Intel Arc Pro B370 vs Intel Arc Pro B65 Comparison

Head-to-Head Benchmarks

The database contains no recorded benchmark scores for either the Intel Arc Pro B370 or the Intel Arc Pro B65. Both entries show an average benchmark score of zero, and the head-to-head benchmark comparison list is empty. This absence of measured data means the two GPUs cannot be ranked against each other through synthetic workload performance metrics. The percentile ranking for both parts sits at 50 against all GPUs in the database, which reflects the lack of recorded results rather than a meaningful performance equivalence. With no wins recorded for either side, any comparative performance analysis must rely entirely on architectural specifications and theoretical throughput figures derived from clock rates and shader counts.

The raw compute figures do provide a starting point for expectations. The Arc Pro B65 delivers 12.29 TFLOPS of FP32 throughput, exactly double the 6.144 TFLOPS of the Arc Pro B370. That same 2:1 ratio appears in FP16 performance, where the B65 reaches 24.58 TFLOPS against the B370's 12.29 TFLOPS. Pixel fill rate follows the same pattern: the B65 outputs 192.0 GPixel/s while the B370 manages 48.00 GPixel/s, a 4x difference. Texture fill rate shows the largest gap, with the B65 at 384.0 GTexel/s versus 96.00 GTexel/s for the B370. These figures indicate that in any future benchmark run, the B65 should dominate throughput-bound workloads, though the absence of measured results prevents confirmation.

Architecture Differences

The two GPUs come from entirely different architectural families. The Arc Pro B370 uses the Panther Lake chip built on Xe3-LPG architecture, manufactured on Intel's 3 nm process node. The Arc Pro B65 uses the BMG-G21 chip built on Xe2-HPG architecture, manufactured by TSMC on a 5 nm node. This generational split places the B370 in the Arc Graphics-WM (Panther Lake) generation, while the B65 belongs to the Battlemage (Pro Series) generation. The B370 functions as an integrated graphics processor (IGP) with no separate die measurements recorded, whereas the B65 is a discrete dual-slot card with a 272 mm² die containing 19,600 million transistors and a transistor density of 72.1M per mm².

Shader resources differ substantially. The B370 packs 1280 shading units, 40 texture mapping units, 20 raster output units, and 10 ray tracing cores. The B65 doubles nearly every count: 2560 shading units, 160 TMUs, 80 ROPs, and 20 RT cores. Clock behavior also diverges. The B370 has a 300 MHz base clock that boosts to 2400 MHz, a wide dynamic range suited to power-constrained integrated operation. The B65 runs at a flat 2400 MHz for both base and boost, with no idle downclocking listed. Both GPUs support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, so API-level feature parity exists despite the internal differences.

Memory architecture creates the most fundamental separation. The B370 uses system shared memory, with its size, type, bus width, and bandwidth all listed as system dependent. The B65 uses 32 GB of dedicated GDDR6 memory on a 256-bit bus, delivering 608.0 GB/s of bandwidth at 2375 MHz (19 Gbps effective). This distinction means the B370's memory performance depends entirely on the host platform's RAM configuration, while the B65 offers fixed, predictable bandwidth. Power delivery reflects this gap: the B370 draws 25 W with no power connectors and an IGP slot width, while the B65 consumes 200 W, requires a single 8-pin power connector, occupies a dual-slot form factor, and needs a 550 W suggested power supply.

FAQ

Q: Which GPU has more raw compute throughput?

A: The Arc Pro B65 delivers 12.29 TFLOPS of FP32 performance, exactly double the 6.144 TFLOPS of the Arc Pro B370. The FP16 figures follow the same 2:1 ratio: 24.58 TFLOPS for the B65 versus 12.29 TFLOPS for the B370.

Q: Do both GPUs support the same graphics APIs?

A: Yes. Both the Arc Pro B370 and the Arc Pro B65 support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The API feature sets are identical, so software compatibility at the API level should match.

Q: How do the memory configurations differ?

A: The B370 relies on system shared memory, with its capacity, type, bus width, and bandwidth all dependent on the host system. The B65 uses 32 GB of dedicated GDDR6 memory on a 256-bit bus with 608.0 GB/s of bandwidth and a 2375 MHz (19 Gbps effective) memory clock.

Q: What are the power requirements for each card?

A: The B370 draws 25 W, uses no power connectors, and fits as an IGP. The B65 draws 200 W, requires one 8-pin power connector, occupies a dual-slot design, and has a suggested power supply rating of 550 W.

Q: Are there any recorded benchmark scores for either GPU?

A: No. Both entries show an average benchmark score of zero, no benchmark results in their respective lists, and no head-to-head benchmark data. The percentile ranking of 50 for each GPU reflects the absence of measured data rather than actual performance parity.

Q: Which GPU was released more recently?

A: The Arc Pro B370 was released on January 26, 2026, while the Arc Pro B65 followed on March 31, 2026. The B370 also lists the HD Graphics-WM as its predecessor, while the B65 has no predecessor recorded.

Specification Differences

| Specification | Intel Arc Pro B370 | Intel Arc Pro B65 |

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

| Chip | Panther Lake | BMG-G21 |

| Architecture | Xe3-LPG | Xe2-HPG |

| Generation | Arc Graphics-WM (Panther Lake) | Battlemage (Pro Series) |

| Process node | 3 nm | 5 nm |

| Foundry | Intel | TSMC |

| Transistors | Unknown | 19,600 million |

| Die size | Unknown | 272 mm² |

| Transistor density | Not recorded | 72.1M / mm² |

| Base clock | 300 MHz | 2400 MHz |

| Boost clock | 2400 MHz | 2400 MHz |

| Memory clock | System Shared | 2375 MHz 19 Gbps effective |

| Memory size | System Shared | 32 GB |

| Memory type | System Shared | GDDR6 |

| Memory bus width | System Shared | 256 bit |

| Memory bandwidth | System Dependent | 608.0 GB/s |

| Shading units | 1280 | 2560 |

| TMUs | 40 | 160 |

| ROPs | 20 | 80 |

| RT cores | 10 | 20 |

| Pixel rate | 48.00 GPixel/s | 192.0 GPixel/s |

| Texture rate | 96.00 GTexel/s | 384.0 GTexel/s |

| FP32 | 6.144 TFLOPS | 12.29 TFLOPS |

| FP16 | 12.29 TFLOPS (2:1) | 24.58 TFLOPS (2:1) |

| TDP | 25 W | 200 W |

| Slot width | IGP | Dual-slot |

| Power connectors | None | 1x 8-pin |

| Suggested PSU | Not recorded | 550 W |

| Bus interface | IGP | PCIe 5.0 x16 |

| Display outputs | Portable Device Dependent | 4x DisplayPort 2.1 |

| Release date | 2026-01-26 | 2026-03-31 |

| Predecessor | HD Graphics-WM | None recorded |

The Verdict

The recorded data draws a clear line between these two products. The Arc Pro B370 is an integrated graphics solution built on Intel's newer 3 nm Xe3-LPG architecture, designed for the Panther Lake platform. Its 25 W power envelope, IGP slot width, and system shared memory mark it as a processor-embedded graphics engine for portable devices, with display outputs dependent on the host device. The Arc Pro B65 is a discrete workstation card on the Xe2-HPG Battlemage architecture, using TSMC's 5 nm process and featuring 32 GB of dedicated GDDR6 memory, a 256-bit bus, PCIe 5.0 x16 connectivity, and four DisplayPort 2.1 outputs.

For users who need a self-contained graphics solution with predictable memory bandwidth and high throughput, the B65 is the only choice that fits. Its 12.29 TFLOPS FP32 performance, 608.0 GB/s memory bandwidth, and 200 W power budget position it for demanding rendering workloads on a desktop workstation. The B370, by contrast, offers a fraction of the compute resources, relies on system memory, and carries no discrete power delivery. Its role is limited to integrated graphics duty inside a Panther Lake-based system.

The release chronology shows the B370 arriving on January 26, 2026, with the B65 following on March 31, 2026. Neither part has recorded benchmark results, so the database cannot confirm real-world performance differences. The specification sheet, however, indicates a decisive advantage for the B65 in every measurable compute category: double the shading units, quadruple the texture and pixel rates, and a dedicated memory subsystem with 608.0 GB/s of bandwidth versus a system-dependent configuration. The B370's advantage lies in power efficiency and integration, not performance. Users on the Panther Lake platform get a capable built-in GPU with no add-in card required. Users who need discrete GPU performance with a fixed memory pool and high-bandwidth output should select the B65. The absence of benchmark data means these conclusions rest on architectural specifications alone, but those specifications point in one direction.