AMD Radeon RX 570 vs Intel Arc A370M Comparison

# Head-to-Head Benchmarks

The benchmark data covers two compute workloads, and in both, the AMD Radeon RX 570 takes the lead. The largest gap appears in the Geekbench Vulkan test, where the RX 570 scores 42,752 against the Intel Arc A370M's 28,673. That is a 32.9% advantage for AMD, a substantial margin that points to a significant difference in graphics API performance under Vulkan. The Intel part trails by nearly a third in this workload, making it the clearest separation between the two cards.

In Geekbench OpenCL, the RX 570 again wins, scoring 32,084 versus the Arc A370M's 29,676. The delta here is 7.5% in AMD's favor, a more modest gap than Vulkan but still a consistent pattern. The Intel Arc A370M averages 29,175 across its two benchmark runs, while the RX 570 averages 28,766 across four tests. Interestingly, despite losing both head-to-head matchups, the Arc A370M's average benchmark score is actually 1.4% higher than the RX 570's (29,175 vs 28,766). This occurs because the RX 570's benchmark suite includes a Geekbench Metal score of 39,349 and a 3DMark Steel Nomad DX12 score of 880, both of which pull its average down relative to its OpenCL and Vulkan results.

The Vulkan delta is particularly notable when viewed against each card's nearest rivals. The Intel Arc A370M sits at the 74th percentile among all GPUs, with its closest competitor being the AMD Radeon RX Vega M GH at an average score of 29,197 (a 0.1% difference) and the AMD FirePro W8000 at 29,211 (also 0.1% away). The RX 570 also holds the 74th percentile, but its nearest rival list includes the AMD Radeon RX 6800M at 28,874 (0.4% lower) and the AMD Radeon R9 M295X at 28,580 (0.7% higher). Both cards therefore occupy a similar performance tier overall, yet the head-to-head data shows the RX 570 with a decisive edge in the two compute APIs where they were directly compared.

The win count is unambiguous: the RX 570 takes both available head-to-head benchmarks, giving it a 2-0 record over the Arc A370M. No test in the data set favors the Intel part. However, the margin of victory varies widely by API, suggesting that the nature of the workload matters more than a simple "faster" label. In OpenCL the two are close enough that real-world differences might be small, but in Vulkan the RX 570's lead is large enough to be felt in any application that relies heavily on that API.

# The Verdict

Based strictly on the benchmark results, the AMD Radeon RX 570 is the faster card in both measured workloads. It wins Geekbench OpenCL by 7.5% and Geekbench Vulkan by 32.9%. Anyone prioritizing raw compute performance in these two APIs should choose the RX 570 without hesitation. The data does not show a single scenario where the Intel Arc A370M comes out ahead.

That said, the Arc A370M is not without its merits in this comparison. Its average benchmark score across its two tests is 29,175, slightly above the RX 570's 28,766 average across four tests. This means that if you broaden the scope beyond the head-to-head tests and include other benchmark types (like Metal or 3DMark), the Intel part actually looks marginally better on average. But those extra benchmarks are not part of the direct head-to-head comparison, so they should be treated as supplementary context rather than evidence of parity.

The RX 570 also carries a much higher power draw at 150 W versus the Arc A370M's 35 W, which could be a deciding factor for users with thermal or power constraints. The Intel part is listed as an IGP (integrated graphics processor) with no power connectors, while the AMD card is a dual-slot design requiring a 6-pin power connector and a 450 W suggested PSU. If the choice is between a low-power integrated solution and a discrete card that needs external power, the Arc A370M's efficiency might win over users who are not chasing maximum compute scores.

For the majority of users who care about raw performance in OpenCL and Vulkan, the RX 570 is the clear pick. For those who need a low-power, integrated graphics solution with a smaller footprint, the Arc A370M offers a credible alternative, despite losing the benchmark head-to-head. The data does not support calling the Arc a better performer, but it does support calling it a more efficient one.

# Architecture Differences

The two GPUs come from different architectural generations and manufacturing processes. The Intel Arc A370M uses the DG2-128 chip built on Xe-HPG architecture, which belongs to the Alchemist generation (Arc 3 Mobile). It is fabricated on a 6 nm process at TSMC, packing 7,200 million transistors into a 157 mm² die, yielding a transistor density of 45.9 million per square millimeter. In contrast, the AMD Radeon RX 570 uses the Polaris 20 chip based on GCN 4.0 architecture, part of the Polaris (RX 500) generation. It is built on a 14 nm process at GlobalFoundries, with 5,700 million transistors on a 232 mm² die, giving a transistor density of 24.6 million per square millimeter.

These process differences translate into significant clock and efficiency gaps. The Arc A370M has a base clock of 1550 MHz and a boost clock of 2050 MHz, while the RX 570 runs at 1168 MHz base and 1244 MHz boost. Despite the higher clocks, the Intel part draws only 35 W, whereas the AMD card consumes 150 W. The transistor density advantage of the Intel chip (nearly double that of AMD) explains how it achieves higher clocks at a fraction of the power.

The memory subsystems also differ markedly. The Arc A370M uses 4 GB of GDDR6 on a 64-bit bus, delivering 112.0 GB/s of bandwidth with memory clocked at 1750 MHz (14 Gbps effective). The RX 570 uses 4 GB of GDDR5 on a 256-bit bus, doubling the bus width and achieving 224.0 GB/s of bandwidth at 1750 MHz (7 Gbps effective). The RX 570's bandwidth is exactly double that of the Intel part, which is a substantial advantage in memory-heavy workloads.

The compute units tell a different story. The RX 570 has 2048 shading units and 128 texture mapping units, while the Arc A370M has 1024 shading units and 64 TMUs. Both have 32 ROPs. The AMD card also leads in FP32 throughput at 5.095 TFLOPS versus 4.198 TFLOPS for Intel, and in texture rate at 159.2 GTexel/s versus 131.2 GTexel/s. However, the Arc A370M has a higher pixel rate at 65.60 GPixel/s versus 39.81 GPixel/s for the RX 570, which is a notable reversal. The Intel part also includes 8 ray tracing cores, while the RX 570 has none. FP16 performance favors Intel at 8.397 TFLOPS (2:1 ratio) versus AMD's 5.095 TFLOPS (1:1 ratio).

API support differs as well. The Arc A370M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The RX 570 supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.3. The Intel part has a higher DirectX feature level and a newer Vulkan version, which may explain its relatively better showing in Vulkan (though it still lost that test). The RX 570 also includes a Metal benchmark result (39,349), which the Intel part lacks.

# FAQ

Q: Which GPU wins in Geekbench OpenCL?

A: The AMD Radeon RX 570 scores 32,084 versus the Intel Arc A370M's 29,676, a 7.5% advantage for AMD.

Q: How large is the Vulkan performance gap?

A: The RX 570 scores 42,752 in Geekbench Vulkan, while the Arc A370M scores 28,673. That is a 32.9% lead for AMD.

Q: Do both cards have the same amount of memory?

A: Yes, both have 4 GB, but the RX 570 uses GDDR5 on a 256-bit bus (224.0 GB/s), while the Arc A370M uses GDDR6 on a 64-bit bus (112.0 GB/s).

Q: Which card has more shading units?

A: The AMD Radeon RX 570 has 2048 shading units, double the Arc A370M's 1024 shading units.

Q: Does the Intel Arc A370M support ray tracing?

A: Yes, it has 8 ray tracing cores. The RX 570 has no ray tracing cores listed.

Q: What is the power consumption difference?

A: The Arc A370M is rated at 35 W and is an IGP, while the RX 570 is rated at 150 W, requires a 6-pin power connector, and a 450 W suggested PSU.

# Where Each One Wins

The AMD Radeon RX 570 wins in both direct benchmark comparisons: Geekbench OpenCL by 7.5% and Geekbench Vulkan by 32.9%. It also leads in raw compute metrics like FP32 throughput (5.095 TFLOPS vs 4.198 TFLOPS), texture rate (159.2 GTexel/s vs 131.2 GTexel/s), and memory bandwidth (224.0 GB/s vs 112.0 GB/s). With double the shading units and TMUs, the RX 570 is clearly the stronger performer in shader-heavy and bandwidth-intensive workloads. Its 256-bit memory bus is a particular strength, giving it twice the bandwidth of the Intel part, which should benefit high-resolution textures and large data sets.

The Intel Arc A370M, despite losing both benchmarks, has its own areas of advantage. It has a higher pixel rate (65.60 GPixel/s vs 39.81 GPixel/s), which suggests better fill-rate-bound performance in certain rendering scenarios. It also supports FP16 at a 2:1 ratio (8.397 TFLOPS) versus the RX 570's 1:1 ratio (5.095 TFLOPS), making it more efficient at half-precision compute tasks. The presence of 8 ray tracing cores gives it a feature that the RX 570 lacks entirely. Its 35 W power draw is a fraction of the RX 570's 150 W, and its IGP form factor means it needs no power connectors, making it suitable for compact or low-power systems. The Arc A370M also has a higher average benchmark score (29,175 vs 28,766) when all available tests are considered, and it posts a higher Geekbench OpenCL score relative to its nearest rivals (it sits 1% above the RX 6800M, for instance).

# Specification Differences

| Specification | Intel Arc A370M | AMD Radeon RX 570 |

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

| Process node | 6 nm | 14 nm |

| Foundry | TSMC | GlobalFoundries |

| Transistors | 7,200 million | 5,700 million |

| Die size | 157 mm² | 232 mm² |

| Transistor density | 45.9M / mm² | 24.6M / mm² |

| Base clock | 1550 MHz | 1168 MHz |

| Boost clock | 2050 MHz | 1244 MHz |

| Memory type | GDDR6 | GDDR5 |

| Memory bus width | 64 bit | 256 bit |

| Memory bandwidth | 112.0 GB/s | 224.0 GB/s |

| Shading units | 1024 | 2048 |

| TMUs | 64 | 128 |

| ROPs | 32 | 32 |

| RT cores | 8 | None |

| FP32 performance | 4.198 TFLOPS | 5.095 TFLOPS |

| FP16 performance | 8.397 TFLOPS (2:1) | 5.095 TFLOPS (1:1) |

| Pixel rate | 65.60 GPixel/s | 39.81 GPixel/s |

| Texture rate | 131.2 GTexel/s | 159.2 GTexel/s |

| TDP | 35 W | 150 W |

| Slot width | IGP | Dual-slot |

| Power connectors | None | 1x 6-pin |

| Suggested PSU | None | 450 W |

| Bus interface | PCIe 4.0 x8 | PCIe 3.0 x16 |

| DirectX support | 12 Ultimate (12_2) | 12 (12_0) |

| Vulkan support | 1.4 | 1.3 |

| Release date | 2022-03-29 | 2017-04-17 |