Intel Arc Pro A60 vs Intel Arc Pro B70 Comparison

Head-to-Head Benchmarks

The recorded data presents an unusual comparison scenario. The Intel Arc Pro A60 carries two benchmark entries in the database, while the Intel Arc Pro B70 currently has no recorded benchmark scores. Direct head-to-head comparisons are therefore limited to architectural and specification analysis rather than measured performance deltas.

The Arc Pro A60 achieves an average benchmark score of 60,326 across its two recorded tests. Its Geekbench OpenCL score stands at 63,485, while its Geekbench Vulkan score is 57,166. These results place the A60 at the 88th percentile among all GPUs in the database. The nearest rival data for the A60 shows a tight competitive cluster: the NVIDIA GeForce RTX 4090 averages 60,347, a delta of 0 percent; the AMD Radeon Pro W6600M averages 61,896, putting it 2.5 percent ahead of the A60; the AMD Radeon Pro Vega 48 averages 60,140, landing 0.3 percent behind; and the AMD Radeon PRO V710 averages 58,657, trailing by 2.8 percent.

The Arc Pro B70, by contrast, has no benchmark scores, no average score, and no nearest rival entries. Its percentile ranking of 50 reflects the absence of measured results rather than actual performance. The database lists the B70 with a launch MSRP of 949 USD, while the A60 has no launch MSRP recorded.

Since the B70 lacks measured scores, the only quantitative comparison available is through raw specification data. The B70 delivers 22.94 TFLOPS of FP32 compute, which is 2.73 times the 8.397 TFLOPS of the A60. Texture rate shows a similar ratio: 716.8 GTexel/s versus 262.4 GTexel/s, a 2.73x advantage. Pixel rate scales to 358.4 GPixel/s against 131.2 GPixel/s. Memory bandwidth improves from 384.0 GB/s on the A60 to 608.0 GB/s on the B70, a 1.58x increase.

Where Each One Wins

Without recorded B70 benchmarks, win attribution in measured workloads cannot be established from the database. The A60 is the only card in this pairing with actual scores. Its 88th percentile placement indicates it sits above the median GPU in the database, and the rival deltas show it operating within a narrow performance band around cards such as the RTX 4090, Radeon Pro W6600M, and Radeon Pro Vega 48.

The B70 wins on every published hardware specification. It doubles the shading unit count from 2,048 to 4,096, doubles TMUs from 128 to 256, and doubles ROPs from 64 to 128. Ray tracing cores increase from 16 to 32. Memory capacity quadruples from 12 GB to 32 GB. The bus width expands from 192 bit to 256 bit. Boost clock rises from 2050 MHz to 2800 MHz, and base clock rises from 900 MHz to 2280 MHz. The B70 also moves to a newer PCIe interface: PCIe 5.0 x16 versus PCIe 4.0 x16.

The A60 wins on physical footprint. It is a single-slot card with no listed power connector requirement, while the B70 is a dual-slot card with a single 8-pin connector. The A60 lists a suggested PSU of 300 W; the B70 lists 550 W. The A60 has a lower TDP of 130 W compared to 230 W for the B70.

Display output differs in configuration. The A60 provides four DisplayPort 2.0 outputs. The B70 provides one HDMI 2.1a and three DisplayPort 2.1 outputs. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Architecture Differences

The two cards represent different Intel GPU generations. The Arc Pro A60 uses the DG2-256 chip built on the Xe-HPG architecture, belonging to the Alchemist (Pro Series) generation. The Arc Pro B70 uses the BMG-G31 chip built on the Xe2-HPG architecture, belonging to the Battlemage (Pro Series) generation.

Process technology moved forward one node step. The A60 is fabricated on a 6 nm process at TSMC. The B70 is fabricated on a 5 nm process at the same foundry. Transistor counts are only recorded for the A60 at 11,500 million, with a die size of 269 mm² and a transistor density of 42.8M per mm². The B70 lists an unknown transistor count but a larger die size of 368 mm², with no density figure recorded.

The wider architecture changes correspond to the generational jump. The B70 doubles the execution resources across shading units, TMUs, ROPs, and ray tracing cores. Memory subsystem upgrades include a larger bus width and higher effective memory speed: 19 Gbps effective for the B70 versus 16 Gbps effective for the A60. The memory clock is listed at 2375 MHz for the B70 and 2000 MHz for the A60.

The B70 adds a physical dimension record: 267 mm in length, 110 mm in height, and 39 mm in width. The A60 has no dimension data recorded. The B70 also records its power connector as 1x 8-pin, while the A60 lists no power connector. The B70 uses a dual-slot cooler design; the A60 uses a single-slot design.

Production status differs. The A60 is listed as Active. The B70 has no production status recorded. Release dates place the A60 at 2023-06-05 and the B70 at 2026-03-25.

FAQ

Q: Which card has a higher average benchmark score?

A: The Intel Arc Pro A60 has an average benchmark score of 60,326. The Intel Arc Pro B70 has no recorded benchmark scores, so its average score is listed as 0.

Q: How much more FP32 compute does the B70 provide?

A: The B70 delivers 22.94 TFLOPS of FP32 compute versus 8.397 TFLOPS for the A60, a 2.73x increase.

Q: What is the memory capacity difference between the two cards?

A: The A60 has 12 GB of GDDR6 memory on a 192 bit bus with 384.0 GB/s bandwidth. The B70 has 32 GB of GDDR6 memory on a 256 bit bus with 608.0 GB/s bandwidth.

Q: Which card supports a newer PCIe interface?

A: The B70 uses PCIe 5.0 x16, while the A60 uses PCIe 4.0 x16.

Q: Do both cards support the same graphics APIs?

A: Yes. Both cards support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: What are the physical size differences?

A: The A60 is a single-slot card with no recorded dimensions. The B70 is a dual-slot card measuring 267 mm in length, 110 mm in height, and 39 mm in width.

Specification Differences

The following fields differ between the Intel Arc Pro A60 and the Intel Arc Pro B70:

| Field | Intel Arc Pro A60 | Intel Arc Pro B70 |

|---|---|---|

| Chip | DG2-256 | BMG-G31 |

| Architecture | Xe-HPG | Xe2-HPG |

| Generation | Alchemist (Pro Series) | Battlemage (Pro Series) |

| Process node | 6 nm | 5 nm |

| Die size | 269 mm² | 368 mm² |

| Transistors | 11,500 million | unknown |

| Transistor density | 42.8M / mm² | null |

| Base clock | 900 MHz | 2280 MHz |

| Boost clock | 2050 MHz | 2800 MHz |

| Memory clock | 2000 MHz, 16 Gbps effective | 2375 MHz, 19 Gbps effective |

| Memory size | 12 GB | 32 GB |

| Memory bus width | 192 bit | 256 bit |

| Memory bandwidth | 384.0 GB/s | 608.0 GB/s |

| Shading units | 2048 | 4096 |

| TMUs | 128 | 256 |

| ROPs | 64 | 128 |

| Ray tracing cores | 16 | 32 |

| Pixel rate | 131.2 GPixel/s | 358.4 GPixel/s |

| Texture rate | 262.4 GTexel/s | 716.8 GTexel/s |

| FP32 | 8.397 TFLOPS | 22.94 TFLOPS |

| FP16 | 16.79 TFLOPS (2:1) | 45.88 TFLOPS (2:1) |

| TDP | 130 W | 230 W |

| Slot width | Single-slot | Dual-slot |

| Power connectors | null | 1x 8-pin |

| Suggested PSU | 300 W | 550 W |

| Bus interface | PCIe 4.0 x16 | PCIe 5.0 x16 |

| Display outputs | 4x DisplayPort 2.0 | 1x HDMI 2.1a, 3x DisplayPort 2.1 |

| Dimensions | null | 267 mm x 110 mm x 39 mm |

| Production status | Active | null |

| Release date | 2023-06-05 | 2026-03-25 |

| Launch MSRP | null | 949 USD |

| Benchmark scores | Geekbench OpenCL 63,485; Geekbench Vulkan 57,166 | none |

| Average benchmark score | 60,326 | 0 |

| Percentile vs all GPUs | 88 | 50 |

| Nearest rivals | RTX 4090, Radeon Pro Vega 48, Radeon Pro W6600M, Radeon PRO V710 | none |