AMD Radeon RX 570 vs NVIDIA GeForce GTX 1060 5 GB Comparison
AMD Radeon RX 570
GeForce GTX 1060 5 GB
PERFORMANCE BENCHMARKS
Analysis: AMD Radeon RX 570 vs NVIDIA GeForce GTX 1060 5 GB
Head-to-Head Benchmarks
The recorded benchmark data splits the two contenders almost perfectly evenly, with each card taking one decisive victory. In the Geekbench OpenCL test, the NVIDIA GeForce GTX 1060 5 GB scores 33,397 against the AMD Radeon RX 570's 32,084, a 4.1% advantage for the NVIDIA card. That is a modest but measurable lead in a compute-oriented workload that favors the Pascal architecture's scheduling efficiency.
The Vulkan test tells a very different story. The AMD Radeon RX 570 posts 42,752, while the GTX 1060 5 GB manages 33,991. That is a 20.5% swing in AMD's favor, a substantial margin that highlights how differently the two architectures respond to modern low-level graphics APIs. The RX 570's GCN 4.0 design clearly extracts more throughput from Vulkan's explicit command model, while the GTX 1060 5 GB falls behind despite its higher OpenCL showing.
The average benchmark score across all recorded tests tells a broader story. The GTX 1060 5 GB sits at 33,694, while the RX 570 averages 28,766. That 4,928-point gap reflects the fact that the NVIDIA card's benchmark profile includes only two tests, both of which are near its peak performance, whereas the RX 570's profile includes a wider set of workloads, including a 3DMark Steel Nomad DX12 run at 880 and a Geekbench Metal score of 39,349. When looking at the percentile rankings, the GTX 1060 5 GB lands in the 78th percentile of all GPUs, while the RX 570 sits at the 74th percentile. The database's nearest-rival data reinforces this positioning: the GTX 1060 5 GB trades blows with the NVIDIA RTX A5000 (0.2% delta), the AMD Radeon RX 7700S (0.5% delta), and the AMD Radeon HD 7950 (0.8% delta). The RX 570, by contrast, sits within 0.4% of the AMD Radeon RX 6800M and within 0.8% of the AMD Radeon RX 470, showing it belongs to a slightly lower performance tier overall.
Architecture Differences
The two cards come from fundamentally different design philosophies. The GTX 1060 5 GB uses NVIDIA's GP106 chip built on the Pascal architecture at TSMC's 16 nm process. The die measures 200 mm² and packs 4,400 million transistors, giving a transistor density of 22.0 million per square millimeter. The RX 570 uses AMD's Polaris 20 chip on GCN 4.0 architecture, produced at GlobalFoundries on a 14 nm node. Its die is larger at 232 mm², with 5,700 million transistors, achieving 24.6 million per square millimeter.
The compute resources differ sharply. The GTX 1060 5 GB has 1,280 shading units, 80 texture mapping units, and 40 raster operations pipelines. The RX 570 counters with 2,048 shading units, 128 TMUs, and 32 ROPs. This explains the texture rate disparity: the RX 570 delivers 159.2 GTexel/s versus the GTX 1060 5 GB's 136.7 GTexel/s. The pixel rate goes the other way, with the GTX 1060 5 GB hitting 68.36 GPixel/s against the RX 570's 39.81 GPixel/s, a consequence of the NVIDIA card's higher ROP count and much higher clock speeds.
Clock behavior is another major split. The GTX 1060 5 GB runs at a 1506 MHz base and 1709 MHz boost, while the RX 570 operates at 1168 MHz base and 1244 MHz boost. The NVIDIA card's higher clocks partially compensate for its smaller compute array, but the raw FP32 throughput still favors AMD: the RX 570 delivers 5.095 TFLOPS against the GTX 1060 5 GB's 4.375 TFLOPS. The FP16 story is even more one-sided. The RX 570 runs FP16 at a 1:1 ratio, matching its FP32 rate at 5.095 TFLOPS. The GTX 1060 5 GB runs FP16 at a 1:64 ratio, producing just 68.36 GFLOPS, a negligible figure that indicates Pascal was never designed for half-precision compute.
Memory configuration completes the architectural picture. The GTX 1060 5 GB uses 5 GB of GDDR5 on a 160-bit bus, yielding 160.2 GB/s of bandwidth. The RX 570 uses 4 GB of GDDR5 on a 256-bit bus, delivering 224.0 GB/s. The AMD card's wider bus gives it a 39.9% bandwidth advantage, which shows up clearly in the Vulkan benchmark. Memory speed differs as well: the GTX 1060 5 GB runs at 2002 MHz (8 Gbps effective), while the RX 570 runs at 1750 MHz (7 Gbps effective), but the wider bus more than compensates for the lower clock.
API support shows a small but notable difference. Both cards support DirectX 12 and OpenGL 4.6. The GTX 1060 5 GB supports DirectX 12_1 and Vulkan 1.4, while the RX 570 supports DirectX 12_0 and Vulkan 1.3. The NVIDIA card has a slight edge in feature-level support, but the Vulkan benchmark results suggest AMD's driver implementation extracts more real-world performance from the API.
FAQ
Q: Why does the RX 570 win the Vulkan benchmark by such a large margin?
A: The RX 570 scores 42,752 in Geekbench Vulkan versus the GTX 1060 5 GB's 33,991, a 20.5% difference. The AMD card's GCN 4.0 architecture, with 2,048 shading units and 224.0 GB/s of memory bandwidth, appears to handle Vulkan's explicit threading and command-buffer model more efficiently than Pascal's 1,280-shader design.
Q: Which card has higher raw compute throughput?
A: The RX 570 leads in FP32 at 5.095 TFLOPS versus 4.375 TFLOPS for the GTX 1060 5 GB. The AMD card also offers FP16 at a 1:1 ratio (5.095 TFLOPS), whereas the GTX 1060 5 GB runs FP16 at a 1:64 ratio, producing only 68.36 GFLOPS.
Q: Does the GTX 1060 5 GB have any memory advantage despite its smaller bus?
A: The GTX 1060 5 GB has 5 GB of GDDR5 versus 4 GB on the RX 570, but its 160-bit bus provides only 160.2 GB/s of bandwidth. The RX 570's 256-bit bus delivers 224.0 GB/s, so the AMD card moves data faster, while the NVIDIA card holds a 1 GB capacity edge.
Q: How do the two cards compare in the OpenCL test?
A: The GTX 1060 5 GB scores 33,397 against the RX 570's 32,084, a 4.1% advantage. This is a narrower margin than the Vulkan gap, suggesting the two architectures trade efficiency based on workload type.
Q: What is the overall performance tier for each card?
A: The GTX 1060 5 GB sits at the 78th percentile of all GPUs, with an average benchmark score of 33,694. The RX 570 sits at the 74th percentile, with an average score of 28,766. The RX 570's nearest rival is the AMD Radeon RX 6800M at a 0.4% delta, while the GTX 1060 5 GB's nearest rival is the NVIDIA RTX A5000 at a 0.2% delta.
Q: Which card has better API feature support?
A: The GTX 1060 5 GB supports DirectX 12_1 and Vulkan 1.4, while the RX 570 supports DirectX 12_0 and Vulkan 1.3. Both cards support OpenGL 4.6.
The Verdict
The data points to a split decision. The GTX 1060 5 GB wins the OpenCL benchmark by 4.1%, has a higher percentile ranking (78th versus 74th), a higher average benchmark score (33,694 versus 28,766), and double the VRAM capacity (5 GB versus 4 GB). It also draws less power, with a 120 W TDP against the RX 570's 150 W, and requires a 300 W PSU versus 450 W.
The RX 570 wins the Vulkan benchmark by 20.5%, a dominant margin, and offers higher raw compute throughput in both FP32 and FP16. Its 224.0 GB/s memory bandwidth is a significant asset for bandwidth-sensitive workloads, and its texture rate of 159.2 GTexel/s exceeds the GTX 1060 5 GB's 136.7 GTexel/s.
For a builder prioritizing OpenCL compute performance, higher VRAM capacity, lower power draw, and a smaller PSU requirement, the GTX 1060 5 GB is the stronger choice. For a builder prioritizing Vulkan gaming performance, raw FP32 throughput, or FP16 compute, the RX 570 is clearly ahead. The choice hinges on the specific application mix.
Specification Differences
| Specification | NVIDIA GeForce GTX 1060 5 GB | AMD Radeon RX 570 |
|---|---|---|
| Architecture | Pascal | GCN 4.0 |
| Process node | 16 nm | 14 nm |
| Transistors | 4,400 million | 5,700 million |
| Die size | 200 mm² | 232 mm² |
| Transistor density | 22.0M / mm² | 24.6M / mm² |
| Base clock | 1506 MHz | 1168 MHz |
| Boost clock | 1709 MHz | 1244 MHz |
| Memory size | 5 GB | 4 GB |
| Memory bus width | 160 bit | 256 bit |
| Memory bandwidth | 160.2 GB/s | 224.0 GB/s |
| Shading units | 1280 | 2048 |
| TMUs | 80 | 128 |
| ROPs | 40 | 32 |
| Pixel rate | 68.36 GPixel/s | 39.81 GPixel/s |
| Texture rate | 136.7 GTexel/s | 159.2 GTexel/s |
| FP32 | 4.375 TFLOPS | 5.095 TFLOPS |
| FP16 | 68.36 GFLOPS (1:64) | 5.095 TFLOPS (1:1) |
| TDP | 120 W | 150 W |
| Suggested PSU | 300 W | 450 W |
| DirectX support | 12 (12_1) | 12 (12_0) |
| Vulkan support | 1.4 | 1.3 |
| Display outputs | 1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.4a | 1x DVI, 1x HDMI 2.0b, 3x DisplayPort 1.4a |
| Length | 250 mm (9.8 inches) | 241 mm (9.5 inches) |
| Release date | 2017-12-25 | 2017-04-17 |
Where Each One Wins
The GTX 1060 5 GB wins in OpenCL compute. Its 33,397 score beats the RX 570's 32,084 by 4.1%. The NVIDIA card also holds a 5 GB versus 4 GB memory capacity advantage, which matters for workloads that exceed 4 GB of frame buffer. Its lower 120 W TDP and 300 W PSU requirement make it the easier fit for smaller power supplies or tighter cases. The higher pixel rate of 68.36 GPixel/s, driven by 40 ROPs and higher clocks, helps in fill-rate-bound scenarios.
The RX 570 wins in Vulkan workloads. Its 42,752 score against 33,991 is a 20.5% advantage, the largest single delta in the entire comparison. The AMD card's 5.095 TFLOPS FP32 throughput and 5.095 TFLOPS FP16 throughput give it a decisive edge in compute-heavy tasks, particularly those that can exploit half-precision. Its 224.0 GB/s memory bandwidth and 159.2 GTexel/s texture rate make it the better choice for texture-heavy rendering and bandwidth-sensitive workloads. The RX 570's 128 TMUs versus 80 on the NVIDIA card further supports this advantage.
The split is clean and practical. If the workload is dominated by OpenCL or requires more than 4 GB of VRAM, the GTX 1060 5 GB is the answer. If the workload is Vulkan-based, uses FP16 compute, or demands maximum memory bandwidth, the RX 570 is the answer. The recorded data gives no single winner across all tests; the correct choice depends entirely on the application.