AMD Radeon RX 5700 vs AMD Radeon RX 6700M Comparison

Head-to-Head Benchmarks

The recorded data shows a close contest between the AMD Radeon RX 6700M and the AMD Radeon RX 5700, with the mobile part winning 6 of the 11 head-to-head tests and the desktop part winning 5. The average benchmark score tells the broader story: the RX 6700M averages 25,633 points against the RX 5700's 22,170, a gap of roughly 15.6%. Yet the individual tests reveal a more nuanced picture, with each card dominating in specific API environments.

The largest single win belongs to the RX 6700M in Geekbench Vulkan, where it scores 90,816 against the RX 5700's 64,166, a commanding 41.5% advantage. This is a decisive result, suggesting that the newer RDNA 2.0 architecture extracts far more performance from the Vulkan API than the older RDNA 1.0 design. The RX 6700M also wins Geekbench OpenCL by 11.6% (77,666 vs. 69,603) and Geekbench Metal by 6.8% (91,911 vs. 86,079), showing a consistent edge in compute-oriented workloads.

In DirectX 11, the RX 6700M leads by 27.7% (129 vs. 101), and in DirectX 12 by 20.4% (65 vs. 54). These are substantial margins in legacy and current gaming APIs alike. The RX 6700M also edges out the RX 5700 in Passmark DirectX 10, scoring 92 against 87, a 5.7% gain.

The RX 5700, however, posts decisive wins in several other tests. The most striking is Passmark DirectX 9, where it scores 209 versus the RX 6700M's 155, a 25.8% margin. In Passmark G2D, the RX 5700 leads by 36.7% (877 vs. 555), a massive gap in 2D graphics throughput. The RX 5700 also wins Passmark GPU Compute by 20.3% (6,517 vs. 5,191) and Passmark G3D by 5.4% (14,314 vs. 13,536). In 3DMark Steel Nomad DX12, the RX 5700 takes a narrow 0.9% victory (1,862 vs. 1,845), a nearly identical result.

The pattern is clear: the RX 6700M dominates in modern, cross-platform APIs (Vulkan, OpenCL, Metal) and newer DirectX versions, while the RX 5700 excels in older DirectX 9, 2D rasterization, and raw compute through Passmark's suite. The 3DMark result being essentially tied suggests that in a pure, modern DirectX 12 gaming scenario, neither card holds a significant edge.

Where Each One Wins

The RX 6700M is the choice for workloads that leverage Vulkan or OpenCL. Its 41.5% Vulkan advantage is the single largest delta in the entire comparison, making it the clear winner for Linux gaming, emulation, and any application built on Vulkan. The 11.6% OpenCL lead and 6.8% Metal lead further cement its position in compute and Apple ecosystem workflows (via Metal on macOS). For DirectX 11 and DirectX 12 gaming, the 27.7% and 20.4% wins respectively mean the RX 6700M is the stronger pick for most modern Windows titles, even if the 3DMark Steel Nomad result shows near-parity in one specific DX12 benchmark.

The RX 5700 wins where legacy compatibility and 2D performance matter. Its 25.8% DirectX 9 advantage makes it the better option for older games running on that API. The 36.7% G2D lead is enormous, indicating superior 2D acceleration for desktop compositing, image processing, or any task relying on 2D rendering. The 20.3% GPU Compute win through Passmark suggests that in certain compute workloads, the RX 5700's higher memory bandwidth (448.0 GB/s) and wider 256-bit bus provide an edge over the RX 6700M's 320.0 GB/s on a 160-bit bus. The 5.4% G3D lead in Passmark's general 3D test also points to the RX 5700 being slightly ahead in that specific rasterization benchmark.

For a user prioritizing modern game compatibility and cross-platform performance, the RX 6700M is the data-backed choice. For someone running older DirectX 9 titles, heavy 2D applications, or specific compute tasks that favor Passmark's implementation, the RX 5700 holds the advantage.

Architecture Differences

The two GPUs come from different architecture generations. The RX 6700M uses RDNA 2.0, built on the Navi 22 chip, while the RX 5700 uses RDNA 1.0, built on the Navi 10 chip. Both are fabricated on TSMC's 7 nm process, but the transistor counts differ substantially: the RX 6700M packs 17,200 million transistors on a 335 mm² die, while the RX 5700 has 10,300 million transistors on a 251 mm² die. This translates to a higher transistor density for the RX 6700M at 51.3M per mm² versus 41.0M per mm² for the RX 5700.

Clock speeds favor the RX 6700M significantly. Its base clock is 1489 MHz, close to the RX 5700's 1465 MHz, but the boost clock tells a different story: 2400 MHz versus 1725 MHz. The RX 6700M also has a game clock of 2300 MHz, which is not listed for the RX 5700 (its game clock is 1625 MHz). Memory clocks also differ, with the RX 6700M running at 2000 MHz (16 Gbps effective) versus the RX 5700's 1750 MHz (14 Gbps effective).

The most important architectural addition is ray tracing. The RX 6700M includes 36 ray tracing cores, while the RX 5700 has none. This is a direct consequence of the RDNA 2.0 upgrade, which added dedicated RT hardware. The RX 6700M also supports DirectX 12 Ultimate (12_2), while the RX 5700 is limited to DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4.

The compute resources are identical in count: both have 2304 shading units, 144 texture mapping units, and 64 raster output units. However, the clock speed difference drives a large performance gap in raw throughput. The RX 6700M delivers 11.06 TFLOPS FP32 versus 7.949 TFLOPS for the RX 5700, a 39% advantage. Pixel rate is 153.6 GPixel/s versus 110.4 GPixel/s, and texture rate is 345.6 GTexel/s versus 248.4 GTexel/s.

Memory configurations diverge sharply. The RX 6700M has 10 GB of GDDR6 on a 160-bit bus, yielding 320.0 GB/s bandwidth. The RX 5700 has 8 GB on a 256-bit bus, yielding 448.0 GB/s. Despite the RX 6700M's higher memory clock, the narrower bus limits its bandwidth, which likely explains the RX 5700's wins in bandwidth-sensitive tests like Passmark GPU Compute and G2D.

FAQ

Q: Which GPU is faster in Vulkan performance?

A: The AMD Radeon RX 6700M wins decisively, scoring 90,816 in Geekbench Vulkan versus 64,166 for the RX 5700, a 41.5% advantage.

Q: Does the RX 5700 beat the RX 6700M in any modern benchmark?

A: Yes, in 3DMark Steel Nomad DX12 the RX 5700 wins narrowly (1,862 vs. 1,845, a 0.9% margin), and it leads in Passmark G3D by 5.4% (14,314 vs. 13,536).

Q: What is the memory bandwidth difference between the two?

A: The RX 5700 has 448.0 GB/s bandwidth from its 256-bit bus, while the RX 6700M has 320.0 GB/s from a 160-bit bus. The RX 5700's wider bus provides 40% more bandwidth.

Q: Which card supports ray tracing?

A: Only the AMD Radeon RX 6700M includes 36 ray tracing cores. The RX 5700 has no ray tracing cores and is limited to DirectX 12 (12_1), while the RX 6700M supports DirectX 12 Ultimate (12_2).

Q: How do the average benchmark scores compare?

A: The RX 6700M averages 25,633 points, placing it in the 71st percentile of all GPUs. The RX 5700 averages 22,170 points, in the 67th percentile. The RX 6700M is about 15.6% higher on average.

Q: Which card has higher clock speeds?

A: The RX 6700M has a boost clock of 2400 MHz and a game clock of 2300 MHz. The RX 5700 has a boost clock of 1725 MHz and a game clock of 1625 MHz. The RX 6700M also has a higher memory clock at 2000 MHz versus 1750 MHz.

Specification Differences

| Specification | AMD Radeon RX 6700M | AMD Radeon RX 5700 |

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

| Architecture | RDNA 2.0 | RDNA 1.0 |

| Chip | Navi 22 | Navi 10 |

| Transistors | 17,200 million | 10,300 million |

| Die Size | 335 mm² | 251 mm² |

| Transistor Density | 51.3M / mm² | 41.0M / mm² |

| Base Clock | 1489 MHz | 1465 MHz |

| Boost Clock | 2400 MHz | 1725 MHz |

| Game Clock | 2300 MHz | 1625 MHz |

| Memory Clock | 2000 MHz (16 Gbps effective) | 1750 MHz (14 Gbps effective) |

| Memory Size | 10 GB | 8 GB |

| Memory Bus Width | 160 bit | 256 bit |

| Memory Bandwidth | 320.0 GB/s | 448.0 GB/s |

| Ray Tracing Cores | 36 | None |

| Pixel Rate | 153.6 GPixel/s | 110.4 GPixel/s |

| Texture Rate | 345.6 GTexel/s | 248.4 GTexel/s |

| FP32 Performance | 11.06 TFLOPS | 7.949 TFLOPS |

| FP16 Performance | 22.12 TFLOPS (2:1) | 15.90 TFLOPS (2:1) |

| TDP | 135 W | 180 W |

| Slot Width | IGP | Dual-slot |

| Power Connectors | None | 1x 6-pin + 1x 8-pin |

| Suggested PSU | Not specified | 450 W |

| Display Outputs | Portable Device Dependent | 1x HDMI 2.0b, 3x DisplayPort 1.4a |

| DirectX Support | 12 Ultimate (12_2) | 12 (12_1) |

| Release Date | 2021-05-30 | 2019-07-06 |

| Dimensions | Not specified | 268 mm length, 111 mm height, 36 mm width |

The Verdict

The data supports a clear split based on use case. The AMD Radeon RX 6700M is the superior choice for modern gaming and cross-platform compute. Its 41.5% Vulkan lead, 27.7% DirectX 11 lead, and 20.4% DirectX 12 lead make it the stronger performer in current and near-future titles. The inclusion of 36 ray tracing cores and DirectX 12 Ultimate support means it can handle ray-traced games and next-generation features that the RX 5700 cannot process at all. Its lower TDP of 135 W versus 180 W also makes it more efficient, though it is an IGP (integrated graphics package) designed for mobile systems, while the RX 5700 is a dual-slot desktop card.

The AMD Radeon RX 5700 remains relevant for specific niches. Its 25.8% DirectX 9 advantage and 36.7% G2D lead make it the better pick for legacy software and 2D-heavy workloads. The 20.3% Passmark GPU Compute win, driven by its 448.0 GB/s memory bandwidth, suggests it excels in bandwidth-bound compute tasks. For users running older games or applications that do not leverage Vulkan or modern DirectX, the RX 5700 is the data-backed choice.

The average benchmark scores and percentiles reflect the overall hierarchy: the RX 6700M sits at the 71st percentile with a 25,633 average, while the RX 5700 sits at the 67th percentile with 22,170. The RX 6700M's nearest rivals include the AMD Radeon Pro W5700 (0.4% behind) and the NVIDIA GeForce RTX 3080 Ti Mobile (0.4% behind), while the RX 5700's nearest rival is the AMD Radeon RX 6700S (0.1% ahead). This places the RX 6700M in a higher performance tier overall.

In summary, choose the RX 6700M for its modern API performance, ray tracing capability, and higher raw throughput. Choose the RX 5700 for legacy DirectX 9 games, 2D performance, and bandwidth-sensitive compute workloads. The 3DMark Steel Nomad result being nearly tied (0.9% apart) suggests that in pure rasterized DirectX 12 gaming, the two are effectively interchangeable, but everywhere else the architecture generation gap is decisive.