Intel Arc A580 vs Intel Arc B390 Comparison

Head-to-Head Benchmarks

The recorded database contains exactly one shared benchmark between the Intel Arc A580 and the Intel Arc B390: the 3DMark Steel Nomad DX12 test. The results are decisive. The Intel Arc A580 scores 2229 points, while the Intel Arc B390 scores 1482 points. The delta is 50.4 percent in favor of the A580. This is not a narrow margin; it is a substantial performance gap that separates these two parts into entirely different performance classes.

Context from the nearest rival data reinforces how wide this gap truly is. The Intel Arc A580 sits at the 87th percentile among all GPUs in the database, with an average benchmark score of 57756. Its nearest rivals include the AMD Radeon RX 5600 OEM at 58085, just 0.6 percent higher, and the AMD Radeon RX 9070 GRE at 57367, 0.7 percent lower. The A580 is effectively trading blows with these desktop discrete GPUs. The Intel Arc B390, by contrast, sits at the 9th percentile, with an average benchmark score of 1482. Its nearest rivals are far older, lower-end parts: the NVIDIA GeForce GT 520MX at 1463, 1.3 percent lower, and the NVIDIA GeForce 800M at 1460, 1.5 percent lower. The B390 is not competing in the same league as the A580 at all.

The single head-to-head result shows a 50.4 percent advantage for the A580 in 3DMark Steel Nomad DX12. Interpreting that number requires understanding what each part is designed for. The A580 is a discrete dual-slot desktop card with its own 8 GB of GDDR6 memory and a 256 bit bus. The B390 is an integrated graphics processor, built into a mobile platform, sharing system memory. The benchmark delta reflects not just raw shader output but also memory architecture, power envelope, and thermal design.

Where Each One Wins

The Intel Arc A580 wins the only directly comparable benchmark, and it wins by a wide margin. In 3DMark Steel Nomad DX12, the A580 delivers 2229 points versus 1482 for the B390. The A580 also holds a decisive advantage in raw compute metrics recorded in the database. Its FP32 throughput is 12.29 TFLOPS, compared to 7.680 TFLOPS for the B390. Pixel rate is 192.0 GPixel/s versus 60.00 GPixel/s. Texture rate is 384.0 GTexel/s versus 120.0 GTexel/s. These figures all point in the same direction: the A580 has roughly double the shading units, four times the texture mapping units, and four times the raster operation units.

The Intel Arc B390 does not win any recorded benchmark. Its wins are structural rather than performance-based. It uses a 3 nm process node from Intel, compared to the 6 nm TSMC node of the A580. Its TDP is 80 W, less than half of the A580's 175 W. It is an IGP with no power connectors and no suggested PSU requirement. The B390 also has a higher boost clock at 2500 MHz versus 2000 MHz for the A580, and a lower base clock at 300 MHz versus 1700 MHz. The boost clock advantage does not translate into benchmark victories in the recorded data.

For use cases, the split is clear. The A580 is for workloads that need sustained discrete GPU performance: desktop gaming, 3D rendering, and compute tasks that can use its 12.29 TFLOPS of FP32 throughput. The B390 is for portable or integrated platforms where power draw, physical space, and thermal output are the primary constraints. Its 80 W TDP and IGP form factor make it suitable for thin-and-light systems, but the benchmark data shows it cannot match the A580 in GPU-bound tasks.

The Verdict

The data points to a straightforward conclusion. The Intel Arc A580 is the stronger GPU by every measurable performance metric in the database. Its 50.4 percent lead in 3DMark Steel Nomad DX12, its 87th percentile ranking, and its average benchmark score of 57756 all place it far above the B390. The Intel Arc B390, with its 9th percentile ranking and average score of 1482, is not a competitor to the A580 in raw performance terms.

The choice between them is not about performance parity. It is about platform constraints. The A580 is a discrete card with a 175 W TDP, dual-slot cooling, and two 8-pin power connectors. It requires a 450 W suggested PSU and a PCIe 4.0 x16 slot. The B390 is an integrated processor with an 80 W TDP, no power connectors, and no dedicated memory. It uses system shared memory and its display outputs are portable device dependent. A user building a desktop tower would pick the A580. A user working on a compact mobile platform would have no choice but the B390, accepting the large performance gap.

The database records the A580 as being released in 2023, with the B390 following in 2026. The B390 is newer, but newer does not mean faster in this comparison. The A580's larger die, dedicated memory, and higher power budget deliver results that the integrated B390 cannot approach. The verdict from the data is unambiguous: the A580 wins on performance, while the B390 wins only on integration and power efficiency.

FAQ

Q: Which GPU has the higher 3DMark Steel Nomad DX12 score?

A: The Intel Arc A580 scores 2229 points, while the Intel Arc B390 scores 1482 points. The A580 leads by 50.4 percent.

Q: How does the Intel Arc A580 compare to its nearest rivals?

A: The A580 has an average benchmark score of 57756. It trails the AMD Radeon RX 5600 OEM by 0.6 percent, trails the Intel Arc A570M by 0.8 percent, trails the AMD Radeon RX 6950 XT by 1.1 percent, and leads the AMD Radeon RX 9070 GRE by 0.7 percent.

Q: How does the Intel Arc B390 compare to its nearest rivals?

A: The B390 has an average benchmark score of 1482. It leads the NVIDIA GeForce GT 520MX by 1.3 percent, the NVIDIA GeForce 800M by 1.5 percent, the NVIDIA GeForce GT 625 OEM by 2.5 percent, and the NVIDIA GeForce GT 710 by 2.7 percent.

Q: What is the FP32 compute difference between the two?

A: The Intel Arc A580 delivers 12.29 TFLOPS of FP32 throughput. The Intel Arc B390 delivers 7.680 TFLOPS. The A580 also offers 24.58 TFLOPS of FP16, compared to 15.36 TFLOPS for the B390.

Q: What memory configuration does each GPU use?

A: The Intel Arc A580 uses 8 GB of GDDR6 memory on a 256 bit bus, with 512.0 GB/s of bandwidth. The Intel Arc B390 uses system shared memory, with system dependent bandwidth.

Q: What is the power draw of each GPU?

A: The Intel Arc A580 has a TDP of 175 W and requires two 8-pin power connectors. The Intel Arc B390 has a TDP of 80 W and uses no power connectors.

Architecture Differences

The two GPUs come from different architectural families within Intel's lineup. The Intel Arc A580 is built on the Xe-HPG architecture, using the DG2-512 chip, and belongs to the Alchemist generation under the Arc 5 series. The Intel Arc B390 uses the Xe3-LPG architecture, based on the Panther Lake chip, and belongs to the Arc Graphics-M generation for Panther Lake platforms.

The manufacturing processes differ substantially. The A580 uses a 6 nm process from TSMC, with 21,700 million transistors on a 406 mm² die, yielding a transistor density of 53.4 million per square millimeter. The B390 uses a 3 nm process from Intel, with transistor count and die size listed as unknown in the database. The shift from an external foundry to Intel's own 3 nm node represents a major change in fabrication strategy.

The compute resources are heavily skewed toward the A580. It has 3072 shading units, 192 texture mapping units, 96 raster operation units, and 24 ray tracing cores. The B390 has 1536 shading units, 48 texture mapping units, 24 raster operation units, and 12 ray tracing cores. This means the A580 has exactly double the shading units and ray tracing cores, and four times the TMUs and ROPs. The pixel rate difference follows: 192.0 GPixel/s for the A580 versus 60.00 GPixel/s for the B390. The texture rate is 384.0 GTexel/s versus 120.0 GTexel/s.

Clock behavior also differs. The A580 runs at a base clock of 1700 MHz and a boost clock of 2000 MHz. The B390 runs at a base clock of 300 MHz and a boost clock of 2500 MHz. The B390's higher boost clock is notable, but it does not compensate for the much smaller execution resource pool. The A580's memory clock is 2000 MHz with 16 Gbps effective, while the B390 relies on system shared memory with no dedicated memory clock.

Both GPUs support the same API feature set: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Neither has a tensor core count listed in the database. The A580 is a dual-slot card with 1x HDMI 2.1 and 3x DisplayPort 2.0 outputs. The B390 is an IGP with portable device dependent display outputs.

Specification Differences

The specification table shows clear divergences between the two parts. The A580 uses the DG2-512 chip, while the B390 uses the Panther Lake chip. The process node is 6 nm TSMC for the A580 and 3 nm Intel for the B390. Transistor count is 21,700 million for the A580 and unknown for the B390. Die size is 406 mm² for the A580 and unknown for the B390.

Memory is a major differentiator. The A580 has 8 GB of GDDR6 with a 256 bit bus and 512.0 GB/s bandwidth. The B390 has system shared memory, system shared type, system shared bus width, and system dependent bandwidth. The A580 has 3072 shading units, 192 TMUs, 96 ROPs, and 24 RT cores. The B390 has 1536 shading units, 48 TMUs, 24 ROPs, and 12 RT cores.

Pixel rate is 192.0 GPixel/s for the A580 and 60.00 GPixel/s for the B390. Texture rate is 384.0 GTexel/s versus 120.0 GTexel/s. FP32 is 12.29 TFLOPS versus 7.680 TFLOPS. FP16 is 24.58 TFLOPS (2:1) versus 15.36 TFLOPS (2:1). TDP is 175 W versus 80 W. The A580 is dual-slot with 2x 8-pin power connectors and a 450 W suggested PSU. The B390 is an IGP with no power connectors and no suggested PSU listed.

The bus interface is PCIe 4.0 x16 for the A580 and IGP for the B390. Display outputs are 1x HDMI 2.1 plus 3x DisplayPort 2.0 for the A580, and portable device dependent for the B390. The A580 was released in 2023 with a predecessor of Xe Graphics and a successor of Battlemage. The B390 was released in 2026 with no predecessor or successor listed. Neither part has a launch MSRP recorded in the database.