AMD Radeon R9 M390X vs NVIDIA RTX A4000 Mobile Comparison
AMD Radeon R9 M390X
RTX A4000 Mobile
PERFORMANCE BENCHMARKS
Analysis: AMD Radeon R9 M390X vs NVIDIA RTX A4000 Mobile
The NVIDIA RTX A4000 Mobile and AMD Radeon R9 M390X represent two distant generations of mobile graphics, and the benchmark data reflects a decisive shift in performance. The sole shared benchmark, Geekbench Vulkan, shows the RTX A4000 Mobile scoring 73,002 against the R9 M390X’s 16,951, a 330.7% lead for the NVIDIA part. This is not a close contest; it is a generational gap measured in multiples, not percentages. The RTX A4000 Mobile also carries a much broader benchmark suite, with results spanning DirectX 9 through 12 and compute workloads, while the AMD card only has two recorded Geekbench entries. The data indicates that these cards should not be considered peers, but rather representatives of what a six-year leap in mobile GPU design accomplishes.
Head-to-Head Benchmarks
The only direct comparison available is Geekbench Vulkan, and it is a landslide. The NVIDIA RTX A4000 Mobile posts 73,002 points, while the AMD Radeon R9 M390X manages 16,951 points. That is a delta of 330.7%, meaning the A4000 delivers more than four times the Vulkan performance. For context, the RTX A4000 Mobile’s nearest rivals in overall average score include the NVIDIA Quadro RTX 5000 at 21,629 (delta -1.2%) and the AMD Radeon HD 8970M at 21,237 (delta 0.7%). The R9 M390X, by contrast, sits near the NVIDIA GeForce RTX 3070 Mobile, which scores 20,534 with a delta of 0.6%. In other words, the R9 M390X’s average benchmark score of 20,662 places it in the same performance tier as modern mid-range parts, but its Vulkan showing is anomalously weak compared to its own average, suggesting that its older architecture scales poorly with modern APIs.
Looking at the A4000’s other benchmark results, it shows a mixed profile across legacy DirectX tests. It scores 157 in Passmark DirectX 9, 127 in DirectX 11, and only 66 in DirectX 12. These low absolute numbers are typical of professional mobile GPUs, which are not optimized for rasterization-heavy gaming workloads. Its Passmark G3D score of 14,796 and GPU compute score of 6,394 are more representative of its compute-oriented design. The R9 M390X, lacking any DirectX or Passmark results, cannot be compared on those metrics. The data strongly suggests that if the two were tested across a broader suite, the A4000 would dominate in compute and modern API tests, while the R9 M390X might hold up better in older DirectX 9 or 11 scenarios—but no such data exists in the pack.
Architecture Differences
The architectural gap between these two GPUs is cavernous. The NVIDIA RTX A4000 Mobile uses the GA104 chip on an 8 nm Samsung process, packing 17,400 million transistors into a 392 mm² die, yielding a transistor density of 44.4 million per mm². The AMD Radeon R9 M390X uses the Amethyst chip on a 28 nm TSMC process, with 5,000 million transistors on a 366 mm² die, giving a density of 13.7 million per mm². The A4000 has over three times the transistor count in a similar die size, which explains its massive compute advantage. The A4000 is built on the Ampere architecture, while the R9 M390X uses GCN 3.0, a design that predates hardware ray tracing and tensor cores entirely.
Shader resources differ dramatically. The A4000 features 5,120 shading units, 160 texture mapping units, and 80 render output units. It also includes 40 dedicated ray tracing cores and 160 tensor cores, which are absent from the R9 M390X, as indicated by null values in the data. The R9 M390X has 2,048 shading units, 128 TMUs, and 32 ROPs. That is fewer than half the shaders and a quarter of the ROPs. Pixel rate tells the story: the A4000 achieves 134.4 GPixel/s versus the R9 M390X’s 23.14 GPixel/s. Texture rate is 268.8 GTexel/s versus 92.54 GTexel/s. FP32 compute is 17.20 TFLOPS against 2.961 TFLOPS, a 5.8x difference. The A4000 even matches its FP32 rate in FP16 at 17.20 TFLOPS (1:1), whereas the R9 M390X also lists FP16 at 2.961 TFLOPS (1:1), but at a much lower absolute level.
Memory is another clear divider. The A4000 has 8 GB of GDDR6 on a 256-bit bus, delivering 384.0 GB/s of bandwidth. The R9 M390X has 4 GB of GDDR5 on a 256-bit bus, with 160.0 GB/s. Effective memory speed is 12 Gbps for the A4000 versus 5 Gbps for the R9 M390X. Process node and memory type alone would suggest the A4000 is several generations ahead, but the addition of ray tracing and tensor cores makes it a fundamentally different product. The A4000 also supports PCIe 4.0 x16, while the R9 M390X is limited to PCIe 3.0 x16. API support follows the same pattern: the A4000 offers DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the R9 M390X tops out at DirectX 12 (12_0) and Vulkan 1.2.170.
FAQ
Q: Which GPU has a higher average benchmark score?
A: The NVIDIA RTX A4000 Mobile has an average benchmark score of 21,379, while the AMD Radeon R9 M390X scores 20,662. The A4000 is roughly 3.5% higher, placing both in the 66th percentile of all GPUs.
Q: Is the R9 M390X competitive with the RTX A4000 Mobile in any metric?
A: No. The R9 M390X has no wins in any head-to-head benchmark, and its single Geekbench Vulkan score is 330.7% lower. Its transistor count, shading units, memory bandwidth, and compute rates are all significantly lower across the board.
Q: Why does the R9 M390X have the same percentile as the A4000 despite much lower specs?
A: Both cards sit in the 66th percentile vs all GPUs, but this is an aggregate ranking. The R9 M390X’s average score of 20,662 is pulled by its nearest rivals, which include the NVIDIA GeForce RTX 3070 Mobile at 20,534 and the Intel Arc A750 at 20,582. The percentile reflects the wide distribution of mobile GPUs, where older parts still outperform low-end modern ones.
Q: Does the R9 M390X support ray tracing?
A: No. The data lists no ray tracing cores (null) for the R9 M390X. The RTX A4000 Mobile includes 40 RT cores and 160 tensor cores, enabling hardware-accelerated ray tracing and AI workloads.
Q: What is the memory bandwidth difference?
A: The RTX A4000 Mobile delivers 384.0 GB/s using 8 GB of GDDR6 on a 256-bit bus. The R9 M390X provides 160.0 GB/s using 4 GB of GDDR5 on a 256-bit bus. The A4000 offers 2.4x the bandwidth.
Q: Which GPU has a smaller manufacturing process?
A: The RTX A4000 Mobile is built on an 8 nm Samsung process, whereas the R9 M390X uses a 28 nm TSMC process. The A4000’s process is considerably more advanced, allowing for higher transistor density.
The Verdict
The data points to one conclusion: the NVIDIA RTX A4000 Mobile is the superior GPU by every measurable metric in the pack. Its 330.7% lead in Geekbench Vulkan, 5.8x advantage in FP32 compute, and 2.4x memory bandwidth make it the clear choice for any workload that leverages modern APIs or compute. The R9 M390X, with its 2,048 shaders and 160.0 GB/s bandwidth, is a product of an older era, suited only for legacy tasks. The A4000 also brings hardware ray tracing and tensor cores, which the R9 M390X lacks entirely. If you are choosing between these two for a new system, the A4000 is the only rational pick based on performance data. The R9 M390X’s only advantage is its 75 W TDP compared to the A4000’s 115 W, which may matter in power-constrained designs, but that is a niche consideration against a 330% performance gap.
Specification Differences
- Chip: GA104 (NVIDIA) vs Amethyst (AMD)
- Architecture: Ampere vs GCN 3.0
- Process Node: 8 nm (Samsung) vs 28 nm (TSMC)
- Transistors: 17,400 million vs 5,000 million
- Die Size: 392 mm² vs 366 mm²
- Transistor Density: 44.4M / mm² vs 13.7M / mm²
- Memory Size: 8 GB GDDR6 vs 4 GB GDDR5
- Memory Speed: 12 Gbps effective vs 5 Gbps effective
- Memory Bandwidth: 384.0 GB/s vs 160.0 GB/s
- Shading Units: 5,120 vs 2,048
- TMUs: 160 vs 128
- ROPs: 80 vs 32
- RT Cores: 40 vs None
- Tensor Cores: 160 vs None
- Pixel Rate: 134.4 GPixel/s vs 23.14 GPixel/s
- Texture Rate: 268.8 GTexel/s vs 92.54 GTexel/s
- FP32 Compute: 17.20 TFLOPS vs 2.961 TFLOPS
- FP16 Compute: 17.20 TFLOPS vs 2.961 TFLOPS
- TDP: 115 W vs 75 W
- Bus Interface: PCIe 4.0 x16 vs PCIe 3.0 x16
- DirectX Support: 12 Ultimate (12_2) vs 12 (12_0)
- Vulkan Support: 1.4 vs 1.2.170
Where Each One Wins
The NVIDIA RTX A4000 Mobile wins in every benchmark where data exists. It is the only card with a Geekbench Vulkan result that is not a footnote, and it dominates compute with a Passmark GPU compute score of 6,394. For professional workloads—3D rendering, machine learning inference, or any task that can use tensor cores—the A4000 is not just better; it is in a different class. The 40 RT cores make it viable for ray-traced content creation, and the 8 GB of GDDR6 memory provides ample capacity for large datasets. Its 384.0 GB/s bandwidth ensures that memory-bound tasks do not bottleneck the 17.20 TFLOPS of FP32 compute.
The AMD Radeon R9 M390X has no benchmark wins in the pack. Its only potential advantage is power consumption: at 75 W, it draws 40 W less than the A4000’s 115 W. In a thin-and-light laptop with poor cooling, that lower TDP could be the deciding factor for sustained operation. Its 4 GB of GDDR5 memory and 160.0 GB/s bandwidth are sufficient for older games or basic productivity, but the data shows its Vulkan performance is catastrophically low relative to its average score. The R9 M390X also uses a PCIe 3.0 interface, which is fine for legacy systems but limits bandwidth for data transfers. In short, the R9 M390X is a fallback for power-constrained builds; the A4000 is the answer for anyone who needs actual graphics horsepower.