Intel Arc A580 vs Intel Arc Pro B370 Comparison

# The Verdict

The Intel Arc A580 and Intel Arc Pro B370 occupy fundamentally different positions in the hardware landscape. The A580 is a discrete, high-throughput graphics card built on the Xe-HPG architecture, while the B370 is an integrated graphics processor embedded in the Panther Lake chip, using the Xe3-LPG architecture. The recorded data shows no shared benchmark results between the two, and the B370 has no benchmark entries at all, which makes a direct performance comparison impossible from the database alone.

The A580 sits at the 87th percentile among all GPUs in the database, with an average benchmark score of 57756. Its nearest rivals include the AMD Radeon RX 5600 OEM at 58085 (0.6% ahead), the Intel Arc A570M at 58239 (0.8% ahead), the AMD Radeon RX 6950 XT at 58392 (1.1% ahead), and the AMD Radeon RX 9070 GRE at 57367 (0.7% behind). These deltas are all within roughly one percentage point, indicating that the A580 clusters tightly with a group of mid-range to upper-mid-range cards. The B370, by contrast, sits at the 50th percentile with an average benchmark score of zero, reflecting the absence of recorded benchmark results rather than a literal performance of zero.

Who should pick which? The data supports the A580 for any workload requiring substantial graphics throughput, dedicated memory bandwidth, and standalone compute capability. The B370, with its 25 W power envelope, integrated design, and system-shared memory, serves a completely different purpose: low-power, portable, or embedded systems where a discrete card is not viable. The A580 delivers 12.29 TFLOPS of FP32 compute, while the B370 delivers 6.144 TFLOPS, roughly half. The A580 offers 512.0 GB/s of dedicated memory bandwidth on an 8 GB GDDR6 pool with a 256-bit bus; the B370 uses system-shared memory with system-dependent bandwidth. For gaming, rendering, or compute tasks that demand sustained throughput, the A580 is the only choice supported by the data. For ultra-low-power integrated graphics in a mobile or compact platform, the B370 exists in a different category entirely.

# Architecture Differences

The two processors derive from different design lineages. The A580 uses the DG2-512 chip built on the Xe-HPG architecture, belonging to the Alchemist generation under the Arc 5 family. It is fabricated on a 6 nm process at TSMC, with 21,700 million transistors on a 406 mm² die, yielding a transistor density of 53.4 million per square millimeter. The B370 uses the Panther Lake chip with the Xe3-LPG architecture, belonging to the Arc Graphics-WM generation. It is fabricated on a 3 nm process at Intel, with transistor count and die size listed as unknown in the database.

The A580 has 3072 shading units, 192 texture mapping units, 96 render output units, and 24 ray tracing cores. The B370 has 1280 shading units, 40 TMUs, 20 ROPs, and 10 ray tracing cores. The A580's pixel rate is 192.0 GPixel/s and its texture rate is 384.0 GTexel/s. The B370's pixel rate is 48.00 GPixel/s and its texture rate is 96.00 GTexel/s. The A580 delivers 12.29 TFLOPS of FP32 and 24.58 TFLOPS of FP16 (2:1). The B370 delivers 6.144 TFLOPS of FP32 and 12.29 TFLOPS of FP16 (2:1).

Clock behavior differs substantially. The A580 runs at a base clock of 1700 MHz with a boost of 2000 MHz, and its GDDR6 memory operates at 2000 MHz with 16 Gbps effective. The B370 has a base clock of 300 MHz and a boost of 2400 MHz, with system-shared memory. The B370's boost clock is higher, but its base clock is drastically lower, and its memory subsystem is entirely dependent on the host system.

Power and physical design diverge sharply. The A580 has a TDP of 175 W, is dual-slot, uses 2x 8-pin power connectors, and recommends a 450 W PSU. It interfaces via PCIe 4.0 x16 and provides 1x HDMI 2.1 and 3x DisplayPort 2.0 outputs. The B370 has a TDP of 25 W, is an integrated graphics processor (IGP), uses no power connectors, has no suggested PSU, and interfaces via IGP. Its display outputs are listed as portable device dependent.

Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The A580 was released on 2023-10-09, with its predecessor listed as Xe Graphics and successor as Battlemage. The B370 was released on 2026-01-26, with its predecessor listed as HD Graphics-WM and no successor recorded. Both are marked as active in production status.

# Head-to-Head Benchmarks

The database contains no head-to-head benchmark entries between the A580 and B370. The wins count for each is zero. The B370 has no benchmark records at all, so no direct score comparison is possible. What the data does offer is the A580's standalone benchmark results and its position relative to other discrete GPUs.

The A580 scores 2229 in 3DMark Steel Nomad DX12. In Geekbench OpenCL, it scores 91657. In Geekbench Vulkan, it scores 79381. These results place it at the 87th percentile among all GPUs. Its average benchmark score of 57756 is nearly identical to its nearest rivals: the AMD Radeon RX 5600 OEM is 0.6% ahead, the Intel Arc A570M is 0.8% ahead, the AMD Radeon RX 6950 XT is 1.1% ahead, and the AMD Radeon RX 9070 GRE is 0.7% behind. The tight clustering indicates that the A580 performs within a narrow band around these cards, with no single rival dominating by more than a percentage point.

The B370's average benchmark score is zero, and it has no nearest rivals listed. The percentile of 50 places it at the median of the database, but without benchmark entries this percentile likely reflects its position in the distribution of all GPUs based on whatever limited data exists. The lack of recorded scores means no performance claims can be made from measurements.

Examining the compute ratios provides context. The A580's FP32 throughput of 12.29 TFLOPS is exactly double the B370's 6.144 TFLOPS. The A580's FP16 of 24.58 TFLOPS is exactly double the B370's 12.29 TFLOPS. The A580's pixel rate of 192.0 GPixel/s is exactly four times the B370's 48.00 GPixel/s. The A580's texture rate of 384.0 GTexel/s is exactly four times the B370's 96.00 GTexel/s. These ratios come directly from the recorded specifications, not from benchmark runs.

The memory systems are not comparable. The A580 has 8 GB of dedicated GDDR6 on a 256-bit bus with 512.0 GB/s of bandwidth. The B370 has system-shared memory of system-shared type, with system-dependent bandwidth. This means the B370's memory performance cannot be quantified from the database; it depends entirely on the host platform's memory configuration.

# FAQ

Q: Which GPU has higher FP32 compute throughput?

A: The Intel Arc A580 delivers 12.29 TFLOPS of FP32, while the Intel Arc Pro B370 delivers 6.144 TFLOPS. The A580 has exactly double the FP32 throughput of the B370.

Q: What are the memory configurations of each GPU?

A: The A580 has 8 GB of GDDR6 memory on a 256-bit bus with 512.0 GB/s of bandwidth. The B370 uses system-shared memory with system-shared type and bus width, and its bandwidth is system dependent.

Q: Which GPU has a higher boost clock?

A: The B370 has a boost clock of 2400 MHz, while the A580 has a boost clock of 2000 MHz. However, the A580 has a base clock of 1700 MHz compared to the B370's 300 MHz base clock.

Q: What is the power consumption difference?

A: The A580 has a TDP of 175 W and requires 2x 8-pin power connectors with a suggested 450 W PSU. The B370 has a TDP of 25 W, uses no power connectors, and has no suggested PSU listed.

Q: How do the benchmark scores of the two compare?

A: The A580 has recorded benchmark scores of 2229 in 3DMark Steel Nomad DX12, 91657 in Geekbench OpenCL, and 79381 in Geekbench Vulkan. The B370 has no benchmark entries in the database, so no direct comparison is possible.

Q: What are the physical form factors?

A: The A580 is a dual-slot discrete card using PCIe 4.0 x16 with 1x HDMI 2.1 and 3x DisplayPort 2.0 outputs. The B370 is an integrated graphics processor with IGP bus interface and portable device dependent display outputs.

# Where Each One Wins

The A580 wins in every measurable performance category that has recorded data. Its 12.29 TFLOPS FP32, 24.58 TFLOPS FP16, 192.0 GPixel/s pixel rate, and 384.0 GTexel/s texture rate all exceed the B370's corresponding figures by exact multiples. Its dedicated 512.0 GB/s memory bandwidth on 8 GB of GDDR6 provides a fixed, high-throughput memory subsystem, whereas the B370's system-shared memory offers no quantifiable bandwidth in the database. The A580's 87th percentile ranking and average benchmark score of 57756 give it a concrete performance position among all GPUs, with nearest rivals within one percentage point.

The B370 wins in power efficiency and integration. Its TDP of 25 W is dramatically lower than the A580's 175 W, a sevenfold difference. It requires no power connectors and no suggested PSU, making it suitable for systems where a discrete power delivery path is unavailable. Its IGP bus interface and portable device dependent display outputs indicate a design target of compact, mobile, or embedded platforms. Its smaller resource pool, 1280 shading units, 40 TMUs, 20 ROPs, and 10 ray tracing cores, consumes far less silicon and power than the A580's 3072 shading units, 192 TMUs, 96 ROPs, and 24 ray tracing cores.

The release timeline also separates the two. The A580 launched on 2023-10-09, while the B370 launched on 2026-01-26, making the B370 a newer design by more than two years. The B370 uses a 3 nm Intel process, while the A580 uses a 6 nm TSMC process. The B370 belongs to the Xe3-LPG architecture, a newer lineage than the A580's Xe-HPG.

For gaming and compute workloads with measurable results, the A580 is the only option with data. For ultra-low-power integrated graphics in a portable device, the B370's specifications align with that role. The database records no benchmark for the B370, so no claim of its actual performance can be made beyond its architecture and clock specifications. The A580's recorded scores and percentile rank provide a clear, quantified performance profile. The B370's lack of benchmark data means it cannot be positioned against any rival, including the A580, on measured performance. Its strengths are structural: lower power, integrated form factor, newer process node, and newer architecture generation.