AMD Radeon RX 570 vs AMD Radeon RX 6800 Comparison

The Verdict

The benchmark data presents a decisive generational split between these two AMD GPUs. The AMD Radeon RX 6800 wins three of four head-to-head tests, with victories ranging from 169.2% to 290.4% over the RX 570. The RX 6800's average benchmark score of 30095 places it in the 75th percentile of all GPUs, while the RX 570's average of 28766 sits at the 74th percentile—a narrow overall gap that masks the dramatic differences in individual workloads. The RX 6800 is the clear choice for modern DirectX 12 and Vulkan gaming, as well as Metal-based compute, where its RDNA 2.0 architecture delivers transformative performance gains. The RX 570 retains relevance only in OpenCL compute workloads, where it actually leads by 23.6%, and for users who prioritize a much lower power draw (150 W vs 250 W) and a simpler single 6-pin power connector. For anyone building a system around current-generation APIs, the RX 6800 is the data-backed selection; for legacy OpenCL compute or a lower-power secondary build, the RX 570's single win matters.

Architecture Differences

The architectural gap between these two cards spans three major process and design generations. The RX 6800 uses the Navi 21 chip on TSMC's 7 nm process, packing 26,800 million transistors into a 520 mm² die with a transistor density of 51.5M per mm². The RX 570 uses the Polaris 20 chip on GlobalFoundries' 14 nm process, with 5,700 million transistors on a 232 mm² die and a density of 24.6M per mm². This represents a 4.7x transistor count advantage for the RX 6800 on a die that is 2.24x larger.

The compute resources scale accordingly. The RX 6800 features 3840 shading units, 240 texture mapping units, and 96 ROPs, compared to the RX 570's 2048 shaders, 128 TMUs, and 32 ROPs. The RX 6800 also introduces 60 dedicated ray tracing cores, a feature entirely absent from the RX 570. Clock speeds differ significantly: the RX 6800 runs at a 1700 MHz base and 2105 MHz boost, while the RX 570 operates at 1168 MHz base and 1244 MHz boost. This clock advantage compounds with the architectural efficiency of RDNA 2.0 versus GCN 4.0.

Memory subsystems further separate the two. The RX 6800 carries 16 GB of GDDR6 on a 256-bit bus, delivering 512.0 GB/s of bandwidth. The RX 570 has 4 GB of GDDR5 on the same 256-bit bus width, but only 224.0 GB/s. The effective memory speed is 16 Gbps versus 7 Gbps. The RX 6800 supports PCIe 4.0 x16, while the RX 570 uses PCIe 3.0 x16. Feature support also diverges: the RX 6800 hits DirectX 12 Ultimate (12_2) and Vulkan 1.4, whereas the RX 570 is limited to DirectX 12 (12_0) and Vulkan 1.3. The RX 6800's display outputs include HDMI 2.1, two DisplayPort 1.4a, and a USB Type-C port, versus the RX 570's DVI, HDMI 2.0b, and three DisplayPort 1.4a connectors. Power requirements reflect the performance gap: 250 W TDP with dual 8-pin connectors and a 600 W suggested PSU for the RX 6800, versus 150 W, a single 6-pin, and a 450 W PSU for the RX 570.

Head-to-Head Benchmarks

The 3DMark Steel Nomad DX12 test shows the most extreme separation. The RX 6800 scores 3188 against the RX 570's 880, a 262.3% advantage. This workload favors the RX 6800's DirectX 12 Ultimate feature set and its 505.2 GTexel/s texture rate versus 159.2 GTexel/s on the RX 570. The pixel rate tells a similar story: 202.1 GPixel/s versus 39.81 GPixel/s, a 5x difference that directly impacts rasterization-heavy scenes.

Geekbench Metal results reinforce the pattern. The RX 6800 scores 153635 versus 39349, a 290.4% lead—the largest single delta in the entire comparison. This test leverages the RX 6800's higher FP32 throughput of 16.17 TFLOPS against the RX 570's 5.095 TFLOPS. The FP16 capability also differs: the RX 6800 delivers 32.33 TFLOPS (2:1 ratio), while the RX 570 manages only 5.095 TFLOPS (1:1), meaning the RX 6800 can double its compute rate on half-precision workloads.

Geekbench Vulkan shows a 169.2% win for the RX 6800, scoring 115107 against 42752. This is the smallest relative win for the RX 6800, but still a massive margin. The Vulkan 1.4 support on the RX 6800 versus Vulkan 1.3 on the RX 570 likely contributes to better driver-level optimization in this test.

The single counter-example is Geekbench OpenCL, where the RX 570 wins 32084 to 24508, a 23.6% margin. This result is anomalous given the RX 6800's superior compute specifications, but it demonstrates that OpenCL optimization on GCN 4.0 can outperform the newer architecture in this specific API. The RX 570's 1:1 FP16 ratio may also play a role in certain OpenCL kernels that do not benefit from the RX 6800's packed math.

FAQ

Q: Which GPU wins more benchmarks overall?

A: The AMD Radeon RX 6800 wins three of four head-to-head tests (3DMark Steel Nomad DX12, Geekbench Metal, Geekbench Vulkan), while the RX 570 wins only Geekbench OpenCL.

Q: How large is the performance gap in the biggest win?

A: The RX 6800 leads by 290.4% in Geekbench Metal, scoring 153635 versus the RX 570's 39349. The smallest RX 6800 win is 169.2% in Geekbench Vulkan (115107 vs 42752).

Q: What is the RX 570's only benchmark victory?

A: The RX 570 wins Geekbench OpenCL with a score of 32084, beating the RX 6800's 24508 by 23.6%. No other test in the head-to-head set favors the RX 570.

Q: How do the memory configurations compare?

A: The RX 6800 has 16 GB of GDDR6 with 512.0 GB/s bandwidth on a 256-bit bus at 16 Gbps effective. The RX 570 has 4 GB of GDDR5 with 224.0 GB/s on the same bus width at 7 Gbps effective.

Q: What are the power requirements for each card?

A: The RX 6800 has a 250 W TDP, requires dual 8-pin power connectors, and recommends a 600 W PSU. The RX 570 has a 150 W TDP, uses a single 6-pin connector, and recommends a 450 W PSU.

Q: Which card supports ray tracing?

A: Only the RX 6800, which includes 60 dedicated ray tracing cores. The RX 570 has no ray tracing hardware.

Where Each One Wins

The RX 6800 dominates across all modern graphics workloads. Its 3DMark Steel Nomad DX12 result of 3188 versus 880 indicates a 262.3% advantage in current-generation DirectX 12 gaming. The RX 6800's DirectX 12 Ultimate support (12_2) versus the RX 570's DirectX 12 (12_0) means the newer card can handle feature levels the older one cannot. For Vulkan-based titles, the 169.2% lead (115107 vs 42752) makes the RX 6800 the definitive choice, especially given its Vulkan 1.4 support versus 1.3 on the RX 570. Metal workloads—relevant for macOS or certain creative applications—show the largest gap at 290.4% (153635 vs 39349), making the RX 6800 the only viable option for serious Metal compute.

The RX 570's sole domain is OpenCL compute, where it wins 32084 to 24508. This 23.6% margin suggests that legacy OpenCL applications, or those with GCN-specific optimizations, may run better on the older architecture. The RX 570's 1:1 FP16 ratio of 5.095 TFLOPS also matches its FP32 throughput, which can be advantageous in certain compute kernels that do not benefit from the RX 6800's packed 2:1 FP16 implementation. The RX 570's lower power draw (150 W vs 250 W), smaller 241 mm length, and single 6-pin connector make it more suitable for compact or lower-wattage builds where the RX 6800's 267 mm length and dual 8-pin requirement would be problematic.

For rasterization-heavy workloads at lower resolutions, the RX 570's 4 GB of GDDR5 may suffice, but the 224.0 GB/s bandwidth versus 512.0 GB/s on the RX 6800 will bottleneck any texture-heavy scene. The RX 570's 39.81 GPixel/s fill rate is less than one-fifth of the RX 6800's 202.1 GPixel/s, making high-resolution rendering impractical on the older card. The data supports the RX 6800 as the winner for every application except OpenCL compute, and even there, the margin is modest compared to the RX 6800's massive leads elsewhere.