AMD Radeon R9 M390X vs AMD Radeon RX 6600 Comparison

The AMD Radeon R9 M390X and the AMD Radeon RX 6600 represent two distinct eras of AMD mobile and desktop graphics. The data shows a substantial generational leap, with the RX 6600 dominating the shared benchmark suite, yet the comparison reveals more than just a raw performance gap. The following analysis breaks down the head-to-head results, architectural shifts, and the specific strengths of each card based solely on the provided benchmark and specification data.

Head-to-Head Benchmarks

The head-to-head comparison is brief but decisive, with the AMD Radeon RX 6600 winning both of the two shared tests. In the Geekbench Metal benchmark, the RX 6600 scores 88,398 points, which is 72.4% higher than the R9 M390X’s score of 24,373. This is a massive margin, indicating that the newer architecture delivers a fundamentally different level of compute throughput in this API. The Geekbench Vulkan result tells a similar story: the RX 6600’s 67,623 points dwarf the R9 M390X’s 16,951, a 74.9% difference in favor of the newer card.

The deltaPct values in the head-to-head table are expressed from the perspective of the loser, meaning the RX 6600 is not just slightly ahead but is roughly three to four times faster in these specific workloads. For context, the R9 M390X’s average benchmark score of 20,662 places it just 0.2% behind the AMD Radeon RX 5600 XT, while the RX 6600’s average of 19,036 is essentially tied with the NVIDIA Quadro K6000. This shows that while the R9 M390X competes with older desktop cards, the RX 6600 sits in a different performance class altogether, despite both cards having similar overall percentile rankings (66th vs 63rd percentile).

It is also worth noting the absence of overlap in their wider benchmark portfolios. The R9 M390X only has Geekbench Metal and Vulkan results, while the RX 6600 has a comprehensive set including 3DMark Steel Nomad, Passmark tests across DirectX 9 through 12, and GPU compute. This makes the direct comparison limited, but the two shared tests are sufficient to establish a clear hierarchy: the RX 6600 is in a different league for both graphics and general compute tasks.

Architecture Differences

The architectural gap between these two GPUs is the primary driver of their performance difference. The R9 M390X is built on the GCN 3.0 architecture, using the Amethyst chip, and is fabricated on a 28 nm process at TSMC. In contrast, the RX 6600 uses the RDNA 2.0 architecture, built on the Navi 23 chip, and is manufactured on a 7 nm process, also at TSMC. This process shrink from 28 nm to 7 nm is a fundamental enabler of the latter’s efficiency and clock speed advantages.

The transistor counts and die sizes reflect this shift. The R9 M390X packs 5,000 million transistors on a 366 mm² die, yielding a transistor density of 13.7M / mm². The RX 6600 contains 11,060 million transistors on a much smaller 237 mm² die, achieving a density of 46.7M / mm². This is over three times the density, allowing for significantly more complex logic in a smaller physical footprint. The R9 M390X’s large die and older process contribute to its 75 W TDP, while the RX 6600, despite having more than double the transistors, has a higher TDP of 132 W but delivers far more performance per watt.

Core configuration also diverges significantly. The R9 M390X has 2048 shading units, 128 TMUs, and 32 ROPs, with no dedicated ray tracing or tensor cores. The RX 6600 has fewer shading units at 1792 and 112 TMUs, but doubles the ROP count to 64 and introduces 28 dedicated ray tracing cores. This is a key architectural change: RDNA 2.0 is designed to be more efficient per shader than GCN, explaining how a lower shader count can deliver dramatically higher performance. The R9 M390X’s FP32 throughput is 2.961 TFLOPS, while the RX 6600 achieves 8.928 TFLOPS, a threefold increase. Furthermore, the RX 6600’s FP16 performance is 17.86 TFLOPS (a 2:1 ratio), whereas the R9 M390X’s FP16 is identical to its FP32 at 2.961 TFLOPS (1:1).

Where Each One Wins

Based on the benchmark data, the AMD Radeon RX 6600 is the clear winner in every measurable category. It wins both shared benchmarks (Geekbench Metal and Vulkan) by margins of over 70%. Its architecture also provides features that the R9 M390X lacks, such as ray tracing cores, which are absent from the older card. The RX 6600’s higher memory bandwidth of 224.0 GB/s, despite a narrower 128-bit bus, outperforms the R9 M390X’s 160.0 GB/s from a 256-bit bus, thanks to faster GDDR6 memory. This makes the RX 6600 the superior choice for any modern workload, including those that leverage DirectX 12 Ultimate features, which the R9 M390X only partially supports with DirectX 12 (12_0).

The AMD Radeon R9 M390X, however, has no benchmark wins to claim. Its strengths lie elsewhere: it is a mobile-specific MXM module with a lower 75 W TDP, making it suitable for laptops where power and space are constrained. Its 4 GB of GDDR5 memory and 256-bit bus were appropriate for its 2015 era, but the data shows it is outclassed in raw compute. The only advantage it holds is a lower power draw, which is a physical property, not a performance metric. For users confined to an older laptop with an MXM slot, it remains a functional option, but it cannot compete on performance.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon R9 M390X has a higher average benchmark score of 20,662 compared to the RX 6600’s 19,036. However, this is misleading, as the R9 M390X only has two benchmark results, while the RX 6600 has eleven, including different workload types.

Q: How much faster is the RX 6600 in the Geekbench Vulkan test?

A: The RX 6600 scores 67,623 in Geekbench Vulkan, which is 74.9% higher than the R9 M390X’s score of 16,951.

Q: Does the R9 M390X support ray tracing?

A: No, the R9 M390X has no ray tracing cores (rtCores: null), while the RX 6600 has 28 dedicated ray tracing cores.

Q: What is the difference in process node between the two cards?

A: The R9 M390X is built on a 28 nm process, while the RX 6600 is built on a 7 nm process, both from TSMC.

Q: Which card has a higher memory bandwidth?

A: The RX 6600 has a higher memory bandwidth of 224.0 GB/s, compared to the R9 M390X’s 160.0 GB/s, despite the RX 6600 having a narrower 128-bit memory bus.

Q: What is the DirectX support level for each GPU?

A: The R9 M390X supports DirectX 12 (12_0), while the RX 6600 supports DirectX 12 Ultimate (12_2), which includes additional features like hardware ray tracing.

The Verdict

The data is unambiguous: the AMD Radeon RX 6600 is the superior GPU for performance. It wins all head-to-head benchmarks by over 70%, offers more than triple the FP32 compute throughput, and includes modern features like ray tracing cores and DirectX 12 Ultimate support. Its 8 GB of GDDR6 memory and higher bandwidth also make it more future-proof for texture-heavy workloads. Any user prioritizing gaming or compute performance should choose the RX 6600.

The R9 M390X is an older, mobile-centric part with a lower TDP of 75 W. It is suitable for a specific niche: upgrading a legacy laptop that uses an MXM module. Its performance, however, is firmly in the past, and it lacks the architectural features needed for modern software. The RX 6600, with a TDP of 132 W, requires a desktop PSU of 300 W and a PCIe 4.0 x8 slot, making it a desktop-only solution. In short, if you can use the RX 6600, the data says it is the only sensible choice.

Specification Differences

| Specification | AMD Radeon R9 M390X | AMD Radeon RX 6600 |

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

| Architecture | GCN 3.0 | RDNA 2.0 |

| Chip | Amethyst | Navi 23 |

| Process Node | 28 nm | 7 nm |

| Transistors | 5,000 million | 11,060 million |

| Die Size | 366 mm² | 237 mm² |

| Transistor Density | 13.7M / mm² | 46.7M / mm² |

| Base Clock | None | 1626 MHz |

| Boost Clock | None | 2491 MHz |

| Game Clock | None | 2044 MHz |

| Memory Clock | 1250 MHz (5 Gbps effective) | 1750 MHz (14 Gbps effective) |

| Memory Size | 4 GB | 8 GB |

| Memory Type | GDDR5 | GDDR6 |

| Memory Bus Width | 256 bit | 128 bit |

| Memory Bandwidth | 160.0 GB/s | 224.0 GB/s |

| Shading Units | 2048 | 1792 |

| TMUs | 128 | 112 |

| ROPs | 32 | 64 |

| Ray Tracing Cores | None | 28 |

| Pixel Rate | 23.14 GPixel/s | 159.4 GPixel/s |

| Texture Rate | 92.54 GTexel/s | 279.0 GTexel/s |

| FP32 Performance | 2.961 TFLOPS | 8.928 TFLOPS |

| FP16 Performance | 2.961 TFLOPS (1:1) | 17.86 TFLOPS (2:1) |

| TDP | 75 W | 132 W |

| Slot Width | MXM Module | Dual-slot |

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

| Suggested PSU | None | 300 W |

| Bus Interface | PCIe 3.0 x16 | PCIe 4.0 x8 |

| Display Outputs | Portable Device Dependent | 1x HDMI 2.13x DisplayPort 1.4a |

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

| Vulkan Support | 1.2.170 | 1.4 |

| Release Date | 2015-05-04 | 2021-10-12 |

| Launch MSRP | None | 329 USD |