Intel Arc B580

The Intel Arc B580 is a Battlemage-generation discrete GPU built on TSMC's 5 nm process, packing 19,600 million transistors into a 272 mm² die. It features 2560 shading units, 160 TMUs, 80 ROPs, and 20 ray tracing cores, with 12 GB of GDDR6 on a 192-bit bus delivering 456.0 GB/s of bandwidth. Clocked at a flat 2670 MHz base and boost, it produces 13.67 TFLOPS FP32 and 27.34 TFLOPS FP16 (2:1). Its launch MSRP is 249 USD. In the benchmark database, it holds a 66th percentile among all GPUs, with an average benchmark score of 23021.

Who Should Consider It

The Arc B580’s average score of 23021 places it squarely in the mainstream performance tier, nearly indistinguishable from the GeForce RTX 4060 (23181) and slightly ahead of the Radeon 760M (22999) and P106-100 (22950). The 3DMark Steel Nomad DX12 score of 3068 and PassMark G3D score of 15748 indicate solid DirectX 12 and general 3D rendering capability. With 12 GB of GDDR6 memory and 456.0 GB/s of bandwidth, the card is well suited for texture-heavy games and modern assets, though the data does not specify resolution-specific performance. The 66th percentile ranking suggests it outperforms the majority of GPUs in the database, making it a viable option for gamers targeting high refresh rates at 1080p or comfortable frame rates at 1440p with adjusted settings. Users who prioritize consistent performance in the RTX 4060 class will find the Arc B580 a direct competitor, while those expecting 4K-class throughput should look at higher-scoring parts.

Ray Tracing and Feature Set

The Arc B580 includes 20 ray tracing cores, enabling hardware-accelerated ray tracing for supported titles. It supports DirectX 12 Ultimate (12_2), which brings features such as variable rate shading and mesh shaders, alongside Vulkan 1.4 and OpenGL 4.6. The card offers 1x HDMI 2.1a and 3x DisplayPort 2.1 outputs, allowing for high-bandwidth display connections. The specification list does not include a dedicated tensor core count, so no AI-accelerated features are quantified here. The 20 RT cores are the sole dedicated accelerator beyond the standard shader array, and the card’s compute performance is driven by its 2560 shading units and 160 TMUs.

Power and Cooling

The Arc B580 has a TDP of 190 W and requires a single 8-pin power connector. Intel recommends a 450 W power supply, which is modest for a card of this performance class. The card is a dual-slot design, measuring 272 mm in length (10.7 inches), 115 mm in height (4.5 inches), and 45 mm in width (1.8 inches). These dimensions indicate a typical dual-slot cooler, suitable for most mid-tower cases. The single 8-pin connector simplifies installation, and the 450 W PSU recommendation leaves ample headroom for typical system configurations.

How It Compares

AMD Radeon 760M

The Arc B580’s average score of 23021 is 0.1% higher than the Radeon 760M’s 22999, making the two effectively identical in overall benchmark performance. The margin is within measurement noise, so the choice between them would hinge on other factors such as driver support, feature set, and memory capacity.

NVIDIA P106-100

Against the P106-100, the Arc B580 leads by 0.3% (23021 vs 22950). This negligible gap places both cards in the same performance bracket. The P106-100 is an older mining-oriented card, but the benchmark data shows no significant advantage for either in average scores.

NVIDIA GeForce RTX 4060

The RTX 4060 scores 23181, which is 0.7% higher than the Arc B580’s 23021. This is a very small difference, indicating that the Arc B580 essentially matches the RTX 4060 in average benchmark performance. In real-world gaming, the difference would be imperceptible, making the Arc B580 a strong alternative to NVIDIA’s mainstream offering.

NVIDIA GeForce RTX 4060 Mobile

The Arc B580 outperforms the RTX 4060 Mobile by 1.3% (23021 vs 22729). While the mobile variant is constrained by power limits, the desktop Arc B580 holds a slight edge in these benchmarks, though the gap is still small enough to be considered comparable.

Benchmark Performance

The Arc B580’s average benchmark score of 23021 sits at the 66th percentile of all GPUs, meaning it outperforms roughly two-thirds of the database. Its closest rival, the RTX 4060, scores 23181, a 0.7% advantage that falls well within typical run-to-run variance. The Radeon 760M and P106-100 are effectively tied with the Arc B580, with deltas of 0.1% and 0.3% respectively. The only notable difference is against the RTX 4060 Mobile, where the Arc B580 leads by 1.3%.

Looking at individual benchmarks, the 3DMark Steel Nomad DX12 score of 3068 reflects strong modern DirectX 12 performance, aligning with the card’s architecture and API support. In Geekbench, the Vulkan score of 109672 is notably higher than the OpenCL score of 92821, suggesting that the card’s driver stack is well optimized for Vulkan workloads. PassMark results show a mixed picture: the G3D score of 15748 is solid for mainstream gaming, while the GPU compute score of 7729 indicates moderate compute throughput. The PassMark DirectX scores—183 in DX9, 128 in DX11, and 76 in DX12—are interesting because the DX12 score is lower than the older API scores, which may reflect driver optimization priorities or the specific test workload. However, the 3DMark Steel Nomad result provides a more representative measure of modern DX12 performance.

The card’s 13.67 TFLOPS FP32 and 27.34 TFLOPS FP16 (2:1) compute figures are consistent with its shading unit count and clock speed. The 456.0 GB/s memory bandwidth, combined with 12 GB of VRAM, supports high-resolution textures and complex scenes without immediate capacity bottlenecks. Overall, the benchmark data positions the Arc B580 as a capable mid-range GPU that trades blows with the RTX 4060 and outpaces integrated and older discrete solutions, while its 66th percentile ranking confirms its place above the majority of the GPU landscape.