Intel Arc Pro B60 Dual vs NVIDIA B200 SXM6 Comparison

Head-to-Head Benchmarks

The database contains no recorded head-to-head benchmark results for the Intel Arc Pro B60 Dual and the NVIDIA B200 SXM6. Both entries show an average benchmark score of 0 and a percentile rank of 50 among all GPUs. The wins columns for both items are set to 0, meaning the dataset does not currently provide any direct performance comparisons between these two accelerators.

This absence of measured data is itself informative. The two products occupy entirely different segments, and the database has not yet accumulated any overlapping workload results. The Intel Arc Pro B60 Dual and the NVIDIA B200 SXM6 are not competitors in the conventional sense; they are aimed at different use cases, with the former being a professional graphics card and the latter a server-grade compute module. Without benchmark scores, any numerical comparison of performance must rely on the architectural specifications recorded in the database.

The fp32 compute figures are the most direct point of comparison. The NVIDIA B200 SXM6 delivers 69.34 TFLOPS of FP32 performance, while the Intel Arc Pro B60 Dual provides 12.29 TFLOPS. That places the NVIDIA part at roughly 5.6 times the FP32 throughput of the Intel card, a ratio derived directly from the recorded values. In FP16, the gap narrows somewhat: the B200 SXM6 sustains 69.34 TFLOPS at a 1:1 ratio, while the Arc Pro B60 Dual reaches 24.58 TFLOPS at a 2:1 rate. The NVIDIA module still leads, but the Intel card's FP16 output is proportionally closer to its FP32 figure.

Memory bandwidth tells a similar story. The B200 SXM6 records 8.19 TB/s of bandwidth from its HBM3e stack, versus 456.0 GB/s for the Arc Pro B60 Dual's GDDR6 memory. That is an 18-fold difference in raw memory throughput, which strongly influences any memory-bound workload. Texture rate also favors NVIDIA: 1,083.4 GTexel/s versus 384.0 GTexel/s. The pixel rate, however, is an outlier. The Intel card records 192.0 GPixel/s, while the B200 SXM6 records only 43.92 GPixel/s. This suggests that the NVIDIA module is not optimized for traditional rasterization output, a point reinforced by its lack of display outputs.

Architecture Differences

The two accelerators are built on fundamentally different architectures. The Intel Arc Pro B60 Dual uses the Xe2-HPG architecture with the BMG-G21 chip, part of the Battlemage (Pro Series) generation. The NVIDIA B200 SXM6 uses the Blackwell architecture with the GB100 chip, part of the Server Blackwell (Bxx) generation. Both are fabricated by TSMC on a 5 nm process node, but the similarity ends there.

The transistor counts diverge sharply. The NVIDIA GB100 packs 208,000 million transistors across a 1628 mm² die, yielding a transistor density of 127.8 million per mm². The Intel BMG-G21 contains 19,600 million transistors on a 272 mm² die, for a density of 72.1 million per mm². The NVIDIA chip is more than 10 times larger in transistor count and roughly 6 times larger in die area, with a notably higher density.

Memory configurations differ in kind, not just capacity. The B200 SXM6 uses 180 GB of HBM3e across an 8192-bit bus, while the Arc Pro B60 Dual uses 24 GB of GDDR6 across a 192-bit bus. The bus width difference is extreme: 8192 bits versus 192 bits. The NVIDIA module also uses a PCIe 6.0 x16 interface, while the Intel card uses PCIe 5.0 x8.

Clock behavior is another differentiator. The Intel part runs at a 2000 MHz base and 2400 MHz boost, with memory at 2375 MHz (19 Gbps effective). The NVIDIA module lists a 120 MHz base and 1830 MHz boost, with memory at 2000 MHz (8 Gbps effective). The low base clock on the B200 SXM6 is typical of a large server chip designed for sustained boost operation under controlled cooling, whereas the Intel card maintains a high base clock across a dual-slot form factor.

The compute unit counts reflect their intended workloads. The B200 SXM6 has 18,944 shading units, 592 TMUs, 24 ROPs, and 592 tensor cores, with no RT core count recorded. The Arc Pro B60 Dual has 2,560 shading units, 160 TMUs, 80 ROPs, and 20 RT cores, with no tensor core count recorded. The NVIDIA part is clearly built around its tensor core array and massive shader count, while the Intel part includes dedicated ray tracing hardware and a more conventional ROP configuration for graphics output.

API support also separates the two. The Intel card lists DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The B200 SXM6 lists N/A for DirectX, OpenGL, and Vulkan, reflecting its compute-only orientation.

Where Each One Wins

Based strictly on the recorded specifications, the NVIDIA B200 SXM6 dominates in raw compute throughput, memory capacity, memory bandwidth, and texture processing. Its 69.34 TFLOPS of FP32 is the clear standout for scientific computing, AI inference, and training workloads that rely on dense matrix operations. The 592 tensor cores provide a dedicated path for such tasks. The 180 GB of HBM3e memory with 8.19 TB/s of bandwidth is suited for extremely large models or datasets that must reside close to the compute units. The texture rate of 1,083.4 GTexel/s also suggests strong performance in texture-heavy shader work, though the low pixel rate and absence of display outputs indicate this is not a primary use case.

The Intel Arc Pro B60 Dual wins in areas tied to graphics output and traditional rasterization. Its pixel rate of 192.0 GPixel/s is more than 4 times that of the B200 SXM6. It also provides four mini-DisplayPort 2.1 outputs, enabling direct display connection, whereas the NVIDIA module has no outputs at all. The Intel card supports the full set of graphics APIs, including DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, while the NVIDIA part supports none of these. For any workload requiring real-time rendering, ray tracing (20 RT cores), or display presentation, the Intel card is the only viable option between the two.

The Intel card also operates within a lower power envelope. Its TDP is 400 W, compared to 1000 W for the B200 SXM6. The suggested PSU for the Intel card is 800 W, versus 1400 W for the NVIDIA module. The Intel card is a dual-slot PCIe card at 300 mm length, while the B200 SXM6 is an SXM module with no recorded dimensions, indicating a different physical integration path.

The Verdict

The data supports a clear split: the NVIDIA B200 SXM6 is a compute accelerator, and the Intel Arc Pro B60 Dual is a graphics workstation card. The B200 SXM6's 69.34 TFLOPS of FP32, 180 GB of HBM3e, and 8.19 TB/s of memory bandwidth position it for large-scale parallel computation. The Arc Pro B60 Dual's 12.29 TFLOPS, 24 GB of GDDR6, and 456.0 GB/s of bandwidth are modest by comparison, but its 192.0 GPixel/s pixel rate, 20 RT cores, and display outputs make it a functional graphics solution.

Release timing also differs. The B200 SXM6 was released on 2024-10-31, while the Arc Pro B60 Dual followed on 2025-09-04. The NVIDIA part's predecessor is listed as Server Hopper, with a successor in Server Rubin. The Intel part has no recorded predecessor or successor. Both are marked as Active in production status.

The launch MSRP for the Intel Arc Pro B60 Dual is 1,199 USD. The launch MSRP for the NVIDIA B200 SXM6 is 34,999 USD. Those figures, recorded in the database, reflect the intended market distance between the two products.

For a professional graphics workload, the Arc Pro B60 Dual is the only choice with display support and graphics APIs. For a server compute workload, the B200 SXM6 offers more than 5 times the FP32 throughput and 18 times the memory bandwidth, along with 7.5 times the memory capacity. The choice between them is not about performance parity but about workload fit, and the database clearly places them in different categories.

FAQ

Q: Which GPU has higher FP32 performance?

A: The NVIDIA B200 SXM6 records 69.34 TFLOPS of FP32 performance, compared to 12.29 TFLOPS for the Intel Arc Pro B60 Dual.

Q: How much memory does each GPU have?

A: The NVIDIA B200 SXM6 has 180 GB of HBM3e memory, while the Intel Arc Pro B60 Dual has 24 GB of GDDR6 memory.

Q: Does either GPU support display output?

A: The Intel Arc Pro B60 Dual has 4x mini-DisplayPort 2.1 outputs. The NVIDIA B200 SXM6 has no display outputs.

Q: What are the power requirements?

A: The Intel Arc Pro B60 Dual has a TDP of 400 W and a suggested PSU of 800 W. The NVIDIA B200 SXM6 has a TDP of 1000 W and a suggested PSU of 1400 W.

Q: Which GPU has more memory bandwidth?

A: The NVIDIA B200 SXM6 records 8.19 TB/s of memory bandwidth, versus 456.0 GB/s for the Intel Arc Pro B60 Dual.

Q: Which GPU supports more graphics APIs?

A: The Intel Arc Pro B60 Dual supports DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4. The NVIDIA B200 SXM6 lists N/A for all three graphics APIs.

Specification Differences

| Specification | Intel Arc Pro B60 Dual | NVIDIA B200 SXM6 |

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

| Chip | BMG-G21 | GB100 |

| Architecture | Xe2-HPG | Blackwell |

| Generation | Battlemage (Pro Series) | Server Blackwell (Bxx) |

| Transistors | 19,600 million | 208,000 million |

| Die Size | 272 mm² | 1628 mm² |

| Transistor Density | 72.1M / mm² | 127.8M / mm² |

| Base Clock | 2000 MHz | 120 MHz |

| Boost Clock | 2400 MHz | 1830 MHz |

| Memory Clock | 2375 MHz, 19 Gbps effective | 2000 MHz, 8 Gbps effective |

| Memory Size | 24 GB | 180 GB |

| Memory Type | GDDR6 | HBM3e |

| Memory Bus Width | 192 bit | 8192 bit |

| Memory Bandwidth | 456.0 GB/s | 8.19 TB/s |

| Shading Units | 2560 | 18944 |

| TMUs | 160 | 592 |

| ROPs | 80 | 24 |

| RT Cores | 20 | null |

| Tensor Cores | null | 592 |

| Pixel Rate | 192.0 GPixel/s | 43.92 GPixel/s |

| Texture Rate | 384.0 GTexel/s | 1,083.4 GTexel/s |

| FP32 | 12.29 TFLOPS | 69.34 TFLOPS |

| FP16 | 24.58 TFLOPS (2:1) | 69.34 TFLOPS (1:1) |

| TDP | 400 W | 1000 W |

| Slot Width | Dual-slot | SXM Module |

| Power Connectors | 1x 16-pin | null |

| Suggested PSU | 800 W | 1400 W |

| Bus Interface | PCIe 5.0 x8 | PCIe 6.0 x16 |

| Display Outputs | 4x mini-DisplayPort 2.1 | No outputs |

| DirectX | 12 Ultimate (12_2) | N/A |

| OpenGL | 4.6 | N/A |

| Vulkan | 1.4 | N/A |

| Dimensions | 300 mm x 110 mm x 40 mm | null |

| Release Date | 2025-09-04 | 2024-10-31 |

| Predecessor | null | Server Hopper |

| Successor | null | Server Rubin |

| Launch MSRP | 1,199 USD | 34,999 USD |