Intel Arc A380E x2 vs Intel Arc Pro B370 Comparison

Head-to-Head Benchmarks

The recorded database contains no direct head-to-head benchmark results for the Intel Arc A380E x2 versus the Intel Arc Pro B370. Both entries list an average benchmark score of zero and a percentile rank of 50 among all GPUs. This means the two cannot be ranked against each other using measured performance data. The absence of benchmark numbers does not imply equivalence; it simply reflects that neither part has accumulated scored runs in the database.

What the data does reveal is a clear split in raw compute capabilities. The Arc A380E x2 delivers 4.096 TFLOPS of FP32 throughput, while the Arc Pro B370 delivers 6.144 TFLOPS. That is a 50% advantage for the Pro B370 in single-precision floating-point work. For FP16, the A380E x2 provides 8.192 TFLOPS, and the Pro B370 provides 12.29 TFLOPS, a 50% lead again. These figures are theoretical peak rates based on the recorded shading unit counts and clock speeds, not application-level results.

Texture and pixel throughput also favor different parts. The A380E x2 has a texture rate of 128.0 GTexel/s versus 96.00 GTexel/s for the Pro B370, a 33% margin for the A380E x2. Pixel rate goes the other direction: the A380E x2 produces 64.00 GPixel/s, while the Pro B370 produces 48.00 GPixel/s, a 33% lead for the A380E x2. So the A380E x2 wins in fill-rate-oriented work, while the Pro B370 wins in raw shader math.

The A380E x2 packs 1024 shading units, 64 texture mapping units, and 32 render output units. The Pro B370 has 1280 shading units, 40 TMUs, and 20 ROPs. The Pro B370 has 25% more shading units, which explains its higher FP32 output per clock. However, it has 37.5% fewer TMUs and 37.5% fewer ROPs, which explains its lower texture and pixel rates despite a higher boost clock.

Ray tracing hardware differs as well. The A380E x2 has 8 ray tracing cores, while the Pro B370 has 10, a 25% increase. No ray tracing benchmark scores exist in the database, so the practical impact cannot be quantified from recorded measurements. The theoretical core count suggests the Pro B370 could handle ray intersection work with more parallelism, but fill-rate constraints could limit real-world results.

Clock behavior is notable. The A380E x2 runs at a fixed 2000 MHz for both base and boost. The Pro B370 has a base clock of 300 MHz and a boost clock of 2400 MHz. The wide gap between base and boost on the Pro B370 indicates a power-managed design that scales heavily under load. The A380E x2 maintains a constant 2000 MHz, which simplifies thermal and power behavior but sacrifices peak clock headroom.

The A380E x2 has dedicated memory: 6 GB of GDDR6 on a 96-bit bus, yielding 186.0 GB/s of bandwidth. The Pro B370 uses system shared memory with system-dependent bandwidth. This is a fundamental architectural split. The A380E x2 guarantees local video memory bandwidth, while the Pro B370 depends entirely on the host platform's memory subsystem.

In the absence of benchmark scores, the theoretical peak rates are the only quantitative comparison available. The Pro B370 holds a 50% advantage in FP32 and FP16 compute. The A380E x2 holds a 33% advantage in texture and pixel fill rates. Neither part has a recorded win in the head-to-head benchmark section; the database lists zero wins for each.

FAQ

Q: Which GPU has higher FP32 compute according to the database?

A: The Intel Arc Pro B370 records 6.144 TFLOPS, which is 50% higher than the Intel Arc A380E x2's 4.096 TFLOPS.

Q: Does the Intel Arc A380E x2 have dedicated video memory?

A: Yes, it uses 6 GB of GDDR6 with a 96-bit bus and 186.0 GB/s bandwidth. The Arc Pro B370 uses system shared memory with system-dependent bandwidth.

Q: What are the shading unit counts for each GPU?

A: The Arc A380E x2 has 1024 shading units. The Arc Pro B370 has 1280 shading units, which is 25% more.

Q: How do the texture rates compare?

A: The Arc A380E x2 achieves 128.0 GTexel/s, while the Arc Pro B370 achieves 96.00 GTexel/s. The A380E x2 is 33% faster in texture fill rate.

Q: Which GPU has more ray tracing cores?

A: The Arc Pro B370 has 10 ray tracing cores, compared to 8 in the Arc A380E x2. This is a 25% higher count.

Q: What is the power draw difference?

A: The Arc A380E x2 has a TDP of 130 W, while the Arc Pro B370 has a TDP of 25 W. The Pro B370 uses 19% of the power of the A380E x2.

Q: Are there any recorded benchmark scores for these GPUs?

A: The database lists an average benchmark score of zero for both parts, and both have a percentile rank of 50. No head-to-head benchmark results are recorded.

Where Each One Wins

The Intel Arc A380E x2 wins in memory bandwidth and fill-rate metrics. Its dedicated 186.0 GB/s of bandwidth is a hard advantage over system-shared memory, which cannot guarantee any specific bandwidth figure. For workloads that stream large textures, render to many pixels, or require consistent memory access patterns, the A380E x2 has a structural edge. Its 64.00 GPixel/s pixel rate and 128.0 GTexel/s texture rate are both 33% higher than the Pro B370's corresponding figures. This makes the A380E x2 better suited for traditional rasterization tasks, where pixel and texel throughput directly limit frame rendering.

The A380E x2 also has more render output units (32 versus 20) and more texture mapping units (64 versus 40). These count differences align with its pixel and texture rate advantages. The fixed 2000 MHz clock means the A380E x2 does not rely on boost behavior; its peak is its sustained rate. For applications that need predictable performance, such as multi-GPU rendering or compute pipelines with consistent load, this fixed clock is a practical benefit.

The Intel Arc Pro B370 wins in compute throughput and efficiency. Its 6.144 TFLOPS FP32 and 12.29 TFLOPS FP16 are 50% higher than the A380E x2. This matters for workloads dominated by shader math, such as compute shaders, machine learning inference with FP16, physics simulations, or any task that scales with shading unit count. The Pro B370's 1280 shading units provide 25% more parallel execution lanes than the A380E x2's 1024.

The Pro B370 also has 10 ray tracing cores versus 8, giving it a theoretical advantage in ray-traced effects. The very low 25 W TDP makes it suitable for power-constrained environments, though the database does not record any efficiency benchmark. The boost clock of 2400 MHz is 20% higher than the A380E x2's 2000 MHz, which could allow burst compute performance when thermal and power limits permit.

The system shared memory design of the Pro B370 is both a weakness and a strength. It eliminates the need for a dedicated memory bus and VRAM chips, reducing power and cost. But it also means performance depends entirely on the host system's memory bandwidth, which the database lists as "System Dependent." In a high-bandwidth system, the Pro B370 could approach or exceed the A380E x2's memory throughput, but in a low-bandwidth system it would fall short. No measurements exist to quantify this dependency.

For workloads that require sustained fill rate, the A380E x2 is the stronger choice. For workloads that require peak FP32/FP16 compute or ray tracing core count, the Pro B370 is the stronger choice. The A380E x2 suits direct rendering and multi-GPU configurations where dedicated memory is essential. The Pro B370 suits embedded or portable devices where power draw is critical and memory bandwidth is provided by the platform.

Specification Differences

The two GPUs differ across nearly every recorded specification category. The Intel Arc A380E x2 uses the DG2-128 chip, while the Intel Arc Pro B370 uses the Panther Lake chip. The A380E x2 belongs to the Alchemist (Arc 3) generation, and the Pro B370 belongs to the Arc Graphics-WM (Panther Lake) generation.

Process node differs: the A380E x2 is built on a 6 nm process by TSMC, with 7,200 million transistors on a 157 mm² die. The Pro B370 is built on a 3 nm process by Intel, with transistor count and die size listed as unknown. The A380E x2 has a transistor density of 45.9M per mm², while the Pro B370 has no recorded density.

Clock speeds differ significantly. The A380E x2 has a base clock of 2000 MHz and a boost clock of 2000 MHz. The Pro B370 has a base clock of 300 MHz and a boost clock of 2400 MHz. Memory clocks also differ: the A380E x2 uses 1937 MHz with 15.5 Gbps effective, while the Pro B370 uses system shared memory.

Memory configuration is entirely different. The A380E x2 has 6 GB of GDDR6 with a 96-bit bus and 186.0 GB/s bandwidth. The Pro B370 has system shared memory with system-dependent bandwidth and no dedicated bus width.

Shading units, TMUs, and ROPs all differ. The A380E x2 has 1024 shading units, 64 TMUs, and 32 ROPs. The Pro B370 has 1280 shading units, 40 TMUs, and 20 ROPs. Ray tracing cores differ: 8 versus 10.

Pixel and texture rates differ as detailed earlier. FP32 and FP16 rates differ as detailed earlier. TDP differs: 130 W for the A380E x2, 25 W for the Pro B370. Slot width differs: single-slot for the A380E x2, IGP (integrated graphics processor) for the Pro B370.

Power connectors differ: the A380E x2 requires one 6-pin connector, while the Pro B370 has none. The A380E x2 has a suggested PSU of 300 W, while the Pro B370 has no suggested PSU recorded. Bus interface differs: PCIe 4.0 x8 for the A380E x2, IGP for the Pro B370.

Display outputs differ: the A380E x2 has 8x mini-DisplayPort 2.0, while the Pro B370 has portable device dependent outputs. Dimensions differ: the A380E x2 measures 265 mm by 127 mm by 20 mm, while the Pro B370 has no recorded dimensions.

Production status differs: the A380E x2 is end-of-life, the Pro B370 is active. Release dates differ: the A380E x2 was released in 2024, the Pro B370 in 2026. The A380E x2's predecessor is Xe Graphics and its successor is Battlemage. The Pro B370's predecessor is HD Graphics-WM and it has no recorded successor.

Architecture Differences

The Intel Arc A380E x2 uses the Xe-HPG architecture, while the Intel Arc Pro B370 uses the Xe3-LPG architecture. These are distinct GPU microarchitectures from different Intel generations. Xe-HPG is the architecture behind the Alchemist discrete GPUs, designed for gaming and general-purpose compute. Xe3-LPG is a low-power architecture for integrated graphics, part of the Panther Lake platform.

The process node differs: 6 nm TSMC for the A380E x2 versus 3 nm Intel for the Pro B370. The A380E x2 is a discrete GPU with its own chip, 7,200 million transistors, and a 157 mm² die. The Pro B370 is an integrated GPU with unknown transistor count and die size, sharing the package with other components.

Core organization differs. The A380E x2 has 1024 shading units, 64 TMUs, and 32 ROPs. The Pro B370 has 1280 shading units, 40 TMUs, and 20 ROPs. The higher shading unit count in the Pro B370 is paired with lower fixed-function unit counts, indicating a design that favors compute over rasterization. The A380E x2 balances more TMUs and ROPs for higher fill rates.

Ray tracing core counts differ: 8 for the A380E x2, 10 for the Pro B370. Neither part has tensor cores recorded in the database. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, so the API feature set is identical.

Memory architecture is fundamentally different. The A380E x2 has dedicated GDDR6 memory on a 96-bit bus. The Pro B370 uses system shared memory, meaning it has no local video memory and relies entirely on the host system's memory controller. This affects how each GPU handles textures, framebuffers, and compute data. The A380E x2 has guaranteed bandwidth of 186.0 GB/s. The Pro B370 has no fixed bandwidth figure.

The power envelope reflects the architectural split. The A380E x2 draws 130 W and needs a 6-pin power connector and a 300 W PSU. The Pro B370 draws 25 W and requires no power connector. The A380E x2 is a single-slot card with dimensions of 265 mm by 127 mm by 20 mm. The Pro B370 is an IGP with no physical slot dimensions.

The A380E x2 uses PCIe 4.0 x8 for host communication. The Pro B370 uses IGP, which means it connects through the processor's internal fabric. Display outputs differ: the A380E x2 provides 8x mini-DisplayPort 2.0 connections, while the Pro B370's outputs depend on the portable device it is integrated into.

Clocks reveal different operating strategies. The A380E x2 runs at a constant 2000 MHz with no boost headroom. The Pro B370 idles at 300 MHz and boosts to 2400 MHz, a 700% increase from base to boost. This suggests the Pro B370 is designed to scale frequency dynamically based on thermal and power budgets, whereas the A380E x2 operates at a fixed frequency.

The transistor density differs: the A380E x2 has 45.9M transistors per mm² on a 6 nm TSMC process. The Pro B370's density is not recorded. The A380E x2's die size of 157 mm² is confirmed, while the Pro B370's die size is unknown. The Pro B370's 3 nm Intel process likely provides higher density, but no numbers exist to confirm this.

The architectural differences point to different use cases. Xe-HPG with dedicated memory suits standalone graphics tasks. Xe3-LPG with shared memory suits integrated systems where power and space are limited. The A380E x2's end-of-life status and the Pro B370's active status indicate that Intel is transitioning from the discrete Alchemist generation to the integrated Panther Lake generation. The Pro B370 has no successor recorded, while the A380E x2's successor is Battlemage, a separate architecture not covered in this comparison.