Intel Iris Pro Graphics P580 vs NVIDIA RTX PRO 5000 72 GB Blackwell Comparison
Intel Iris Pro Graphics P580
RTX PRO 5000 72 GB Blackwell
PERFORMANCE BENCHMARKS
Analysis: Intel Iris Pro Graphics P580 vs NVIDIA RTX PRO 5000 72 GB Blackwell
Intel Iris Pro Graphics P580 is an end-of-life integrated graphics solution from Intel’s Skylake generation, built on a 14 nm+ process for the HD Graphics-W family. NVIDIA RTX PRO 5000 72 GB Blackwell is an active, dual-slot workstation discrete GPU based on the GB202 chip, fabricated by TSMC on a 5 nm node. The two occupy opposite ends of the graphics spectrum: the P580 is a 15 W IGP with system-shared memory, while the RTX PRO 5000 is a 300 W card with 72 GB of dedicated GDDR7 memory and a 384-bit bus. Benchmark data shows a stark contrast in raw compute, but the average scores and percentile rankings tell a more nuanced story that requires careful reading of specific tests rather than a single aggregate number.
Where Each One Wins
The Intel Iris Pro Graphics P580 wins in the two benchmark entries recorded in its dataset: Geekbench OpenCL and Geekbench Vulkan. Its Geekbench OpenCL score is 9082, and its Geekbench Vulkan score is 5258. These are the only two data points available for the Intel part, and both represent compute-oriented workloads rather than traditional rasterization. The P580’s average benchmark score across these two tests is 7170, which places it at the 39th percentile of all GPUs tracked by the database.
The NVIDIA RTX PRO 5000 72 GB Blackwell wins in the Passmark suite, with a G3D score of 24310, a GPU Compute score of 15667, and a G2D score of 914. It also posts DirectX-specific Passmark scores: 311 for DirectX 9, 219 for DirectX 11, 175 for DirectX 10, and 78 for DirectX 12. Its 3DMark Steel Nomad DX12 score is 9579.5. However, its average benchmark score across all listed tests is only 6407, which is lower than the Intel part’s average of 7170. This is a critical divergence: the NVIDIA card’s average is dragged down by its low Passmark DirectX scores, while its G3D score is far higher than anything the Intel IGP can produce.
The data indicates a split based on workload type. For synthetic compute benchmarks like Geekbench OpenCL and Vulkan, the Intel P580 holds a lead in average score. For DirectX-based gaming and graphics throughput tests, the NVIDIA RTX PRO 5000 dominates with scores that are orders of magnitude higher in the G3D category. The RTX PRO 5000’s percentile ranking is 37, slightly below the Intel part’s 39, reflecting the distortion caused by its inconsistent Passmark sub-scores.
Architecture Differences
The architectural gap between these two GPUs is generational and fundamental. The Intel Iris Pro Graphics P580 uses the Skylake GT4e chip, belonging to Generation 9.0 graphics architecture, with a process node of 14 nm+ from Intel’s own foundry. It has 576 shading units, 72 texture mapping units, and only 9 raster operations pipelines. Its base clock is 350 MHz with a boost clock of 1000 MHz. The memory subsystem is entirely system-shared, with no dedicated VRAM, no fixed bus width, and bandwidth described as system dependent. The pixel rate is 9.000 GPixel/s, texture rate is 72.00 GTexel/s, and FP32 performance is 1,152.0 GFLOPS. FP16 performance is 2.304 TFLOPS with a 2:1 ratio. It supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3, and connects via Ring Bus.
The NVIDIA RTX PRO 5000 72 GB Blackwell is built on the Blackwell 2.0 architecture with the GB202 chip. It uses a 5 nm process from TSMC, packing 92,200 million transistors on a 750 mm² die, for a transistor density of 122.9 million per square millimeter. The GPU has 14,080 shading units, 440 TMUs, and 160 ROPs. It includes 110 ray tracing cores and 440 tensor cores, neither of which exists on the Intel part. Base clock is 1740 MHz, boost clock is 2377 MHz, and memory runs at 1750 MHz with 28 Gbps effective speed. Memory is 72 GB of GDDR7 on a 384-bit bus, delivering 1.34 TB/s of bandwidth. Pixel rate is 380.3 GPixel/s, texture rate is 1,045.9 GTexel/s, and FP32 performance is 66.94 TFLOPS. FP16 is also 66.94 TFLOPS with a 1:1 ratio, indicating no rate penalty for half-precision. It supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, and uses a PCIe 5.0 x16 interface. Power delivery requires a single 16-pin connector with a 300 W TDP and a suggested 700 W power supply.
The RTX PRO 5000 has 24.4 times the shading units, 6.1 times the TMUs, and 17.8 times the ROPs of the Intel part. Its FP32 throughput is 58.1 times higher. The clock speeds are also vastly different: the NVIDIA base clock is 5 times higher than the Intel boost clock, and the NVIDIA boost clock is 2.4 times the Intel boost. The memory bandwidth difference is effectively unbounded since the Intel part has no dedicated bandwidth figure, but the NVIDIA’s 1.34 TB/s is a fixed, dedicated resource while the Intel IGP depends entirely on system memory performance.
The Verdict
Based strictly on the benchmark data, the NVIDIA RTX PRO 5000 72 GB Blackwell is the superior choice for any workload that involves DirectX-based rendering or general 3D graphics throughput. Its Passmark G3D score of 24310 is the single highest recorded score in either dataset, and its GPU Compute score of 15667 demonstrates strong parallel processing capability. The presence of dedicated ray tracing cores and tensor cores, along with 72 GB of GDDR7 memory, positions it for professional workstation tasks that require large memory capacity and hardware-accelerated features. The data shows no scenario where the Intel P580 matches or exceeds these specific metrics.
However, the average benchmark score comparison is inverted. The Intel P580’s average of 7170 exceeds the NVIDIA’s 6407 average. This is because the NVIDIA card’s Passmark DirectX sub-scores (78 to 311) are anomalously low relative to its G3D score, and these low values pull the average down. The Intel part’s two Geekbench scores are both in the 5000 to 9000 range, so its average is more representative of consistent compute performance. A buyer prioritizing Geekbench OpenCL or Vulkan scores specifically would find the Intel part ahead, with 9082 and 5258 respectively, compared to no equivalent Geekbench scores for the NVIDIA card.
The verdict depends entirely on workload. For DirectX gaming or professional 3D rendering, the RTX PRO 5000 is the clear choice, despite its lower average. For compute-oriented synthetic tests, the Intel P580 wins on average, but this is a narrow victory based on a limited test set. The RTX PRO 5000’s percentile ranking of 37 versus the Intel’s 39 suggests that in the broader GPU landscape, neither is near the top, but the NVIDIA part’s raw capabilities in G3D and memory are unmatched by the Intel IGP. The Intel part is end-of-life, while the NVIDIA part is active, which further suggests the RTX PRO 5000 is the forward-looking option.
FAQ
Q: Which GPU has a higher average benchmark score?
A: The Intel Iris Pro Graphics P580 has an average benchmark score of 7170, while the NVIDIA RTX PRO 5000 72 GB Blackwell has an average of 6407.
Q: What is the highest single benchmark score recorded for each GPU?
A: The Intel P580’s highest score is 9082 in Geekbench OpenCL. The NVIDIA RTX PRO 5000’s highest score is 24310 in Passmark G3D.
Q: Does the NVIDIA GPU have ray tracing support?
A: Yes, the RTX PRO 5000 includes 110 ray tracing cores. The Intel P580 has no ray tracing cores listed.
Q: How do the memory configurations differ?
A: The Intel P580 uses system-shared memory with no dedicated size, type, or bus width. The NVIDIA RTX PRO 5000 has 72 GB of GDDR7 memory on a 384-bit bus with 1.34 TB/s bandwidth.
Q: What are the DirectX version support levels?
A: The Intel P580 supports DirectX 12 (12_1), while the NVIDIA RTX PRO 5000 supports DirectX 12 Ultimate (12_2).
Q: Which GPU has a higher percentile ranking among all GPUs?
A: The Intel P580 ranks at the 39th percentile, while the NVIDIA RTX PRO 5000 ranks at the 37th percentile.
Head-to-Head Benchmarks
The head-to-head benchmark list is empty, so direct comparisons must be drawn from the individual benchmark records. The most dramatic win for the NVIDIA RTX PRO 5000 is in Passmark G3D, where it scores 24310. This is the only score in either dataset that exceeds 10,000, and it represents a synthetic 3D graphics throughput test that heavily favors discrete GPUs with dedicated memory and high fill rates. The Intel P580 has no comparable G3D score, so the margin is effectively complete dominance in this category.
The NVIDIA card also wins in GPU Compute with 15667, a score that reflects general-purpose compute workloads. The Intel P580’s closest compute score is its Geekbench OpenCL result of 9082, which is 41.9% lower than the NVIDIA’s GPU Compute score. In DirectX-specific tests, the NVIDIA card’s scores are 311 (DX9), 219 (DX11), 175 (DX10), and 78 (DX12). These are low absolute numbers, but they are the only DirectX scores available in the dataset; the Intel P580 has no DirectX benchmark entries at all.
The Intel P580’s wins come in Geekbench OpenCL (9082) and Geekbench Vulkan (5258). The NVIDIA RTX PRO 5000 has no Geekbench scores recorded, so there is no direct comparison. However, the Intel’s Vulkan score of 5258 is lower than its OpenCL score, suggesting a performance gap between those APIs on the same hardware. The average benchmark scores favor the Intel part (7170 vs 6407), but this is misleading because the NVIDIA card’s average is depressed by its Passmark DirectX sub-scores, which are all below 311.
In terms of raw throughput metrics, the NVIDIA card’s FP32 of 66.94 TFLOPS dwarfs the Intel’s 1,152.0 GFLOPS. The pixel rate difference is 380.3 GPixel/s versus 9.000 GPixel/s, a 42.3 times advantage. Texture rate is 1,045.9 GTexel/s versus 72.00 GTexel/s, a 14.5 times advantage. These hardware specifications align with the benchmark scores, showing that the NVIDIA card is designed for high-throughput graphics, while the Intel IGP is a basic integrated solution for low-power systems.
Specification Differences
The two GPUs differ in nearly every measurable specification. Process node is 14 nm+ for Intel versus 5 nm for NVIDIA, with foundries of Intel and TSMC respectively. The NVIDIA chip contains 92,200 million transistors on a 750 mm² die, while the Intel chip has no transistor count or die size listed. Clock speeds differ substantially: Intel base is 350 MHz with a 1000 MHz boost, while NVIDIA base is 1740 MHz with a 2377 MHz boost. The Intel memory is system-shared with no size, type, or bus width, whereas the NVIDIA card has 72 GB GDDR7 on a 384-bit bus with 1.34 TB/s bandwidth.
Shading units are 576 on Intel versus 14,080 on NVIDIA. TMUs are 72 versus 440, and ROPs are 9 versus 160. The NVIDIA card has 110 RT cores and 440 tensor cores; the Intel card has none. Pixel rate is 9.000 GPixel/s versus 380.3 GPixel/s. Texture rate is 72.00 GTexel/s versus 1,045.9 GTexel/s. FP32 is 1,152.0 GFLOPS versus 66.94 TFLOPS. FP16 is 2.304 TFLOPS (2:1) versus 66.94 TFLOPS (1:1). TDP is 15 W versus 300 W. Slot width is IGP versus dual-slot. Power connectors are absent on Intel, while NVIDIA requires 1x 16-pin with a suggested 700 W PSU. Bus interface is Ring Bus versus PCIe 5.0 x16. Display outputs are motherboard dependent versus 4x DisplayPort 2.1b. API support shows DirectX 12 (12_1) versus 12 Ultimate (12_2), Vulkan 1.3 versus 1.4, and OpenGL 4.6 for both. Production status is end-of-life versus active. Release dates are 2015-08-31 versus 2025-10-20. The NVIDIA card has a predecessor named Workstation Ada, while the Intel part has no predecessor or successor listed.