AMD Radeon RX 6800S vs AMD Radeon RX 9070 Comparison

The AMD Radeon RX 6800S and AMD Radeon RX 9070 represent two distinct generations of AMD graphics, separated by a significant architectural leap and a clear performance gulf. The data shows a decisive overall victory for the RX 9070, which claims 9 out of 10 head-to-head benchmarks, yet the RX 6800S manages to secure a notable single win that highlights a specific strength. The average benchmark scores are remarkably close—24063 for the RX 6800S and 23877 for the RX 9070—but this aggregate figure masks a wide divergence in individual test results, with the newer card often winning by large margins while losing one test by a substantial percentage.

Head-to-Head Benchmarks

The most dramatic difference appears in the 3DMark Steel Nomad DX12 test, which is a modern, demanding workload. The RX 9070 scores 6290, while the RX 6800S manages only 1689, resulting in a delta of -73.1% in favor of the RX 9070. This is not a marginal improvement; it is a generational stomping that indicates the RX 9070 is in a different performance class for contemporary rendering tasks. The RX 9070 also shows a massive lead in compute-oriented workloads, with the Passmark GPU Compute score of 14737 versus 6144 for the RX 6800S, a -58.3% delta. Similarly, Geekbench OpenCL results show the RX 9070 at 131539 against 73202, a -44.3% difference, reinforcing that the newer card's raw processing power is far superior.

In legacy DirectX tests, the RX 9070 continues its dominance. In Passmark DirectX 11, it scores 281 compared to 139 for the RX 6800S, a -50.5% delta. The DirectX 9 test shows 343 versus 200, a -41.7% delta, and DirectX 10 shows 141 versus 91, a -35.5% delta. Even in the Passmark G3D test, which is a more general metric, the RX 9070 wins 25381 to 15908, a -37.3% delta. The 2D performance is also not close, with the RX 9070 posting a Passmark G2D score of 1280 against 732, a -42.8% delta. The smallest win for the RX 9070 is in DirectX 12, where it scores 74 versus 65, a -12.2% delta, showing a more modest but still decisive advantage in that specific API.

However, the RX 6800S has a single, significant victory. In the Geekbench Vulkan test, it scores 81272, while the RX 9070 scores 58705, a +38.4% delta. This is a substantial win and is the only test where the older card comes out ahead. This suggests that for workloads specifically optimized for the Vulkan API, the RX 6800S has a unique advantage that is not replicated in other benchmarks. It is a stark reminder that raw specs do not always translate to universal superiority, and that API-level optimizations can still favor older architectures in specific scenarios. The data indicates that while the RX 9070 is the clear overall winner, the RX 6800S is not without its own niche where it can outperform its successor.

FAQ

Q: Which card has the higher average benchmark score?

A: The AMD Radeon RX 6800S has a slightly higher average benchmark score of 24063, compared to the RX 9070's average of 23877. The RX 9070's nearest rival data shows a -0.8% delta relative to the RX 6800S, indicating the performance difference in the aggregate is very small, despite the large variance in individual tests.

Q: How does the RX 9070 compare to the RX 6800S in the 3DMark Steel Nomad test?

A: The RX 9070 is dramatically faster in this test, scoring 6290 against the RX 6800S's 1689. This represents a -73.1% delta in favor of the RX 9070, making it the largest performance gap between the two cards in any benchmark.

Q: In which benchmark does the RX 6800S beat the RX 9070?

A: The RX 6800S wins the Geekbench Vulkan test with a score of 81272, while the RX 9070 scores 58705. This is a +38.4% delta, making it the only test out of ten where the RX 6800S comes out ahead.

Q: What is the performance difference in the Passmark G3D test?

A: The RX 9070 wins the Passmark G3D test with a score of 25381, compared to 15908 for the RX 6800S. This is a -37.3% delta, showing a clear and significant advantage for the newer card in this general 3D performance metric.

Q: Are these two cards considered to be in the same performance percentile?

A: Yes, both the AMD Radeon RX 6800S and the AMD Radeon RX 9070 have a percentile ranking of 69 against all GPUs. This places them in the same tier when considering their overall standing, despite the RX 9070 winning the majority of direct comparisons.

Q: What is the difference in their Passmark DirectX 12 scores?

A: The RX 9070 scores 74, while the RX 6800S scores 65. The delta is -12.2%, which is the smallest margin of victory for the RX 9070 in any of its wins, indicating a closer performance level in this specific API test.

The Verdict

The data unequivocally points to the AMD Radeon RX 9070 as the superior performer. It wins 9 out of 10 benchmarks, with victories ranging from a modest 12.2% in DirectX 12 to a crushing 73.1% in 3DMark Steel Nomad. For any user prioritizing modern gaming performance, compute workloads, or general 3D rendering, the RX 9070 is the clear choice based on benchmark results. Its wins in OpenCL, G3D, and compute tests show it is not just faster in one area but across the board.

The case for the RX 6800S is narrow but real. Its single victory in the Geekbench Vulkan test is substantial, with a +38.4% delta over the RX 9070. This suggests that for specific applications or games that are heavily optimized for the Vulkan API and perhaps favor the older RDNA 2.0 architecture's characteristics, the RX 6800S could offer a better experience. However, this is a single data point against a broad trend of defeat. The RX 6800S's higher average benchmark score of 24063 versus 23877 is a statistical curiosity, but it is heavily influenced by its strong Vulkan result and does not reflect its overall performance in the majority of tests.

Given the benchmark data, a user should pick the RX 9070 unless they have a specific, confirmed workload that relies on Vulkan and shows a preference for the RX 6800S's architecture. The RX 9070's performance in the 3DMark Steel Nomad test alone, which is likely indicative of future game demands, makes it the more future-proof and capable option. The RX 6800S is a legacy product that retains a niche advantage, while the RX 9070 is the dominant choice for almost all other scenarios.

Specification Differences

The two cards differ in nearly every core specification. The RX 9070 is built on a 4 nm process node, while the RX 6800S uses a 7 nm node. The RX 9070 has a larger die size of 357 mm² compared to 237 mm², and it packs 53,900 million transistors versus 11,060 million. The RX 9070 also has a much higher transistor density of 151.0M / mm², compared to 46.7M / mm² for the RX 6800S. Memory configurations are also different: the RX 9070 has 16 GB of GDDR6 memory on a 256-bit bus with 644.6 GB/s of bandwidth, while the RX 6800S has 8 GB of GDDR6 on a 128-bit bus with 256.0 GB/s of bandwidth.

The compute units scale up significantly on the RX 9070, which has 3584 shading units, 224 TMUs, and 128 ROPs, versus 2048 shading units, 128 TMUs, and 64 ROPs on the RX 6800S. The RX 9070 also has 56 ray tracing cores compared to 32. Clock speeds vary, with the RX 6800S having a higher base clock of 1800 MHz versus 1330 MHz, but the RX 9070 has a higher boost clock of 2520 MHz versus 2100 MHz, and a higher game clock of 2070 MHz versus 1975 MHz. The power profiles are also distinct, with the RX 9070 being a dual-slot card with a 220 W TDP and requiring two 8-pin power connectors, while the RX 6800S is an IGP with a 100 W TDP and no power connectors. The PCIe interface differs, with the RX 9070 using PCIe 5.0 x16 and the RX 6800S using PCIe 4.0 x8. The RX 9070 also features 1x HDMI 2.1b and 3x DisplayPort 2.1a outputs, while the RX 6800S has portable-device-dependent outputs.

Architecture Differences

The architectural gap is fundamental. The RX 6800S is based on the RDNA 2.0 architecture with the Navi 23 chip, part of the Navi Mobile (RX 6000M) generation. The RX 9070, in contrast, uses the newer RDNA 4.0 architecture on the Navi 48 chip, from the Navi IV (RX 9000) generation. This generational leap is reflected in their process nodes, with the RX 9070 using a more advanced 4 nm process from TSMC, versus the 7 nm process for the RX 6800S. The RX 9070's architecture also delivers a different compute capability, with FP32 performance of 36.13 TFLOPS compared to 8.602 TFLOPS on the RX 6800S. Notably, the RX 9070 achieves FP16 performance at a 1:1 ratio (36.13 TFLOPS), while the RX 6800S does so at a 2:1 ratio (17.20 TFLOPS), indicating a different design philosophy for compute workloads. Both cards support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, but the underlying hardware implementation is from different eras, explaining the performance disparities. The RX 9070 is an active product with a launch MSRP of 549 USD, while the RX 6800S is end-of-life.

Where Each One Wins

The RX 9070 wins in almost every measurable category. It is the clear victor in modern DX12 workloads as shown by the 3DMark Steel Nomad test, in general compute via OpenCL and Passmark GPU Compute, in legacy DirectX APIs from 9 to 11, in general 3D performance via Passmark G3D, and in 2D performance. For a user building a new system for gaming or content creation, the RX 9070's wins in these areas make it the obvious choice. Its performance in the 3DMark Steel Nomad test is particularly important, as it suggests a strong ability to handle future, more demanding game engines.

The RX 6800S wins exclusively in the Geekbench Vulkan test. This makes it a specialist card for Vulkan-specific applications. If a user's primary workload is a particular game or application that is known to be bottlenecked by other APIs but performs exceptionally well on Vulkan, the RX 6800S could be a viable option despite its age. However, this is a high-risk bet, as it has no other benchmark wins to fall back on. The data suggests that for a general-purpose GPU, the RX 9070 is the superior choice, while the RX 6800S is only preferable in a very narrow, Vulkan-centric use case.