Intel UHD Graphics P750 vs NVIDIA GeForce GTX 580M Comparison

Intel UHD Graphics P750 and NVIDIA GeForce GTX 580M are both end-of-life mobile graphics solutions, but they represent fundamentally different design philosophies. The Intel part is a 15 W integrated GPU from the Rocket Lake generation, while the NVIDIA chip is a 100 W discrete MXM module from the Fermi 2.0 era. Benchmark data shows the Intel UHD Graphics P750 edges out the GTX 580M in the single available compute test, but the underlying specifications tell a more nuanced story about which part suits different workloads.

Head-to-Head Benchmarks

The only direct benchmark comparison available is the Geekbench OpenCL test, and the results are remarkably close. The Intel UHD Graphics P750 scores 6,554 points, while the NVIDIA GeForce GTX 580M scores 6,389 points. That gives Intel a 2.6% advantage in this single test, a margin that is statistically minor but consistent with the percentile rankings — Intel sits at the 38th percentile of all GPUs, while NVIDIA sits just one point behind at the 37th percentile.

Looking at the nearest rivals for each card puts these scores in context. The Intel P750’s closest competitor is the AMD Radeon HD 7730M, which scores 6,581 points — a mere 0.4% difference. On the other side, the NVIDIA GTX 580M is virtually tied with the NVIDIA GeForce GTX 460 SE at 6,389 points, showing a 0% delta. The Intel part also beats the NVIDIA GeForce GTX 670M (6,513 points) by 0.6%, while the GTX 580M trails the AMD Radeon Pro WX 4100 (6,330 points) by 0.9% and the AMD Radeon R7 M350 (6,327 points) by 1%. These are all sub-2% deltas, meaning neither GPU has a decisive raw compute advantage in this OpenCL workload.

What is notable is the comparison between the two cards’ nearest rival lists. The Intel P750 sits alongside mid-range mobile parts like the GTX 670M and GT 555M, while the GTX 580M’s rivals include a professional workstation card (Radeon Pro WX 4100) and even an enterprise data center GPU (RTX PRO 5000 72 GB Blackwell). This suggests the GTX 580M’s score is more impressive relative to its age, as it competes with much newer and more expensive hardware. However, the Intel part’s 2.6% win is the only head-to-head data point, so the verdict remains narrow.

Where Each One Wins

The Intel UHD Graphics P750 wins the Geekbench OpenCL test outright, and it does so while consuming far less power — 15 W versus 100 W for the GTX 580M. This makes the Intel part the clear choice for scenarios where power efficiency matters, such as thin-and-light laptops or systems with tight thermal budgets. The Intel GPU also supports DirectX 12 (12_1) and Vulkan 1.4, whereas the GTX 580M only manages DirectX 12 (11_0) and has no Vulkan support listed. For modern API compatibility, Intel is the winner.

The NVIDIA GeForce GTX 580M, despite losing the single benchmark, holds advantages in several hardware categories that the OpenCL score does not capture. It has 384 shading units compared to Intel’s 256, and its FP32 throughput is 952.3 GFLOPS versus Intel’s 665.6 GFLOPS — a 43% raw compute advantage. The GTX 580M also has dedicated 2 GB GDDR5 memory on a 256-bit bus with 96.00 GB/s bandwidth, while the Intel P750 relies on system shared memory with bandwidth described as “System Dependent.” For memory-intensive tasks like texture streaming or large frame buffers, the NVIDIA part is clearly better equipped.

The GTX 580M also has a higher boost clock in practice, though the FACT PACK does not list a specific boost frequency for NVIDIA. Its 40 nm TSMC process node is older than Intel’s 14 nm+++, but the larger die size (332 mm² versus unspecified for Intel) and 1,950 million transistors suggest a more complex, higher-throughput design. The GTX 580M’s MXM-B (3.0) bus interface and MXM Module slot width indicate it was designed for upgradeable gaming laptops, while the Intel P750 is an IGP soldered to the motherboard.

Architecture Differences

The architectural divide between these two GPUs is stark. The Intel UHD Graphics P750 is built on Intel’s Generation 12.1 architecture, fabricated on a 14 nm+++ process at Intel’s own foundry. It features 256 shading units, 64 TMUs, and 32 ROPs, with a base clock of 350 MHz and a boost clock of 1300 MHz. Its pixel rate is 41.60 GPixel/s and texture rate is 83.20 GTexel/s, both significantly higher than the NVIDIA part. The Intel GPU’s FP16 performance is 1,331.2 GFLOPS (2:1 ratio), which suggests some compute flexibility for mixed-precision workloads.

The NVIDIA GeForce GTX 580M uses the Fermi 2.0 architecture on a 40 nm process at TSMC. It has 384 shading units, 64 TMUs, and 32 ROPs, with memory clocked at 750 MHz (3 Gbps effective). Its pixel rate is only 9.920 GPixel/s and texture rate is 39.68 GTexel/s — roughly a quarter and half of Intel’s rates, respectively. However, its FP32 output of 952.3 GFLOPS exceeds Intel’s 665.6 GFLOPS. The GTX 580M has no listed FP16 support, while Intel provides a 2:1 FP16 ratio.

Memory architecture is the biggest differentiator. The GTX 580M has a 256-bit memory bus with dedicated GDDR5, delivering 96.00 GB/s of bandwidth. The Intel P750 uses system shared memory with no dedicated VRAM, and its bandwidth is dependent on the host system’s memory configuration. This means the GTX 580M will maintain consistent performance regardless of system RAM, while the Intel part’s performance can vary widely based on the laptop’s memory setup. The GTX 580M also lists transistor count (1,950 million) and die size (332 mm²) with a density of 5.9M / mm², while Intel provides none of those figures.

The Verdict

From the data, the Intel UHD Graphics P750 is the better choice for users who prioritize modern API support, power efficiency, and integrated simplicity. It wins the only benchmark test, supports DirectX 12_1 and Vulkan 1.4, and consumes just 15 W — a fraction of the GTX 580M’s 100 W TDP. Its pixel and texture fill rates are dramatically higher, and its 14 nm+++ process node is more advanced than the 40 nm node used by NVIDIA. For general productivity, light gaming, or media playback in a modern ultraportable, the Intel P750 is the more practical option.

However, the NVIDIA GeForce GTX 580M is the stronger choice for compute-heavy tasks that require sustained FP32 performance. With 384 shading units and 952.3 GFLOPS of FP32 throughput, it outperforms the Intel part by 43% in raw compute — a gap that the OpenCL benchmark does not reflect, likely due to driver optimizations or memory bandwidth limitations. The GTX 580M’s dedicated 2 GB GDDR5 memory with 96.00 GB/s bandwidth is a massive advantage for texture-heavy games or GPU-accelerated applications that cannot rely on shared system memory. Its MXM form factor also means it was designed for high-performance gaming laptops where upgradability was expected.

For most modern users, the Intel P750 wins on compatibility and efficiency. For retro gaming laptops or legacy CUDA workloads, the GTX 580M retains a niche advantage in raw compute and memory bandwidth. The 2.6% benchmark delta is too small to declare a decisive winner on performance alone — the choice comes down to system context and workload.

FAQ

Q: Which GPU scores higher in Geekbench OpenCL?

A: The Intel UHD Graphics P750 scores 6,554 points, which is 2.6% higher than the NVIDIA GeForce GTX 580M’s 6,389 points.

Q: Does the NVIDIA GTX 580M have more shading units?

A: Yes, the GTX 580M has 384 shading units, while the Intel UHD Graphics P750 has 256 shading units.

Q: What is the TDP difference between the two cards?

A: The Intel UHD Graphics P750 has a 15 W TDP, while the NVIDIA GeForce GTX 580M has a 100 W TDP.

Q: Which GPU supports Vulkan?

A: The Intel UHD Graphics P750 supports Vulkan 1.4, but the NVIDIA GeForce GTX 580M has no Vulkan support listed.

Q: How much dedicated memory does the GTX 580M have?

A: The GTX 580M has 2 GB of GDDR5 memory on a 256-bit bus, with 96.00 GB/s bandwidth, while the Intel P750 uses system shared memory.

Q: What is the FP32 performance difference?

A: The NVIDIA GTX 580M delivers 952.3 GFLOPS of FP32 performance, which is about 43% higher than the Intel P750’s 665.6 GFLOPS.

Specification Differences

| Specification | Intel UHD Graphics P750 | NVIDIA GeForce GTX 580M |

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

| Architecture | Generation 12.1 | Fermi 2.0 |

| Process Node | 14 nm+++ | 40 nm |

| Foundry | Intel | TSMC |

| Transistors | Not listed | 1,950 million |

| Die Size | Not listed | 332 mm² |

| Transistor Density | Not listed | 5.9M / mm² |

| Base Clock | 350 MHz | Not listed |

| Boost Clock | 1300 MHz | Not listed |

| Memory Size | System Shared | 2 GB |

| Memory Type | System Shared | GDDR5 |

| Memory Bus Width | System Shared | 256 bit |

| Memory Bandwidth | System Dependent | 96.00 GB/s |

| Shading Units | 256 | 384 |

| TMUs | 64 | 64 |

| ROPs | 32 | 32 |

| Pixel Rate | 41.60 GPixel/s | 9.920 GPixel/s |

| Texture Rate | 83.20 GTexel/s | 39.68 GTexel/s |

| FP32 Performance | 665.6 GFLOPS | 952.3 GFLOPS |

| FP16 Performance | 1,331.2 GFLOPS (2:1) | Not listed |

| TDP | 15 W | 100 W |

| Slot Width | IGP | MXM Module |

| Bus Interface | Ring Bus | MXM-B (3.0) |

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

| OpenGL Support | 4.6 | 4.6 |

| Vulkan Support | 1.4 | Not listed |

| Power Connectors | Not listed | None |

| Display Outputs | Motherboard Dependent | Portable Device Dependent |

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

| Release Date | Not listed | 2011-06-27 |

| Predecessor | Not listed | GeForce 400M |

| Successor | Not listed | GeForce 600M |