AMD Radeon R9 M390X vs Intel Arc B580 Comparison

Where Each One Wins

The Intel Arc B580 is the clear winner in the only directly comparable benchmark recorded in the database. In Geekbench Vulkan, the Arc B580 scores 109,672 against the Radeon R9 M390X's 16,951, a 547% advantage. This is not a marginal lead; it is a generational gap that places the two cards in entirely different performance tiers.

The Arc B580 also holds a substantial lead in the broader benchmark suite. Its average benchmark score across all recorded tests is 23,021, compared to 20,662 for the R9 M390X. The B580 sits at the 68th percentile of all GPUs in the database, while the M390X sits at the 66th percentile. These percentile figures may look close, but the actual score gap is meaningful, and the B580's lead is consistent across every workload where both cards have data.

The Radeon R9 M390X does have one recorded benchmark that the Arc B580 does not share: Geekbench Metal, where it scores 24,373. However, this is a single API-specific result on an older mobile-oriented card, and it does not translate into any head-to-head victory against the B580, since the B580 does not have a Metal score recorded. In practical terms, the M390X wins no category where both cards are measured.

The Arc B580's wins are spread across DirectX 10, 11, and 12 tests, OpenCL, Vulkan, and compute workloads. Its Passmark G3D score of 15,748 and GPU Compute score of 7,729 indicate strong general-purpose and compute performance. The R9 M390X, by contrast, has only two recorded benchmarks, both from Geekbench, and neither approaches the B580's numbers.

Architecture Differences

The two cards come from different architectural eras entirely. The Intel Arc B580 is built on Xe2-HPG architecture, part of the Battlemage generation, and uses the BMG-G21 chip fabricated on a 5 nm process at TSMC. The AMD Radeon R9 M390X uses GCN 3.0 architecture, part of the Gem System generation, and relies on the Amethyst chip on a 28 nm process, also at TSMC. The process node difference alone is massive: 5 nm versus 28 nm, which explains a great deal about the efficiency and density gap.

Transistor counts reflect this. The B580 packs 19,600 million transistors on a 272 mm² die, for a density of 72.1 million transistors per square millimeter. The M390X has 5,000 million transistors on a 366 mm² die, which is actually a larger physical die but with a density of only 13.7 million transistors per square millimeter. The B580 is denser, newer, and more efficient in how it uses silicon area.

The memory architectures are also fundamentally different. The B580 uses 12 GB of GDDR6 on a 192-bit bus, delivering 456.0 GB/s of bandwidth. The M390X uses 4 GB of GDDR5 on a 256-bit bus, delivering 160.0 GB/s. Despite the narrower bus on the B580, the much faster memory type and higher effective clock rate (19 Gbps effective versus 5 Gbps effective) give it nearly three times the bandwidth.

Feature support diverges as well. The B580 has 20 dedicated ray tracing cores, supports DirectX 12 Ultimate (12_2), Vulkan 1.4, and OpenGL 4.6. The M390X has no ray tracing cores, supports DirectX 12 (12_0) only, Vulkan 1.2.170, and OpenGL 4.6. The B580 also supports PCIe 4.0 x8, while the M390X is limited to PCIe 3.0 x16. The B580's display outputs include HDMI 2.1a and three DisplayPort 2.1 connections, whereas the M390X's outputs are portable-device dependent, reflecting its MXM module form factor for laptops.

Head-to-Head Benchmarks

The database records only one benchmark where both cards are measured under the same test: Geekbench Vulkan. The Intel Arc B580 scores 109,672, while the AMD Radeon R9 M390X scores 16,951. The delta is 547% in favor of the B580. That is not a close race; it is a complete mismatch. The B580 is more than six times faster in this API workload.

Looking at other recorded benchmarks for each card individually, the pattern holds. The B580's Geekbench OpenCL score is 92,821, and its Vulkan score of 109,672 is its strongest recorded result. The M390X's Vulkan score of 16,951 is its weakest recorded result, while its Metal score of 24,373 is stronger but still far below the B580's numbers. The B580 also shows a strong Passmark G3D score of 15,748, which is more than double what the M390X would need to approach parity; the M390X has no Passmark scores recorded.

The nearest rivals in the database further contextualize the B580's position. The Arc B580's average score of 23,021 places it within 1% of the AMD Radeon RX 580 2048SP (23,061, delta -0.2%), the NVIDIA GeForce RTX 2080 (22,895, delta 0.6%), the NVIDIA GeForce RTX 3080 (23,172, delta -0.7%), and the NVIDIA P106-100 (23,249, delta -1%). The M390X's average of 20,662 places it near the AMD Radeon RX 5600 XT (20,713, delta -0.2%), Intel Arc A750 (20,582, delta 0.4%), Intel Arc B570 (20,556, delta 0.5%), and NVIDIA GeForce RTX 3070 Mobile (20,534, delta 0.6%). The B580 competes with desktop GPUs from the RTX 20 and 30 series, while the M390X competes with mid-range desktop and mobile parts.

Specification Differences

The two cards differ in nearly every specification category recorded in the database.

Process node: the B580 is on 5 nm, the M390X on 28 nm. Transistors: 19,600 million versus 5,000 million. Die size: 272 mm² versus 366 mm². Transistor density: 72.1M per mm² versus 13.7M per mm².

Clock speeds: the B580 has a base and boost clock of 2670 MHz (both identical), while the M390X has no base or boost clock recorded. Memory clock: 2375 MHz (19 Gbps effective) versus 1250 MHz (5 Gbps effective).

Memory: 12 GB GDDR6 on a 192-bit bus with 456.0 GB/s bandwidth versus 4 GB GDDR5 on a 256-bit bus with 160.0 GB/s bandwidth.

Compute units: the B580 has 2560 shading units, 160 TMUs, 80 ROPs, and 20 ray tracing cores. The M390X has 2048 shading units, 128 TMUs, and 32 ROPs, with no ray tracing cores. Pixel rate: 213.6 GPixel/s versus 23.14 GPixel/s. Texture rate: 427.2 GTexel/s versus 92.54 GTexel/s. FP32: 13.67 TFLOPS versus 2.961 TFLOPS. FP16: 27.34 TFLOPS (2:1) versus 2.961 TFLOPS (1:1).

Power and form factor: the B580 has a 190 W TDP, is dual-slot, uses a single 8-pin power connector, and requires a 450 W suggested PSU. The M390X has a 75 W TDP, is an MXM module, and uses no power connectors. The B580 uses PCIe 4.0 x8; the M390X uses PCIe 3.0 x16.

API support: the B580 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The M390X supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170.

Production status: the B580 is Active; the M390X is End-of-life. Release dates: the B580 launched on 2024-12-12, the M390X on 2015-05-04. The B580's predecessor is Alchemist; the M390X's predecessor is Solar System, and its successor is Polaris Mobile. The B580 has a launch MSRP of 249 USD; the M390X has no recorded launch MSRP.

FAQ

Q: Which card is faster in Vulkan?

A: The Intel Arc B580 scores 109,672 in Geekbench Vulkan, compared to 16,951 for the AMD Radeon R9 M390X, a 547% advantage for the B580.

Q: How do the memory configurations compare?

A: The B580 has 12 GB of GDDR6 on a 192-bit bus with 456.0 GB/s bandwidth. The M390X has 4 GB of GDDR5 on a 256-bit bus with 160.0 GB/s bandwidth. The B580 delivers nearly three times the bandwidth despite the narrower bus.

Q: Does the Radeon R9 M390X support ray tracing?

A: No. The M390X has no ray tracing cores. The Arc B580 has 20 dedicated ray tracing cores.

Q: What is the average benchmark score for each card?

A: The Arc B580 has an average benchmark score of 23,021, placing it at the 68th percentile of all GPUs. The R9 M390X has an average benchmark score of 20,662, placing it at the 66th percentile.

Q: What are the closest competitors to each card in the database?

A: The B580's nearest rivals are the AMD Radeon RX 580 2048SP (23,061, -0.2%), NVIDIA GeForce RTX 2080 (22,895, 0.6%), NVIDIA GeForce RTX 3080 (23,172, -0.7%), and NVIDIA P106-100 (23,249, -1%). The M390X's nearest rivals are the AMD Radeon RX 5600 XT (20,713, -0.2%), Intel Arc A750 (20,582, 0.4%), Intel Arc B570 (20,556, 0.5%), and NVIDIA GeForce RTX 3070 Mobile (20,534, 0.6%).

Q: Which card is more power efficient?

A: The M390X has a 75 W TDP compared to the B580's 190 W TDP, making the M390X much lower power. However, the B580 delivers dramatically higher performance per frame of reference, with 13.67 TFLOPS of FP32 compute versus 2.961 TFLOPS for the M390X, so the efficiency trade-off depends on whether absolute performance or absolute power draw matters more.