AMD Radeon 660M vs AMD Radeon Pro 555X Comparison
AMD Radeon 660M
Radeon Pro 555X
PERFORMANCE BENCHMARKS
Analysis: AMD Radeon 660M vs AMD Radeon Pro 555X
The AMD Radeon 660M and AMD Radeon Pro 555X represent two very different approaches to integrated graphics, separated by four years of architectural evolution. The data shows the 660M, built on RDNA 2.0, edges out the older GCN 4.0-based Pro 555X in average benchmark score, 13,812 to 13,321, a margin of roughly 3.7%. While the Pro 555X retains a lead in one API-specific test, the 660M demonstrates superior efficiency and modern feature support, making it the stronger all-around choice for most workloads.
The Verdict
From the benchmark data alone, the AMD Radeon 660M is the recommended pick for users prioritizing raw compute performance and modern API compatibility. It wins both head-to-head benchmark comparisons where both GPUs were tested: Geekbench OpenCL (12,876 vs 12,628, a 2% advantage) and Geekbench Vulkan (14,748 vs 12,855, a substantial 14.7% lead). The 660M also holds a higher percentile ranking among all GPUs (55th vs 54th) and a higher average benchmark score (13,812 vs 13,321). For users running Vulkan-based applications or games, the 660M’s advantage is decisive. The Radeon Pro 555X, however, is not without merit. It delivers a higher Geekbench Metal score (14,479 vs no Metal result for the 660M), making it the relevant option for macOS environments where Metal is the primary graphics API. The Pro 555X also offers dedicated 4 GB GDDR5 memory, which can be advantageous in scenarios where system memory bandwidth is a bottleneck, despite the 660M’s overall compute superiority. In short: choose the 660M for general compute and Vulkan workloads; choose the Pro 555X only if Metal API support or dedicated VRAM is a strict requirement.
Architecture Differences
The architectural gap between these two GPUs is significant. The AMD Radeon 660M is built on the RDNA 2.0 architecture, manufactured on a 6 nm process at TSMC, while the Radeon Pro 555X uses the older GCN 4.0 architecture on a 14 nm process from GlobalFoundries. This process advantage translates directly into transistor density: the 660M packs 13,100 million transistors into a 208 mm² die (63.0 million transistors per mm²), whereas the Pro 555X contains only 3,000 million transistors on a 123 mm² die (24.4 million per mm²). The 660M’s chip is codenamed Rembrandt, while the Pro 555X uses Polaris 21.
The 660M features 384 shading units, 24 texture mapping units (TMUs), and 16 raster operation pipelines (ROPs), along with 6 dedicated ray tracing cores. The Pro 555X, by contrast, has double the shading units (768) and TMUs (48), but also 16 ROPs and no ray tracing support. The 660M compensates for its lower shader count with a much higher clock speed: a base of 1500 MHz and boost of 1900 MHz, versus unspecified clock speeds for the Pro 555X (the data only lists its memory clock at 1470 MHz, or 5.9 Gbps effective). Memory configurations differ fundamentally: the 660M uses system-shared memory with system-dependent bandwidth, while the Pro 555X has 4 GB of dedicated GDDR5 on a 128-bit bus with 94.08 GB/s bandwidth.
API support is another key differentiator. The 660M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The Pro 555X is limited to DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.3. The 660M also uses a PCIe 4.0 x8 interface, while the Pro 555X relies on the older PCIe 3.0 x8. Power consumption favors the 660M dramatically: its TDP is 40 W versus 75 W for the Pro 555X, representing a 46.7% reduction in power draw for a higher average score.
FAQ
Q: Which GPU has the higher average benchmark score?
A: The AMD Radeon 660M, with an average benchmark score of 13,812 compared to the Radeon Pro 555X’s 13,321.
Q: What is the most significant performance gap between the two in head-to-head tests?
A: In Geekbench Vulkan, the 660M scores 14,748 versus 12,855 for the Pro 555X, a 14.7% difference in favor of the 660M.
Q: Does the Radeon Pro 555X have any benchmark where it outperforms the 660M?
A: The Pro 555X has a Geekbench Metal score of 14,479, but the 660M has no Metal benchmark result recorded, so no direct comparison exists. In tests where both were measured (OpenCL and Vulkan), the 660M wins both.
Q: What is the memory configuration difference?
A: The Pro 555X has 4 GB of dedicated GDDR5 memory on a 128-bit bus with 94.08 GB/s bandwidth. The 660M uses system-shared memory with bandwidth described as "System Dependent."
Q: How do their power requirements compare?
A: The 660M has a TDP of 40 W, while the Pro 555X has a TDP of 75 W. The 660M is significantly more power-efficient.
Q: Which GPU supports newer graphics APIs?
A: The 660M supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, whereas the Pro 555X is limited to DirectX 12 (12_0) and Vulkan 1.3.
Specification Differences
The two GPUs differ across nearly every core specification. The 660M uses a 6 nm process at TSMC, while the Pro 555X uses 14 nm at GlobalFoundries. Transistor counts are 13,100 million versus 3,000 million, on die sizes of 208 mm² versus 123 mm². The 660M has a base clock of 1500 MHz and a boost clock of 1900 MHz; the Pro 555X has no base or boost clock listed. Memory clocks are listed as "System Shared" for the 660M and 1470 MHz (5.9 Gbps effective) for the Pro 555X.
Shading units: 384 (660M) vs 768 (Pro 555X). TMUs: 24 vs 48. ROPs are equal at 16. The 660M has 6 ray tracing cores; the Pro 555X has none. Pixel rate is 30.40 GPixel/s for the 660M versus 14.51 GPixel/s for the Pro 555X. Texture rate is 45.60 GTexel/s versus 43.54 GTexel/s. FP32 performance is 1,459.2 GFLOPS versus 1,393.2 GFLOPS. FP16 performance is 2.918 TFLOPS (2:1) versus 1,393.2 GFLOPS (1:1). TDP is 40 W versus 75 W. Bus interface is PCIe 4.0 x8 versus PCIe 3.0 x8. DirectX support is 12 Ultimate (12_2) versus 12 (12_0); Vulkan support is 1.4 versus 1.3. Memory size/type/bus width are "System Shared" for the 660M versus 4 GB GDDR5 on a 128-bit bus for the Pro 555X. Bandwidth is "System Dependent" versus 94.08 GB/s. Release dates differ: 2022-01-03 for the 660M versus 2018-07-15 for the Pro 555X.
Head-to-Head Benchmarks
The two GPUs were directly compared in two benchmark tests, both won by the Radeon 660M. In Geekbench OpenCL, the 660M scored 12,876 against the Pro 555X’s 12,628, a 2% difference. This is a narrow margin, indicating that in OpenCL compute workloads, the two GPUs are nearly equivalent despite their architectural differences. The 660M’s higher clocks and newer architecture offset the Pro 555X’s double shader count (384 vs 768) and higher TMU count (24 vs 48).
The second test, Geekbench Vulkan, shows a much larger gap. The 660M scored 14,748, while the Pro 555X managed only 12,855—a 14.7% advantage for the 660M. This substantial delta likely stems from the 660M’s RDNA 2.0 architecture being optimized for modern asynchronous compute and Vulkan’s low-level access, combined with its much higher pixel rate (30.40 GPixel/s vs 14.51 GPixel/s) and texture rate (45.60 GTexel/s vs 43.54 GTexel/s). The 660M’s support for Vulkan 1.4 versus 1.3 on the Pro 555X also contributes to its stronger performance in this API.
Beyond the head-to-head tests, the 660M’s average benchmark score of 13,812 places it within 0.1% of the NVIDIA RTX A2000 Mobile (13,821) and 0.5% ahead of the AMD Radeon RX 7900 XT (13,745), according to nearest rival data. The Pro 555X’s average of 13,321 sits 0.2% above the NVIDIA GeForce GTX 480 (13,300) and 0.3% behind the AMD Radeon RX 5500M (13,356). These rival comparisons contextualize the 660M’s lead: it performs at a level comparable to dedicated mobile workstation GPUs, while the Pro 555X aligns more closely with older desktop cards.
Where Each One Wins
The AMD Radeon 660M wins in every workload where both GPUs were benchmarked. In OpenCL compute, its 2% edge is modest but consistent, making it the safer choice for general-purpose GPU computing tasks like rendering, video encoding, or scientific simulations that leverage OpenCL. In Vulkan-based gaming or compute, the 660M’s 14.7% lead is decisive, and its support for DirectX 12 Ultimate (12_2) and Vulkan 1.4 ensures compatibility with the latest graphics features, including ray tracing (via its 6 RT cores). The 660M also wins on efficiency: at 40 W TDP versus 75 W, it delivers more performance per watt, which is critical for thin-and-light laptops where thermal headroom is limited. Its 6 nm process and higher transistor density (63.0M/mm² vs 24.4M/mm²) indicate a more modern, future-proof design.
The Radeon Pro 555X’s wins are more situational. Its only recorded benchmark advantage is in Geekbench Metal (14,479), but since the 660M has no Metal result, it cannot be directly compared. However, this Metal support makes the Pro 555X the only viable option in this pairing for macOS-based systems, as Metal is Apple’s proprietary graphics API. The Pro 555X also offers dedicated 4 GB GDDR5 memory with 94.08 GB/s bandwidth, which is a fixed resource not shared with the system. In scenarios where system RAM is limited or slow, the dedicated VRAM could provide more predictable performance, particularly at higher resolutions with large textures. Finally, the Pro 555X’s double shading unit and TMU counts (768 and 48) suggest potential strengths in workloads that are highly parallel but not clock-sensitive—though the benchmark data does not substantiate this, as it loses both recorded tests. For users locked into the Mac ecosystem, the Pro 555X is the clear choice; for everyone else, the 660M’s benchmark wins, API support, and power efficiency make it the superior GPU.