AMD Radeon R7 350 vs NVIDIA Quadro K620 Comparison

Head-to-Head Benchmarks

The benchmark data is unequivocal: the AMD Radeon R7 350 wins both recorded head-to-head tests against the NVIDIA Quadro K620. In the Geekbench OpenCL test, the R7 350 scores 7792, while the Quadro K620 scores 6693. That is a 16.4% advantage for the AMD card. In the Geekbench Vulkan test, the R7 350 scores 7057 versus 5870 for the Quadro K620, a 20.2% lead.

The OpenCL result is the more straightforward compute comparison, since both cards are being pushed through the same API. The R7 350’s 16.4% margin is substantial, not a minor edge. Meanwhile, the Vulkan gap is even larger, with the R7 350 stretching its lead to over a fifth faster than the Quadro.

Looking at the broader database context, the R7 350’s average benchmark score is 7425, placing it at the 40th percentile of all GPUs. The Quadro K620’s average is 6282, at the 36th percentile. That is a meaningful gap in overall standing. The R7 350’s nearest rivals on average score are the Intel UHD Graphics 750 (7441, only 0.2% behind), the AMD Radeon HD 8850M (7447, 0.3% behind), and the NVIDIA GeForce GTX 1650 (7472, 0.6% behind). The Quadro K620, in contrast, sits right next to the AMD Radeon R7 M350 (6327, only 0.7% ahead of the K620) and the AMD Radeon Pro WX 4100 (6330, 0.8% ahead). The R7 350’s score is nearly 18% higher than the Quadro K620’s average, which is a decisive spread.

The wins are not split; the R7 350 takes both tests outright. The Quadro K620 does not record a single victory in the head-to-head data. This is a clean sweep for the AMD part.

Architecture Differences

The two cards come from different architectural generations. The AMD Radeon R7 350 uses the Cape Verde chip based on GCN 1.0 architecture, belonging to the Pirate Islands (R7 300) generation. The NVIDIA Quadro K620 uses the GM107 chip on the Maxwell architecture, listed under the Quadro Kepler (Kx200) generation despite the Maxwell core.

Both are built on the same 28 nm process at TSMC, but the transistor counts differ. The R7 350 packs 1,500 million transistors on a 123 mm² die, giving a transistor density of 12.2 million per mm². The Quadro K620 has 1,870 million transistors on a 148 mm² die, density at 12.6 million per mm². So the NVIDIA chip is physically larger and has more transistors, but the AMD card still wins in compute benchmarks.

The memory subsystems are very different. The R7 350 uses 2 GB of GDDR5 with a 128-bit bus and 72.00 GB/s of bandwidth. The Quadro K620 also has 2 GB, but it is DDR3 on the same 128-bit bus, yielding only 28.80 GB/s. That is a massive bandwidth advantage for the AMD card, more than double the NVIDIA’s throughput. The memory clock figures reflect this: the R7 350 runs its memory at 1125 MHz (4.5 Gbps effective), while the Quadro K620’s memory runs at 900 MHz (1800 Mbps effective).

The compute resources also differ. The R7 350 has 512 shading units, 32 texture mapping units, and 16 ROPs. The Quadro K620 has 384 shading units, 24 TMUs, and 16 ROPs. Despite fewer cores, the Quadro K620 has a higher pixel rate at 17.98 GPixel/s versus 12.80 GPixel/s for the AMD card. Its texture rate is also slightly higher at 26.98 GTexel/s versus 25.60 GTexel/s. The FP32 throughput tells a different story: the R7 350 delivers 819.2 GFLOPS, while the Quadro K620 delivers 863.2 GFLOPS, so the NVIDIA card has a small theoretical compute edge. Yet the real-world benchmark scores favor AMD, likely due to memory bandwidth and driver efficiency.

Clock speeds are another divergence. The Quadro K620 has a base clock of 1058 MHz and a boost of 1124 MHz, while the R7 350 has no recorded base or boost clock in the database. The Quadro runs at a higher absolute frequency, but the R7 350’s GDDR5 memory compensates.

The bus interface differs: the R7 350 uses PCIe 3.0 x16, while the Quadro K620 is on PCIe 2.0 x16. The AMD card also has more display outputs: 1x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2, versus the Quadro’s 1x DVI and 1x DisplayPort 1.2.

API support is close but not identical. Both support DirectX 12, but the R7 350 is at 12 (11_1) while the Quadro is at 12 (11_0). OpenGL is 4.6 for both. The Vulkan version differs: the R7 350 has Vulkan 1.2.170, while the Quadro K620 lists Vulkan 1.4, which is a newer API revision.

Power requirements favor the Quadro. The R7 350 has a TDP of 55 W and a suggested PSU of 250 W. The Quadro K620 has a TDP of 45 W and a suggested PSU of 200 W. Both are single-slot cards with no power connectors. The R7 350 is slightly longer at 168 mm (6.6 inches) versus the Quadro’s 160 mm (6.3 inches), and the Quadro has a recorded height of 69 mm (2.7 inches).

Release dates are far apart: the R7 350 launched on 2016-07-05, while the Quadro K620 launched on 2014-07-21. The R7 350’s predecessor is Volcanic Islands and successor is Arctic Islands. The Quadro’s predecessor is Quadro Fermi and successor is Quadro Maxwell.

FAQ

Q: Which card is faster in the recorded benchmarks?

A: The AMD Radeon R7 350 wins both tests. It leads by 16.4% in Geekbench OpenCL and by 20.2% in Geekbench Vulkan.

Q: How do their average benchmark scores compare?

A: The R7 350 has an average benchmark score of 7425, while the Quadro K620 sits at 6282. The R7 350 sits at the 40th percentile of all GPUs, the Quadro at the 36th.

Q: Why does the R7 350 win despite the Quadro having a higher FP32 rating?

A: The Quadro K620 has a higher theoretical FP32 at 863.2 GFLOPS versus 819.2 GFLOPS for the R7 350. However, the R7 350 has far more memory bandwidth: 72.00 GB/s versus 28.80 GB/s. The benchmark results indicate that memory bandwidth plays a larger role in these workloads.

Q: What memory do these cards use?

A: The R7 350 uses 2 GB of GDDR5 on a 128-bit bus. The Quadro K620 uses 2 GB of DDR3 on the same 128-bit bus.

Q: Are both cards single-slot designs?

A: Yes, both are single-slot, and neither has a power connector. The R7 350 has a TDP of 55 W with a suggested 250 W PSU; the Quadro K620 has a TDP of 45 W with a suggested 200 W PSU.

Q: Do the cards support modern APIs?

A: Both support DirectX 12 and OpenGL 4.6. The R7 350 lists Vulkan 1.2.170, while the Quadro K620 lists Vulkan 1.4.

The Verdict

The data points to a clear choice for compute workloads: the AMD Radeon R7 350 is the stronger performer. It wins both head-to-head benchmarks by double-digit margins, has a higher average benchmark score, and sits at a higher percentile. The 20.2% Vulkan lead is particularly decisive.

The Quadro K620’s only advantages are in power draw and physical dimensions. It runs at 45 W versus 55 W, requires a 200 W PSU instead of 250 W, and is 8 mm shorter. It also has a higher pixel rate and texture rate, but those do not translate into wins in the recorded tests.

For anyone choosing between these two end-of-life cards, the R7 350 is the better investment for general compute and Vulkan-based tasks. The Quadro K620 is only preferable if power consumption is the absolute priority and the performance deficit is acceptable.

Specification Differences

| Field | AMD Radeon R7 350 | NVIDIA Quadro K620 |

|-------|-------------------|-------------------|

| Chip | Cape Verde | GM107 |

| Architecture | GCN 1.0 | Maxwell |

| Generation | Pirate Islands (R7 300) | Quadro Kepler (Kx200) |

| Process Node | 28 nm | 28 nm |

| Transistors | 1,500 million | 1,870 million |

| Die Size | 123 mm² | 148 mm² |

| Transistor Density | 12.2M / mm² | 12.6M / mm² |

| Base Clock | Not recorded | 1058 MHz |

| Boost Clock | Not recorded | 1124 MHz |

| Memory Clock | 1125 MHz (4.5 Gbps effective) | 900 MHz (1800 Mbps effective) |

| Memory Size | 2 GB | 2 GB |

| Memory Type | GDDR5 | DDR3 |

| Memory Bus Width | 128 bit | 128 bit |

| Memory Bandwidth | 72.00 GB/s | 28.80 GB/s |

| Shading Units | 512 | 384 |

| TMUs | 32 | 24 |

| ROPs | 16 | 16 |

| Pixel Rate | 12.80 GPixel/s | 17.98 GPixel/s |

| Texture Rate | 25.60 GTexel/s | 26.98 GTexel/s |

| FP32 | 819.2 GFLOPS | 863.2 GFLOPS |

| TDP | 55 W | 45 W |

| Suggested PSU | 250 W | 200 W |

| Bus Interface | PCIe 3.0 x16 | PCIe 2.0 x16 |

| Display Outputs | 1x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 | 1x DVI, 1x DisplayPort 1.2 |

| DirectX | 12 (11_1) | 12 (11_0) |

| OpenGL | 4.6 | 4.6 |

| Vulkan | 1.2.170 | 1.4 |

| Length | 168 mm (6.6 inches) | 160 mm (6.3 inches) |

| Height | Not recorded | 69 mm (2.7 inches) |

| Release Date | 2016-07-05 | 2014-07-21 |

| Predecessor | Volcanic Islands | Quadro Fermi |

| Successor | Arctic Islands | Quadro Maxwell |

Where Each One Wins

The R7 350 wins in every recorded benchmark, so its strengths are clear: OpenCL compute, Vulkan compute, and memory bandwidth. The 72.00 GB/s of GDDR5 bandwidth is the standout spec, and the benchmark scores reflect it. This card is the pick for tasks that stress memory throughput or use Vulkan APIs.

The Quadro K620 wins in power efficiency and physical footprint. It draws 10 W less, needs a smaller PSU, and is shorter by 8 mm. It also has higher pixel and texture rates, which could matter in specific rasterization scenarios, but those do not show up as wins in the database tests.

For a builder with a tight power budget or a small case, the Quadro K620 is the safer choice. For anyone who wants raw performance in the recorded workloads, the R7 350 is the clear winner. There is no scenario in the data where the Quadro K620 overtakes the R7 350 in compute speed.