Intel Iris Xe Graphics 80EU Mobile vs NVIDIA GeForce RTX 5090 SE Comparison

Head-to-Head Benchmarks

The recorded benchmark data for this comparison is sparse, with no head-to-head benchmark entries, no wins recorded for either side, and no average benchmark scores populated. Both the Intel Iris Xe Graphics 80EU Mobile and the NVIDIA GeForce RTX 5090 SE sit at the 50th percentile against all GPUs in the database, a placeholder value that indicates neither part has been measured yet in the current dataset. Without concrete benchmark scores, the analysis must rely entirely on the architectural specifications provided.

The raw compute figures show an enormous gulf between the two. The Intel Iris Xe delivers 1.856 TFLOPS of FP32 throughput, while the NVIDIA RTX 5090 SE delivers 66.94 TFLOPS. That places the NVIDIA part at roughly 36 times the FP32 output of the Intel integrated solution, a gap that reflects the difference between an IGP designed for a 15 W envelope and a discrete dual-slot card rated for 500 W. Pixel throughput tells a similar story: the Intel part renders 29.00 GPixel/s, while the RTX 5090 SE reaches 380.3 GPixel/s, a factor of about 13. Texture rate widens the divide further, with Intel at 58.00 GTexel/s versus the NVIDIA card’s 1,045.9 GTexel/s, a difference of roughly 18 times.

The FP16 comparison is asymmetric. Intel delivers 3.712 TFLOPS at a 2:1 ratio, meaning half-rate FP16 relative to FP32. NVIDIA delivers 66.94 TFLOPS at a 1:1 ratio, meaning full-rate FP16 matching its FP32 throughput. In absolute terms, NVIDIA’s FP16 output is about 18 times higher than Intel’s. This matters for workloads that leverage half-precision math, such as certain machine learning inference tasks, although the RTX 5090 SE also carries dedicated tensor cores that the Intel part lacks entirely.

Clock speeds differ substantially. The Intel IGP runs at a base of 300 MHz with a boost of 1450 MHz. The RTX 5090 SE operates at a base of 1740 MHz and boosts to 2377 MHz. Even at boost, the NVIDIA card runs over 900 MHz higher than the Intel part’s maximum, and its base clock exceeds Intel’s boost by 290 MHz. The NVIDIA card’s memory clock is listed at 1750 MHz with 28 Gbps effective data rate, while Intel uses system-shared memory with no dedicated clock information provided.

Memory capacity and bandwidth are not comparable in any practical sense. Intel uses system-shared memory with a system-dependent bandwidth figure, which means performance scales with the host platform’s RAM configuration. NVIDIA pairs 24 GB of GDDR7 on a 384-bit bus with 1.34 TB/s of bandwidth. The dedicated memory allocation removes dependency on system RAM, and the bandwidth figure is orders of magnitude beyond what any shared-memory IGP can achieve, though the exact system-shared bandwidth for Intel is not recorded in the database.

The Verdict

The data indicates two products with fundamentally different purposes. The Intel Iris Xe Graphics 80EU Mobile is an integrated graphics processor within a Raptor Lake mobile chip, built on Intel’s 10 nm process, with a 15 W TDP and no power connectors. It is designed for portable devices, uses system-shared memory, and has no dedicated RT or tensor cores. The NVIDIA GeForce RTX 5090 SE is a discrete, dual-slot card built on TSMC’s 5 nm process with a 500 W TDP, a 1x 16-pin power connector, a suggested 900 W power supply, and a PCIe 5.0 x16 interface. It includes 110 RT cores and 440 tensor cores, features absent from the Intel part.

For workloads involving real-time ray tracing, tensor-based computation, or high-bandwidth memory access, the RTX 5090 SE is the only viable option among these two. The Intel part has no RT cores and no tensor cores, and its system-dependent memory bandwidth cannot match a dedicated 1.34 TB/s GDDR7 implementation. For general mobile graphics, light productivity, and display output in a portable device, the Intel IGP fits that role by virtue of its integration, 15 W power draw, and shared memory architecture.

The benchmark database shows no measured performance for either product, so percentile rankings and average scores are not informative. The specification sheet alone distinguishes them: one is an integrated solution for low-power mobile systems, the other is a high-end discrete accelerator for desktop workloads. Users seeking a discrete GPU with dedicated VRAM, high FP32 throughput, and hardware ray tracing should select the RTX 5090 SE. Users constrained to a mobile platform with integrated graphics and minimal power draw will find the Iris Xe as the only applicable part in this pairing.

FAQ

Q: Which GPU has higher FP32 compute throughput?

A: The NVIDIA GeForce RTX 5090 SE delivers 66.94 TFLOPS FP32, while the Intel Iris Xe Graphics 80EU Mobile delivers 1.856 TFLOPS, a difference of roughly 36 times in favor of NVIDIA.

Q: Does the Intel Iris Xe support hardware ray tracing?

A: No. The Intel part has no RT cores listed. The NVIDIA RTX 5090 SE includes 110 RT cores.

Q: What memory configuration does each GPU use?

A: Intel uses system-shared memory with system-dependent bandwidth. NVIDIA uses 24 GB of GDDR7 on a 384-bit bus with 1.34 TB/s bandwidth.

Q: Are the FP16 rates the same between the two GPUs?

A: No. Intel delivers 3.712 TFLOPS FP16 at a 2:1 ratio (half rate relative to FP32). NVIDIA delivers 66.94 TFLOPS FP16 at a 1:1 ratio (full rate matching FP32).

Q: What is the power requirement for each?

A: Intel has a 15 W TDP and uses no power connectors. NVIDIA has a 500 W TDP, requires a 1x 16-pin power connector, and a suggested 900 W power supply.

Q: Which GPU has tensor cores?

A: Only the NVIDIA RTX 5090 SE has tensor cores, totaling 440. The Intel Iris Xe has no tensor cores listed.

Specification Differences

| Field | Intel Iris Xe Graphics 80EU Mobile | NVIDIA GeForce RTX 5090 SE |

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

| Manufacturer | Intel | NVIDIA |

| Chip | Raptor Lake | GB202 |

| Architecture | Generation 12.2 | Blackwell 2.0 |

| Process Node | 10 nm | 5 nm |

| Foundry | Intel | TSMC |

| Transistors | Not listed | 92,200 million |

| Die Size | Not listed | 750 mm² |

| Base Clock | 300 MHz | 1740 MHz |

| Boost Clock | 1450 MHz | 2377 MHz |

| Memory Clock | System Shared | 1750 MHz 28 Gbps effective |

| Memory Size | System Shared | 24 GB |

| Memory Type | System Shared | GDDR7 |

| Memory Bus Width | System Shared | 384 bit |

| Memory Bandwidth | System Dependent | 1.34 TB/s |

| Shading Units | 640 | 14080 |

| TMUs | 40 | 440 |

| ROPs | 20 | 160 |

| RT Cores | Not listed | 110 |

| Tensor Cores | Not listed | 440 |

| Pixel Rate | 29.00 GPixel/s | 380.3 GPixel/s |

| Texture Rate | 58.00 GTexel/s | 1,045.9 GTexel/s |

| FP32 | 1.856 TFLOPS | 66.94 TFLOPS |

| FP16 | 3.712 TFLOPS (2:1) | 66.94 TFLOPS (1:1) |

| TDP | 15 W | 500 W |

| Slot Width | IGP | Dual-slot |

| Power Connectors | Not listed | 1x 16-pin |

| Suggested PSU | Not listed | 900 W |

| Bus Interface | Ring Bus | PCIe 5.0 x16 |

| Display Outputs | Portable Device Dependent | 1x HDMI 2.1b, 3x DisplayPort 2.1b |

| DirectX | 12 (12_1) | 12 Ultimate (12_2) |

| OpenGL | 4.6 | 4.6 |

| Vulkan | 1.4 | 1.4 |

| Dimensions | Not listed | 267 mm length, 111 mm height, 40 mm width |

| Release Date | 2023-01-03 | 2025-12-31 |

| Predecessor | Not listed | GeForce 40 |

| Successor | Arc Graphics-M | GeForce 60 |

| Launch MSRP | Not listed | 1,499 USD |

Architecture Differences

The Intel Iris Xe Graphics 80EU Mobile uses Intel’s Generation 12.2 architecture on a 10 nm process at Intel’s foundry. The chip is Raptor Lake, and the generation is listed as HD Graphics-M (Raptor Lake). It integrates 640 shading units, 40 TMUs, and 20 ROPs, with no RT cores or tensor cores present. The part runs on a Ring Bus interface and is classified as an IGP with a slot width of IGP. Its display outputs are dependent on the portable device in which it is installed. The successor to this part is Arc Graphics-M.

The NVIDIA GeForce RTX 5090 SE uses Blackwell 2.0 architecture on TSMC’s 5 nm process. The chip is GB202, with 92,200 million transistors on a 750 mm² die, yielding a transistor density of 122.9 million transistors per mm². It contains 14,080 shading units, 440 TMUs, 160 ROPs, 110 RT cores, and 440 tensor cores. The card uses a PCIe 5.0 x16 interface and is a dual-slot design with dimensions of 267 mm length, 111 mm height, and 40 mm width. It offers 1x HDMI 2.1b and 3x DisplayPort 2.1b outputs. The predecessor in the product line is GeForce 40, and the successor is GeForce 60.

DirectX feature levels differ. Intel supports DirectX 12 (12_1), while NVIDIA supports DirectX 12 Ultimate (12_2). Both parts support OpenGL 4.6 and Vulkan 1.4. FP16 execution differs in ratio: Intel runs at 2:1 (half rate), NVIDIA at 1:1 (full rate). The RTX 5090 SE’s 440 tensor cores enable matrix math acceleration that the Iris Xe cannot perform in hardware. The RT core count of 110 on NVIDIA enables dedicated ray tracing traversal, a feature entirely absent from the Intel IGP. Memory architecture is the largest structural divide: Intel relies on system-shared DRAM with bandwidth dependent on the host platform, while NVIDIA uses 24 GB of dedicated GDDR7 on a 384-bit bus with 1.34 TB/s throughput.