Intel Arc A380E vs NVIDIA GeForce RTX 5090 Comparison

Head-to-Head Benchmarks

The recorded data contains no direct head-to-head benchmark comparisons between the Intel Arc A380E and the NVIDIA GeForce RTX 5090. The database lists no shared test scores for both products in any common workload, so a direct numerical comparison of identical benchmarks cannot be made from the available facts.

The Intel Arc A380E has no benchmark entries in the database, and its average benchmark score is recorded as zero. This absence of recorded performance data means the A380E contributes no measurable results to any comparison. Its percentile ranking of 50 among all GPUs reflects its position based on architectural classification and hardware specifications, not on observed test outcomes.

The NVIDIA GeForce RTX 5090, by contrast, has a substantial benchmark profile. Its recorded scores include a 3DMark Steel Nomad DX12 result of 18,355, a Geekbench OpenCL score of 334,370, and a Geekbench Vulkan score of 376,728. Additional results cover legacy DirectX paths: PassMark DirectX 9 at 395, DirectX 11 at 341, and DirectX 10 at 226, while DirectX 12 shows 185. The PassMark G3D score reaches 39,650, and the GPU compute score sits at 26,756.

The RTX 5090's average benchmark score of 79,842 places it at the 92nd percentile among all GPUs in the database. Its nearest rivals, as recorded, are the NVIDIA Tesla P100 PCIe 16 GB with an average score of 79,605 (0.3% behind), the Tesla P100 PCIe 12 GB at 79,396 (0.6% behind), and the AMD Radeon RX 6850M XT at 78,940 (1.1% behind). The AMD Radeon Pro Vega 64X sits slightly ahead at 80,959, a 1.4% margin over the RTX 5090.

Since the A380E lacks any recorded benchmark scores, the data cannot substantiate a claim of wins in either direction. The comparison rests entirely on specification differences and the distinct performance envelopes implied by those specifications.

Architecture Differences

The two GPUs diverge fundamentally in every architectural layer. The Intel Arc A380E uses the DG2-128 chip built on the Xe-HPG architecture, belonging to the Alchemist generation (Arc 3). It is fabricated on TSMC's 6 nm process and contains 7,200 million transistors on a 157 mm² die, yielding a transistor density of 45.9 million per square millimeter. The NVIDIA GeForce RTX 5090 uses the GB202 chip on the Blackwell 2.0 architecture, part of the GeForce 50 series. It is built on TSMC's 5 nm process with 92,200 million transistors across a 750 mm² die, achieving a density of 122.9 million per square millimeter.

Compute resources differ by an order of magnitude. The A380E provides 1,024 shading units, 64 texture mapping units, and 32 render output units. It includes 8 ray tracing cores and no tensor cores in the recorded data. The RTX 5090 delivers 21,760 shading units, 680 TMUs, and 176 ROPs, with 170 ray tracing cores and 680 tensor cores.

Clock behavior also separates the two. The A380E runs at a fixed 2000 MHz base and boost clock, while the RTX 5090 starts at 2017 MHz base and boosts to 2407 MHz. The Intel part's memory clock is 1937 MHz (15.5 Gbps effective), while the NVIDIA part runs at 1750 MHz but achieves 28 Gbps effective due to GDDR7 signaling.

Memory configurations are starkly different. The A380E uses 6 GB of GDDR6 on a 96-bit bus, delivering 186.0 GB/s of bandwidth. The RTX 5090 uses 32 GB of GDDR7 on a 512-bit bus, delivering 1.79 TB/s, roughly 9.6 times the bandwidth of the Intel card.

Throughput figures reflect the scale difference. The A380E reaches 64.00 GPixel/s pixel fill rate and 128.0 GTexel/s texture rate, with FP32 compute at 4.096 TFLOPS and FP16 at 8.192 TFLOPS (2:1 ratio). The RTX 5090 reaches 423.6 GPixel/s and 1,636.8 GTexel/s, with FP32 at 104.8 TFLOPS and FP16 at the same 104.8 TFLOPS (1:1 ratio). The NVIDIA part thus delivers roughly 25.6 times the FP32 throughput of the Intel part.

Power and physical design differ accordingly. The A380E carries a 75 W TDP, draws no power connectors, and fits a single-slot profile. The RTX 5090 has a 575 W TDP, requires a single 16-pin connector, and occupies a dual-slot design. The Intel card measures 254 mm by 127 mm by 20 mm, while the NVIDIA card measures 304 mm by 137 mm by 40 mm.

Interface and output capabilities also differ. The A380E uses PCIe 4.0 x8 and provides four DisplayPort 2.0 outputs. The RTX 5090 uses PCIe 5.0 x16 and provides one HDMI 2.1b output plus three DisplayPort 2.1b outputs. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4 per the recorded API data.

FAQ

Q: Which GPU has more memory bandwidth?

A: The NVIDIA GeForce RTX 5090 has 1.79 TB/s of bandwidth from 32 GB of GDDR7 on a 512-bit bus. The Intel Arc A380E has 186.0 GB/s from 6 GB of GDDR6 on a 96-bit bus.

Q: How do the process nodes compare?

A: The Intel Arc A380E uses TSMC's 6 nm process, while the NVIDIA GeForce RTX 5090 uses TSMC's 5 nm process. Transistor density is 45.9M per square millimeter for Intel versus 122.9M per square millimeter for NVIDIA.

Q: What is the power requirement difference?

A: The Intel Arc A380E has a 75 W TDP and needs no power connectors, with a suggested PSU of 250 W. The NVIDIA GeForce RTX 5090 has a 575 W TDP, requires a 16-pin connector, and suggests a 950 W PSU.

Q: Does the RTX 5090 support ray tracing and tensor operations?

A: Yes, the RTX 5090 has 170 ray tracing cores and 680 tensor cores. The Intel Arc A380E has 8 ray tracing cores and no tensor cores recorded.

Q: Which GPU has a higher boost clock?

A: The NVIDIA GeForce RTX 5090 boosts to 2407 MHz from a 2017 MHz base. The Intel Arc A380E runs at a fixed 2000 MHz for both base and boost.

Q: What are the production statuses?

A: The Intel Arc A380E is end-of-life, released on 2024-03-31. The NVIDIA GeForce RTX 5090 is active, released on 2025-01-29.

The Verdict

The data supports a clear separation of roles. The Intel Arc A380E is a low-power, compact solution. Its 75 W TDP, single-slot footprint, and lack of power connectors position it for systems where thermal and spatial constraints dominate. Its 4.096 TFLOPS FP32 throughput, 6 GB memory, and 186.0 GB/s bandwidth establish a baseline capable of modest graphics workloads. Its end-of-life status and 50th percentile ranking among all GPUs indicate it occupies a lower tier of the performance spectrum.

The NVIDIA GeForce RTX 5090 is a high-performance part by every recorded metric. Its 104.8 TFLOPS FP32 output, 32 GB of GDDR7, and 1.79 TB/s bandwidth place it at the 92nd percentile among all GPUs. Its average benchmark score of 79,842 sits within 1.4% of its nearest recorded rival, the AMD Radeon Pro Vega 64X, and ahead of the Tesla P100 variants and the RX 6850M XT by margins of 0.3% to 1.1%. That aggregate score represents a broad workload profile, and the individual benchmark results show strength across modern APIs, including a 334,370 OpenCL score and a 376,728 Vulkan score.

The selection guidance follows directly from the specifications. Systems requiring minimal power draw, single-slot cooling, and a compact 254 mm length align with the A380E. Workloads demanding high compute throughput, large memory capacity, and high bandwidth align with the RTX 5090, which also requires dual-slot cooling and a 950 W suggested PSU. The RTX 5090's launch MSRP is 1,999 USD, stated once here for reference. The A380E has no recorded launch MSRP.

The production statuses reinforce the distinction. The A380E is end-of-life, with its successor listed as Battlemage. The RTX 5090 is active, with its predecessor recorded as GeForce 40 and its successor as GeForce 60. The benchmark evidence, while one-sided due to the A380E's lack of recorded scores, confirms the RTX 5090's position as a high-percentile performer.

Specification Differences

| Specification | Intel Arc A380E | NVIDIA GeForce RTX 5090 |

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

| Chip | DG2-128 | GB202 |

| Architecture | Xe-HPG | Blackwell 2.0 |

| Generation | Alchemist (Arc 3) | GeForce 50 |

| Process Node | 6 nm | 5 nm |

| Transistors | 7,200 million | 92,200 million |

| Die Size | 157 mm² | 750 mm² |

| Transistor Density | 45.9M / mm² | 122.9M / mm² |

| Base Clock | 2000 MHz | 2017 MHz |

| Boost Clock | 2000 MHz | 2407 MHz |

| Memory Clock | 1937 MHz 15.5 Gbps effective | 1750 MHz 28 Gbps effective |

| Memory Size | 6 GB | 32 GB |

| Memory Type | GDDR6 | GDDR7 |

| Memory Bus Width | 96 bit | 512 bit |

| Memory Bandwidth | 186.0 GB/s | 1.79 TB/s |

| Shading Units | 1024 | 21760 |

| TMUs | 64 | 680 |

| ROPs | 32 | 176 |

| Ray Tracing Cores | 8 | 170 |

| Tensor Cores | None | 680 |

| Pixel Rate | 64.00 GPixel/s | 423.6 GPixel/s |

| Texture Rate | 128.0 GTexel/s | 1,636.8 GTexel/s |

| FP32 | 4.096 TFLOPS | 104.8 TFLOPS |

| FP16 | 8.192 TFLOPS (2:1) | 104.8 TFLOPS (1:1) |

| TDP | 75 W | 575 W |

| Slot Width | Single-slot | Dual-slot |

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

| Suggested PSU | 250 W | 950 W |

| Bus Interface | PCIe 4.0 x8 | PCIe 5.0 x16 |

| Display Outputs | 4x DisplayPort 2.0 | 1x HDMI 2.1b, 3x DisplayPort 2.1b |

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

| OpenGL | 4.6 | 4.6 |

| Vulkan | 1.4 | 1.4 |

| Length | 254 mm | 304 mm |

| Height | 127 mm | 137 mm |

| Width | 20 mm | 40 mm |

| Production Status | End-of-life | Active |

| Release Date | 2024-03-31 | 2025-01-29 |

| Predecessor | Xe Graphics | GeForce 40 |

| Successor | Battlemage | GeForce 60 |

| Launch MSRP | None recorded | 1,999 USD |

| Percentile vs All GPUs | 50 | 92 |

| Average Benchmark Score | 0 | 79,842 |