NVIDIA GeForce MX570 vs NVIDIA P104-100 Comparison

The NVIDIA GeForce MX570 and NVIDIA P104-100 occupy very different corners of the GPU landscape. The MX570 is a low-power Ampere-based mobile chip, while the P104-100 is a Pascal-based mining card with no display outputs. The recorded data shows a single common benchmark, Geekbench OpenCL, and the results there are decisive. This analysis walks through where each part wins, what the numbers mean, and who should consider which, based strictly on the measurements in the database.

Where Each One Wins

The P104-100 wins the only head-to-head benchmark available. In Geekbench OpenCL, the P104-100 scores 52368, while the MX570 scores 38299. That is a delta of -26.9% from the perspective of the MX570, meaning the P104-100 is about 27% ahead in this specific OpenCL workload. This is a substantial margin, not a marginal one. The P104-100 also has additional benchmark entries in the database, including a 3DMark Steel Nomad DX12 score of 1413 and a Geekbench Vulkan score of 45165, while the MX570 has no recorded scores in those tests. So, in terms of pure performance breadth and peak OpenCL throughput, the P104-100 is the clear winner.

However, the MX570 is not without its own advantages, though they are not visible in raw benchmark scores. The MX570 is an integrated graphics package (IGP) with a 15 W TDP, no power connectors, and slot width listed as IGP. The P104-100 is a dual-slot card with a 1x 8-pin power connector and a suggested PSU of 200 W. The MX570 draws a fraction of the power, and its portable device dependent display outputs mean it is designed for laptops or small form factor systems. The P104-100 has no display outputs at all, so it cannot drive a monitor. For any workload that requires visual output, the MX570 is the only viable option.

The benchmark data also places each card within a percentile ranking. The MX570 sits at the 81st percentile of all GPUs, while the P104-100 sits at the 77th percentile. Despite the P104-100 winning the head-to-head OpenCL test, the MX570 has a higher percentile ranking, which suggests that the MX570's single benchmark score of 38299 is relatively strong compared to the entire database, while the P104-100's average benchmark score of 32982 is dragged down by its lower scores in other tests. The P104-100's average score includes all three of its benchmarks, and the 3DMark Steel Nomad score of 1413 is much lower than its OpenCL score, pulling the average down.

The Verdict

For compute throughput in OpenCL, the P104-100 is the stronger part. Its score of 52368 is 27% higher than the MX570's 38299, and it also shows a Vulkan score of 45165, which the MX570 does not have recorded. The P104-100 also offers 4 GB of GDDR5X memory on a 256-bit bus, providing 320.3 GB/s of bandwidth, compared to the MX570's 2 GB of GDDR6 on a 64-bit bus with 96.00 GB/s. If the task is raw number crunching and memory bandwidth is critical, the P104-100 is the choice.

For any system that needs a display output, the P104-100 is disqualified. Its display outputs are listed as "No outputs," so it cannot be used as a primary or secondary graphics card for a monitor. The MX570, with its portable device dependent outputs, is designed to drive displays. Additionally, the MX570's power profile is dramatically lower: 15 W TDP versus the P104-100's suggested 200 W PSU. The MX570 requires no external power connector, while the P104-100 needs a single 8-pin connector. For a laptop or a compact system with limited power delivery, the MX570 is the only realistic option.

The percentile data complicates the verdict. The MX570 ranks at the 81st percentile, above the P104-100's 77th percentile. This means that in the broader context of all GPUs in the database, the MX570's single OpenCL score of 38299 is relatively better positioned than the P104-100's average score of 32982. The P104-100's average is pulled down by its 3DMark Steel Nomad score of 1413, which is far lower than its OpenCL score. So, while the P104-100 wins the direct comparison, the MX570 is more consistently strong across the database's scoring system.

The verdict is straightforward: the P104-100 wins on raw compute and memory bandwidth, but only for headless compute tasks. The MX570 wins on power efficiency, display capability, and overall percentile ranking. The user who needs a mining or compute card with no display will prefer the P104-100. The user who needs a mobile GPU for a laptop with a screen will prefer the MX570.

Head-to-Head Benchmarks

The database contains exactly one head-to-head benchmark between these two GPUs: Geekbench OpenCL. In that test, the P104-100 scores 52368, and the MX570 scores 38299. The delta is -26.9% from the MX570's side, meaning the MX570 trails by roughly 27%. This is a significant gap. To put it in context, the MX570's nearest rival is the NVIDIA GeForce RTX 5080 Mobile with an average score of 38349, which is only 0.1% ahead, and the NVIDIA GeForce RTX 4080 Mobile with an average score of 38135, which is 0.4% behind. The MX570's OpenCL score sits right between these two high-end mobile parts. This indicates that the MX570's OpenCL performance is competitive with flagship mobile GPUs, despite its low power draw.

The P104-100's nearest rivals tell a different story. Its average benchmark score of 32982 is close to the NVIDIA T600 Mobile (32849, 0.4% ahead), the NVIDIA T550 Mobile (33161, 0.5% behind), the NVIDIA GeForce RTX 3050 Mobile (33170, 0.6% behind), and the AMD Radeon Pro 570 (33207, 0.7% behind). However, the P104-100's OpenCL score of 52368 is far above its own average, which means that in the OpenCL-specific test, it is much faster than its overall average suggests. The P104-100's 3DMark Steel Nomad score of 1413 and Vulkan score of 45165 are the reason its average is lower.

The key takeaway from the head-to-head is that the P104-100's advantage in OpenCL is not a small edge. A 27% delta is substantial and likely reflects the difference in memory bandwidth (320.3 GB/s versus 96.00 GB/s) and raw FP32 throughput (6.655 TFLOPS versus 4.731 TFLOPS). The MX570 does have a higher percentile ranking (81st versus 77th), but that is based on the MX570's single score, not on a comparison of the two in the same test.

FAQ

Q: Which GPU has the higher Geekbench OpenCL score?

A: The NVIDIA P104-100 scores 52368, while the NVIDIA GeForce MX570 scores 38299. The P104-100 is 26.9% ahead.

Q: Can the NVIDIA P104-100 connect to a monitor?

A: No. The P104-100 lists "No outputs" for display outputs, so it cannot drive any display. The MX570 has "Portable Device Dependent" outputs, meaning it is designed for portable devices with integrated displays.

Q: How do the memory configurations compare?

A: The P104-100 has 4 GB of GDDR5X memory on a 256-bit bus, providing 320.3 GB/s of bandwidth. The MX570 has 2 GB of GDDR6 memory on a 64-bit bus, providing 96.00 GB/s of bandwidth.

Q: What are the power requirements for each?

A: The MX570 has a TDP of 15 W and requires no power connectors. The P104-100 has no listed TDP, but it requires a 1x 8-pin power connector and a suggested PSU of 200 W.

Q: Which GPU has a higher percentile ranking among all GPUs?

A: The MX570 ranks at the 81st percentile, while the P104-100 ranks at the 77th percentile.

Q: Does the P104-100 support ray tracing?

A: No. The P104-100 has no RT cores and no tensor cores. The MX570, based on the Ampere architecture, has 16 RT cores and 64 tensor cores.

Architecture Differences

The two GPUs are built on different architectures and process nodes. The MX570 uses the GA107S chip, which is based on the Ampere architecture, manufactured on an 8 nm process at Samsung. The P104-100 uses the GP104 chip, based on the older Pascal architecture, manufactured on a 16 nm process at TSMC.

The transistor counts are close but not identical. The MX570 has 8,700 million transistors on a 200 mm² die, giving a transistor density of 43.5 million per mm². The P104-100 has 7,200 million transistors on a larger 314 mm² die, giving a transistor density of 22.9 million per mm². The MX570 packs more transistors into a smaller area, which is a direct result of the more advanced 8 nm process.

The compute resources differ significantly. The MX570 has 2048 shading units, 64 texture mapping units (TMUs), and 32 raster output units (ROPs). It also includes 16 RT cores and 64 tensor cores, which are absent from the P104-100. The P104-100 has 1920 shading units, 120 TMUs, and 64 ROPs. So, while the MX570 has more shading units, the P104-100 has nearly double the TMUs and double the ROPs. This explains the P104-100's higher texture rate (208.0 GTexel/s versus 73.92 GTexel/s) and pixel rate (110.9 GPixel/s versus 36.96 GPixel/s).

The FP32 performance favors the P104-100: 6.655 TFLOPS versus 4.731 TFLOPS. The FP16 performance is starkly different. The MX570 achieves 4.731 TFLOPS with a 1:1 ratio, meaning it can do FP16 at the same rate as FP32. The P104-100 achieves only 104.0 GFLOPS with a 1:64 ratio, meaning its FP16 performance is severely crippled. This is a major architectural difference, as the MX570's Ampere design supports full-rate FP16, while the Pascal-based P104-100 does not.

The API support also differs. The MX570 supports DirectX 12 Ultimate (12_2), while the P104-100 only supports DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4. The MX570's newer architecture allows for DirectX 12 Ultimate features, which the P104-100 cannot offer.

Specification Differences

The memory subsystem is a major differentiator. The MX570 has 2 GB of GDDR6 memory on a 64-bit bus, with a bandwidth of 96.00 GB/s and a memory clock of 1500 MHz (12 Gbps effective). The P104-100 has 4 GB of GDDR5X memory on a 256-bit bus, with a bandwidth of 320.3 GB/s and a memory clock of 1251 MHz (10 Gbps effective). The P104-100 has double the memory capacity, four times the bus width, and over three times the bandwidth.

The clock speeds are also different. The MX570 has a base clock of 832 MHz and a boost clock of 1155 MHz. The P104-100 has a base clock of 1607 MHz and a boost clock of 1733 MHz. Despite the MX570's lower clocks, its higher shading unit count (2048 versus 1920) does not compensate for the clock deficit in FP32 throughput.

The physical specifications are clearly different. The MX570 is an IGP with a slot width of "IGP," no power connectors, and a TDP of 15 W. The P104-100 is a dual-slot card with a length of 267 mm (10.5 inches), a single 8-pin power connector, and a suggested PSU of 200 W. The bus interface also differs: the MX570 uses PCIe 4.0 x8, while the P104-100 uses PCIe 1.0 x4, which is a much older and slower interface.

The release dates are far apart. The MX570 was released on December 16, 2021, while the P104-100 was released on December 11, 2017, a four-year gap. Both are end-of-life products. The P104-100 belongs to the "Mining GPUs" generation, while the MX570 belongs to the "GeForce MX (5xx)" generation. The P104-100 has no display outputs, while the MX570 has portable device dependent outputs. These specification differences directly explain the benchmark results and the intended use cases for each card.