Intel Arc A380E x2 vs NVIDIA GeForce RTX 5090 Mobile Comparison

Head-to-Head Benchmarks

The recorded data presents an asymmetric comparison. The Intel Arc A380E x2 has no individual benchmark scores in the database, while the NVIDIA GeForce RTX 5090 Mobile has ten recorded tests. As a result, the head-to-head benchmark section is defined entirely by the NVIDIA part’s performance profile, with no direct numerical wins available for the Intel side.

The RTX 5090 Mobile’s strongest result comes from Geekbench OpenCL, where it scores 201,834 points. Its Vulkan score is nearly as high at 198,405, indicating consistent compute performance across both major graphics APIs. In Passmark’s suite, the G3D score of 30,034 dominates the legacy DirectX tests, which range from 138 in DirectX 12 to 324 in DirectX 9. The GPU compute score of 13,401 further confirms strong general-purpose throughput.

The database places the RTX 5090 Mobile in the 84th percentile of all GPUs, with an average benchmark score of 45,152. Its nearest rivals bracket this figure closely: the AMD Radeon Pro 5500 XT averages 45,384, a 0.5% advantage, while the NVIDIA RTX 5880 Ada Generation averages 45,972, sitting 1.8% higher. The Intel Arc A730M averages 45,592, 1% above the RTX 5090 Mobile, and the NVIDIA GeForce RTX 4070 Ti averages 44,795, 0.8% below. These deltas are small, meaning the RTX 5090 Mobile sits in a tightly contested performance band despite its high percentile ranking.

The DirectX 10 score of 183 and DirectX 11 score of 269 show that the RTX 5090 Mobile handles older API workloads, while the DirectX 12 score of 138 is notably lower, a pattern that may reflect driver optimization priorities rather than raw capability. The Passmark G2D score of 1,057 covers 2D operations, a separate metric from the 3D and compute results.

Without any benchmark entries for the Intel Arc A380E x2, the database records zero wins for it and zero wins for the NVIDIA part in head-to-head tests. The Intel GPU’s percentile ranking of 50 places it at the median of all GPUs, but no average score is recorded. This lack of data prevents any direct numerical comparison between the two products in this section.

Where Each One Wins

The NVIDIA GeForce RTX 5090 Mobile wins every benchmark category where data exists. Its OpenCL and Vulkan scores are within 1.7% of each other, showing balanced performance across heterogeneous compute workloads. The Passmark G3D result of 30,034 is the standout 3D metric, while the compute score of 13,401 is less than half of the G3D figure, indicating that the card’s 3D rendering pipeline is its primary strength.

The Intel Arc A380E x2 has no recorded wins in any test, but its hardware profile suggests a different usage envelope. It uses 6 GB of GDDR6 memory on a 96-bit bus, delivering 186.0 GB/s of bandwidth. The RTX 5090 Mobile uses 24 GB of GDDR7 on a 256-bit bus, delivering 896.0 GB/s. For memory-bound workloads, the NVIDIA part has a clear theoretical advantage, with roughly 4.8 times the bandwidth.

Compute throughput also favors the NVIDIA part decisively. The RTX 5090 Mobile delivers 31.80 TFLOPS of FP32 performance, while the Intel part delivers 4.096 TFLOPS. This is a 7.8x difference in raw floating-point throughput. The RTX 5090 Mobile’s FP16 output is identical to its FP32 at 31.80 TFLOPS, indicating a 1:1 ratio, whereas the Intel part doubles its FP32 to 8.192 TFLOPS with a 2:1 ratio. For FP16 workloads, the NVIDIA part still leads by a factor of 3.9.

The Intel part’s 1024 shading units, 64 texture mapping units, and 32 ROPs are far smaller than the NVIDIA part’s 10,496 shading units, 328 TMUs, and 112 ROPs. Pixel fill rates are 64.00 GPixel/s for Intel versus 169.7 GPixel/s for NVIDIA, and texture fill rates are 128.0 GTexel/s versus 496.9 GTexel/s. These figures translate directly to resolution and detail-level headroom in 3D rendering.

For ray tracing, the Intel part has 8 RT cores, while the NVIDIA part has 82. The NVIDIA part also includes 328 tensor cores, which the Intel part lacks entirely. This makes the RTX 5090 Mobile the only one of the two with dedicated hardware for AI-accelerated workloads such as DLSS or neural rendering features.

FAQ

Q: How does the RTX 5090 Mobile’s average benchmark score compare to its closest rivals?

A: The RTX 5090 Mobile averages 45,152 points. The AMD Radeon Pro 5500 XT scores 45,384, which is 0.5% higher. The Intel Arc A730M scores 45,592, 1% higher. The NVIDIA RTX 5880 Ada Generation scores 45,972, 1.8% higher. The NVIDIA GeForce RTX 4070 Ti scores 44,795, 0.8% lower.

Q: What is the memory configuration difference between the two GPUs?

A: The Intel Arc A380E x2 has 6 GB of GDDR6 on a 96-bit bus with 186.0 GB/s bandwidth. The RTX 5090 Mobile has 24 GB of GDDR7 on a 256-bit bus with 896.0 GB/s bandwidth.

Q: Which GPU has higher FP32 compute throughput?

A: The RTX 5090 Mobile delivers 31.80 TFLOPS of FP32 performance. The Intel Arc A380E x2 delivers 4.096 TFLOPS. The NVIDIA part is roughly 7.8 times faster in this metric.

Q: Are there any benchmarks where the Intel Arc A380E x2 outperforms the RTX 5090 Mobile?

A: The database records no benchmark scores for the Intel Arc A380E x2, so no wins can be attributed to it. All ten recorded benchmark scores belong to the RTX 5090 Mobile.

Q: How do the two GPUs compare in terms of production status?

A: The Intel Arc A380E x2 is marked as end-of-life, with a release date of 2024-03-31. The RTX 5090 Mobile is marked as active, with a release date of 2025-03-26.

Q: What is the transistor density difference between the two chips?

A: The Intel DG2-128 chip contains 7,200 million transistors on a 157 mm² die, yielding 45.9M transistors per mm². The NVIDIA GB203 chip contains 45,600 million transistors on a 378 mm² die, yielding 120.6M transistors per mm².

Specification Differences

The two GPUs differ across nearly every recorded specification category. The Intel part uses a 6 nm process node, while the NVIDIA part uses 5 nm, both fabricated by TSMC. Transistor counts diverge sharply: Intel packs 7,200 million transistors, while NVIDIA packs 45,600 million. Die sizes are 157 mm² for Intel and 378 mm² for NVIDIA, resulting in transistor densities of 45.9M per mm² versus 120.6M per mm².

Clock speeds show an unusual inversion. The Intel GPU runs at a base and boost clock of 2000 MHz, while the NVIDIA GPU runs at a base of 990 MHz and a boost of 1515 MHz. Despite lower clocks, the NVIDIA part achieves far higher throughput due to its larger execution resource count. Memory clocks also differ: Intel runs at 1937 MHz with 15.5 Gbps effective speed, while NVIDIA runs at 1750 MHz with 28 Gbps effective speed.

Memory size, type, bus width, and bandwidth all favor NVIDIA. Intel uses 6 GB GDDR6 on a 96-bit bus with 186.0 GB/s. NVIDIA uses 24 GB GDDR7 on a 256-bit bus with 896.0 GB/s. Shading units number 1024 for Intel versus 10,496 for NVIDIA. TMUs are 64 versus 328, and ROPs are 32 versus 112. RT cores are 8 versus 82, and tensor cores are absent on Intel while NVIDIA has 328.

Pixel rate is 64.00 GPixel/s for Intel and 169.7 GPixel/s for NVIDIA. Texture rate is 128.0 GTexel/s versus 496.9 GTexel/s. FP32 throughput is 4.096 TFLOPS versus 31.80 TFLOPS. FP16 throughput is 8.192 TFLOPS with a 2:1 ratio for Intel, versus 31.80 TFLOPS with a 1:1 ratio for NVIDIA.

Power and physical specifications differ substantially. The Intel card has a TDP of 130 W, is single-slot, uses a 1x 6-pin power connector, and requires a 300 W suggested PSU. It measures 265 mm in length, 127 mm in height, and 20 mm in width. The NVIDIA part has a TDP of 95 W, is listed as IGP (integrated graphics processor), has no power connectors, and has no recorded dimensions. Its bus interface is PCIe 5.0 x16, while Intel uses PCIe 4.0 x8.

Display outputs differ as well: Intel provides 8x mini-DisplayPort 2.0, while NVIDIA’s outputs are listed as portable device dependent. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The Intel part is end-of-life with a predecessor of Xe Graphics and a successor of Battlemage. The NVIDIA part is active, with a predecessor of GeForce 40 Mobile and no recorded successor.

Architecture Differences

The Intel Arc A380E x2 is built on the Xe-HPG architecture, specifically the DG2-128 chip, and belongs to the Alchemist generation under the Arc 3 product line. The NVIDIA GeForce RTX 5090 Mobile uses the Blackwell 2.0 architecture on the GB203 chip, part of the GeForce 50-series and GeForce 50 Mobile generation.

The process nodes differ by one nanometer: Intel uses 6 nm, NVIDIA uses 5 nm, both at TSMC. The transistor density gap is more than 2.6x in NVIDIA’s favor, reflecting both the larger die and the tighter process. Intel’s DG2-128 has 7,200 million transistors on 157 mm², while NVIDIA’s GB203 has 45,600 million on 378 mm².

Cache and core organization are not fully specified in the database, but the available figures show a massive execution resource disparity. NVIDIA’s 10,496 shading units, 328 TMUs, and 112 ROPs dwarf Intel’s 1024, 64, and 32. RT core counts are 82 versus 8, and NVIDIA adds 328 tensor cores while Intel has none.

The FP16 ratio differs: Intel uses a 2:1 ratio, meaning FP16 throughput is double FP32, while NVIDIA uses a 1:1 ratio, meaning FP16 and FP32 throughput are identical. This architectural choice affects mixed-precision workloads. NVIDIA’s tensor cores enable dedicated AI acceleration, a feature absent from Intel’s design.

Memory architecture also differs fundamentally. Intel uses GDDR6 with a 96-bit bus, while NVIDIA uses GDDR7 with a 256-bit bus. The interface widths are PCIe 4.0 x8 for Intel and PCIe 5.0 x16 for NVIDIA, giving the NVIDIA part more host bandwidth for data transfers.

The Intel part is designated as end-of-life, released on 2024-03-31, with Battlemage as its successor. The NVIDIA part remains active, released on 2025-03-26, with no successor listed. The Intel GPU’s predecessor is Xe Graphics, while NVIDIA’s predecessor is GeForce 40 Mobile.

The Verdict

The data supports a clear distinction between these two products. The NVIDIA GeForce RTX 5090 Mobile holds a decisive advantage in every recorded benchmark and nearly every architectural metric. Its 84th percentile ranking, average score of 45,152, and 31.80 TFLOPS of FP32 throughput place it in a performance class that the Intel Arc A380E x2 cannot approach based on the available figures.

The Intel Arc A380E x2, with its 50th percentile ranking and no recorded benchmarks, appears positioned for a different role. Its 130 W TDP, single-slot design, and 8x mini-DisplayPort 2.0 outputs suggest a multi-display or embedded use case, while its 6 GB memory and 96-bit bus limit its bandwidth to 186.0 GB/s. The RTX 5090 Mobile, despite a lower 95 W TDP, delivers 896.0 GB/s of bandwidth and 24 GB of memory, making it suitable for large datasets and high-resolution rendering.

For users prioritizing raw performance, the RTX 5090 Mobile is the only choice with recorded data to support it. Its nearest rivals, including the RTX 4070 Ti and RTX 5880 Ada Generation, sit within 1.8% of its average score, confirming that it competes at the high end of the database’s GPU rankings. The Intel part has no such validation in the benchmark suite.

For users prioritizing low power consumption, the RTX 5090 Mobile also wins, drawing 95 W versus Intel’s 130 W. The NVIDIA part achieves higher performance at lower power, a result of its denser transistor layout and larger execution resource pool. The Intel part’s higher clock speed of 2000 MHz does not compensate for its smaller core count.

The production status differs as well: the Intel GPU is end-of-life, while the NVIDIA GPU is active. This suggests ongoing driver and software support for NVIDIA, while Intel’s part has moved to its Battlemage successor. The RTX 5090 Mobile is the only one of the two with tensor cores, making it the sole option for AI-accelerated workloads within this comparison.

The verdict from the recorded data is unambiguous. The RTX 5090 Mobile outperforms the Intel Arc A380E x2 in every measurable category, from compute throughput to memory bandwidth to benchmark scores. The Intel part’s strengths, such as its multiple DisplayPort outputs and single-slot form factor, are not captured in the benchmark data and cannot be quantified against NVIDIA’s portable-device-dependent output configuration. For any workload represented in the database, the NVIDIA GeForce RTX 5090 Mobile is the superior product.