Intel Arc Graphics 24EU vs NVIDIA GeForce RTX 5070 SUPER Comparison

Head-to-Head Benchmarks

The single recorded benchmark comparison is decisive. In the 3DMark Steel Nomad DX12 test, the NVIDIA GeForce RTX 5070 SUPER scores 2690, while the Intel Arc Graphics 24EU scores 733. The NVIDIA part delivers a 72.8% higher score, meaning it outperforms the Intel integrated solution by a factor of roughly 3.7x in this specific workload. This is the only head-to-head data point in the database, so the win count stands at 1 for NVIDIA and 0 for Intel.

Context from the nearest rivals list reinforces the magnitude of this gap. The RTX 5070 SUPER sits at the 18th percentile among all GPUs, with an average score of 2690. Its closest competitors in the database, the NVIDIA Quadro K1100M (2664), NVIDIA GeForce GT 1030 (2662), Intel Arc Pro B50 (2660), and NVIDIA GeForce GT 440 (2645), all trail by 1% to 1.7%. The Intel Arc Graphics 24EU, by contrast, sits at the 3rd percentile with its 733 average score. Its nearest rivals, the Intel Arc Graphics 32EU (733), Intel Arc Graphics 64EU (733), and AMD Radeon HD 6470M (723), are effectively tied or within 1.4%, while the NVIDIA GeForce GT 415M (751) leads the Intel part by 2.4%.

The data indicates that the RTX 5070 SUPER is not merely faster; it occupies a completely different performance tier. The Intel iGPU's score of 733 places it among entry-level or legacy discrete parts, whereas the RTX 5070 SUPER's 2690 places it squarely in the midrange discrete segment, despite its rivals list including older low-end cards. The 72.8% delta in the head-to-head is the single most important metric here.

Architecture Differences

The two products share almost nothing in architectural DNA. The Intel Arc Graphics 24EU uses the Xe-LPG architecture on the Arrow Lake-S chip, built on a 3 nm process at TSMC. The chip integrates 17,800 million transistors across a 243 mm² die, yielding a transistor density of 73.3 million per square millimeter. It is an integrated graphics processor (IGP) with a ring bus interface, meaning it shares system memory and has no dedicated VRAM.

The NVIDIA GeForce RTX 5070 SUPER uses the Blackwell 2.0 architecture on the GB205 chip, manufactured on a 5 nm process, also at TSMC. The die is slightly larger at 263 mm² but packs 31,100 million transistors, for a density of 118.3 million per square millimeter. This is a discrete card with a PCIe 5.0 x16 interface and its own 18 GB of GDDR7 memory on a 192-bit bus.

Core resource differences are stark. The Intel part has 192 shading units, 12 texture mapping units (TMUs), and 6 raster output units (ROPs). It has no dedicated ray tracing cores and no tensor cores. The NVIDIA part has 6400 shading units, 200 TMUs, and 80 ROPs, plus 50 ray tracing cores and 200 tensor cores. The NVIDIA part's shading unit count is more than 33x higher, and its TMU and ROP counts are roughly 16.7x and 13.3x higher, respectively.

Clock behavior also differs. The Intel IGP runs at a base of 300 MHz and boosts to 2000 MHz. The RTX 5070 SUPER runs at a base of 2325 MHz and boosts to 2512 MHz. Despite the NVIDIA card's higher absolute clocks, its boost clock is only about 25.6% higher than the Intel part's boost; the performance gap is driven overwhelmingly by the massive core count advantage.

Memory architecture is fundamentally different. The Intel part uses system shared memory, with bandwidth described as system dependent. The NVIDIA card uses 18 GB of GDDR7 at 1750 MHz (28 Gbps effective), on a 192-bit bus, delivering 672.0 GB/s of bandwidth. This dedicated high-bandwidth memory is a critical differentiator for bandwidth-bound workloads.

Compute rates confirm the disparity. The Intel part delivers 768.0 GFLOPS FP32 and 1.536 TFLOPS FP16 (2:1 ratio). The NVIDIA part delivers 32.15 TFLOPS FP32 and 32.15 TFLOPS FP16 (1:1 ratio). The NVIDIA card's FP32 throughput is roughly 41.9x higher, and its FP16 throughput is roughly 20.9x higher. Pixel rate is 12.00 GPixel/s for Intel versus 201.0 GPixel/s for NVIDIA, a 16.75x difference. Texture rate is 24.00 GTexel/s versus 502.4 GTexel/s, a 20.9x difference.

Power and physical design differ as expected. The Intel IGP is rated at 65 W TDP and occupies an IGP slot width with no power connectors. The NVIDIA card is rated at 275 W TDP, is dual-slot, and requires a single 16-pin power connector. The NVIDIA card measures 245 mm in length, 115 mm in height, and 40 mm in width.

API support is identical: both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Display outputs differ, with Intel being motherboard dependent and NVIDIA offering 1x HDMI 2.1b and 3x DisplayPort 2.1b.

Where Each One Wins

Based on the recorded data, the NVIDIA GeForce RTX 5070 SUPER wins in every measurable category. Its 2690 score in 3DMark Steel Nomad DX12 dwarfs the Intel Arc Graphics 24EU's 733. The NVIDIA card also wins on every architectural metric: shading units, TMUs, ROPs, ray tracing cores, tensor cores, memory bandwidth, pixel rate, texture rate, FP32 throughput, and FP16 throughput.

The Intel Arc Graphics 24EU has no benchmark wins in the database. Its advantages are limited to power consumption (65 W versus 275 W), physical footprint (IGP versus dual-slot), and the absence of a power connector requirement. It also uses system shared memory, which can be an advantage in cost-constrained integrated systems, but the database provides no performance data to support that as a win.

For use cases, the data suggests the NVIDIA card is suited for any workload requiring high compute throughput, dedicated memory bandwidth, or ray tracing capability. The 50 ray tracing cores and 200 tensor cores are present only on the NVIDIA side. The Intel part, with its 65 W TDP and integrated form factor, is suited for basic display output and low-intensity tasks, but the benchmark data shows it operates at the 3rd percentile, which indicates minimal 3D performance headroom.

The RTX 5070 SUPER's placement at the 18th percentile, despite its high absolute score, indicates that the database contains many faster GPUs. However, its nearest rivals (Quadro K1100M, GT 1030, Arc Pro B50, GT 440) all score lower, confirming that within its immediate comparison group, it is the performance leader.

FAQ

Q: What is the score difference in the 3DMark Steel Nomad DX12 test?

A: The NVIDIA GeForce RTX 5070 SUPER scores 2690, while the Intel Arc Graphics 24EU scores 733. The NVIDIA card leads by 72.8%.

Q: Which GPU has more shading units?

A: The NVIDIA GeForce RTX 5070 SUPER has 6400 shading units, compared to 192 on the Intel Arc Graphics 24EU.

Q: Does the Intel Arc Graphics 24EU support ray tracing?

A: No. The Intel part has no dedicated ray tracing cores. The NVIDIA GeForce RTX 5070 SUPER has 50 ray tracing cores.

Q: What type of memory does each GPU use?

A: The Intel Arc Graphics 24EU uses system shared memory with system dependent bandwidth. The NVIDIA GeForce RTX 5070 SUPER uses 18 GB of GDDR7 on a 192-bit bus with 672.0 GB/s bandwidth.

Q: How do their power requirements compare?

A: The Intel Arc Graphics 24EU is rated at 65 W TDP and requires no power connectors. The NVIDIA GeForce RTX 5070 SUPER is rated at 275 W TDP and requires a single 16-pin power connector.

Q: Are their API supports the same?

A: Yes. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: What are the FP32 performance figures?

A: The Intel Arc Graphics 24EU delivers 768.0 GFLOPS FP32. The NVIDIA GeForce RTX 5070 SUPER delivers 32.15 TFLOPS FP32.

Specification Differences

| Specification | Intel Arc Graphics 24EU | NVIDIA GeForce RTX 5070 SUPER |

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

| Architecture | Xe-LPG | Blackwell 2.0 |

| Process Node | 3 nm | 5 nm |

| Transistors | 17,800 million | 31,100 million |

| Die Size | 243 mm² | 263 mm² |

| Transistor Density | 73.3M / mm² | 118.3M / mm² |

| Base Clock | 300 MHz | 2325 MHz |

| Boost Clock | 2000 MHz | 2512 MHz |

| Memory Size | System Shared | 18 GB |

| Memory Type | System Shared | GDDR7 |

| Memory Bus Width | System Shared | 192 bit |

| Memory Bandwidth | System Dependent | 672.0 GB/s |

| Shading Units | 192 | 6400 |

| TMUs | 12 | 200 |

| ROPs | 6 | 80 |

| Ray Tracing Cores | None | 50 |

| Tensor Cores | None | 200 |

| Pixel Rate | 12.00 GPixel/s | 201.0 GPixel/s |

| Texture Rate | 24.00 GTexel/s | 502.4 GTexel/s |

| FP32 | 768.0 GFLOPS | 32.15 TFLOPS |

| FP16 | 1.536 TFLOPS (2:1) | 32.15 TFLOPS (1:1) |

| TDP | 65 W | 275 W |

| Slot Width | IGP | Dual-slot |

| Power Connectors | None | 1x 16-pin |

| Bus Interface | Ring Bus | PCIe 5.0 x16 |

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

| Dimensions | Not recorded | 245 mm x 115 mm x 40 mm |

| Release Date | 2024-10-23 | 2025-12-31 |

The Verdict

The data is unambiguous. The NVIDIA GeForce RTX 5070 SUPER is the superior performer in every recorded benchmark and architectural metric. Its 2690 average benchmark score is 72.8% higher than the Intel Arc Graphics 24EU's 733. The NVIDIA card offers 33x more shading units, 50 dedicated ray tracing cores, 200 tensor cores, and 672.0 GB/s of dedicated GDDR7 bandwidth versus system shared memory. Its FP32 throughput of 32.15 TFLOPS is more than 40x higher than the Intel part's 768.0 GFLOPS.

The Intel Arc Graphics 24EU is an integrated solution with a 65 W TDP, no power connectors, and an IGP slot width. It is designed for systems where discrete graphics are not an option. Its 3rd percentile ranking indicates it handles only the most basic 3D workloads. The RTX 5070 SUPER, at the 18th percentile, is a discrete dual-slot card with a 275 W TDP and a 16-pin connector, built for sustained high-performance rendering.

For any user prioritizing 3D performance, ray tracing, tensor operations, or high-bandwidth memory, the NVIDIA GeForce RTX 5070 SUPER is the clear choice. The Intel part's only advantages are lower power draw and integrated form factor, which are relevant for compact or low-power systems. The benchmark data shows no scenario where the Intel part wins on performance. Selection between these two should be based entirely on whether the system can accommodate a discrete 275 W dual-slot card; if it can, the NVIDIA part delivers the recorded performance advantage.