Intel Arc Pro A60 vs Intel Arc Pro B390 Comparison

The Verdict

The data presents a clear split between two Intel Arc Pro parts with very different design goals. The Intel Arc Pro A60 is a discrete, single-slot workstation GPU built on the Xe-HPG architecture with a 6 nm TSMC process, while the Intel Arc Pro B390 is an integrated graphics processor (IGP) part of the Arc Graphics-WM (Panther Lake) generation, built on Intel's 3 nm process. The A60 targets users who need a dedicated, high-throughput compute device with substantial memory bandwidth. The B390, meanwhile, serves as a power-efficient integrated solution for portable devices where space and power are at a premium.

Benchmark data in the database shows the A60 has a recorded average benchmark score of 60326, placing it in the 88th percentile of all GPUs. Its nearest rivals in the database include the NVIDIA GeForce RTX 4090 (60347, 0% delta), the AMD Radeon Pro W6600M (61896, -2.5% delta), the AMD Radeon Pro Vega 48 (60140, 0.3% delta), and the AMD Radeon PRO V710 (58657, 2.8% delta). These figures indicate the A60 sits in a competitive performance band for professional workloads, essentially matching the RTX 4090 in the database's average scoring, though the RTX 4090's score is likely dominated by different workload characteristics. The B390 has no recorded benchmarks and no nearest rivals in the database, with a 50th percentile ranking, which places it as a mid-tier performer among all GPUs but without any direct comparative data.

For a user deciding between these two, the choice is stark. The A60 is the only one with measurable performance data, and it is the obvious pick for anyone needing a dedicated GPU with 12 GB of GDDR6 memory and 384.0 GB/s of bandwidth. The B390, with system-shared memory and no dedicated VRAM, is designed for a fundamentally different use case: integrated graphics for portable or low-power systems, where the 80 W TDP and lack of discrete power connectors are defining features. The verdict from the data is that the A60 is the performance option, while the B390 is the efficiency and integration option, and they do not overlap in their intended roles.

Where Each One Wins

The Intel Arc Pro A60 wins decisively in every measured performance category in the database. Its FP32 compute rating is 8.397 TFLOPS, which is 9.3% higher than the B390's 7.680 TFLOPS. Its texture rate is 262.4 GTexel/s, more than double the B390's 120.0 GTexel/s. Its pixel rate is 131.2 GPixel/s versus 60.00 GPixel/s for the B390, a 118.7% advantage. The A60 has 2048 shading units against 1536 for the B390, 128 TMUs versus 48, 64 ROPs versus 24, and 16 ray tracing cores versus 12. Memory bandwidth is a categorical victory: the A60 has 384.0 GB/s of dedicated GDDR6 bandwidth on a 192-bit bus, while the B390 relies on system-shared memory with bandwidth that is system dependent.

The B390 wins in efficiency and integration metrics. Its TDP is 80 W, which is 38.5% lower than the A60's 130 W. It is an IGP, meaning it requires no power connectors, no discrete slot, and its display outputs are portable device dependent. The B390 also has a higher boost clock at 2500 MHz versus 2050 MHz for the A60, and it is built on a more advanced 3 nm Intel process versus the 6 nm TSMC node of the A60. The B390's base clock is much lower at 300 MHz versus 900 MHz, but the boost clock advantage suggests it can scale higher when thermal and power conditions allow, which is typical for integrated parts in portable devices.

In terms of raw workload execution, the A60 is the clear winner for compute-intensive tasks. The B390 wins in scenarios where power draw, physical footprint, and system integration are the primary constraints. There are no head-to-head benchmark records in the database, so all comparisons here are derived from the individual specifications and the A60's recorded benchmark scores.

Architecture Differences

The two GPUs represent different architectural generations and design philosophies from Intel. The Intel Arc Pro A60 uses the Xe-HPG architecture with the DG2-256 chip, classified as part of the Alchemist (Pro Series) generation. It is fabricated on a 6 nm process at TSMC, with 11,500 million transistors on a 269 mm² die, yielding a transistor density of 42.8 million transistors per mm². The B390 uses the Xe3-LPG architecture with the Panther Lake chip, part of the Arc Graphics-WM (Panther Lake) generation. It is fabricated on Intel's 3 nm process, though the database does not list its transistor count or die size.

The memory subsystems are fundamentally different. The A60 has 12 GB of GDDR6 memory on a 192-bit bus, with a memory clock of 2000 MHz (16 Gbps effective) and a bandwidth of 384.0 GB/s. The B390 has no dedicated memory; it uses system-shared memory, with the type, bus width, and bandwidth all listed as system shared or system dependent. This is the defining architectural difference: the A60 is a discrete GPU with its own VRAM, while the B390 is an integrated part that borrows from the host system's memory.

The compute configurations differ notably. The A60 has 2048 shading units, 128 TMUs, 64 ROPs, and 16 ray tracing cores. The B390 has 1536 shading units, 48 TMUs, 24 ROPs, and 12 ray tracing cores. The A60's pixel rate is 131.2 GPixel/s and its texture rate is 262.4 GTexel/s. The B390's pixel rate is 60.00 GPixel/s and its texture rate is 120.0 GTexel/s. FP32 throughput is 8.397 TFLOPS for the A60 and 7.680 TFLOPS for the B390. Both GPUs support FP16 at a 2:1 ratio, with 16.79 TFLOPS for the A60 and 15.36 TFLOPS for the B390.

Both parts support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The A60 uses a PCIe 4.0 x16 bus interface and has four DisplayPort 2.0 outputs. The B390 is an IGP with no discrete bus interface and display outputs that are portable device dependent. The A60 has a suggested PSU of 300 W and a TDP of 130 W, while the B390 has an 80 W TDP and no power connectors. The A60's release date is 2023-06-05, while the B390's release date is 2026-01-26. The B390's predecessor is listed as the HD Graphics-WM, while the A60 has no predecessor or successor listed. Both are production status active.

FAQ

Q: Which GPU has higher FP32 compute performance?

A: The Intel Arc Pro A60 delivers 8.397 TFLOPS compared to 7.680 TFLOPS for the Intel Arc Pro B390, a difference of roughly 9.3% in favor of the A60.

Q: Does the Intel Arc Pro B390 have dedicated video memory?

A: No. The B390 uses system-shared memory for its memory size, type, bus width, and bandwidth. Its bandwidth is listed as system dependent, whereas the A60 has 12 GB of GDDR6 with 384.0 GB/s of dedicated bandwidth.

Q: How do the power requirements compare?

A: The A60 has a 130 W TDP and a suggested PSU of 300 W. The B390 has an 80 W TDP, no power connectors, and no suggested PSU listed, consistent with its IGP classification.

Q: What is the process node difference?

A: The A60 uses a 6 nm process at TSMC with 11,500 million transistors on a 269 mm² die. The B390 uses Intel's 3 nm process, but its transistor count and die size are not listed in the database.

Q: Which GPU has more shading units and ray tracing cores?

A: The A60 has 2048 shading units and 16 ray tracing cores. The B390 has 1536 shading units and 12 ray tracing cores.

Q: Are there any recorded benchmark scores for the B390?

A: No. The database shows no benchmarks for the B390 and no nearest rivals. The A60 has recorded scores of 63485 in Geekbench OpenCL and 57166 in Geekbench Vulkan, with an average benchmark score of 60326.

Head-to-Head Benchmarks

There are no direct head-to-head benchmark records in the database between the Intel Arc Pro A60 and the Intel Arc Pro B390. The B390 has no benchmark entries at all, so a direct comparison of measured scores is not possible. However, the recorded data for the A60 and the listed specifications for both parts allow for a comparative analysis of their expected performance.

The A60's average benchmark score of 60326 places it at the 88th percentile of all GPUs in the database. Its nearest rival comparisons show it within 2.8% of the AMD Radeon PRO V710 (58657, 2.8% delta), essentially tied with the AMD Radeon Pro Vega 48 (60140, 0.3% delta) and the NVIDIA GeForce RTX 4090 (60347, 0% delta), and 2.5% behind the AMD Radeon Pro W6600M (61896, -2.5% delta). These figures indicate the A60 is a solid mid-to-high performer in the database's ranking system, competitive with a range of professional and high-end consumer GPUs.

The B390, with no benchmarks and a 50th percentile ranking, cannot be compared on the same scale. Its specifications do provide some insight. The B390's FP32 throughput is 7.680 TFLOPS, which is 91.5% of the A60's 8.397 TFLOPS. This is a relatively small gap in raw compute, but the memory situation is the differentiator. The A60's 384.0 GB/s of dedicated bandwidth is a massive advantage over a system-shared memory configuration, where bandwidth is dependent on the host system's memory architecture. For workloads that are memory bound, the A60 would be expected to outperform the B390 by a significant margin.

The texture rate comparison is stark: the A60's 262.4 GTexel/s is 2.19 times the B390's 120.0 GTexel/s. The pixel rate shows a similar pattern, with the A60 at 131.2 GPixel/s versus 60.00 GPixel/s for the B390, a 2.19x difference as well. These ratios reflect the A60's 128 TMUs and 64 ROPs versus 48 TMUs and 24 ROPs on the B390. The shading unit count is closer in relative terms: 2048 versus 1536, a 1.33x ratio, which aligns with the smaller gap in FP32 throughput.

The A60's boost clock is 2050 MHz, while the B390's boost clock is 2500 MHz. The B390's higher boost clock partially compensates for its lower unit counts, but the A60's larger number of execution resources still gives it the overall throughput advantage. The A60's base clock of 900 MHz versus 300 MHz for the B390 also indicates that the A60 is designed for sustained performance, while the B390's low base clock suggests a power-saving idle or low-load state.

Given the absence of B390 benchmark scores, the biggest wins in this comparison are derived from the A60's recorded performance data and the specification deltas. The A60 wins on every measured throughput metric: FP32, FP16, texture rate, pixel rate, shading units, TMUs, ROPs, and ray tracing cores. The B390 wins on power efficiency, with an 80 W TDP versus 130 W, and on process technology, with a 3 nm Intel node versus a 6 nm TSMC node. The data shows that for any workload where the A60's dedicated memory and higher throughput can be utilized, it is the superior part. The B390's advantages are in integration and power, making it suitable for portable devices where those factors are the primary constraints.