GPU Comparison

The AMD Radeon RX 6800S and Intel Arc B580 are two graphics cards from different eras and design philosophies, with the former being an end-of-life mobile part and the latter an active desktop solution. The benchmark data presents a fascinating split, where the two cards split their ten head-to-head benchmark wins exactly evenly, yet the nature of those wins reveals starkly different performance profiles. The RX 6800S, a RDNA 2.0 part, holds its own in several legacy and 2D tests, while the Arc B580, built on the newer Xe2-HPG architecture, dominates in modern DirectX 12 and compute-heavy workloads.

Head-to-Head Benchmarks

The most decisive victory in the entire comparison belongs to the Intel Arc B580 in the 3DMark Steel Nomad DX12 test. Here, the Arc B580 scores 3068 points against the RX 6800S’s 1689, a massive 44.9% difference in favor of Intel. This is not a marginal lead; it is a generational gap in raw modern gaming performance. The data suggests that for any workload leveraging DirectX 12's advanced features, the Arc B580 is in a completely different performance class.

The Arc B580 also shows substantial leads in other modern API tests. In Geekbench Vulkan, it scores 109672 compared to the RX 6800S’s 81272, a 25.9% advantage. Similarly, in Geekbench OpenCL, the Intel card posts 92821 versus 73202, a 21.1% lead. These results paint a consistent picture: when the workload is built around contemporary graphics APIs, the Arc B580’s newer architecture provides a clear and significant performance advantage.

However, the AMD Radeon RX 6800S is not without its own victories. Its most notable win comes in the Passmark DirectX 10 test, where it scores 91 against the Arc B580’s 76, a 19.7% margin. This is a substantial lead in an older API, suggesting that the RX 6800S retains strong performance in legacy DirectX titles. The AMD card also wins the Passmark DirectX 9 test, scoring 200 versus 183, a 9.3% advantage. These wins indicate that for older games or applications using these APIs, the RX 6800S is the more capable card.

The remaining wins are closer. In Passmark DirectX 11, the RX 6800S takes a modest 8.6% lead (139 vs 128). The Passmark G2D test sees the AMD card win by a slim 3.2% margin (732 vs 709). The Passmark G3D score is nearly identical, with the RX 6800S edging out a 1% victory (15908 vs 15748). Conversely, the Arc B580 wins the Passmark DirectX 12 test by 14.5% (76 vs 65) and the Passmark GPU Compute test by a significant 20.5% (7729 vs 6144). The overall average benchmark score slightly favors the RX 6800S at 24063 versus 23021, a difference of about 4.3%. This places the RX 6800S in the 69th percentile of all GPUs, while the Arc B580 sits just behind in the 68th percentile.

Where Each One Wins

Based on the benchmark results, the Intel Arc B580 is the clear winner for modern gaming and compute tasks. Its massive 44.9% lead in 3DMark Steel Nomad DX12 is the most telling indicator, showing that it is far more adept at handling the demands of current and future game engines. The 25.9% lead in Geekbench Vulkan and 21.1% lead in OpenCL further cement its status as the superior choice for applications that utilize these cross-platform APIs. The 20.5% win in Passmark GPU Compute also highlights its strength in general-purpose GPU computing, making it the better option for tasks like rendering or machine learning inference.

Conversely, the AMD Radeon RX 6800S is the winner for users with a library of older games or applications that rely on legacy DirectX APIs. Its 19.7% win in DirectX 10 and 9.3% win in DirectX 9 are significant, indicating that it will provide a smoother experience in classic titles that do not support newer APIs. Its 8.6% win in DirectX 11 also gives it an edge in that still-common API. The near-tie in Passmark G3D (1% difference) suggests that for overall rasterization performance in a mixed workload, the two cards are essentially equivalent, but the RX 6800S’s wins in older APIs make it the safer bet for backward compatibility.

Architecture Differences

The two cards are built on fundamentally different architectures. The AMD Radeon RX 6800S utilizes the RDNA 2.0 architecture, implemented with the Navi 23 chip, which is part of the Navi Mobile (RX 6000M) generation. In contrast, the Intel Arc B580 uses the Xe2-HPG architecture, built around the BMG-G21 chip from the Battlemage (Arc 5) generation. This architectural difference is the primary driver behind the performance split seen in the benchmarks.

The manufacturing process also differs significantly. The RX 6800S is built on a 7 nm process at TSMC, while the Arc B580 uses a more advanced 5 nm process, also at TSMC. This allows the Intel card to pack 19,600 million transistors onto a 272 mm² die, resulting in a transistor density of 72.1M per mm². The AMD chip, by contrast, has 11,060 million transistors on a 237 mm² die, for a density of 46.7M per mm². The Arc B580’s higher transistor count and density are key enablers of its superior modern performance.

The architectural differences extend to the compute units. The Arc B580 features 2560 shading units, 160 TMUs, and 80 ROPs, while the RX 6800S has 2048 shading units, 128 TMUs, and 64 ROPs. The Intel card also has 20 ray tracing cores, compared to 32 on the AMD card. Interestingly, despite having fewer RT cores, the Intel card’s newer architecture delivers better overall performance in modern tests. The theoretical pixel and texture rates also favor the Arc B580, with 213.6 GPixel/s and 427.2 GTexel/s, respectively, versus 134.4 GPixel/s and 268.8 GTexel/s for the RX 6800S.

Specification Differences

The most apparent specification difference is the memory configuration. The AMD Radeon RX 6800S comes with 8 GB of GDDR6 memory on a 128-bit bus, providing a bandwidth of 256.0 GB/s. The Intel Arc B580, on the other hand, offers 12 GB of GDDR6 memory on a wider 192-bit bus, yielding a significantly higher bandwidth of 456.0 GB/s. This 78% increase in bandwidth is a major factor in the Arc B580’s strong performance in modern, memory-intensive workloads.

Clock speeds also differ. The RX 6800S has a base clock of 1800 MHz and a boost clock of 2100 MHz, with a game clock of 1975 MHz. The Arc B580 runs at a higher and fixed clock speed of 2670 MHz for both base and boost. This higher clock speed contributes to the Intel card’s superior compute throughput. The FP32 performance reflects this, with the Arc B580 delivering 13.67 TFLOPS versus the RX 6800S’s 8.602 TFLOPS. FP16 performance is similarly higher on the Intel card at 27.34 TFLOPS versus 17.20 TFLOPS.

Power and physical characteristics are starkly different. The RX 6800S has a TDP of 100 W and is an IGP (integrated graphics processor) with no power connectors, designed for portable devices. The Arc B580 is a dual-slot desktop card with a 190 W TDP, requiring a single 8-pin power connector and a 450 W suggested PSU. Its dimensions are 272 mm in length, 115 mm in height, and 45 mm in width. The RX 6800S has no listed dimensions. The AMD card’s display outputs are described as "Portable Device Dependent," while the Intel card offers 1x HDMI 2.1a and 3x DisplayPort 2.1.

FAQ

Q: Which card is faster in modern DirectX 12 games?

A: The Intel Arc B580 is significantly faster. In the 3DMark Steel Nomad DX12 benchmark, it scores 3068 points versus the AMD Radeon RX 6800S’s 1689, a 44.9% advantage.

Q: Does the AMD Radeon RX 6800S have any performance advantages?

A: Yes, it wins in several legacy API tests. It leads by 19.7% in Passmark DirectX 10 (91 vs 76), by 9.3% in DirectX 9 (200 vs 183), and by 8.6% in DirectX 11 (139 vs 128). It also narrowly wins the Passmark G3D test by 1% (15908 vs 15748).

Q: How do their memory configurations compare?

A: The Intel Arc B580 has 12 GB of GDDR6 memory on a 192-bit bus, providing 456.0 GB/s of bandwidth. The AMD Radeon RX 6800S has 8 GB of GDDR6 on a 128-bit bus, with 256.0 GB/s of bandwidth.

Q: Which card has a higher transistor count?

A: The Intel Arc B580 has 19,600 million transistors, compared to 11,060 million on the AMD Radeon RX 6800S. The Intel card is also built on a smaller 5 nm process versus the 7 nm process used for the AMD card.

Q: What is the performance difference in compute workloads?

A: The Intel Arc B580 is stronger in compute. It wins the Passmark GPU Compute test by 20.5% (7729 vs 6144) and also leads in Geekbench OpenCL by 21.1% (92821 vs 73202).

Q: What are their overall performance rankings?

A: The AMD Radeon RX 6800S has an average benchmark score of 24063 and sits in the 69th percentile of all GPUs. The Intel Arc B580 has an average score of 23021 and is in the 68th percentile.

The Verdict

The data presents a clear choice based on use case. The Intel Arc B580 is the superior card for anyone building a desktop system focused on modern gaming and productivity. Its massive 44.9% lead in 3DMark Steel Nomad DX12, coupled with 25.9% and 21.1% leads in Vulkan and OpenCL respectively, demonstrates that it is the far more future-proof and capable option for contemporary software. Its 12 GB of VRAM and higher memory bandwidth further solidify its position for high-resolution textures and complex scenes. The higher TDP and desktop form factor are acceptable trade-offs for this level of modern performance.

The AMD Radeon RX 6800S, despite being an end-of-life mobile part, still holds value in a very specific niche. Its wins in DirectX 9, 10, and 11 benchmarks make it a better choice for users who primarily play older games that rely on these legacy APIs. Its significantly lower TDP of 100 W also makes it suitable for power-constrained systems. However, for a general-purpose desktop GPU, the Arc B580 is the stronger recommendation, as its performance in modern APIs is decisively better, and its overall average benchmark score is only 4.3% lower, a gap that is easily overshadowed by its generational advantages in current workloads. The choice is therefore between modern dominance with the Arc B580 and legacy compatibility with the RX 6800S.