Intel Arc A580 vs Intel Arc Pro B65 Comparison

The Verdict

The Intel Arc A580 and Intel Arc Pro B65 occupy different corners of the GPU market, and the data supports that split clearly. The Arc A580 is the better choice for general-purpose and gaming-oriented workloads, sitting at the 87th percentile among all GPUs in the database, with an average benchmark score of 57,756. The Arc Pro B65, by contrast, sits at the 50th percentile, and no benchmark scores are recorded for it in the database, making direct performance comparisons impossible from measured results alone.

The Arc A580 is the pick when you need proven, measurable performance. Its recorded benchmarks include a 3DMark Steel Nomad DX12 score of 2,229, a Geekbench OpenCL score of 91,657, and a Geekbench Vulkan score of 79,381. These numbers place it within a tight competitive band: it trails the AMD Radeon RX 5600 OEM by 0.6%, leads the AMD Radeon RX 9070 GRE by 0.7%, trails the Intel Arc A570M by 0.8%, and trails the AMD Radeon RX 6950 XT by 1.1%. That is a remarkably narrow spread, meaning the Arc A580 performs almost identically to several well-known cards from both AMD and Intel.

The Arc Pro B65 is the pick when you need memory capacity, modern interface support, and professional display output. It carries 32 GB of GDDR6 memory, which is four times the 8 GB found on the Arc A580. It also uses a PCIe 5.0 x16 interface, while the Arc A580 uses PCIe 4.0 x16. For display connectivity, the Arc Pro B65 offers four DisplayPort 2.1 outputs, while the Arc A580 offers one HDMI 2.1 and three DisplayPort 2.0 outputs. If your workload demands large datasets in VRAM or multiple high-resolution displays, the Arc Pro B65 is the logical choice, despite having no recorded benchmark scores.

Neither card has a launch MSRP in the database, so pricing is not a factor in this comparison. Both are currently marked as Active in production status.

Where Each One Wins

The Arc A580 wins on raw computed performance metrics. Both cards deliver the same FP32 throughput at 12.29 TFLOPS and the same FP16 throughput at 24.58 TFLOPS (2:1), so in pure compute terms they are equal on paper. The Arc A580 also matches the Arc Pro B65 in pixel rate at 192.0 GPixel/s and texture rate at 384.0 GTexel/s. However, the Arc A580 has more hardware units: 3,072 shading units versus 2,560, 192 texture mapping units versus 160, 96 render output units versus 80, and 24 ray tracing cores versus 20. That hardware advantage, combined with its recorded benchmark scores, means the Arc A580 is the card with demonstrable performance data behind it.

The Arc A580 wins on clock behavior as well. Its base clock is 1700 MHz with a boost clock of 2000 MHz, while the Arc Pro B65 runs at a flat 2400 MHz for both base and boost. The Arc A580's boost clock is lower than the Arc Pro B65's sustained clock, but the Arc A580 has more execution units to compensate. The recorded benchmark results confirm that the Arc A580 performs at a level competitive with the RX 5600 OEM, RX 9070 GRE, Arc A570M, and RX 6950 XT, all within roughly one percent.

The Arc Pro B65 wins on memory capacity and bandwidth. It has 32 GB of GDDR6 memory on a 256-bit bus, yielding 608.0 GB/s of bandwidth. The Arc A580 has 8 GB of GDDR6 on the same 256-bit bus, yielding 512.0 GB/s. That is a 96 GB/s bandwidth advantage and a 24 GB capacity advantage for the Arc Pro B65. For workloads that are memory-bound, such as large model inference or high-resolution texture processing, the Arc Pro B65 is the stronger candidate.

The Arc Pro B65 also wins on power connector simplicity. It requires a single 8-pin power connector and suggests a 550 W power supply, while the Arc A580 requires two 8-pin connectors and suggests a 450 W power supply. The Arc Pro B65 has a higher TDP at 200 W versus 175 W for the Arc A580. The Arc Pro B65 uses more power but needs fewer physical power cables, which can simplify installation in systems with limited PSU cabling.

Architecture Differences

The two cards come from different Intel graphics architectures. The Arc A580 is built on Xe-HPG architecture, part of the Alchemist generation in the Arc 5 series, using the DG2-512 chip. The Arc Pro B65 is built on Xe2-HPG architecture, part of the Battlemage generation in the Pro Series, using the BMG-G21 chip.

The manufacturing process differs. The Arc A580 uses a 6 nm process at TSMC, while the Arc Pro B65 uses a 5 nm process at TSMC. The smaller node allows the Arc Pro B65 to pack more transistors per square millimeter: 72.1 million per mm² versus 53.4 million per mm² for the Arc A580. Despite the denser packing, the Arc Pro B65 has fewer total transistors at 19,600 million versus 21,700 million for the Arc A580. The die size also shrinks from 406 mm² on the Arc A580 to 272 mm² on the Arc Pro B65.

The memory clock differs substantially. The Arc A580 runs its GDDR6 memory at 2000 MHz with 16 Gbps effective data rate. The Arc Pro B65 runs its GDDR6 memory at 2375 MHz with 19 Gbps effective data rate. The higher memory clock, combined with the same 256-bit bus width, produces the Arc Pro B65's higher bandwidth of 608.0 GB/s versus 512.0 GB/s.

The bus interface advances from PCIe 4.0 x16 on the Arc A580 to PCIe 5.0 x16 on the Arc Pro B65. The display outputs also differ: the Arc A580 provides one HDMI 2.1 and three DisplayPort 2.0 ports, while the Arc Pro B65 provides four DisplayPort 2.1 ports, with no HDMI output listed.

Both cards support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Neither card lists tensor cores in the database. The Arc A580 has a predecessor listed as Xe Graphics and a successor listed as Battlemage, while the Arc Pro B65 has no predecessor or successor listed.

FAQ

Q: Which card has more shading units?

A: The Intel Arc A580 has 3,072 shading units, while the Intel Arc Pro B65 has 2,560 shading units.

Q: Does the Arc Pro B65 have any recorded benchmark scores?

A: No. The database lists no benchmark scores for the Arc Pro B65, and its average benchmark score is recorded as 0. The Arc A580 has three recorded scores: 2,229 in 3DMark Steel Nomad DX12, 91,657 in Geekbench OpenCL, and 79,381 in Geekbench Vulkan.

Q: What is the memory capacity difference?

A: The Arc Pro B65 has 32 GB of GDDR6 memory, while the Arc A580 has 8 GB of GDDR6 memory. Both use a 256-bit memory bus.

Q: Which card has higher memory bandwidth?

A: The Arc Pro B65 delivers 608.0 GB/s, compared to 512.0 GB/s for the Arc A580. The difference comes from the Arc Pro B65's higher memory clock of 2375 MHz with 19 Gbps effective rate, versus 2000 MHz with 16 Gbps effective rate on the Arc A580.

Q: Do both cards have the same FP32 compute throughput?

A: Yes. Both cards are recorded at 12.29 TFLOPS for FP32 and 24.58 TFLOPS for FP16 (2:1).

Q: What power supply is suggested for each card?

A: The Arc A580 suggests a 450 W power supply and uses two 8-pin connectors. The Arc Pro B65 suggests a 550 W power supply and uses one 8-pin connector.

Head-to-Head Benchmarks

The database contains no direct head-to-head benchmark entries between the Arc A580 and the Arc Pro B65. The headToHeadBenchmarks field is empty, and both the winsA and winsB counters are zero. Therefore, any comparison must rely on the Arc A580's recorded scores and the Arc Pro B65's specification-level advantages.

The Arc A580's 3DMark Steel Nomad DX12 score of 2,229 places it in a specific performance tier. Its nearest rivals in the database are all within a narrow range. The AMD Radeon RX 5600 OEM averages 58,085, which is 0.6% higher than the Arc A580's average of 57,756. The AMD Radeon RX 9070 GRE averages 57,367, which is 0.7% lower. The Intel Arc A570M averages 58,239, which is 0.8% higher. The AMD Radeon RX 6950 XT averages 58,392, which is 1.1% higher. These deltas show that the Arc A580 performs within one percent of four different GPUs, a tight clustering that indicates consistent performance rather than outlier behavior.

The Geekbench results reinforce this picture. The Arc A580 scores 91,657 in OpenCL and 79,381 in Vulkan. The gap between those two scores, roughly 12,276 points, indicates that the card handles OpenCL workloads better than Vulkan workloads in that benchmark. No comparable data exists for the Arc Pro B65, so the database cannot confirm whether it follows a similar pattern.

The Arc Pro B65's advantages are architectural and memory-related, not benchmark-derived. Its FP32 throughput matches the Arc A580 exactly at 12.29 TFLOPS, and its pixel rate and texture rate also match at 192.0 GPixel/s and 384.0 GTexel/s respectively. The Arc Pro B65 achieves identical throughput with fewer shading units, texture mapping units, render output units, and ray tracing cores. That implies higher per-unit efficiency, likely from the 5 nm process and the Xe2-HPG architecture, though the database does not provide per-unit efficiency numbers.

The Arc Pro B65's 32 GB memory capacity is the largest functional difference. No benchmark in the database measures how that capacity translates into real-world performance, but the bandwidth figure of 608.0 GB/s is 18.75% higher than the Arc A580's 512.0 GB/s. For workloads that saturate memory bandwidth, that advantage could be significant, but the database does not include a test that demonstrates it.

Specification Differences

The following fields differ between the two cards:

• Chip: DG2-512 for the Arc A580, BMG-G21 for the Arc Pro B65.
• Architecture: Xe-HPG for the Arc A580, Xe2-HPG for the Arc Pro B65.
• Generation: Alchemist (Arc 5) for the Arc A580, Battlemage (Pro Series) for the Arc Pro B65.
• Process node: 6 nm for the Arc A580, 5 nm for the Arc Pro B65, both at TSMC.
• Transistors: 21,700 million for the Arc A580, 19,600 million for the Arc Pro B65.
• Die size: 406 mm² for the Arc A580, 272 mm² for the Arc Pro B65.
• Transistor density: 53.4M per mm² for the Arc A580, 72.1M per mm² for the Arc Pro B65.
• Base clock: 1700 MHz for the Arc A580, 2400 MHz for the Arc Pro B65.
• Boost clock: 2000 MHz for the Arc A580, 2400 MHz for the Arc Pro B65.
• Memory clock: 2000 MHz with 16 Gbps effective for the Arc A580, 2375 MHz with 19 Gbps effective for the Arc Pro B65.
• Memory size: 8 GB for the Arc A580, 32 GB for the Arc Pro B65.
• Memory bandwidth: 512.0 GB/s for the Arc A580, 608.0 GB/s for the Arc Pro B65.
• Shading units: 3,072 for the Arc A580, 2,560 for the Arc Pro B65.
• Texture mapping units: 192 for the Arc A580, 160 for the Arc Pro B65.
• Render output units: 96 for the Arc A580, 80 for the Arc Pro B65.
• Ray tracing cores: 24 for the Arc A580, 20 for the Arc Pro B65.
• TDP: 175 W for the Arc A580, 200 W for the Arc Pro B65.
• Power connectors: 2x 8-pin for the Arc A580, 1x 8-pin for the Arc Pro B65.
• Suggested PSU: 450 W for the Arc A580, 550 W for the Arc Pro B65.
• Bus interface: PCIe 4.0 x16 for the Arc A580, PCIe 5.0 x16 for the Arc Pro B65.
• Display outputs: 1x HDMI 2.1 plus 3x DisplayPort 2.0 for the Arc A580, 4x DisplayPort 2.1 for the Arc Pro B65.
• Release date: 2023-10-09 for the Arc A580, 2026-03-31 for the Arc Pro B65.
• Predecessor: Xe Graphics for the Arc A580, none for the Arc Pro B65.
• Successor: Battlemage for the Arc A580, none for the Arc Pro B65.
• Percentile vs all GPUs: 87 for the Arc A580, 50 for the Arc Pro B65.
• Average benchmark score: 57,756 for the Arc A580, 0 for the Arc Pro B65.

Fields that are identical include FP32 throughput at 12.29 TFLOPS, FP16 throughput at 24.58 TFLOPS (2:1), pixel rate at 192.0 GPixel/s, texture rate at 384.0 GTexel/s, memory type as GDDR6, memory bus width as 256-bit, slot width as dual-slot, DirectX support as 12 Ultimate (12_2), OpenGL support as 4.6, Vulkan support as 1.4, and production status as Active. Neither card has a launch MSRP in the database.