GPU Comparison

Head-to-Head Benchmarks

The only directly comparable benchmark between the NVIDIA GeForce MX130 and the NVIDIA Quadro K620M is Geekbench OpenCL. The MX130 scores 6102, while the K620M scores 5957. That gives the MX130 a 2.4% advantage in this test. This is a narrow margin, but it is consistent across the board, the MX130 wins the sole head-to-head comparison, with 1 win against 0 for the K620M.

When placed against their respective nearest rivals, the two cards occupy similar performance tiers. The MX130’s average benchmark score sits at 5508, which places it in the 32nd percentile among all GPUs. Its nearest listed rival, the NVIDIA GeForce GTX 765M, scores 5501, a delta of only 0.1%. The AMD Radeon R7 M440 is also within 0.5%, scoring 5483. On the other side, the AMD FirePro M4000 scores 5537, putting it 0.5% ahead of the MX130. The NVIDIA Quadro M4000 is 0.7% ahead with a score of 5467. These deltas are all under 1%, meaning the MX130 is essentially in a dead heat with its immediate competitors.

The K620M’s average benchmark score is 5957, which places it in the 34th percentile, two percentile points higher than the MX130. Its nearest rival, the AMD Radeon HD 8730M, scores 5955, a delta of 0.0%. The AMD Radeon HD 8750M is 0.2% ahead at 5970, while the NVIDIA Quadro K4000 is 0.4% ahead at 5982. The Intel UHD Graphics 730 trails by 0.5%, scoring 5929. Again, all deltas are below 1%, indicating the K620M is tightly clustered with its peers.

The notable discrepancy here is that the K620M’s average score (5957) is higher than the MX130’s average score (5508), yet in the direct OpenCL comparison, the MX130 wins. The difference lies in what each average represents. The MX130’s average is pulled down by its second benchmark, Geekbench Vulkan, where it scores 4914. The K620M has no Vulkan result listed, so its average is based solely on the OpenCL score. This means the MX130’s Vulkan performance is significantly lower than its OpenCL performance, while the K620M’s single data point tells us nothing about its Vulkan capability.

Looking at raw compute metrics, the MX130 has a higher FP32 throughput at 913.2 GFLOPS versus the K620M’s 863.2 GFLOPS. That is roughly a 5.8% difference in raw floating-point horsepower, which aligns with the 2.4% OpenCL delta, though the gap in real-world benchmarks is smaller than the theoretical peak would suggest. The MX130 also leads in pixel rate (9.512 GPixel/s vs 8.992 GPixel/s) and texture rate (28.54 GTexel/s vs 17.98 GTexel/s). The texture rate difference is particularly stark: the MX130 is nearly 59% faster in texel throughput, driven by its higher TMU count (24 vs 16).

Where Each One Wins

The MX130 wins the only head-to-head benchmark, but the data suggests a broader pattern. Its higher texture rate and TMU count make it better suited for fill-rate-bound workloads, scenarios where many textured pixels must be processed simultaneously. The 28.54 GTexel/s texture rate versus 17.98 GTexel/s is a substantial advantage that would manifest in games or 3D applications that rely heavily on texture sampling.

The MX130 also holds the edge in memory bandwidth at 40.10 GB/s versus the K620M’s 16.02 GB/s. This is a 2.5x difference, and it stems from the memory type: the MX130 uses GDDR5 at 5 Gbps effective, while the K620M uses DDR3 at 2 Gbps effective. Both have a 64-bit bus and 2 GB capacity, but the memory technology alone gives the MX130 a decisive bandwidth advantage. For workloads that are bandwidth-sensitive, such as large data transfers, high-resolution texture streaming, or compute kernels that repeatedly access memory, the MX130 is clearly ahead.

The K620M’s wins are more subtle. Its average benchmark score is higher (5957 vs 5508), and it sits in a higher percentile (34 vs 32). Its nearest rivals are also slightly faster relative to it, the Quadro K4000 is 0.4% ahead, whereas the MX130’s closest rival, the GTX 765M, is only 0.1% ahead. This suggests that while the K620M is in a slightly higher performance tier overall, its margin over its own peers is similar to the MX130’s margin over its peers.

In terms of clock speeds, the MX130 runs at 1109 MHz base and 1189 MHz boost, while the K620M runs at 1029 MHz base and 1124 MHz boost. The MX130’s clocks are 80 MHz higher at base and 65 MHz higher at boost. This contributes to its higher FP32 and texture rates, though the K620M’s lower clocks are partially offset by its different memory configuration, though that configuration is inferior in bandwidth.

For professional or workstation use, the Quadro branding traditionally implies different driver validation, but the data here does not show any benchmark advantage from that. The K620M’s single OpenCL score is lower than the MX130’s, and its texture and bandwidth metrics are lower. The only areas where the K620M does not lose are those where the two are identical, shading units (384), ROPs (8), pixel rate (nearly identical at 8.99 vs 9.51 GPixel/s), and TDP (30 W).

The Verdict

From the data alone, the NVIDIA GeForce MX130 is the stronger performer in the single benchmark where both are tested. Its 2.4% lead in Geekbench OpenCL is modest, but it is backed by superior specifications across the board: higher clocks, more TMUs, faster memory, greater bandwidth, and higher pixel and texture rates. For anyone choosing between these two for general-purpose GPU compute or graphics work, the MX130 is the safer pick.

However, the K620M is not without merit. Its average benchmark score is higher than the MX130’s because the MX130’s Vulkan score drags its average down. If a workload relies on Vulkan, the MX130’s 4914 score suggests it will underperform relative to its OpenCL result, while the K620M has no listed Vulkan data, meaning its Vulkan performance is unknown. For users who prioritize consistency across APIs, the K620M’s single benchmark result is at least predictable.

The percentile figures tell a nuanced story. The K620M sits in the 34th percentile, two points above the MX130’s 32nd. This indicates that, relative to all GPUs, the K620M is slightly better placed. But this ranking is based on average scores, and the MX130’s average is depressed by its Vulkan result. In the context of the only shared benchmark, the MX130 wins.

For gaming or media applications that leverage OpenCL, the MX130 is the clear choice. Its memory bandwidth advantage (40.10 vs 16.02 GB/s) is enormous and will be felt in any texture-heavy or bandwidth-limited scenario. For a workstation that runs OpenCL-based professional applications, the MX130 still leads, but the K620M’s similar shading unit count (384) and ROP count (8) mean that compute kernels which are not bandwidth-limited may see more comparable results.

The tiebreaker is the specification sheet. The MX130 uses GDDR5 memory, has 24 TMUs, and boosts to 1189 MHz. The K620M uses DDR3, has 16 TMUs, and boosts to 1124 MHz. Every one of these differences favors the MX130. The only specification where the K620M matches or exceeds is the transistor count and die size, both are identical at 1,020 million transistors and 77 mm². This makes sense because both are built on the GM108S chip, but that does not change the fact that the MX130’s implementation is more capable.

In short, the MX130 wins the head-to-head, wins on raw specifications, and offers superior memory bandwidth. The K620M has a higher average score and percentile, but that is an artifact of missing Vulkan data. For most users, the MX130 is the better GPU. The K620M is only preferable if a specific application requires its particular feature set, which the data does not identify.

FAQ

Q: Which GPU wins the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce MX130 wins with a score of 6102, versus the NVIDIA Quadro K620M’s 5957, a 2.4% advantage.

Q: What is the average benchmark score for each GPU?

A: The MX130 has an average benchmark score of 5508, while the K620M has an average benchmark score of 5957.

Q: How do the two GPUs compare in memory bandwidth?

A: The MX130 has a memory bandwidth of 40.10 GB/s using GDDR5, while the K620M has 16.02 GB/s using DDR3. Both have a 64-bit bus and 2 GB of memory.

Q: What are the clock speeds of the two GPUs?

A: The MX130 runs at 1109 MHz base and 1189 MHz boost. The K620M runs at 1029 MHz base and 1124 MHz boost.

Q: Which GPU has more texture mapping units?

A: The MX130 has 24 TMUs, whereas the K620M has 16 TMUs. This contributes to the MX130’s higher texture rate of 28.54 GTexel/s versus 17.98 GTexel/s.

Q: What is the TDP of each GPU?

A: Both GPUs have a TDP of 30 W.

Architecture Differences

Both the NVIDIA GeForce MX130 and the NVIDIA Quadro K620M are built on the same GM108S chip, using the Maxwell architecture. The process node is identical at 28 nm, fabricated by TSMC. Both have exactly 1,020 million transistors and a die size of 77 mm², yielding the same transistor density of 13.2M / mm².

Despite sharing the same physical chip, the two GPUs are configured differently. Both have 384 shading units and 8 ROPs, but the MX130 has 24 TMUs while the K620M has 16 TMUs. The MX130 also runs at higher clocks: 1109 MHz base and 1189 MHz boost, compared to the K620M’s 1029 MHz base and 1124 MHz boost. The MX130’s FP32 throughput is 913.2 GFLOPS, versus 863.2 GFLOPS for the K620M. Pixel rates are close, 9.512 GPixel/s for the MX130 and 8.992 GPixel/s for the K620M, but the texture rate diverges significantly: 28.54 GTexel/s versus 17.98 GTexel/s.

The memory subsystems differ sharply. The MX130 uses 2 GB of GDDR5 with a 64-bit bus and memory clock of 1253 MHz (5 Gbps effective), yielding 40.10 GB/s bandwidth. The K620M uses 2 GB of DDR3 with the same 64-bit bus but a memory clock of 1001 MHz (2 Gbps effective), yielding only 16.02 GB/s bandwidth. Neither GPU has RT cores or tensor cores, and both lack FP16 support. Both support DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4.

The slot width differs: the MX130 is an IGP, while the K620M is an MXM Module. The bus interface also differs, with the MX130 using PCIe 3.0 x4 and the K620M using MXM-A (3.0). Neither requires external power connectors. The display outputs are labeled "Portable Device Dependent" for both. The production status is "End-of-life" for both, and neither has a launch MSRP listed.

Specification Differences

| Specification | NVIDIA GeForce MX130 | NVIDIA Quadro K620M |

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

| Generation | GeForce MX (1xx) | Quadro Kepler-M (Kx200M) |

| Base Clock | 1109 MHz | 1029 MHz |

| Boost Clock | 1189 MHz | 1124 MHz |

| Memory Clock | 1253 MHz (5 Gbps effective) | 1001 MHz (2 Gbps effective) |

| Memory Type | GDDR5 | DDR3 |

| Memory Bandwidth | 40.10 GB/s | 16.02 GB/s |

| TMUs | 24 | 16 |

| Pixel Rate | 9.512 GPixel/s | 8.992 GPixel/s |

| Texture Rate | 28.54 GTexel/s | 17.98 GTexel/s |

| FP32 | 913.2 GFLOPS | 863.2 GFLOPS |

| Slot Width | IGP | MXM Module |

| Bus Interface | PCIe 3.0 x4 | MXM-A (3.0) |

| Release Date | 2017-11-16 | 2015-02-28 |

| Predecessor | None | Quadro Fermi-M |

| Successor | None | Quadro Maxwell-M |

| Geekbench OpenCL Score | 6102 | 5957 |

| Geekbench Vulkan Score | 4914 | None |

| Average Benchmark Score | 5508 | 5957 |

| Percentile vs All GPUs | 32 | 34 |

The two GPUs share identical specifications in several areas: process node (28 nm), foundry (TSMC), transistors (1,020 million), die size (77 mm²), transistor density (13.2M / mm²), shading units (384), ROPs (8), memory size (2 GB), memory bus width (64 bit), TDP (30 W), power connectors (None), display outputs (Portable Device Dependent), DirectX (12 (11_0)), OpenGL (4.6), Vulkan (1.4), and production status (End-of-life).