Intel Iris Pro Graphics P580 vs Intel UHD Graphics 750 Comparison

Head-to-Head Benchmarks

The benchmark data splits cleanly by API. In Geekbench OpenCL, the Intel Iris Pro Graphics P580 dominates with a score of 9,082 against the UHD Graphics 750's 6,743. That is a 25.8% deficit for the UHD 750—a substantial gap that reflects the P580's much larger execution resource pool. The P580 wins this test outright, and the margin is large enough that OpenCL workloads will feel materially faster on the older part.

Flip to Geekbench Vulkan, and the story reverses completely. The UHD Graphics 750 scores 8,138, while the Iris Pro P580 manages only 5,258. That is a 54.8% advantage for the UHD 750—a crushing win in the newer graphics API. Where the P580 leads by roughly a quarter in OpenCL, the UHD 750 leads by more than half in Vulkan. The delta is asymmetric, and it points to a fundamental generational shift in how these two iGPUs handle modern compute paths.

The average benchmark scores tell a more nuanced story. The UHD 750's average is 7,441, while the P580's is 7,170. That puts the UHD 750 about 3.8% ahead overall, despite losing the OpenCL test by a wide margin. The UHD 750's Vulkan win is so large that it drags the average above the P580's, even though the P580 has the higher single-test peak. If you weight both APIs equally, the UHD 750 is the better all-rounder, but only just.

Nearest-rival context reinforces the picture. The UHD 750 sits within 0.4% of the NVIDIA GeForce GTX 1650 and within 1.4% of the Intel Arc A310, based on average scores. The P580, by contrast, ties the NVIDIA GeForce GTX 560 SE at 0% delta and trails the AMD Radeon Vega 8 Mobile by 0.5%. Neither iGPU is breaking into upper-tier discrete territory, but the UHD 750 is trading blows with much more recent hardware, while the P580 is competing with older or lower-end parts.

In short: the P580 wins the compute-throughput test by a mile, the UHD 750 wins the modern-graphics test by a bigger mile, and the average score slightly favors the newer chip. Pick your API, and you pick your winner.

FAQ

Q: Which GPU has the higher peak benchmark score?

A: The Intel Iris Pro Graphics P580, with a Geekbench OpenCL score of 9,082. The UHD Graphics 750's best single score is 8,138 in Vulkan.

Q: Why does the UHD 750 win the Vulkan test so decisively?

A: The UHD 750 scores 8,138 versus 5,258 for the P580, a 54.8% advantage. This aligns with its Generation 12.1 architecture, which supports Vulkan 1.4, whereas the P580's Generation 9.0 architecture supports only Vulkan 1.3.

Q: Is the Iris Pro P580 faster in every test?

A: No. It wins Geekbench OpenCL by 25.8%, but loses Geekbench Vulkan by 54.8%. Each GPU wins exactly one of the two head-to-head tests.

Q: How do these iGPUs compare to discrete graphics cards?

A: The UHD 750's average score of 7,441 puts it 0.4% behind the NVIDIA GeForce GTX 1650 and 1.4% behind the Intel Arc A310. The P580's average of 7,170 ties the NVIDIA GeForce GTX 560 SE and trails the NVIDIA GeForce GTX 750 by 0.7%.

Q: Which GPU has a higher average benchmark score?

A: The UHD Graphics 750, with 7,441 versus 7,170 for the Iris Pro P580. That is a roughly 3.8% edge in combined performance.

Q: Do both GPUs support the same DirectX and OpenGL versions?

A: Yes. Both support DirectX 12 (12_1) and OpenGL 4.6. The difference is Vulkan: UHD 750 supports 1.4, P580 supports 1.3.

Architecture Differences

The two iGPUs come from different Intel generations entirely. The UHD Graphics 750 is built on the Rocket Lake chip with Generation 12.1 architecture, fabricated on Intel's 14 nm+++ process. The Iris Pro P580 uses the Skylake GT4e chip with Generation 9.0 architecture, on a slightly less refined 14 nm+ node. Both are 15 W parts with a Ring Bus interface and motherboard-dependent display outputs, but the internal designs diverge sharply.

The P580 packs significantly more execution hardware. It has 576 shading units, 72 texture mapping units, and 9 ROPs. The UHD 750, by contrast, offers 256 shading units, 16 TMUs, and 8 ROPs. That is more than double the shader count and over four times the TMU count for the P580. This explains its OpenCL dominance: raw throughput is much higher. The P580's pixel rate is 9.000 GPixel/s, while the UHD 750's is 10.40 GPixel/s—the newer part actually wins on pixel fill despite fewer ROPs. Texture rate goes the other way: 72.00 GTexel/s for the P580 versus 20.80 GTexel/s for the UHD 750, a 3.5x gap in the P580's favor.

Compute throughput follows the same pattern. The P580 delivers 1,152.0 GFLOPS FP32 and 2.304 TFLOPS FP16 (2:1). The UHD 750 manages 665.6 GFLOPS FP32 and 1,331.2 GFLOPS FP16 (2:1). The P580 leads by roughly 73% in FP32 and 73% in FP16, which is consistent with its shader count advantage. Clock speeds partially compensate: the UHD 750 boosts to 1300 MHz from a 300 MHz base, while the P580 boosts to 1000 MHz from a 350 MHz base. The UHD 750's higher boost clock helps, but not enough to close the execution-unit gap.

Memory is system shared on both, with system-dependent bandwidth and a system-shared bus width. Neither has dedicated VRAM. The UHD 750's memory clock is also listed as "System Shared," meaning there is no fixed memory speed to compare. Both are end-of-life IGP solutions, with the UHD 750 released in March 2021 and the P580 in August 2015. The UHD 750 is the newer architecture by nearly six years, which shows in its Vulkan support (1.4 versus 1.3) and its modern API performance.

The architectural takeaway: the P580 is a brute-force compute engine with a massive shader array, while the UHD 750 is a leaner, more modern design that trades raw throughput for better API efficiency and higher clock speeds. The 14 nm+++ process of the UHD 750 and its Generation 12.1 design are the key differentiators, not raw transistor counts or die size—those are not specified.

The Verdict

The data supports a clear split decision. If your workload is OpenCL-heavy—think compute tasks, older game engines, or productivity apps that still use that API—the Intel Iris Pro P580 is the stronger choice. Its 9,082 OpenCL score beats the UHD 750 by 25.8%, and its 576 shading units provide the raw muscle for parallel compute. The P580 also wins on texture throughput (72.00 GTexel/s) and FP32 compute (1,152.0 GFLOPS), so any task that scales with shader count will favor it.

If your workload leans on Vulkan—modern games, newer emulators, or any recent graphics stack—the Intel UHD Graphics 750 is the clear winner. Its 8,138 Vulkan score beats the P580 by 54.8%, and its Vulkan 1.4 support is a generation ahead of the P580's 1.3. The UHD 750 also has the higher average benchmark score (7,441 versus 7,170), higher pixel rate (10.40 GPixel/s versus 9.000 GPixel/s), and a higher boost clock (1300 MHz versus 1000 MHz).

Who should pick which? If you are running a system that uses Vulkan as the primary graphics API, the UHD 750 is the only sensible pick—the 54.8% delta is too large to ignore. If you are stuck with OpenCL-only software, the P580's 25.8% lead makes it the better workhorse, despite its age. For mixed use, the UHD 750's higher average score tips the balance, but only slightly. Neither GPU will satisfy a gamer seeking discrete-level performance—the UHD 750 sits within 0.4% of a GeForce GTX 1650, and the P580 ties a GTX 560 SE—so manage expectations accordingly.

The verdict is not a blanket recommendation. It is a conditional one: modern API workloads go to the UHD 750, legacy compute workloads go to the P580. Choose based on your software stack, not on brand loyalty or release date.

Specification Differences

• Shading Units: 256 (UHD 750) vs 576 (P580)
• Texture Mapping Units: 16 (UHD 750) vs 72 (P580)
• ROPs: 8 (UHD 750) vs 9 (P580)
• Base Clock: 300 MHz (UHD 750) vs 350 MHz (P580)
• Boost Clock: 1300 MHz (UHD 750) vs 1000 MHz (P580)
• Pixel Rate: 10.40 GPixel/s (UHD 750) vs 9.000 GPixel/s (P580)
• Texture Rate: 20.80 GTexel/s (UHD 750) vs 72.00 GTexel/s (P580)
• FP32 Performance: 665.6 GFLOPS (UHD 750) vs 1,152.0 GFLOPS (P580)
• FP16 Performance: 1,331.2 GFLOPS (UHD 750) vs 2.304 TFLOPS (P580)
• Architecture: Generation 12.1 (UHD 750) vs Generation 9.0 (P580)
• Process Node: 14 nm+++ (UHD 750) vs 14 nm+ (P580)
• Chip: Rocket Lake (UHD 750) vs Skylake GT4e (P580)
• Vulkan Support: 1.4 (UHD 750) vs 1.3 (P580)
• Release Date: March 2021 (UHD 750) vs August 2015 (P580)

Both share: 15 W TDP, IGP slot width, Ring Bus interface, motherboard-dependent display outputs, system-shared memory type/size/bus width, system-dependent bandwidth, DirectX 12 (12_1), OpenGL 4.6, and end-of-life production status. Neither has a launch MSRP listed.

Where Each One Wins

Intel UHD Graphics 750 wins in: Vulkan workloads, by a 54.8% margin. It also wins on pixel fill rate (10.40 GPixel/s), boost clock (1300 MHz), average benchmark score (7,441), and overall percentile ranking (40th vs 39th). Its nearest-rival comparison is more flattering: it sits within 0.4% of a GeForce GTX 1650, a much more modern discrete card than the P580's nearest rivals. This is the GPU for anyone running Vulkan-based games or applications, where its 1.4 API support gives it a decisive edge.

Intel Iris Pro P580 wins in: OpenCL workloads, by a 25.8% margin. It also wins on shading units (576 vs 256), TMUs (72 vs 16), texture rate (72.00 GTexel/s), FP32 compute (1,152.0 GFLOPS), FP16 compute (2.304 TFLOPS), and base clock (350 MHz). Its OpenCL score of 9,082 is the highest single benchmark result of either GPU. This is the GPU for compute-heavy tasks that still rely on OpenCL, where its massive shader array and texture throughput provide a clear advantage. It also edges out the UHD 750 on ROPs (9 vs 8), though the impact is minor.

Tie or near-tie: Both have 15 W TDP, system-shared memory, and identical DirectX/OpenGL support. The average score difference is small enough (3.8%) that neither GPU is a runaway winner for general-purpose use. The P580's nearest rival is a GTX 560 SE at 0% delta, while the UHD 750's nearest rival is an HD 8850M at -0.1%—both are effectively at parity with their closest competitors. Choose by workload, not by spec sheet.