Intel Arc B580 vs Intel Arc Pro B65 Comparison

The Verdict

The recorded data presents a clear split between two Intel Battlemage parts built on the same BMG-G21 chip and Xe2-HPG architecture. The Arc B580 is an active consumer-focused card with a full benchmark suite, while the Arc Pro B65 has no recorded benchmark results in the database, meaning its performance must be inferred from its specification differences rather than measured scores. For anyone prioritizing proven, quantifiable performance, the Arc B580 is the only option with data to support a decision. The Arc Pro B65 targets a different role: it carries 32 GB of memory, a 256 bit bus, 608.0 GB/s of bandwidth, and PCIe 5.0 x16 connectivity, all of which point toward memory-heavy professional workloads rather than raw rasterization speed. The data indicates the B580 is the choice for gaming and general compute based on its measured scores and 68th percentile ranking, while the B65 is positioned for tasks where capacity and bandwidth matter more than clock speed or FP32 throughput.

Architecture Differences

Both cards share the same fundamental silicon. The chip is BMG-G21, built on a 5 nm process at TSMC, with 19,600 million transistors on a 272 mm² die, yielding a transistor density of 72.1M per mm². The architecture is Xe2-HPG for both, with the B580 classified under Battlemage (Arc 5) and the B65 under Battlemage (Pro Series). The shading units, texture mapping units, raster operation units, and ray tracing cores are identical: 2560 shading units, 160 TMUs, 80 ROPs, and 20 RT cores. Neither card lists tensor cores. The API support matches exactly, with DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4 on both.

The differences begin with clock behavior. The B580 runs at a base and boost of 2670 MHz, while the B65 is locked at 2400 MHz for both base and boost. That 270 MHz gap directly affects throughput rates. The B580 achieves a pixel rate of 213.6 GPixel/s and a texture rate of 427.2 GTexel/s, compared to 192.0 GPixel/s and 384.0 GTexel/s on the B65. FP32 compute is 13.67 TFLOPS on the B580 versus 12.29 TFLOPS on the B65, and FP16 follows the same pattern with 27.34 TFLOPS versus 24.58 TFLOPS, both at a 2:1 ratio. Memory clock is identical at 2375 MHz with 19 Gbps effective, but the memory subsystem diverges sharply: the B580 uses 12 GB of GDDR6 on a 192 bit bus with 456.0 GB/s of bandwidth, while the B65 uses 32 GB of GDDR6 on a 256 bit bus with 608.0 GB/s of bandwidth.

The bus interface also differs. The B580 uses PCIe 4.0 x8, while the B65 uses PCIe 5.0 x16. Display outputs are not the same either: the B580 offers 1x HDMI 2.1a and 3x DisplayPort 2.1, while the B65 provides 4x DisplayPort 2.1 with no HDMI. The B580 has recorded dimensions of 272 mm in length, 115 mm in height, and 45 mm in width, while the B65 has no dimension data in the database. Both are dual-slot cards with a single 8-pin power connector. The suggested PSU rating differs, with the B580 at 450 W and the B65 at 550 W, and the TDP is 190 W for the B580 versus 200 W for the B65.

Head-to-Head Benchmarks

The head-to-head benchmark array is empty, and the wins counter shows zero for both cards. Consequently, the only measured performance data comes from the B580's individual benchmark suite. In 3DMark Steel Nomad DX12, the B580 scores 3068. Geekbench results show 92821 in OpenCL and 109672 in Vulkan. Passmark results include a G3D score of 15748, a GPU compute score of 7729, and a G2D score of 709. The older DirectX tests in Passmark are lower: DirectX 10 scores 76, DirectX 11 scores 128, DirectX 12 scores 76, and DirectX 9 scores 183.

The B580's average benchmark score is 23021, placing it at the 68th percentile among all GPUs in the database. Its nearest rivals demonstrate how tightly grouped this performance tier is. The AMD Radeon RX 580 2048SP averages 23061, which is 0.2% higher than the B580. The NVIDIA GeForce RTX 2080 averages 22895, putting the B580 0.6% ahead. The NVIDIA GeForce RTX 3080 averages 23172, meaning the B580 trails by 0.7%. The NVIDIA P106-100 averages 23249, leaving the B580 1% behind. These deltas are remarkably small, indicating that the B580 sits in a narrow performance band where the closest competitors are within a single percentage point in either direction. The B65 has no benchmarks and no rivals listed, so no direct comparison is possible from the recorded data.

Specification Differences

The specification table shows where the two cards diverge. The B580 is the older release, with a launch date of 2024-12-12, while the B65 is dated 2026-03-31. The B580 has a launch MSRP of 249 USD, while the B65 has no launch MSRP recorded. The B580's predecessor is listed as Alchemist, while the B65 has no predecessor in the database. Neither card has a successor listed.

Clock speeds differ as noted: 2670 MHz base and boost for the B580 versus 2400 MHz for the B65. Memory capacity is 12 GB versus 32 GB. Bus width is 192 bit versus 256 bit. Bandwidth is 456.0 GB/s versus 608.0 GB/s. Pixel rate is 213.6 GPixel/s versus 192.0 GPixel/s. Texture rate is 427.2 GTexel/s versus 384.0 GTexel/s. FP32 is 13.67 TFLOPS versus 12.29 TFLOPS. FP16 is 27.34 TFLOPS versus 24.58 TFLOPS. TDP is 190 W versus 200 W. Suggested PSU is 450 W versus 550 W. Bus interface is PCIe 4.0 x8 versus PCIe 5.0 x16. Display outputs are 1x HDMI 2.1a plus 3x DisplayPort 2.1 versus 4x DisplayPort 2.1. Physical dimensions are recorded for the B580 only. The B580 has a full benchmark set, while the B65 has none. The B580 sits at the 68th percentile, while the B65 sits at the 50th percentile with an average benchmark score of zero.

The shared specifications are worth noting as well: same chip, same architecture, same process node, same foundry, same transistor count, same die size, same transistor density, same shading units, same TMUs, same ROPs, same RT cores, same memory type, same memory clock, same effective memory speed, same slot width, same power connector, and identical API support.

FAQ

Q: Which card has higher raw compute performance?

A: The Arc B580. Its FP32 throughput is 13.67 TFLOPS versus 12.29 TFLOPS on the Arc Pro B65, and its FP16 throughput is 27.34 TFLOPS versus 24.58 TFLOPS. The B580 also has higher pixel and texture rates.

Q: Why does the Arc Pro B65 have a higher suggested PSU rating?

A: The B65 has a TDP of 200 W compared to 190 W on the B580, and its suggested PSU is 550 W versus 450 W. The higher power draw and PSU recommendation likely relate to the larger memory configuration and different bus interface.

Q: What is the main advantage of the Arc Pro B65 over the B580?

A: The memory subsystem. The B65 has 32 GB of GDDR6 on a 256 bit bus with 608.0 GB/s of bandwidth, while the B580 has 12 GB on a 192 bit bus with 456.0 GB/s. The B65 also uses PCIe 5.0 x16 instead of PCIe 4.0 x8.

Q: How does the B580 compare to its nearest rivals?

A: The B580's average benchmark score of 23021 is 0.2% below the AMD Radeon RX 580 2048SP, 0.6% above the NVIDIA GeForce RTX 2080, 0.7% below the NVIDIA GeForce RTX 3080, and 1% below the NVIDIA P106-100. All four rivals are within one percentage point of the B580.

Q: Does the Arc Pro B65 have any benchmark results?

A: No. The database lists no benchmarks for the B65, and its average benchmark score is recorded as zero. Its performance can only be inferred from specifications, not measured results.

Q: What is the release timeline for these two cards?

A: The Arc B580 was released on 2024-12-12, while the Arc Pro B65 is dated 2026-03-31. The B580 is listed as having the Alchemist predecessor, while the B65 has no predecessor recorded.

Where Each One Wins

The Arc B580 wins on every measured performance metric in the database. Its 2670 MHz clocks produce higher pixel, texture, and compute throughput than the B65's 2400 MHz. The B580's average benchmark score of 23021 places it at the 68th percentile, and its nearest rivals are all within a single percentage point, showing that it competes effectively in a tight performance band. The B580 also has the advantage of being a proven, tested product with results across 3DMark, Geekbench, and Passmark suites. For gaming, DirectX 12 workloads, and general GPU compute, the B580's measured data indicates it delivers competitive performance with an average score that edges out the RTX 2080 by 0.6% and trails the RTX 3080 by only 0.7%.

The Arc Pro B65 wins on memory capacity, memory bandwidth, and connectivity. Its 32 GB of GDDR6 is more than double the B580's 12 GB, and its 608.0 GB/s of bandwidth is 33% higher than the B580's 456.0 GB/s. The 256 bit bus provides a wider path for data movement, and the PCIe 5.0 x16 interface offers a newer, wider connection to the host system compared to the B580's PCIe 4.0 x8. The B65 also provides four DisplayPort 2.1 outputs, which suits multi-monitor professional setups, whereas the B580 mixes one HDMI 2.1a port with three DisplayPort outputs. The B65's higher TDP of 200 W and suggested PSU of 550 W indicate it draws more power, which aligns with its larger memory array and professional positioning.

The data suggests a clear division of roles. The B580 is the card for users who need verified performance and have benchmark results to rely on. Its 68th percentile standing and tight competition with the RX 580 2048SP, RTX 2080, RTX 3080, and P106-100 confirm it as a solid mid-range performer. The B65 is the card for workloads that demand large memory capacity and high bandwidth, such as rendering large scenes or processing big datasets, where the 32 GB frame buffer and 608.0 GB/s bandwidth provide headroom the B580 cannot match. The B65's lack of benchmark data means its real-world performance remains unquantified, but its specification profile points toward memory-bound professional applications rather than raw compute or gaming. Each card wins in its intended domain: the B580 in measured, general-purpose performance, and the B65 in memory capacity and system integration features.