AMD Radeon HD 7970 vs AMD Radeon RX 570 Comparison

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon HD 7970 holds the higher average benchmark score at 34,541, while the AMD Radeon RX 570 averages 28,766. This places the HD 7970 in the 79th percentile of all GPUs, versus the 74th percentile for the RX 570.

Q: How do the two cards compare in the Geekbench OpenCL test?

A: The HD 7970 wins the Geekbench OpenCL test with a score of 34,541 against 32,084 for the RX 570, a margin of 7.7%. This is the only head-to-head benchmark recorded in the database.

Q: What are the closest rivals to the RX 570 in the database?

A: The RX 570's nearest rivals by average score are the AMD Radeon RX 6800M (28,874, 0.4% higher), the AMD Radeon R9 M295X (28,580, 0.7% lower), the AMD Radeon RX 470 (28,996, 0.8% higher), and the NVIDIA Quadro RTX 8000 (28,421, 1.2% lower).

Q: Which card has a higher FP32 performance rating?

A: The RX 570 is rated at 5.095 TFLOPS FP32, notably higher than the HD 7970's 3.789 TFLOPS. The RX 570 also lists FP16 performance at 5.095 TFLOPS (1:1), while the HD 7970 has no FP16 rating.

Q: Which card has more video memory bandwidth?

A: The HD 7970 provides 264.0 GB/s of bandwidth over a 384-bit bus with 3 GB GDDR5. The RX 570 provides 224.0 GB/s over a 256-bit bus with 4 GB GDDR5.

Q: How do their power requirements differ?

A: The HD 7970 is rated at 250 W TDP with a suggested 600 W PSU and uses 1x 6-pin plus 1x 8-pin connectors. The RX 570 is rated at 150 W TDP with a suggested 450 W PSU and uses a single 6-pin connector.

Architecture Differences

The AMD Radeon HD 7970 is built on the Tahiti chip using the GCN 1.0 architecture, part of the Southern Islands (HD 7900) generation. It is fabricated on a 28 nm process at TSMC, with 4,313 million transistors on a 352 mm² die, giving a transistor density of 12.3 million per mm². It supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170.

The AMD Radeon RX 570 uses the Polaris 20 chip with GCN 4.0 architecture, belonging to the Polaris (RX 500) generation. It is manufactured on a 14 nm process at GlobalFoundries, with 5,700 million transistors on a 232 mm² die, for a transistor density of 24.6 million per mm². It supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.3.

The architectural leap from GCN 1.0 to GCN 4.0 brings a mature feature set. The RX 570 supports DirectX 12 (12_0), a full level above the HD 7970's 12 (11_1). Vulkan support is also newer on the RX 570, at version 1.3 versus 1.2.170. Both cards have identical shading unit counts, texture mapping units, and ROPs: 2,048 shaders, 128 TMUs, and 32 ROPs. However, the RX 570 sustains higher pixel and texture rates at 39.81 GPixel/s and 159.2 GTexel/s, compared to 29.60 GPixel/s and 118.4 GTexel/s on the HD 7970. The RX 570 also adds FP16 compute capability at a 1:1 ratio to FP32, while the older card has no listed FP16 throughput.

Memory subsystems diverge significantly. The HD 7970 uses 3 GB of GDDR5 on a 384-bit bus at 1,375 MHz (5.5 Gbps effective), yielding 264.0 GB/s bandwidth. The RX 570 uses 4 GB of GDDR5 on a 256-bit bus at 1,750 MHz (7 Gbps effective), for 224.0 GB/s. The RX 570's newer process node compensates with higher clocks and higher density on a smaller die area.

Display outputs also differ by era. The HD 7970 offers 1x DVI, 1x HDMI 4a, and 2x mini-DisplayPort 1.2. The RX 570 provides 1x DVI, 1x HDMI 2.0b, and 3x DisplayPort 1.4a, reflecting updated display standards.

Physical size and power profile separate the two as well. The HD 7970 measures 275 mm (10.8 inches) in length, 111 mm (4.4 inches) in height, and 38 mm (1.5 inches) in width. The RX 570 is shorter at 241 mm (9.5 inches) with no recorded height or width. The RX 570 draws less power, 150 W versus 250 W, and uses a smaller suggested PSU, 450 W versus 600 W.

The Verdict

The data supports the AMD Radeon HD 7970 as the stronger overall GPU in the recorded average benchmark, leading the RX 570 by 5,775 points in average score. In the only direct head-to-head test, Geekbench OpenCL, the HD 7970 wins by 7.7%. It also sits five percentile points higher in the database's all-GPU ranking.

The RX 570 is not without advantages in raw specifications. It delivers 34.5% more FP32 compute (5.095 TFLOPS versus 3.789 TFLOPS), higher pixel and texture rates, and more video memory at 4 GB versus 3 GB. It also draws 100 W less power and requires a 150 W lower suggested PSU. For workloads that rely on compute throughput, the RX 570's measured specification sheet is superior.

However, benchmark results indicate that the HD 7970's memory bandwidth advantage, 264.0 GB/s versus 224.0 GB/s, and its wider 384-bit bus help it overcome the RX 570's compute lead in the OpenCL test. The HD 7970's nearest rivals include the NVIDIA T1000 8 GB (34,561, 0.1% higher), NVIDIA A2 (34,690, 0.4% higher), NVIDIA TITAN V (34,355, 0.5% lower), and NVIDIA RTX A1000 (34,207, 1% lower). The RX 570's nearest rivals include the RX 6800M (28,874, 0.4% higher), R9 M295X (28,580, 0.7% lower), RX 470 (28,996, 0.8% higher), and Quadro RTX 8000 (28,421, 1.2% lower).

Who should pick which: the HD 7970 suits users prioritizing raw OpenCL performance and wider memory bandwidth in the database's measurements. The RX 570 suits users who need higher compute throughput, more VRAM, and lower power consumption, accepting a lower average benchmark score. The RX 570 also provides newer API support, including DirectX 12 (12_0) and Vulkan 1.3.

Specification Differences

| Specification | AMD Radeon HD 7970 | AMD Radeon RX 570 |

|---|---|---|

| Chip | Tahiti | Polaris 20 |

| Architecture | GCN 1.0 | GCN 4.0 |

| Generation | Southern Islands (HD 7900) | Polaris (RX 500) |

| Process Node | 28 nm | 14 nm |

| Foundry | TSMC | GlobalFoundries |

| Transistors | 4,313 million | 5,700 million |

| Die Size | 352 mm² | 232 mm² |

| Transistor Density | 12.3M / mm² | 24.6M / mm² |

| Base Clock | Not recorded | 1168 MHz |

| Boost Clock | Not recorded | 1244 MHz |

| Memory Clock | 1375 MHz (5.5 Gbps effective) | 1750 MHz (7 Gbps effective) |

| Memory Size | 3 GB | 4 GB |

| Memory Bus Width | 384 bit | 256 bit |

| Memory Bandwidth | 264.0 GB/s | 224.0 GB/s |

| Pixel Rate | 29.60 GPixel/s | 39.81 GPixel/s |

| Texture Rate | 118.4 GTexel/s | 159.2 GTexel/s |

| FP32 | 3.789 TFLOPS | 5.095 TFLOPS |

| FP16 | Not recorded | 5.095 TFLOPS (1:1) |

| TDP | 250 W | 150 W |

| Power Connectors | 1x 6-pin + 1x 8-pin | 1x 6-pin |

| Suggested PSU | 600 W | 450 W |

| Display Outputs | 1x DVI, 1x HDMI 1.4a, 2x mini-DisplayPort 1.2 | 1x DVI, 1x HDMI 2.0b, 3x DisplayPort 1.4a |

| DirectX | 12 (11_1) | 12 (12_0) |

| Vulkan | 1.2.170 | 1.3 |

| Length | 275 mm (10.8 inches) | 241 mm (9.5 inches) |

| Height | 111 mm (4.4 inches) | Not recorded |

| Width | 38 mm (1.5 inches) | Not recorded |

| Release Date | 2012-01-08 | 2017-04-17 |

| Predecessor | Northern Islands | Arctic Islands |

| Successor | Sea Islands | Vega |

| Launch MSRP | 549 USD | 169 USD |

Both cards share identical shading unit counts (2,048), TMUs (128), ROPs (32), PCIe 3.0 x16 bus interface, dual-slot width, and OpenGL 4.6 support. The RX 570 is end-of-life, as is the HD 7970.

Head-to-Head Benchmarks

The database records a single direct comparison between these two GPUs: Geekbench OpenCL. The AMD Radeon HD 7970 scores 34,541 against the AMD Radeon RX 570's 32,084, a 7.7% advantage for the older card.

This result is striking given the RX 570's specification advantages. The RX 570 carries 5.095 TFLOPS FP32 compute, 34.5% higher than the HD 7970's 3.789 TFLOPS. It also renders pixels at 39.81 GPixel/s and textures at 159.2 GTexel/s, versus 29.60 GPixel/s and 118.4 GTexel/s. Yet the HD 7970 reclaims the win in OpenCL, likely due to its 264.0 GB/s bandwidth and 384-bit interface, which substantially exceed the RX 570's 224.0 GB/s and 256-bit bus.

The HD 7970's nearest rival in the database, the NVIDIA T1000 8 GB, scores 34,561, just 0.1% ahead. The NVIDIA A2 sits 0.4% higher at 34,690, while the NVIDIA TITAN V trails by 0.5% at 34,355 and the RTX A1000 is 1% behind at 34,207. This places the HD 7970 in a tight cluster near the top of its performance tier.

The RX 570's nearest rivals form a lower cluster. The AMD Radeon RX 470 leads it by 0.8% at 28,996, the RX 6800M by 0.4% at 28,874, the R9 M295X trails by 0.7% at 28,580, and the Quadro RTX 8000 trails by 1.2% at 28,421. The RX 570's average score of 28,766 puts it roughly 20.1% behind the HD 7970's 34,541 in the database's aggregate metric.

The RX 570 does record additional benchmark entries beyond OpenCL: 3DMark Steel Nomad DX12 at 880, Geekbench Metal at 39,349, and Geekbench Vulkan at 42,752. The HD 7970 has no comparable results in those tests, so the RX 570's newer API support is demonstrated by its ability to run Vulkan and Metal workloads at all. Its Vulkan score of 42,752 is its highest recorded result, exceeding its OpenCL score by 33.3%. Its Metal score of 39,349 also outpaces OpenCL by 22.6%. The HD 7970, lacking Vulkan and Metal entries, cannot be compared on those APIs.

The head-to-head tally in the database stands at one win for the HD 7970 and zero for the RX 570. In the OpenCL test, the margin is clear: 7.7%. For users reading the database, the HD 7970 is the higher-scoring card, while the RX 570 offers superior compute specifications and broader API coverage in its recorded results.