AMD Radeon RX 470 vs AMD Radeon RX 6700 XT Comparison

AMD Radeon RX 470 and AMD Radeon RX 6700 XT represent two distinct eras of AMD graphics, separated by nearly five years of architectural evolution. The data shows a decisive performance hierarchy, with the RX 6700 XT dominating every shared benchmark, yet the RX 470 holds a unique position in the historical context of the RX 400 series. Benchmark results indicate that while the RX 6700 XT is the clear performance king, the RX 470’s legacy as a capable 1080p card from its generation remains intact, though it is thoroughly outclassed in raw throughput by its modern counterpart.

Head-to-Head Benchmarks

The RX 6700 XT wins all four directly comparable benchmark tests, with margins ranging from substantial to overwhelming. The largest single-test gap appears in Geekbench Metal, where the RX 6700 XT scores 130,896 against the RX 470’s 41,690, a delta of -68.2% from the RX 470’s perspective. That means the RX 6700 XT delivers roughly 3.1 times the Metal performance, an enormous generational leap that reflects both architectural and clock speed advancements. In 3DMark Steel Nomad DX12, the RX 6700 XT posts 2,435 points versus 842 for the RX 470, a -65.4% delta, translating to about 2.9 times the performance. These two tests highlight the RX 6700 XT’s dominance in both modern DirectX 12 workloads and Apple’s Metal API.

Geekbench Vulkan shows a slightly narrower but still decisive gap: 93,500 for the RX 6700 XT versus 39,884 for the RX 470, a -57.3% delta. This indicates the RX 6700 XT is roughly 2.3 times faster in Vulkan compute tasks, a meaningful advantage for cross-platform gaming and compute applications. The smallest margin appears in Geekbench OpenCL, where the RX 6700 XT scores 44,152 against 33,568 for the RX 470, a -24% delta. While still a clear win for the RX 6700 XT, this relatively modest 1.3x improvement suggests the RX 470’s GCN architecture held up better in OpenCL compute workloads than in other APIs. Overall, the head-to-head record stands at 0 wins for the RX 470 and 4 for the RX 6700 XT.

When placed against their respective nearest rivals, both cards show competitive positioning within their own performance tiers. The RX 470’s average benchmark score of 28,996 places it just 0.4% above the AMD Radeon RX 6800M (28,874) and 0.6% below the Intel Arc A370M (29,175), indicating it sits in a tightly contested mid-range laptop segment. The RX 6700 XT’s average score of 27,425 is essentially tied with the NVIDIA GeForce RTX 4070 Mobile (27,435, 0% delta) and only 0.5% below the GeForce RTX 3090 (27,565), which is remarkable given the RTX 3090’s desktop flagship status. These percentile comparisons show that while the RX 6700 XT wins head-to-head, its overall average is dragged down by its lower OpenCL and Passmark scores, whereas the RX 470’s average is buoyed by relatively consistent results across its fewer benchmark entries.

FAQ

Q: Which card has a higher average benchmark score?

A: The AMD Radeon RX 470 has a higher average benchmark score of 28,996 compared to the RX 6700 XT’s 27,425, despite losing every head-to-head test. This is because the RX 470’s benchmark set (4 tests) and the RX 6700 XT’s set (11 tests) differ, and the RX 6700 XT’s Passmark scores are disproportionately low relative to its Geekbench and 3DMark results.

Q: How much faster is the RX 6700 XT in 3DMark Steel Nomad DX12?

A: The RX 6700 XT scores 2,435 points versus 842 for the RX 470, a -65.4% delta, meaning the RX 6700 XT is roughly 2.9 times faster in this DirectX 12 benchmark.

Q: Is the RX 6700 XT’s advantage consistent across all APIs?

A: No. The RX 6700 XT’s lead varies by API: largest in Metal (-68.2%), then 3DMark DX12 (-65.4%), Vulkan (-57.3%), and smallest in OpenCL (-24%). The RX 470’s GCN architecture appears relatively stronger in OpenCL compute tasks.

Q: How do these cards compare to their nearest rivals?

A: The RX 470 is within 0.7% of three rivals (RX 6800M, Arc A370M, RX Vega M GH) and the FirePro W8000, showing tight mid-range competition. The RX 6700 XT is within 1.5% of the RTX 4070 Mobile, RTX 3090, RTX PRO 4000 Blackwell, and Radeon Pro Vega 20, with the RTX 3090 comparison being particularly notable.

Q: What is the percentile rank for each card among all GPUs?

A: The RX 470 ranks at the 74th percentile, while the RX 6700 XT ranks at the 72nd percentile. Despite the RX 6700 XT’s head-to-head dominance, its lower average score places it slightly below the RX 470 in overall percentile ranking.

Q: Which card has more shading units and texture mapping units?

A: The RX 6700 XT has 2,560 shading units and 160 TMUs, versus 2,048 shading units and 128 TMUs for the RX 470. The RX 6700 XT also doubles the ROP count from 32 to 64.

Architecture Differences

The architectural gap between these two GPUs is generational and profound. The RX 470 uses the Ellesmere chip built on GCN 4.0 architecture, fabricated on a 14 nm process at GlobalFoundries, with 5,700 million transistors on a 232 mm² die. The RX 6700 XT uses the Navi 22 chip built on RDNA 2.0 architecture, fabricated on a 7 nm process at TSMC, with 17,200 million transistors on a 335 mm² die. The transistor density tells the story of process advancement: 24.6 million transistors per mm² for the RX 470 versus 51.3 million per mm² for the RX 6700 XT, more than double the density.

Clock speeds represent another massive divergence. The RX 470’s base clock is 926 MHz with a boost of 1206 MHz, while the RX 6700 XT’s base clock is 2321 MHz with a boost of 2581 MHz and a game clock of 2424 MHz. This over 2x clock advantage, combined with the newer architecture, drives the performance gap. The RX 6700 XT also introduces 40 ray tracing cores, a feature entirely absent from the RX 470, reflecting the RDNA 2.0 focus on real-time ray tracing. The FP16 compute ratio differs as well: the RX 470 offers 4.940 TFLOPS FP16 (1:1 ratio with FP32), while the RX 6700 XT offers 26.43 TFLOPS FP16 (2:1 ratio), giving it a significant advantage in workloads that leverage half-precision math.

Memory architecture also diverges significantly. The RX 470 uses 4 GB of GDDR5 on a 256-bit bus with 211.2 GB/s bandwidth, while the RX 6700 XT uses 12 GB of GDDR6 on a 192-bit bus with 384.0 GB/s bandwidth. Despite the narrower bus, the RX 6700 XT achieves 82% more bandwidth thanks to faster memory clocks (2000 MHz versus 1650 MHz, or 16 Gbps effective versus 6.6 Gbps effective) and newer memory technology. The RX 6700 XT also supports PCIe 4.0 x16, double the bandwidth of the RX 470’s PCIe 3.0 x16 interface, which matters for data transfer in memory-constrained scenarios. API support reflects the newer card’s modernity: DirectX 12 Ultimate (12_2) versus DirectX 12 (12_0), and Vulkan 1.4 versus 1.3, while both support OpenGL 4.6.

The Verdict

The data points to a clear, unambiguous winner for anyone choosing between these two cards today: the AMD Radeon RX 6700 XT. It wins every shared benchmark, often by margins exceeding 2x, and offers modern features like ray tracing cores, 12 GB of VRAM, and PCIe 4.0 support that the RX 470 simply cannot match. The RX 6700 XT’s 13.21 TFLOPS FP32 performance versus the RX 470’s 4.940 TFLOPS represents a 2.7x raw compute advantage, which manifests in the benchmark deltas. Its 384.0 GB/s memory bandwidth and 12 GB capacity make it far more suitable for modern game textures and resolutions beyond 1080p.

For the RX 470, the verdict is more nuanced. Its higher average benchmark score (28,996) and 74th percentile ranking versus the RX 6700 XT’s 72nd percentile suggest that in a broader database context, the RX 470’s results are more consistent relative to its peers. However, this is a statistical artifact of different test sets rather than a performance endorsement. The RX 470’s nearest rivals—the RX 6800M, Arc A370M, and RX Vega M GH—are all mobile or older parts, indicating it sits in a legacy laptop or low-end desktop segment. Its 4 GB of VRAM and 6.6 Gbps effective memory speed are severe limitations for current software. The RX 470 does hold one tangible advantage: a lower TDP of 120 W versus 230 W for the RX 6700 XT, and a lower suggested PSU requirement of 300 W versus 550 W. For users with older or smaller power supplies, this could be a practical consideration, but the performance trade-off is extreme.

The launch MSRP figures frame the generational gap: the RX 470 launched at 179 USD, while the RX 6700 XT launched at 479 USD. That price difference is justified by the performance data, as the RX 6700 XT delivers roughly 2.9x the 3DMark score and 3.1x the Metal score for 2.7x the launch price. Users seeking maximum performance per dollar at the time of each card’s launch would likely find the RX 6700 XT the better value, though the RX 470’s lower absolute price made it accessible to a wider audience. In the current used market, the RX 470 might serve as a low-cost entry point for legacy 1080p gaming, but the RX 6700 XT is the only sensible choice for anyone wanting to run modern titles at high settings with ray tracing enabled.

Specification Differences

| Specification | AMD Radeon RX 470 | AMD Radeon RX 6700 XT |

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

| Chip | Ellesmere | Navi 22 |

| Architecture | GCN 4.0 | RDNA 2.0 |

| Process Node | 14 nm | 7 nm |

| Foundry | GlobalFoundries | TSMC |

| Transistors | 5,700 million | 17,200 million |

| Die Size | 232 mm² | 335 mm² |

| Base Clock | 926 MHz | 2321 MHz |

| Boost Clock | 1206 MHz | 2581 MHz |

| Game Clock | N/A | 2424 MHz |

| Memory Size | 4 GB | 12 GB |

| Memory Type | GDDR5 | GDDR6 |

| Memory Bus | 256 bit | 192 bit |

| Memory Clock | 1650 MHz / 6.6 Gbps effective | 2000 MHz / 16 Gbps effective |

| Bandwidth | 211.2 GB/s | 384.0 GB/s |

| Shading Units | 2048 | 2560 |

| TMUs | 128 | 160 |

| ROPs | 32 | 64 |

| Ray Tracing Cores | N/A | 40 |

| Pixel Rate | 38.59 GPixel/s | 165.2 GPixel/s |

| Texture Rate | 154.4 GTexel/s | 413.0 GTexel/s |

| FP32 | 4.940 TFLOPS | 13.21 TFLOPS |

| FP16 | 4.940 TFLOPS (1:1) | 26.43 TFLOPS (2:1) |

| TDP | 120 W | 230 W |

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

| Suggested PSU | 300 W | 550 W |

| Bus Interface | PCIe 3.0 x16 | PCIe 4.0 x16 |

| Display Outputs | 1x HDMI 2.0b, 3x DisplayPort 1.4a | 1x HDMI 2.1, 3x DisplayPort 1.4a |

| DirectX | 12 (12_0) | 12 Ultimate (12_2) |

| Vulkan | 1.3 | 1.4 |

| Dimensions | 240 mm x 95 mm x 35 mm | 267 mm x 110 mm x 40 mm |

| Release Date | 2016-08-03 | 2021-03-02 |

| Launch MSRP | 179 USD | 479 USD |

| Predecessor | Pirate Islands | Navi |

| Successor | Polaris | Navi III |