support?

A: The Quadro K620 supports Vulkan 1.4 and has a Geekbench Vulkan score of 5870, while the HD 8730M only supports Vulkan 1.2.170 and has no Vulkan score listed.

Q: How does the Quadro K620 compare to the AMD Radeon R7 M350?

A: The Quadro K620’s average score of 6282 is 0.7% lower than the R7 M350’s 6327, making them near-identical performers.

Where Each One Wins

The NVIDIA Quadro K620 wins in every measurable compute benchmark and in most architectural specifications. Its 12.4% OpenCL lead is the clearest win, but the advantages extend to FP32 throughput (863.2 vs 537.6 GFLOPS), pixel rate (17.98 vs 5.600 GPixel/s), and texture rate (26.98 vs 16.80 GTexel/s). It also has double the ROP count (16 vs 8) and a significantly higher transistor count (1,870 million vs 950 million). For any workload that stresses raw compute, OpenCL acceleration, or high-resolution rendering, the Quadro K620 is the clear winner. Its Vulkan 1.4 support and higher boost clock (1124 MHz vs 700 MHz) further cement its position.

The AMD Radeon HD 8730M does not win any benchmarks, but it has structural advantages that may matter in specific deployments. Its die is nearly half the size (77 mm² vs 148 mm²), which suggests lower manufacturing complexity and potentially lower power draw, although no TDP is listed for the AMD part. It uses a PCIe 3.0 x8 interface, which is a newer bus standard than the Quadro K620’s PCIe 2.0 x16, potentially offering better bandwidth efficiency in compatible systems. Its DirectX 12 support is slightly newer (11_1 vs 11_0), which could matter for certain legacy applications. The HD 8730M also sits in a different performance neighborhood, trading blows with the Quadro K620M and the Quadro K4000, so it is not a weak part—it is simply a tier below the Quadro K620.

For users who need maximum performance in a single-slot professional card, the Quadro K620 is the only choice. For users who need a compact mobile solution with a smaller chip and newer PCIe interface, the HD 8730M has its place. But the data is unambiguous: the Quadro K620 wins on performance, and the HD 8730M wins only on physical size and bus generation.