Intel Iris Xe Graphics 80EU Mobile vs NVIDIA N1X 40SM Comparison

Intel Iris Xe Graphics 80EU Mobile and NVIDIA N1X 40SM represent two very different approaches to integrated graphics, separated by process technology, architecture generation, and raw compute scale. The database records no direct head-to-head benchmark matches between them, so the comparison below relies entirely on their recorded specifications, derived rates, and architectural traits.

Where Each One Wins

The Intel Iris Xe Graphics 80EU Mobile is built for efficiency and integration within a Raptor Lake mobile processor. Its 15 W TDP, system-shared memory, and Ring Bus interface make it suitable for thin-and-light laptops where power draw is tightly constrained. The recorded data shows a base clock of 300 MHz and a boost clock of 1450 MHz, with pixel rate of 29.00 GPixel/s and texture rate of 58.00 GTexel/s. Those figures place it in the lower tier of integrated graphics, but the GPU supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, which means it can run modern API workloads at modest settings. Its percentile ranking versus all GPUs is 50, indicating it sits near the midpoint of the database's performance distribution, though that percentile is based on no recorded benchmark scores.

The NVIDIA N1X 40SM is a fundamentally larger part. It uses a GB20B chip built on a 5 nm process at TSMC, with a die size of 382 mm², and it is also classified as an IGP, meaning it is designed for integration rather than as a discrete card. The data shows 5120 shading units, 320 texture mapping units, 40 ROPs, 40 ray tracing cores, and 160 tensor cores. Its boost clock reaches 2346 MHz, and its memory subsystem is far more substantial: 128 GB of LPDDR5X on a 256 bit bus, with bandwidth of 273.2 GB/s. The resulting pixel rate is 93.84 GPixel/s and texture rate is 750.7 GTexel/s. Those numbers are roughly 3.2 times the pixel rate and 12.9 times the texture rate of the Intel part. The NVIDIA GPU also delivers 24.02 TFLOPS of FP32 compute, versus 1.856 TFLOPS for Intel. In every raw throughput category recorded in the database, the NVIDIA part wins by a wide margin.

Where the Intel part wins is in API compatibility and form factor flexibility. Intel lists DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, while NVIDIA lists N/A for all three major APIs. That is not a performance win, but it is a functional win for software environments that require those APIs. Intel also has a Ring Bus interface and system-shared memory, which means it can operate in any laptop that provides the host processor, without dedicated memory allocation. NVIDIA uses PCIe 5.0 x16, which requires a matching host interface. The Intel part has no power connectors listed, and NVIDIA also lists none, but Intel's TDP is recorded as 15 W, while NVIDIA's TDP is unknown. The Intel GPU is the only one of the two with a clear power envelope, and that envelope is low enough for passive cooling in many portable designs.

Architecture Differences

The two GPUs come from entirely different architectural lineages. Intel's Iris Xe Graphics 80EU Mobile uses Generation 12.2 architecture, specifically the HD Graphics-M (Raptor Lake) generation. It is manufactured on a 10 nm process at Intel's own foundry. The chip has 640 shading units, 40 TMUs, and 20 ROPs, with no ray tracing cores and no tensor cores listed. The FP16 rate is 3.712 TFLOPS with a 2:1 ratio, meaning FP16 compute is double FP32 compute, a common design for consumer integrated GPUs that do not prioritize double-rate FP16 workloads. The memory path is entirely system shared, with no dedicated VRAM, and the bus width is also system shared, so bandwidth is system dependent. The GPU connects via Ring Bus, which ties it directly into the CPU's ring interconnect.

NVIDIA's N1X 40SM uses Blackwell 2.0 architecture, in the Blackwell IGP (N1x) generation. It is built on a 5 nm process at TSMC, with a die size of 382 mm². The chip has 5120 shading units, 320 TMUs, 40 ROPs, 40 ray tracing cores, and 160 tensor cores. The FP16 rate is 24.02 TFLOPS with a 1:1 ratio, meaning FP16 and FP32 throughput are identical, a design choice typical for GPUs that rely on tensor cores for mixed-precision work and do not dedicate separate FP16 units. The memory is 128 GB of LPDDR5X on a 256 bit bus, with bandwidth of 273.2 GB/s. The memory clock is listed as 1067 MHz with 8.5 Gbps effective data rate. The bus interface is PCIe 5.0 x16, and the display output is a single HDMI port. The transistor count is listed as "unknown," though the die size is recorded.

The node difference is significant. Intel's 10 nm process is one generation behind NVIDIA's 5 nm process, and that gap shows in clock rates and density. Intel's boost clock is 1450 MHz, while NVIDIA's boost clock is 2346 MHz, a 61.8% higher clock. The NVIDIA die is 382 mm², which is large for an integrated part, and it packs 5120 shading units into that space. Intel's die size is not recorded, but the shading unit count is only 640, so the density difference is substantial. The memory architecture also diverges: Intel uses system shared memory with no dedicated bandwidth figure, while NVIDIA uses a dedicated 256 bit LPDDR5X interface with a fixed bandwidth of 273.2 GB/s. That dedicated memory path is a major architectural advantage for bandwidth-sensitive workloads.

Ray tracing and tensor cores are present only on the NVIDIA side. The N1X 40SM has 40 RT cores and 160 tensor cores, which means it can handle hardware-accelerated ray tracing and AI inference workloads. Intel's Iris Xe has no such units recorded, so any ray tracing or tensor operations would run on general-purpose shaders, which is far less efficient. The API support also differs, with Intel listing DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, while NVIDIA lists N/A for all three. That suggests the NVIDIA part may rely on proprietary or future API paths, or the database simply has not recorded its API support yet.

Head-to-Head Benchmarks

The database contains no recorded head-to-head benchmark matches between these two GPUs. That means there are no measured scores showing how they perform in specific games or compute workloads. Instead, the comparison must be made from the recorded specification-derived rates and compute figures.

The largest win for NVIDIA is in FP32 compute. The N1X 40SM delivers 24.02 TFLOPS, while the Intel Iris Xe delivers 1.856 TFLOPS. That is a 12.9 times difference in favor of NVIDIA. For FP16, NVIDIA delivers 24.02 TFLOPS, and Intel delivers 3.712 TFLOPS, a 6.5 times difference. In texture rate, NVIDIA reaches 750.7 GTexel/s versus 58.00 GTexel/s, again a 12.9 times gap. Pixel rate shows NVIDIA at 93.84 GPixel/s versus 29.00 GPixel/s, a 3.2 times difference. The shading unit count is 5120 versus 640, exactly 8 times, and the TMU count is 320 versus 40, also 8 times. The ROP count is 40 versus 20, a 2 times difference.

Memory bandwidth is another major divider. NVIDIA has a fixed 273.2 GB/s from its 256 bit LPDDR5X interface, while Intel's bandwidth is listed as "System Dependent," meaning it varies with the host memory configuration. In a typical dual-channel DDR5 laptop configuration, system shared bandwidth might reach tens of GB/s, but even the best case would fall far short of 273.2 GB/s. The NVIDIA part also has 128 GB of dedicated LPDDR5X memory, while Intel has no dedicated memory at all.

Clock speeds favor NVIDIA as well. The boost clock of 2346 MHz is 61.8% higher than Intel's 1450 MHz. The base clock is also higher: 741 MHz versus 300 MHz, a 147% difference. These clock advantages compound with the shading unit advantage, producing the large compute gap.

The only categories where Intel shows a number that is not lower are the API support fields. Intel lists DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, while NVIDIA lists N/A for all three. In a purely functional sense, Intel wins on API availability. Also, Intel's TDP is recorded as 15 W, while NVIDIA's TDP is unknown, so Intel wins on power disclosure, though not necessarily on actual power draw. The NVIDIA part likely consumes far more power given its compute scale, but that figure is not in the database, so it cannot be stated.

The production status for both is Active. Intel's release date is 2023-01-03, while NVIDIA's release date is 2026-05-31. Intel has a successor listed as Arc Graphics-M, while NVIDIA has no successor recorded. The Intel part is in the HD Graphics-M generation, and the NVIDIA part is in the Blackwell IGP (N1x) generation.

The Verdict

From the recorded data, the NVIDIA N1X 40SM is the far stronger compute device. It leads in shading units, TMUs, ROPs, ray tracing cores, tensor cores, clock speeds, memory capacity, memory bus width, memory bandwidth, pixel rate, texture rate, FP32 throughput, and FP16 throughput. Any workload that depends on raw GPU compute, texture fetching, pixel fill, or memory bandwidth will favor the NVIDIA part by a factor of roughly 3 to 13, depending on the metric. The availability of 40 ray tracing cores and 160 tensor cores makes it suitable for hardware-accelerated ray tracing and AI inference, neither of which the Intel part can do in hardware.

The Intel Iris Xe Graphics 80EU Mobile wins on API compatibility and power envelope. It supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, which are the standard graphics APIs for consumer software. NVIDIA lists N/A for all three, so software that requires those APIs may not run on the N1X 40SM, or at least the database does not confirm support. Intel also has a recorded TDP of 15 W, which is low enough for fanless or low-power designs, while NVIDIA's TDP is unknown. The Intel part uses system shared memory, which simplifies system design, while NVIDIA uses a dedicated 128 GB LPDDR5X pool, which adds cost and complexity.

For a user choosing between these two, the decision depends on workload. If the task is general-purpose gaming or compute on standard APIs, the Intel part is the only one with confirmed API support. If the task is maximum throughput, especially in FP32 or FP16 compute, ray tracing, or tensor operations, the NVIDIA part is overwhelmingly faster, provided the software can interface with its API stack. The NVIDIA part also has a later release date, a smaller process node, and a larger die, all of which point to a more advanced design. The Intel part has a successor already in the database, while NVIDIA does not, suggesting Intel has moved on to newer architecture. The database records no benchmark scores for either GPU, so the percentile ranking of 50 for both is identical and carries no comparative weight.

FAQ

Q: Which GPU has higher FP32 compute?

A: The NVIDIA N1X 40SM delivers 24.02 TFLOPS, while the Intel Iris Xe Graphics 80EU Mobile delivers 1.856 TFLOPS. The NVIDIA part is approximately 12.9 times faster in FP32.

Q: Does the Intel Iris Xe support DirectX 12?

A: Yes. The database lists DirectX 12 (12_1) support for the Intel part, along with OpenGL 4.6 and Vulkan 1.4. The NVIDIA N1X 40SM lists N/A for all three APIs.

Q: What memory configuration does each GPU use?

A: The Intel part uses system shared memory with no dedicated size, type, or bus width; bandwidth is listed as system dependent. The NVIDIA part uses 128 GB of LPDDR5X on a 256 bit bus, with bandwidth of 273.2 GB/s.

Q: Does the NVIDIA N1X 40SM have ray tracing cores?

A: Yes. It has 40 ray tracing cores and 160 tensor cores. The Intel Iris Xe has no ray tracing or tensor cores recorded.

Q: What are the process nodes for these GPUs?

A: The Intel part is manufactured on a 10 nm process at Intel foundry. The NVIDIA part is manufactured on a 5 nm process at TSMC.

Q: Which GPU has a higher boost clock?

A: The NVIDIA N1X 40SM boosts to 2346 MHz, while the Intel Iris Xe boosts to 1450 MHz. The NVIDIA boost clock is 61.8% higher.

Specification Differences

| Specification | Intel Iris Xe Graphics 80EU Mobile | NVIDIA N1X 40SM |

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

| Architecture | Generation 12.2 | Blackwell 2.0 |

| Process Node | 10 nm | 5 nm |

| Foundry | Intel | TSMC |

| Die Size | Not recorded | 382 mm² |

| Base Clock | 300 MHz | 741 MHz |

| Boost Clock | 1450 MHz | 2346 MHz |

| Memory Size | System Shared | 128 GB |

| Memory Type | System Shared | LPDDR5X |

| Memory Bus Width | System Shared | 256 bit |

| Memory Bandwidth | System Dependent | 273.2 GB/s |

| Shading Units | 640 | 5120 |

| TMUs | 40 | 320 |

| ROPs | 20 | 40 |

| Ray Tracing Cores | Not recorded | 40 |

| Tensor Cores | Not recorded | 160 |

| Pixel Rate | 29.00 GPixel/s | 93.84 GPixel/s |

| Texture Rate | 58.00 GTexel/s | 750.7 GTexel/s |

| FP32 Compute | 1.856 TFLOPS | 24.02 TFLOPS |

| FP16 Compute | 3.712 TFLOPS (2:1) | 24.02 TFLOPS (1:1) |

| TDP | 15 W | Unknown |

| Bus Interface | Ring Bus | PCIe 5.0 x16 |

| Display Outputs | Portable Device Dependent | 1x HDMI |

| DirectX Support | 12 (12_1) | N/A |

| OpenGL Support | 4.6 | N/A |

| Vulkan Support | 1.4 | N/A |

| Release Date | 2023-01-03 | 2026-05-31 |

| Successor | Arc Graphics-M | Not recorded |