AMD Ryzen AI Z2 Extreme GPU vs Intel Arc B370 Comparison

FAQ

Q: What is the primary architectural difference between the AMD Ryzen AI Z2 Extreme GPU and the Intel Arc B370?

A: The AMD Ryzen AI Z2 Extreme GPU uses the RDNA 3.5 architecture on a 4 nm TSMC process with a Strix Point chip, while the Intel Arc B370 uses the Xe3-LPG architecture on a 3 nm Intel process with a Panther Lake chip.

Q: How do the memory configurations compare between the two GPUs?

A: The AMD Ryzen AI Z2 Extreme GPU has 16 GB of LPDDR5X memory on a 256-bit bus with 256.0 GB/s bandwidth. The Intel Arc B370 uses System Shared memory with System Dependent bandwidth, meaning its memory configuration is tied to the host system.

Q: Which GPU has the higher boost clock?

A: The AMD Ryzen AI Z2 Extreme GPU boosts to 2700 MHz, while the Intel Arc B370 boosts to 2400 MHz. The AMD part also has a higher base clock at 800 MHz versus 300 MHz for the Intel part.

Q: What does the 3DMark Steel Nomad DX12 benchmark show for the Intel Arc B370?

A: The Intel Arc B370 scores 1184 in 3DMark Steel Nomad DX12, placing it at the 5th percentile of all GPUs. Its closest rival is the ATI Mobility Radeon HD 5570 at 1186, a 0.2% difference, and the AMD FirePro M2000 at 1168, where the Arc B370 is 1.4% ahead.

Q: What are the power consumption figures for both GPUs?

A: The AMD Ryzen AI Z2 Extreme GPU has a TDP of 28 W, while the Intel Arc B370 has a TDP of 25 W. Both use no external power connectors.

Q: Which GPU has more shading units and what is the FP32 throughput difference?

A: The Intel Arc B370 has 1280 shading units versus 1024 for the AMD Ryzen AI Z2 Extreme GPU. The Intel part delivers 6.144 TFLOPS FP32, while the AMD part delivers 5.530 TFLOPS FP32.

Architecture Differences

The AMD Ryzen AI Z2 Extreme GPU is built on the RDNA 3.5 architecture using a Strix Point chip, manufactured on a 4 nm process at TSMC. The die contains 34,000 million transistors across a 233 mm² area, yielding a transistor density of 145.9M per mm². This part belongs to the Console GPU (AMD) generation.

The Intel Arc B370 uses the Xe3-LPG architecture with a Panther Lake chip, manufactured on a 3 nm process at Intel's foundry. The transistor count and die size are listed as unknown in the database, and no transistor density figure is recorded. This part belongs to the Arc Graphics-M (Panther Lake) generation.

The AMD GPU integrates 16 GB of LPDDR5X memory on a 256-bit bus, delivering 256.0 GB/s of bandwidth. The memory clock is 1000 MHz with 8 Gbps effective. The Intel GPU instead relies entirely on System Shared memory, with bandwidth described as System Dependent. This is a fundamental difference: the AMD part has dedicated memory with fixed bandwidth, while the Intel part's memory performance depends on the host platform.

Compute resources differ significantly. The Intel Arc B370 has 1280 shading units, 40 texture mapping units, and 20 raster output units. The AMD Ryzen AI Z2 Extreme GPU has 1024 shading units, 64 TMUs, and 48 ROPs. The AMD part thus has more texture units and ROPs despite fewer shaders. Ray tracing cores count 16 on the AMD part versus 10 on the Intel part.

Clock behavior also differs. The AMD GPU runs at a base clock of 800 MHz and boosts to 2700 MHz. The Intel GPU has a much lower base of 300 MHz but boosts to 2400 MHz. The AMD part achieves a pixel rate of 129.6 GPixel/s and a texture rate of 172.8 GTexel/s. The Intel part delivers 48.00 GPixel/s and 96.00 GTexel/s.

FP16 throughput shows a notable divergence. The AMD part delivers 5.530 TFLOPS FP16 at a 1:1 ratio with FP32. The Intel part delivers 12.29 TFLOPS FP16 at a 2:1 ratio, indicating that the Intel architecture has dedicated FP16 acceleration that doubles throughput relative to FP32. Both GPUs support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

The AMD GPU is an integrated part with a single USB Type-C display output. The Intel GPU is also an integrated part, with display outputs described as Portable Device Dependent and a bus interface listed as IGP. Both have no power connectors and are currently Active in production status.

The Verdict

The recorded data shows two integrated GPUs with different strengths. The AMD Ryzen AI Z2 Extreme GPU provides dedicated 16 GB memory with 256.0 GB/s bandwidth, which guarantees consistent memory performance regardless of host system. It also has more texture units, more ROPs, more ray tracing cores, and a higher boost clock at 2700 MHz.

The Intel Arc B370 offers more shading units at 1280, higher FP32 throughput at 6.144 TFLOPS, and substantially higher FP16 throughput at 12.29 TFLOPS. It also has a lower TDP at 25 W versus 28 W for the AMD part. The Intel part's System Shared memory means its performance is dependent on the host system's memory configuration, which introduces variability.

The only benchmark data available is for the Intel Arc B370, which scores 1184 in 3DMark Steel Nomad DX12. This places it at the 5th percentile of all GPUs, with nearest rivals including the ATI Mobility Radeon HD 5570 (1186, 0.2% behind), ATI Radeon HD 5770 (1190, 0.5% behind), AMD Radeon HD 7650M (1192, 0.7% behind), and AMD FirePro M2000 (1168, 1.4% ahead). No benchmark scores are recorded for the AMD Ryzen AI Z2 Extreme GPU, so direct comparison is limited.

The AMD part is preferable when fixed memory bandwidth and higher pixel throughput matter. The Intel part is preferable when raw shader throughput and FP16 compute are the priority, provided the host system can supply adequate memory bandwidth. The AMD part's higher ROP count and texture rate suggest stronger fill-rate performance, while the Intel part's higher shading unit count and FP32 throughput suggest stronger compute performance.

Specification Differences

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

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

| Architecture | RDNA 3.5 | Xe3-LPG |

| Chip | Strix Point | Panther Lake |

| Process Node | 4 nm | 3 nm |

| Foundry | TSMC | Intel |

| Transistors | 34,000 million | unknown |

| Die Size | 233 mm² | unknown |

| Transistor Density | 145.9M / mm² | null |

| Base Clock | 800 MHz | 300 MHz |

| Boost Clock | 2700 MHz | 2400 MHz |

| Memory Clock | 1000 MHz 8 Gbps effective | System Shared |

| Memory Size | 16 GB | System Shared |

| Memory Type | LPDDR5X | System Shared |

| Memory Bus Width | 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 | null | IGP |

| Bus Interface | null | IGP |

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

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

Head-to-Head Benchmarks

The database contains no head-to-head benchmark entries between these two GPUs. The only recorded benchmark score is for the Intel Arc B370 in 3DMark Steel Nomad DX12, where it scores 1184. This score places the Intel part at the 5th percentile of all GPUs, which indicates that this is a low-performing part in the overall GPU landscape.

The nearest rivals for the Intel Arc B370 provide context. The ATI Mobility Radeon HD 5570 scores 1186, putting the Arc B370 0.2% behind. The ATI Radeon HD 5770 scores 1190, putting the Arc B370 0.5% behind. The AMD Radeon HD 7650M scores 1192, putting the Arc B370 0.7% behind. The AMD FirePro M2000 scores 1168, putting the Arc B370 1.4% ahead. These deltas are all small, within a 2.1% spread across the four rivals.

Without benchmark data for the AMD Ryzen AI Z2 Extreme GPU, the comparison must rely on architectural specifications. The AMD part's pixel rate of 129.6 GPixel/s is 2.7 times the Intel part's 48.00 GPixel/s. The AMD part's texture rate of 172.8 GTexel/s is 1.8 times the Intel part's 96.00 GTexel/s. These fill-rate advantages come from the AMD part's higher ROP count (48 versus 20) and TMU count (64 versus 40), combined with a higher boost clock.

The Intel part's FP32 throughput of 6.144 TFLOPS is 11% higher than the AMD part's 5.530 TFLOPS. The Intel part's FP16 throughput of 12.29 TFLOPS is 2.2 times the AMD part's 5.530 TFLOPS. These compute advantages come from the Intel part's higher shading unit count (1280 versus 1024) and its 2:1 FP16 ratio.

The AMD part's memory bandwidth of 256.0 GB/s is fixed and dedicated. The Intel part's bandwidth is System Dependent, so no fixed comparison is possible. In a host system with high-bandwidth memory, the Intel part could narrow or overcome this gap; in a system with modest memory bandwidth, the AMD part would maintain a clear advantage.

Where Each One Wins

The AMD Ryzen AI Z2 Extreme GPU wins in scenarios that depend on fill rate and memory bandwidth. Its 48 ROPs deliver 129.6 GPixel/s, which is 2.7 times the Intel Arc B370's 48.00 GPixel/s. Its 64 TMUs deliver 172.8 GTexel/s, which is 1.8 times the Intel part's 96.00 GTexel/s. Applications that are pixel-bound or texture-bound, such as high-resolution rendering or games with heavy texture sampling, would favor the AMD part.

The AMD part also wins in memory-sensitive workloads. Its 16 GB of LPDDR5X on a 256-bit bus provides 256.0 GB/s of dedicated bandwidth, eliminating any dependence on host memory configuration. The Intel part's System Shared memory means its effective bandwidth varies by platform, making its performance less predictable.

The AMD part's 16 ray tracing cores versus 10 on the Intel part suggest an advantage in ray-traced workloads, though no benchmark data confirms this. The AMD part's higher boost clock of 2700 MHz versus 2400 MHz also contributes to its fill-rate advantages.

The Intel Arc B370 wins in compute-bound scenarios. Its 1280 shading units deliver 6.144 TFLOPS FP32, which is 11% higher than the AMD part's 5.530 TFLOPS. Its FP16 throughput of 12.29 TFLOPS is 2.2 times the AMD part's 5.530 TFLOPS, making it substantially stronger for workloads that use FP16 arithmetic, such as certain AI inference or compute shaders.

The Intel part also wins on power efficiency in terms of TDP. It draws 25 W versus 28 W for the AMD part, a 3 W difference. The Intel part achieves its higher compute throughput at a lower TDP, indicating better compute-per-watt for FP32 and FP16 workloads.

The Intel part's 3 nm process node versus 4 nm for the AMD part may contribute to its efficiency, though the Intel part's transistor count and die size are unknown, so a direct density comparison is not possible.

The Intel Arc B370's position at the 5th percentile of all GPUs, with a 3DMark Steel Nomad DX12 score of 1184, indicates it is a low-end part. Its nearest rivals are all older ATI and AMD parts from a different era. The AMD Ryzen AI Z2 Extreme GPU has no recorded benchmark score, so its percentile position is unknown, though its 50th percentile ranking in the database suggests it sits in the middle of the GPU performance distribution.

For use cases where the host system has limited memory bandwidth, the AMD part's dedicated 256.0 GB/s is a clear advantage. For use cases where the host system has high-bandwidth memory and the workload is compute-heavy, the Intel part's higher shader throughput and FP16 capability would be preferable.