AMD Radeon R9 M390X vs Intel Arc A770M Comparison

FAQ

Q: How does the AMD Radeon R9 M390X compare to the Intel Arc A770M in average benchmark score?

A: The AMD Radeon R9 M390X records an average benchmark score of 20,662 points, while the Intel Arc A770M scores 18,383 points. This places the AMD part 12.4% higher in the database's aggregated metric.

Q: What is the performance difference in the Geekbench Vulkan test?

A: The Intel Arc A770M scores 74,422 points in Geekbench Vulkan, compared to 16,951 points for the AMD Radeon R9 M390X. The data shows the Intel part is 77.2% ahead in this specific API workload.

Q: Which GPU has a higher transistor density?

A: The Intel Arc A770M, built on TSMC's 6 nm process, reaches a density of 53.4 million transistors per square millimeter. The AMD Radeon R9 M390X, on a 28 nm node, achieves only 13.7 million transistors per square millimeter.

Q: How do the memory subsystems differ between these two mobile GPUs?

A: The AMD Radeon R9 M390X has 4 GB of GDDR5 memory on a 256-bit bus with 160.0 GB/s bandwidth. The Intel Arc A770M features 16 GB of GDDR6 memory on a 256-bit bus with 512.0 GB/s bandwidth, offering four times the capacity and 3.2 times the bandwidth.

Q: What are the pixel and texture rate differences?

A: The Intel Arc A770M outputs 262.4 GPixel/s and 524.8 GTexel/s. The AMD Radeon R9 M390X achieves 23.14 GPixel/s and 92.54 GTexel/s. Intel's part is 11.3 times faster in pixel throughput and 5.7 times faster in texture fill.

Q: Which GPU supports more advanced DirectX features?

A: The Intel Arc A770M supports DirectX 12 Ultimate (12_2), including ray tracing capabilities with 32 dedicated RT cores. The AMD Radeon R9 M390X supports DirectX 12 (12_0) without hardware ray tracing cores.

Architecture Differences

The AMD Radeon R9 M390X uses the GCN 3.0 architecture, built around the Amethyst chip. This architecture dates to the Gem System generation (R9 M300 series), released in May 2015. The silicon is manufactured on a 28 nm process at TSMC, with a die size of 366 mm² and 5,000 million transistors. GCN 3.0 is a unified shader design with 2,048 shading units, 128 texture mapping units, and 32 render output units. It supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170. The design predates dedicated ray tracing hardware and tensor core acceleration, relying on traditional compute units for all workloads.

The Intel Arc A770M employs the Xe-HPG architecture, based on the DG2-512 chip. This is part of the Alchemist generation (Arc 7 Mobile), built on TSMC's 6 nm process. The die measures 406 mm² and contains 21,700 million transistors, giving it a transistor density of 53.4M per square millimeter, roughly four times higher than the AMD part. Xe-HPG is a modern architecture with 4,096 shading units, 256 TMUs, 128 ROPs, and 32 dedicated ray tracing cores. It supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, with the full feature set expected from a contemporary GPU.

The architectural divide spans more than a decade of design philosophy. GCN 3.0 is a compute-oriented design with a 1:1 FP16 to FP32 ratio, meaning it processes half-precision and single-precision at the same rate. Xe-HPG implements a 2:1 ratio, doubling FP16 throughput relative to FP32, which benefits AI inference and certain media workloads. The Intel part also includes hardware ray tracing, a feature entirely absent from the AMD silicon.

Clock behavior differs significantly. The AMD Radeon R9 M390X has no recorded base or boost clock in the database; only the memory clock is listed at 1250 MHz (5 Gbps effective). The Intel Arc A770M has a base clock of 1650 MHz and a boost clock of 2050 MHz, with memory at 2000 MHz (16 Gbps effective). This clock advantage, combined with the newer process node, drives the massive throughput gap between the two.

Head-to-Head Benchmarks

The only directly comparable benchmark in the database is Geekbench Vulkan. The Intel Arc A770M scores 74,422 points against the AMD Radeon R9 M390X's 16,951 points. This represents a 77.2% deficit for the AMD part, a decisive margin that reflects the generational gap in raw compute capability, memory bandwidth, and driver optimization for modern APIs.

In terms of FP32 compute, the Intel Arc A770M delivers 16.79 TFLOPS versus 2.961 TFLOPS for the AMD Radeon R9 M390X. The Intel part is 5.7 times faster in single-precision throughput. For FP16 workloads, the gap widens further: Intel achieves 33.59 TFLOPS (2:1 ratio) while AMD manages 2.961 TFLOPS (1:1 ratio), making Intel 11.3 times faster in half-precision tasks.

Memory bandwidth tells a similar story. The Intel Arc A770M's 512.0 GB/s is 3.2 times the AMD Radeon R9 M390X's 160.0 GB/s. This bandwidth advantage directly impacts texture-heavy scenes, frame buffer operations, and large dataset processing. Texture fill rate follows: Intel's 524.8 GTexel/s is 5.7 times AMD's 92.54 GTexel/s. Pixel fill rate shows the largest relative gap, with Intel at 262.4 GPixel/s versus AMD's 23.14 GPixel/s, an 11.3 times advantage.

Despite these overwhelming compute and memory advantages, the average benchmark scores tell a more nuanced story. The AMD Radeon R9 M390X averages 20,662 points, while the Intel Arc A770M averages 18,383 points. The AMD part's percentile ranking is 66 against all GPUs, while Intel sits at 62. This discrepancy suggests that the Intel Arc A770M's performance is not fully captured by the database's aggregate metrics, possibly due to driver maturity or workload-specific behavior in the older benchmarks that dominate the average.

Specification Differences

| Specification | AMD Radeon R9 M390X | Intel Arc A770M |

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

| Architecture | GCN 3.0 | Xe-HPG |

| Process Node | 28 nm | 6 nm |

| Transistors | 5,000 million | 21,700 million |

| Die Size | 366 mm² | 406 mm² |

| Transistor Density | 13.7M / mm² | 53.4M / mm² |

| Base Clock | Not recorded | 1650 MHz |

| Boost Clock | Not recorded | 2050 MHz |

| Memory Clock | 1250 MHz (5 Gbps effective) | 2000 MHz (16 Gbps effective) |

| Memory Size | 4 GB | 16 GB |

| Memory Type | GDDR5 | GDDR6 |

| Memory Bandwidth | 160.0 GB/s | 512.0 GB/s |

| Shading Units | 2,048 | 4,096 |

| TMUs | 128 | 256 |

| ROPs | 32 | 128 |

| Ray Tracing Cores | None | 32 |

| Pixel Rate | 23.14 GPixel/s | 262.4 GPixel/s |

| Texture Rate | 92.54 GTexel/s | 524.8 GTexel/s |

| FP32 | 2.961 TFLOPS | 16.79 TFLOPS |

| FP16 | 2.961 TFLOPS (1:1) | 33.59 TFLOPS (2:1) |

| TDP | 75 W | 120 W |

| Slot Width | MXM Module | IGP |

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

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

| Vulkan Support | 1.2.170 | 1.4 |

| Release Date | May 2015 | Not recorded |

The power envelope also differs: the AMD Radeon R9 M390X draws 75 W in a MXM module form factor with no external power connectors, while the Intel Arc A770M consumes 120 W as an integrated graphics processor (IGP). The AMD part uses PCIe 3.0 x16, while Intel supports PCIe 4.0 x16. Both use portable device dependent display outputs, meaning the actual connectors depend on the laptop design.

Where Each One Wins

The Intel Arc A770M wins decisively in raw computational throughput. Its FP32 output of 16.79 TFLOPS dwarfs the AMD Radeon R9 M390X's 2.961 TFLOPS, making it suitable for compute-heavy tasks like 3D rendering, scientific simulation, and machine learning inference. The 2:1 FP16 ratio further extends this lead in half-precision workloads, where the Intel part reaches 33.59 TFLOPS. The 32 ray tracing cores enable hardware-accelerated ray tracing in supported games and DCC applications, a capability the AMD part lacks entirely.

Memory capacity is another clear Intel advantage. The 16 GB GDDR6 frame buffer, paired with 512.0 GB/s bandwidth, allows the Arc A770M to handle large textures, high-resolution assets, and multi-tasking scenarios without hitting memory limits. The 4 GB GDDR5 on the AMD part, with 160.0 GB/s, will constrain modern games at high settings and can become a bottleneck in content creation workloads. The 3.2 times bandwidth advantage also benefits bandwidth-sensitive compute kernels and frame buffer operations.

The AMD Radeon R9 M390X wins in the average benchmark metric and percentile ranking. Its 20,662 average score against Intel's 18,383, coupled with a 66th percentile versus 62nd, indicates that in the database's broader set of legacy benchmarks, the AMD part performs competitively or better. This likely reflects its maturity: GCN 3.0 had years of driver optimization across a wide range of applications, while the Arc A770M's newer architecture may not be fully optimized in all legacy test scenarios. The AMD part also draws less power at 75 W versus 120 W, which can matter in thin-and-light laptops where thermal and battery constraints are strict.

For gaming, the picture is split. The Intel Arc A770M's massive pixel rate (262.4 GPixel/s) and texture rate (524.8 GTexel/s) make it the clear choice for high-resolution gaming with modern APIs, especially with ray tracing enabled. The AMD Radeon R9 M390X, with its 23.14 GPixel/s and 92.54 GTexel/s, will struggle with modern titles at high settings but remains serviceable for older games and esports titles that match its era. The Vulkan benchmark gap (77.2% in Intel's favor) shows that API-modern workloads strongly favor the Intel part.

For compute and content creation, the Intel Arc A770M is the superior tool. The 5.7 times FP32 advantage, 11.3 times FP16 advantage, and larger memory pool make it suitable for video encoding, 3D rendering, and data processing. The AMD part's lower TDP and mature driver stack make it a reasonable choice for legacy applications or systems with tight power budgets. The data supports a clear verdict: the Intel Arc A770M is the more capable GPU for modern, demanding workloads, while the AMD Radeon R9 M390X holds its own in the database's aggregate metrics due to its long history of optimization.