NVIDIA GeForce MX130 vs NVIDIA RTX PRO 6000 Blackwell Server Comparison

Where Each One Wins

The recorded benchmark data splits these two NVIDIA parts into entirely different performance universes, though neither lands far from the other in overall percentile standing. The RTX PRO 6000 Blackwell Server sits at the 34th percentile among all GPUs in the database, while the GeForce MX130 rests at the 32nd percentile. That near-tie in percentile ranking is misleading, because the two products were measured with different test suites and occupy opposite ends of the hardware spectrum.

The RTX PRO 6000 Blackwell Server posts a single recorded result: 5,996 points in 3DMark Steel Nomad DX12. That score places it within a hair of several older or lower-tier parts, including the GeForce GTX 770M at 6,000 (0.1% ahead), the Radeon RX 6400 at 6,001 (0.1% ahead), the FirePro W4100 at 5,987 (0.2% behind), and the Quadro K4000M at 5,986 (0.2% behind). The data shows a GPU whose measured workload performance sits in a narrow band around the 6,000 mark, effectively indistinguishable from those rivals in this single test.

The GeForce MX130, by contrast, shows two recorded results: 6,102 in Geekbench OpenCL and 4,914 in Geekbench Vulkan. Its average benchmark score is 5,508. Its nearest rivals bracket it closely: the GTX 765M at 5,501 (0.1% ahead of the MX130), the Radeon R7 M440 at 5,483 (0.5% ahead), the FirePro M4000 at 5,537 (0.5% behind), and the Quadro M4000 at 5,467 (0.7% behind). The MX130’s average sits just above those comparably modest parts, with its OpenCL score notably higher than its Vulkan score.

The use-case split is stark. The RTX PRO 6000 Blackwell Server is a server-class accelerator with 96 GB of GDDR7 memory, a 512-bit bus, and 1.79 TB/s of bandwidth. Its single benchmark reflects a DirectX 12 workload, which is a narrow lens but consistent with a part aimed at compute-heavy or professional rendering tasks. The MX130, with 2 GB of GDDR5, a 64-bit bus, and 40.10 GB/s of bandwidth, is a mobile integrated-class part. Its benchmark results cover OpenCL and Vulkan, suggesting a focus on general-purpose compute and cross-platform graphics in thin laptops.

Where each wins depends on the workload context. In the Steel Nomad DX12 test, the RTX PRO 6000 scores 5996, which is 88 points above the MX130’s average of 5508 but 106 points below the MX130’s best OpenCL result of 6102. That is not a clean victory for either. The RTX PRO 6000 wins on raw hardware specifications by every metric that matters for server workloads, but the recorded benchmark scores tell a more tangled story. The MX130 wins in OpenCL throughput, at least as measured by Geekbench, and it wins on power efficiency by a wide margin: 30 W versus 600 W TDP.

The RTX PRO 6000 wins in memory capacity, bandwidth, shading units, texture units, ROPs, ray tracing cores, tensor cores, and pixel and texture fill rates. It also wins on process node (5 nm versus 28 nm), transistor count (92,200 million versus 1,020 million), and die size (750 mm² versus 77 mm²). Those are not benchmark scores, but they define the class of work each part can handle. The MX130 cannot address server-scale datasets; the RTX PRO 6000 cannot fit in a thin-and-light chassis.

The Verdict

The data points to two different buyers. The RTX PRO 6000 Blackwell Server is for anyone running large-scale rendering, AI inference, or simulation workloads that need 96 GB of memory and 1.79 TB/s of bandwidth. Its 600 W TDP, dual-slot cooler, and 16-pin power connector make it a workstation or server component, not a desktop gaming card. Its single 3DMark result of 5,996 shows it performs in line with parts like the GTX 770M and RX 6400 in that specific test, which is surprising given its massive hardware resources, but that test may not exercise the features that differentiate it, such as ray tracing cores or tensor cores.

The GeForce MX130 is for users of older or entry-level laptops who need basic GPU acceleration for OpenCL or Vulkan workloads. Its 30 W TDP means no external power connector, and its integrated form factor (IGP) means it is soldered into portable devices. Its average score of 5,508, with an OpenCL result of 6,102, makes it competitive with parts like the GTX 765M and R7 M440. Anyone choosing between these two is not actually comparing like-for-like products. The RTX PRO 6000 is a server accelerator; the MX130 is a laptop GPU from 2017. The former is active production, the latter is end-of-life.

If the choice is forced, the RTX PRO 6000 is the only one that can handle modern professional workloads, and the MX130 is the only one that fits in a low-power mobile device. The benchmark data alone does not settle the debate, because the tests do not overlap. The RTX PRO 6000 has no OpenCL or Vulkan scores, and the MX130 has no 3DMark score. That missing overlap means the database cannot directly compare them on identical workloads, so the verdict rests on specifications and intended use.

Head-to-Head Benchmarks

There are no direct head-to-head benchmark entries in the database for these two parts. The recorded results are separate tests with separate methodologies. The RTX PRO 6000’s only score is 5,996 in 3DMark Steel Nomad DX12. The MX130’s scores are 6,102 in Geekbench OpenCL and 4,914 in Geekbench Vulkan. Because these tests measure different things, any cross-comparison is indirect.

The largest gap in raw numbers appears in the OpenCL result. The MX130 scores 6,102, which is 106 points higher than the RTX PRO 6000’s 3DMark result. That is a 1.8% lead for the MX130 in a completely different benchmark. In Vulkan, the MX130 scores 4,914, which is 1,082 points lower than the RTX PRO 6000’s 3DMark score, a 18.0% deficit. But again, these are not the same workload.

Looking at the rivals, the RTX PRO 6000’s closest competitor is the RX 6400 at 6,001, which beats it by 0.1%. The MX130’s closest rival is the GTX 765M at 5,501, which the MX130 beats by 0.1%. The data suggests that each part sits at a performance plateau relative to its peers, but the plateaus are not comparable.

What the data does show is that the RTX PRO 6000’s massive compute resources do not translate into a dominant score in the one test recorded. Its 126.0 TFLOPS of FP32 and 502.5 GPixel/s pixel rate do not appear to be fully utilized in Steel Nomad, or the test is simply not representative of its intended workloads. The MX130’s 913.2 GFLOPS FP32 and 9.512 GPixel/s pixel rate are dwarfed by those figures, yet its OpenCL score is higher than the RTX PRO 6000’s 3DMark score. That inversion is a reminder that benchmark scores are workload-specific.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The RTX PRO 6000 Blackwell Server has an average benchmark score of 5,996, which is 488 points higher than the GeForce MX130’s average of 5,508.

Q: What is the RTX PRO 6000’s closest rival in the database?

A: The AMD Radeon RX 6400 has an average score of 6,001, which is 0.1% ahead of the RTX PRO 6000’s 5,996. The GeForce GTX 770M also sits 0.1% ahead at 6,000.

Q: What is the MX130’s best recorded benchmark result?

A: The MX130 scores 6,102 in Geekbench OpenCL, which is its higher result. Its Geekbench Vulkan score is 4,914.

Q: How do the memory configurations differ?

A: The RTX PRO 6000 has 96 GB of GDDR7 on a 512-bit bus with 1.79 TB/s bandwidth. The MX130 has 2 GB of GDDR5 on a 64-bit bus with 40.10 GB/s bandwidth.

Q: Which GPU has ray tracing and tensor cores?

A: The RTX PRO 6000 has 188 ray tracing cores and 752 tensor cores. The MX130 has no ray tracing cores and no tensor cores listed.

Q: What are the power requirements?

A: The RTX PRO 6000 has a 600 W TDP and requires a 16-pin power connector with a suggested 1000 W PSU. The MX130 has a 30 W TDP, no power connector, and no suggested PSU.

Architecture Differences

The two GPUs come from different architectural generations. The RTX PRO 6000 Blackwell Server uses the GB202 chip on the Blackwell 2.0 architecture, built on a 5 nm process at TSMC. The MX130 uses the GM108S chip on the Maxwell architecture, built on a 28 nm process at TSMC. The process node difference alone explains a large part of the capability gap: 5 nm versus 28 nm means higher transistor density and lower power per transistor.

Transistor counts differ by an order of magnitude. The RTX PRO 6000 packs 92,200 million transistors on a 750 mm² die, yielding a density of 122.9 million transistors per square millimeter. The MX130 has 1,020 million transistors on a 77 mm² die, for a density of 13.2 million per square millimeter. That is a 90-fold difference in transistor count and a 9.3-fold difference in density.

Memory architectures are also generations apart. The RTX PRO 6000 uses GDDR7 with a 512-bit bus and 1.79 TB/s bandwidth. The MX130 uses GDDR5 with a 64-bit bus and 40.10 GB/s bandwidth. The RTX PRO 6000 has 96 GB of memory, the MX130 has 2 GB. The bus width difference means the RTX PRO 6000 can move 8 times more data per clock, and the newer memory type doubles the effective data rate per pin.

The RTX PRO 6000 includes dedicated hardware that the MX130 lacks entirely: 188 ray tracing cores and 752 tensor cores. That makes it suitable for real-time ray tracing and AI-accelerated workloads. The MX130 has no such units, relying purely on its 384 shading units for all compute. The RTX PRO 6000 has 24,064 shading units, 752 texture mapping units, and 192 ROPs. The MX130 has 384 shading units, 24 TMUs, and 8 ROPs.

The API support differs as well. The RTX PRO 6000 supports DirectX 12 Ultimate (12_2), while the MX130 supports DirectX 12 (11_0). Both support OpenGL 4.6 and Vulkan 1.4. The RTX PRO 6000 also uses PCIe 5.0 x16, while the MX130 uses PCIe 3.0 x4. That means the RTX PRO 6000 has 4 times the lanes and a newer protocol, offering far more host bandwidth.

Specification Differences

The table below lists only the fields where the two parts differ, based on the recorded data.

| Specification | RTX PRO 6000 Blackwell Server | GeForce MX130 |

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

| Chip | GB202 | GM108S |

| Architecture | Blackwell 2.0 | Maxwell |

| Generation | Server Blackwell (Bxx) | GeForce MX (1xx) |

| Process Node | 5 nm | 28 nm |

| Transistors | 92,200 million | 1,020 million |

| Die Size | 750 mm² | 77 mm² |

| Transistor Density | 122.9M / mm² | 13.2M / mm² |

| Base Clock | 1590 MHz | 1109 MHz |

| Boost Clock | 2617 MHz | 1189 MHz |

| Memory Clock | 1750 MHz, 28 Gbps effective | 1253 MHz, 5 Gbps effective |

| Memory Size | 96 GB | 2 GB |

| Memory Type | GDDR7 | GDDR5 |

| Memory Bus Width | 512 bit | 64 bit |

| Memory Bandwidth | 1.79 TB/s | 40.10 GB/s |

| Shading Units | 24064 | 384 |

| TMUs | 752 | 24 |

| ROPs | 192 | 8 |

| RT Cores | 188 | None |

| Tensor Cores | 752 | None |

| Pixel Rate | 502.5 GPixel/s | 9.512 GPixel/s |

| Texture Rate | 1,968.0 GTexel/s | 28.54 GTexel/s |

| FP32 | 126.0 TFLOPS | 913.2 GFLOPS |

| FP16 | 126.0 TFLOPS (1:1) | None |

| TDP | 600 W | 30 W |

| Slot Width | Dual-slot | IGP |

| Power Connectors | 1x 16-pin | None |

| Suggested PSU | 1000 W | None |

| Bus Interface | PCIe 5.0 x16 | PCIe 3.0 x4 |

| Display Outputs | 4x DisplayPort 2.1b | Portable Device Dependent |

| DirectX | 12 Ultimate (12_2) | 12 (11_0) |

| Dimensions | 267 mm x 111 mm x 40 mm | None recorded |

| Production Status | Active | End-of-life |

| Release Date | 2025-03-17 | 2017-11-16 |

| Predecessor | Server Hopper | None |

| Successor | Server Rubin | None |

| Average Benchmark Score | 5996 | 5508 |

| Percentile vs All GPUs | 34 | 32 |