AMD Radeon RX 6800M vs AMD Radeon RX 6800S Comparison

The Verdict

The benchmark data presents a split decision. The AMD Radeon RX 6800M wins 6 of 11 head-to-head tests, while the AMD Radeon RX 6800S wins 5. However, the nature of those wins tells a clear story. The 6800M dominates in modern, API-heavy workloads: it is 32.5% ahead in 3DMark Steel Nomad DX12, 33.4% ahead in Geekbench Metal, 19.7% ahead in Geekbench OpenCL, and 16.6% ahead in Geekbench Vulkan. These are the tests that reflect current gaming and compute demands.

The 6800S, conversely, posts wins in legacy DirectX tests and in Passmark's G3D and compute suites. It is 26.5% faster in DirectX 9 and 26.5% faster in G2D, plus 16.6% faster in Passmark G3D and 18.1% faster in GPU compute. The overall average benchmark score favors the 6800M at 28874 versus 24063, a difference of roughly 20%. The 6800M also sits at the 74th percentile of all GPUs, while the 6800S sits at the 69th percentile.

Who should pick which? Strictly from the data, the 6800M is the choice for users prioritizing DirectX 12, Vulkan, Metal, and OpenCL performance. The 6800S is the choice for users whose workloads are dominated by legacy DirectX 9/10 titles or who value the Passmark G3D and compute metrics, where it holds a decisive edge. The 6800S also wins the Passmark DirectX 11 test by 8.6%, so it is not purely a legacy part. The two are tied in DirectX 12 Passmark at 65 points each.

Architecture Differences

Both GPUs share the RDNA 2.0 architecture and are built on TSMC's 7 nm process. The 6800M uses the Navi 22 chip, while the 6800S uses Navi 23. The transistor counts differ substantially: the 6800M packs 17,200 million transistors on a 335 mm² die, while the 6800S has 11,060 million on a 237 mm² die. Transistor density is higher on the 6800M at 51.3M per mm² versus 46.7M per mm² on the 6800S. Both are from the Navi Mobile (RX 6000M) generation.

The compute resources scale accordingly. The 6800M has 2560 shading units, 160 texture mapping units, and 64 ROPs. The 6800S has 2048 shading units, 128 TMUs, and the same 64 ROPs. Ray tracing cores follow the same pattern: 40 on the 6800M versus 32 on the 6800S. Neither has tensor cores.

Clock speeds tell a different part of the story. The 6800M runs a base clock of 2116 MHz, a boost of 2390 MHz, and a game clock of 2300 MHz. The 6800S runs lower: 1800 MHz base, 2100 MHz boost, and 1975 MHz game clock. Despite the 6800S having fewer cores and lower clocks, it wins several Passmark tests. This suggests architectural efficiency differences at the same RDNA 2.0 node, or perhaps driver optimization for the Passmark suite. The memory subsystems also differ, which is covered in the specification section below.

Head-to-Head Benchmarks

The largest margin of victory belongs to the 6800M in Geekbench Metal, where it scores 113721 versus 85256, a 33.4% lead. The 3DMark Steel Nomad DX12 test shows a 32.5% lead, with scores of 2238 versus 1689. These are substantial gaps in modern graphics APIs. The 6800M also leads Geekbench OpenCL by 19.7% (87621 versus 73202) and Geekbench Vulkan by 16.6% (94766 versus 81272).

The 6800S fights back in the Passmark suite. In DirectX 9, it scores 200 versus 147, a 26.5% advantage. It matches that 26.5% margin in G2D, scoring 732 versus 538. In Passmark G3D, the 6800S scores 15908 against 13261, a 16.6% lead. The GPU compute test shows 6144 versus 5032, an 18.1% lead for the 6800S. The DirectX 11 test is closer, with the 6800S ahead by 8.6% (139 versus 127). The DirectX 10 and DirectX 12 tests are near ties: the 6800M wins DirectX 10 by 11% (101 versus 91) and DirectX 12 is a dead heat at 65 each.

This pattern is curious. The 6800M, with more cores, higher clocks, and more bandwidth, loses the Passmark G3D and compute tests by double digits. The data suggests the Passmark suite may favor the 6800S's specific configuration, or the 6800S drivers are tuned differently for those workloads. The 6800M's wins in Geekbench and 3DMark are more representative of modern gaming loads.

The nearest rival data adds context. The 6800M's average score of 28874 places it just 0.4% above the AMD Radeon RX 570 and 1% above the Radeon R9 M295X, while sitting 0.4% below the Radeon RX 470 and 1% below the Intel Arc A370M. The 6800S's average of 24063 places it 0.8% above the Radeon RX 9070, but 0.4% below the GeForce RTX 2080 SUPER, 0.7% below the GTX 780 Ti, and 0.9% below the GTX 1630. These are tight groupings, meaning both mobile GPUs are competitive with a broad range of desktop and mobile parts from several generations.

Specification Differences

The two GPUs differ in several key fields. The chip is different: Navi 22 for the 6800M versus Navi 23 for the 6800S. Transistors are 17,200 million versus 11,060 million. Die size is 335 mm² versus 237 mm². Transistor density is 51.3M per mm² versus 46.7M per mm².

Clock speeds differ across the board. Base clock is 2116 MHz versus 1800 MHz. Boost is 2390 MHz versus 2100 MHz. Game clock is 2300 MHz versus 1975 MHz. Memory clocks are identical at 2000 MHz with 16 Gbps effective.

Memory configuration diverges significantly. The 6800M has 12 GB of GDDR6 on a 192-bit bus, yielding 384.0 GB/s bandwidth. The 6800S has 8 GB of GDDR6 on a 128-bit bus, yielding 256.0 GB/s bandwidth. That is a 50% bandwidth advantage for the 6800M.

Compute unit counts differ: 2560 shading units versus 2048, 160 TMUs versus 128, and 40 RT cores versus 32. ROPs are equal at 64. Pixel rates are 153.0 GPixel/s for the 6800M versus 134.4 GPixel/s for the 6800S. Texture rates are 382.4 GTexel/s versus 268.8 GTexel/s. FP32 throughput is 12.24 TFLOPS versus 8.602 TFLOPS. FP16 is 24.47 TFLOPS (2:1) versus 17.20 TFLOPS (2:1).

TDP is another major difference: 145 W for the 6800M versus 100 W for the 6800S. Both are IGP slot width with no power connectors and portable device dependent display outputs. The 6800M uses PCIe 4.0 x16, while the 6800S uses PCIe 4.0 x8. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Release dates differ: the 6800M launched on 2021-05-30, while the 6800S launched on 2022-01-03. Both are end-of-life and have Polaris Mobile as their predecessor.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon RX 6800M has an average benchmark score of 28874, compared to 24063 for the AMD Radeon RX 6800S. This places the 6800M at the 74th percentile of all GPUs and the 6800S at the 69th percentile.

Q: How do the two compare in DirectX 12 performance?

A: In the 3DMark Steel Nomad DX12 test, the 6800M scores 2238 versus 1689, a 32.5% lead. However, in the Passmark DirectX 12 test, both score exactly 65, resulting in a 0% delta.

Q: Is the 6800S better at any major modern API?

A: No. The 6800S loses in Geekbench Metal (33.4% behind), OpenCL (19.7% behind), and Vulkan (16.6% behind). Its wins are confined to Passmark tests: DirectX 9, DirectX 11, G2D, G3D, and GPU compute.

Q: What are the memory differences?

A: The 6800M has 12 GB of GDDR6 on a 192-bit bus with 384.0 GB/s bandwidth. The 6800S has 8 GB of GDDR6 on a 128-bit bus with 256.0 GB/s bandwidth. Memory clock is identical at 2000 MHz with 16 Gbps effective.

Q: Which GPU has a higher TDP?

A: The 6800M has a TDP of 145 W, while the 6800S has a TDP of 100 W. This is a 45 W difference.

Q: How close are the two in Passmark G3D?

A: The 6800S wins by 16.6%, scoring 15908 versus 13261 for the 6800M. This is one of the largest margins in the Passmark suite, along with the 26.5% leads for the 6800S in DirectX 9 and G2D.

Where Each One Wins

The 6800M is the clear winner in modern graphics APIs. It leads by 33.4% in Geekbench Metal, 32.5% in 3DMark Steel Nomad DX12, 19.7% in OpenCL, and 16.6% in Vulkan. These tests are the closest proxies for current gaming engines and compute applications. The 6800M also wins Passmark DirectX 10 by 11% and ties in DirectX 12. For users running DirectX 12 titles, Vulkan games, or Metal-based applications on compatible platforms, the 6800M is the stronger part by a wide margin. Its 12 GB memory and 384.0 GB/s bandwidth provide a structural advantage for high-resolution textures and data-heavy workloads.

The 6800S wins the legacy and compute-oriented Passmark tests. It is 26.5% faster in DirectX 9 and G2D, 18.1% faster in GPU compute, 16.6% faster in G3D, and 8.6% faster in DirectX 11. The G3D and compute wins are particularly notable because they are not ancient API tests; they represent a broader workload mix. The 6800S achieves this with a 100 W TDP, 45 W lower than the 6800M, and with fewer cores and less bandwidth. This suggests the 6800S is more efficient in certain workloads, or that its driver stack handles the Passmark suite better. For users running older DirectX 9/10/11 titles, or for compute tasks that resemble the Passmark GPU compute test, the 6800S delivers better raw scores.

The use-case split is therefore clean. The 6800M is for modern gaming, high-resolution work, and API-heavy applications. The 6800S is for legacy titles, specific compute tasks, and scenarios where the Passmark G3D and compute scores matter more than Geekbench or 3DMark results. The 6800S also offers the lower TDP, which may matter in thin-and-light designs, though the data does not quantify chassis-level thermal effects. The tie in DirectX 12 Passmark at 65 points means neither has a universal advantage in that specific test. Overall, the 6800M's higher average score and modern API dominance make it the more future-proof choice, while the 6800S's Passmark wins make it a specialist for certain legacy and compute workloads.