AMD Ryzen AI Z2 Extreme GPU vs Intel Arc Pro B370 Comparison

FAQ

Q: What are the core architectural differences between the AMD Ryzen AI Z2 Extreme GPU and the Intel Arc Pro B370?

A: The AMD part uses a Strix Point chip built on RDNA 3.5 architecture, manufactured on a 4 nm process by TSMC. The Intel part uses a Panther Lake chip with Xe3-LPG architecture, manufactured on a 3 nm process by Intel itself.

Q: Which GPU has higher raw compute throughput in FP32 operations?

A: The Intel Arc Pro B370 delivers 6.144 TFLOPS FP32, which is about 11% higher than the AMD Ryzen AI Z2 Extreme GPU's 5.530 TFLOPS.

Q: How do the memory subsystems differ between these two GPUs?

A: The AMD Ryzen AI Z2 Extreme GPU has a dedicated 16 GB LPDDR5X memory pool on a 256-bit bus, providing 256.0 GB/s bandwidth. The Intel Arc Pro B370 uses system shared memory, with bandwidth marked as system dependent.

Q: Which GPU has more shading units?

A: The Intel Arc Pro B370 has 1280 shading units, while the AMD Ryzen AI Z2 Extreme GPU has 1024 shading units.

Q: What are the power requirements for each GPU?

A: The AMD Ryzen AI Z2 Extreme GPU has a TDP of 28 W, and the Intel Arc Pro B370 has a TDP of 25 W. Neither requires external power connectors.

Q: Which GPU has a higher pixel fill rate?

A: The AMD Ryzen AI Z2 Extreme GPU achieves 129.6 GPixel/s, while the Intel Arc Pro B370 achieves 48.00 GPixel/s.

Architecture Differences

The AMD Ryzen AI Z2 Extreme GPU and the Intel Arc Pro B370 represent two distinct approaches to integrated graphics. The AMD chip, codenamed Strix Point, is built on RDNA 3.5 architecture and manufactured on a 4 nm process at TSMC. The transistor count is recorded at 34,000 million, with a die size of 233 mm². This yields a transistor density of 145.9M per mm². The Intel part, codenamed Panther Lake, uses Xe3-LPG architecture on a 3 nm process from Intel's own fabs. Its transistor count and die size are not recorded in the database.

The AMD GPU uses a dedicated memory configuration: 16 GB of LPDDR5X on a 256-bit bus, delivering 256.0 GB/s of bandwidth. The Intel GPU instead relies on system shared memory, with the bus width and bandwidth both marked as system dependent. This is a fundamental architectural split, with the AMD part having a fixed, dedicated memory pool versus the Intel part's flexible but variable shared memory approach.

Shader resource allocation differs notably. The Intel Arc Pro B370 carries 1280 shading units, 40 texture mapping units, and 20 raster output units. The AMD Ryzen AI Z2 Extreme GPU has fewer shading units at 1024, but more TMUs at 64 and more ROPs at 48. The AMD part also has 16 ray tracing cores versus 10 on the Intel part.

Clock behavior separates the two as well. The AMD GPU runs at a base clock of 800 MHz and boosts to 2700 MHz, while its memory runs at 1000 MHz with 8 Gbps effective transfer. The Intel GPU has a lower base clock of 300 MHz and a boost clock of 2400 MHz, with system shared memory clocking.

Compute throughput metrics show different strengths. The Intel part reaches 6.144 TFLOPS FP32, while the AMD part reaches 5.530 TFLOPS. In FP16, the Intel GPU delivers 12.29 TFLOPS with a 2:1 ratio, whereas the AMD GPU delivers 5.530 TFLOPS with a 1:1 ratio. This gives Intel a substantial lead in half-precision workloads. However, the AMD part dominates in pixel throughput at 129.6 GPixel/s versus 48.00 GPixel/s, and in texture throughput at 172.8 GTexel/s versus 96.00 GTexel/s.

Both GPUs support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The AMD GPU has a single USB Type-C display output, while the Intel GPU's display outputs are listed as portable device dependent. The Intel part is an IGP with a bus interface of IGP, whereas the AMD part's bus interface is not recorded. Neither GPU requires external power connectors. The AMD part has a TDP of 28 W, and the Intel part has a TDP of 25 W.

Where Each One Wins

The AMD Ryzen AI Z2 Extreme GPU shows strength in rasterization-heavy tasks. Its pixel rate of 129.6 GPixel/s is more than 2.7 times the Intel Arc Pro B370's 48.00 GPixel/s. The texture rate of 172.8 GTexel/s versus 96.00 GTexel/s gives the AMD part a 1.8x advantage in texture-heavy workloads. The larger ROP count of 48 versus 20 reinforces this rasterization lead.

The Intel Arc Pro B370 wins in compute-heavy scenarios. Its FP32 throughput of 6.144 TFLOPS exceeds the AMD part's 5.530 TFLOPS, and its FP16 throughput of 12.29 TFLOPS is more than double the AMD part's 5.530 TFLOPS. The 1280 shading units versus 1024 gives Intel a raw ALU advantage. These metrics suggest the Intel GPU handles general compute and half-precision tasks more efficiently.

Memory configuration favors the AMD part for predictability and bandwidth. The dedicated 16 GB LPDDR5X pool with 256.0 GB/s bandwidth ensures consistent performance regardless of system memory pressure. The Intel part's system shared memory means its effective bandwidth depends on the host platform, which could vary widely.

Ray tracing capability favors the AMD part on core count, with 16 RT cores versus 10 on the Intel GPU. However, the architectural implementation differs between RDNA 3.5 and Xe3-LPG, and the database does not record direct RT benchmark comparisons.

For power efficiency, the Intel part has a lower TDP at 25 W versus 28 W, but the AMD part delivers significantly higher pixel and texture rates within that slightly higher envelope. The Intel part delivers higher FP32 and FP16 throughput at a lower power draw.

Specification Differences

| Specification | AMD Ryzen AI Z2 Extreme GPU | Intel Arc Pro B370 |

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

| Chip | Strix Point | Panther Lake |

| Architecture | RDNA 3.5 | Xe3-LPG |

| Process Node | 4 nm | 3 nm |

| Foundry | TSMC | Intel |

| Transistors | 34,000 million | unknown |

| Die Size | 233 mm² | unknown |

| Base Clock | 800 MHz | 300 MHz |

| Boost Clock | 2700 MHz | 2400 MHz |

| Memory Size | 16 GB | System Shared |

| Memory Type | LPDDR5X | System Shared |

| Memory Bus | 256 bit | System Shared |

| Memory Bandwidth | 256.0 GB/s | System Dependent |

| Shading Units | 1024 | 1280 |

| TMUs | 64 | 40 |

| ROPs | 48 | 20 |

| RT Cores | 16 | 10 |

| Pixel Rate | 129.6 GPixel/s | 48.00 GPixel/s |

| Texture Rate | 172.8 GTexel/s | 96.00 GTexel/s |

| FP32 | 5.530 TFLOPS | 6.144 TFLOPS |

| FP16 | 5.530 TFLOPS (1:1) | 12.29 TFLOPS (2:1) |

| TDP | 28 W | 25 W |

| Slot Width | Not recorded | IGP |

| Bus Interface | Not recorded | IGP |

| Display Outputs | 1x USB Type-C | Portable Device Dependent |

| Release Date | 2025-10-15 | 2026-01-26 |

| Predecessor | None recorded | HD Graphics-WM |

Head-to-Head Benchmarks

The database records no direct benchmark scores for either GPU, and the head-to-head benchmark array is empty. Both GPUs sit at the 50th percentile among all GPUs in the database, with average benchmark scores of zero recorded. The nearest rivals lists are also empty, which limits direct comparative analysis.

Despite the absence of measured benchmark results, the specification data provides clear performance indicators. The AMD Ryzen AI Z2 Extreme GPU holds a decisive lead in fill rates. Its pixel rate of 129.6 GPixel/s compares to 48.00 GPixel/s for the Intel Arc Pro B370, a 2.7x advantage. Texture rate follows a similar pattern, with 172.8 GTexel/s versus 96.00 GTexel/s, giving AMD an 80% lead. These metrics indicate the AMD part processes rasterized scenes substantially faster.

The Intel Arc Pro B370 counters in compute throughput. Its FP32 output of 6.144 TFLOPS beats the AMD part's 5.530 TFLOPS by about 11%. The FP16 comparison is more dramatic: Intel delivers 12.29 TFLOPS versus AMD's 5.530 TFLOPS, a 2.2x advantage. The Intel GPU's 1280 shading units provide the raw parallelism for these compute wins.

Memory bandwidth favors the AMD part with a fixed 256.0 GB/s, while the Intel part's bandwidth is system dependent. In a best-case system, the Intel GPU could match or exceed this figure, but the AMD part guarantees its bandwidth regardless of platform.

Clock speeds tell a mixed story. The AMD GPU has a higher base clock at 800 MHz versus 300 MHz, and a higher boost clock at 2700 MHz versus 2400 MHz. Despite lower clocks, the Intel GPU achieves higher FP32 throughput due to its larger shader array. The AMD part's higher clocks contribute to its fill rate advantage, as pixel and texture throughput scale with clock speed.

Ray tracing core counts favor AMD at 16 versus 10, but without benchmark data, the real-world RT performance gap remains unquantified. Both GPUs support DirectX 12 Ultimate (12_2), so ray tracing features are available on both platforms.

The release timeline shows the AMD part launched on 2025-10-15, while the Intel part followed on 2026-01-26. Both are marked as active production. The Intel part succeeds HD Graphics-WM, while the AMD part has no recorded predecessor.

Power draw is close, with the AMD part at 28 W and the Intel part at 25 W. The AMD GPU delivers its higher fill rates within a 3 W higher envelope, while the Intel GPU delivers higher compute throughput at a lower power draw. Neither requires external power connectors, and the AMD part has no recorded slot width while the Intel part is an IGP.

The display output situation differs: the AMD GPU provides a single USB Type-C output, while the Intel GPU's outputs are portable device dependent. This suggests the Intel part is designed for mobile or embedded implementations where the display routing is determined by the host device.

Both GPUs share identical API support: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. This ensures feature parity in modern graphics APIs, with the performance differences determined by the underlying hardware specifications.

The transistor data is only available for the AMD part, with 34,000 million transistors on a 233 mm² die. The Intel part's transistor count and die size are unknown, which limits density comparisons. The AMD part's density of 145.9M transistors per mm² reflects its 4 nm TSMC process, while the Intel part's 3 nm process likely yields different density characteristics, though no figures are recorded.