Intel Arc A380E vs Intel Arc A570M Comparison

Head-to-Head Benchmarks

The Intel Arc A570M holds a decisive advantage in the recorded benchmark data. The sole available measurement, Geekbench OpenCL, places the A570M at a score of 58,239, while the Intel Arc A380E has no recorded benchmark scores in the database. This absence of data for the A380E means the comparison is entirely one-sided in terms of measured performance. The A570M's score places it in the 88th percentile among all GPUs, indicating strong overall performance relative to the broader market.

The A570M's nearest rivals in the database provide context for its standing. It trails the AMD Radeon RX 6950 XT by a narrow 0.3 percent, with the rival scoring 58,392. It also sits 0.5 percent behind the NVIDIA P102-100, which scores 58,528, and 0.7 percent behind the AMD Radeon PRO V710 at 58,657. Conversely, the A570M leads the AMD Radeon RX 5600 OEM by 0.3 percent, as that card scores 58,085. These margins are razor-thin, showing that the A570M operates in a tightly contested performance band where a single percentage point separates it from several established desktop and workstation parts.

The A380E, with no entries in the benchmark field, cannot be positioned against these rivals. Its percentile ranking of 50 places it at the median of all GPUs in the database, but this is a static classification rather than a measured performance result. The data suggests that the A570M delivers substantially more compute throughput in the OpenCL workload, with a raw score more than double what would be expected from a median-ranked part. The A380E's lack of recorded scores means no direct percentage delta can be calculated between the two Intel products.

Architecture Differences

Both GPUs are built on Intel's Xe-HPG architecture and belong to the Alchemist generation, but they occupy different tiers within that family. The A380E uses the DG2-128 chip, while the A570M uses the larger DG2-256 die. Both are fabricated on TSMC's 6 nm process, yet the physical implementations differ significantly. The DG2-128 die measures 157 mm² and contains 7,200 million transistors, yielding a transistor density of 45.9 million per square millimeter. The DG2-256 die is substantially larger at 269 mm², housing 11,500 million transistors at a density of 42.8 million per square millimeter.

The A570M's larger chip translates into a nearly doubled execution resource pool. It features 2,048 shading units, 128 texture mapping units, 64 render output units, and 16 ray tracing cores. The A380E, by contrast, provides 1,024 shading units, 64 TMUs, 32 ROPs, and 8 ray tracing cores. This 2:1 ratio across all major compute blocks explains the A570M's higher theoretical throughput. The A570M reaches 5.325 TFLOPS of FP32 performance and 10.65 TFLOPS of FP16 with a 2:1 ratio, while the A380E delivers 4.096 TFLOPS and 8.192 TFLOPS respectively. Despite having half the shading units, the A380E's FP32 figure is closer to the A570M than the core count suggests, because the A380E runs at a substantially higher clock speed.

Clock behavior marks another major architectural divergence. The A380E operates with a base clock of 2000 MHz and a boost clock of 2000 MHz, meaning it runs at a flat, sustained frequency. The A570M has a much lower base of 900 MHz but boosts to 1300 MHz, representing a 44 percent uplift over its base. This difference reflects their intended environments: the A380E is a desktop add-in board with a fixed 75 W power envelope, while the A570M is a mobile part designed for laptops, where thermal and power constraints require more aggressive dynamic frequency scaling.

Memory configurations also differ. The A380E ships with 6 GB of GDDR6 on a 96-bit bus, producing 186.0 GB/s of bandwidth at an effective 15.5 Gbps. The A570M offers 8 GB of GDDR6 on a 128-bit bus, yielding 224.0 GB/s at 14 Gbps effective. The A570M has a wider bus and more capacity, giving it both higher peak bandwidth and more headroom for large datasets, though the A380E runs its memory at a higher effective speed.

FAQ

Q: Which GPU has higher FP32 compute throughput?

A: The Intel Arc A570M leads with 5.325 TFLOPS, compared to the Intel Arc A380E's 4.096 TFLOPS. The A570M also doubles the FP16 rate with 10.65 TFLOPS versus 8.192 TFLOPS.

Q: Are both cards based on the same architecture?

A: Yes, both use Intel's Xe-HPG architecture and belong to the Alchemist family. The A380E is classified as Arc 3, while the A570M is an Arc 5 Mobile part.

Q: What is the memory bandwidth difference?

A: The A570M provides 224.0 GB/s over a 128-bit bus with 8 GB of GDDR6. The A380E provides 186.0 GB/s over a 96-bit bus with 6 GB of GDDR6.

Q: How does the A570M compare to its nearest rivals?

A: In Geekbench OpenCL, the A570M scores 58,239. It trails the AMD Radeon RX 6950 XT by 0.3 percent, the NVIDIA P102-100 by 0.5 percent, and the AMD Radeon PRO V710 by 0.7 percent. It leads the AMD Radeon RX 5600 OEM by 0.3 percent.

Q: What is the process node and die size for each?

A: Both use TSMC's 6 nm process. The A380E's DG2-128 die is 157 mm² with 7,200 million transistors. The A570M's DG2-256 die is 269 mm² with 11,500 million transistors.

Q: What are the power requirements?

A: The A380E has a 75 W TDP and suggests a 250 W power supply. The A570M also has a 75 W TDP, but the database lists no suggested PSU for the mobile part.

Specification Differences

The two Intel GPUs differ across nearly every major specification category. The chip design separates them immediately: the A380E uses DG2-128, while the A570M uses DG2-256. The A380E is labeled as Alchemist (Arc 3), while the A570M is Alchemist (Arc 5 Mobile). Transistor counts diverge sharply, with the A380E at 7,200 million and the A570M at 11,500 million. Die size follows suit at 157 mm² versus 269 mm². Transistor density is slightly higher on the smaller chip at 45.9M per mm², compared to 42.8M per mm² on the larger one.

Clock speeds show the most dramatic gap. The A380E runs at a flat 2000 MHz for both base and boost, while the A570M starts at 900 MHz and boosts to 1300 MHz. Memory clocks also differ, with the A380E at 1937 MHz (15.5 Gbps effective) and the A570M at 1750 MHz (14 Gbps effective). Memory capacity and bus width favor the A570M, which has 8 GB on a 128-bit interface versus 6 GB on 96-bit. Bandwidth follows at 224.0 GB/s versus 186.0 GB/s.

Execution resources double on the A570M across the board: 2,048 shading units versus 1,024, 128 TMUs versus 64, 64 ROPs versus 32, and 16 ray tracing cores versus 8. Pixel and texture rates scale accordingly. The A380E achieves 64.00 GPixel/s and 128.0 GTexel/s, while the A570M reaches 83.20 GPixel/s and 166.4 GTexel/s. FP32 and FP16 throughput also favor the A570M, as noted previously.

Physical form and connectivity differ by design intent. The A380E is a single-slot desktop card measuring 254 mm by 127 mm by 20 mm, with four DisplayPort 2.0 outputs and no power connectors, drawing power entirely from its 75 W slot budget. The A570M is an integrated mobile part with no dimensions listed, no slot width, and display outputs described as portable device dependent. Both use a PCIe 4.0 x8 bus interface and support identical API levels: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Production status and release timing also diverge. The A380E is end-of-life, released on 2024-03-31, with a predecessor listed as Xe Graphics and a successor of Battlemage. The A570M is active, released earlier on 2023-07-31, with no predecessor or successor recorded. Neither part has a launch MSRP in the database.

The Verdict

The recorded data paints a clear picture for compute-heavy workloads. The Intel Arc A570M outperforms the A380E in every measured and theoretical metric, with a Geekbench OpenCL score of 58,239 against no recorded score for the A380E. Its 88th percentile ranking among all GPUs places it in the upper tier of the database, while the A380E sits at the 50th percentile. The A570M's nearest rival comparisons show it trading blows with the AMD Radeon RX 6950 XT and NVIDIA P102-100, all within a single percentage point.

For raw compute, the A570M is the obvious choice. Its 2,048 shading units, 16 ray tracing cores, and 224.0 GB/s of memory bandwidth give it a structural advantage that the A380E cannot offset with its higher clocks. The A380E's flat 2000 MHz operation is notable for consistency, but a 2000 MHz clock on half the execution resources still yields fewer operations per second. The A570M's FP32 output of 5.325 TFLOPS versus 4.096 TFLOPS confirms this.

The A380E remains relevant only in specific deployment scenarios. As a single-slot, 75 W desktop card with four DisplayPort 2.0 outputs and no external power connector, it suits compact or embedded systems where physical footprint and output flexibility matter more than peak throughput. Its end-of-life status, however, limits long-term availability. The A570M, being an active mobile part, targets laptops and portable devices where the A570M's higher performance can be delivered within the same 75 W TDP envelope.

The data does not support a neutral verdict. The A570M wins on compute, memory bandwidth, architectural resources, and measured benchmark performance. The A380E wins only on clock speed, pixel fill rate per clock, and desktop-specific features like fixed dimensions and dedicated display outputs. For any workload that scales with shading units or memory bandwidth, the A570M is the stronger part. For a constrained desktop build requiring a low-profile, connector-rich card, the A380E serves a narrower purpose, but its missing benchmark results and median percentile ranking offer no evidence of competitive performance.