Intel Arc Graphics 64EU Mobile vs Intel Arc Pro B390 Comparison

Head-to-Head Benchmarks

The database contains no recorded head-to-head benchmark results between the Intel Arc Graphics 64EU Mobile and the Intel Arc Pro B390. Both entries have empty benchmark arrays, zero average benchmark scores, and no nearest rival data. This means a direct performance comparison must be derived entirely from the recorded specifications, which show a substantial gap in capability.

The most decisive difference appears in raw compute throughput. The Arc Pro B390 delivers 7.680 TFLOPS of FP32 performance, while the Arc Graphics 64EU Mobile manages 1.792 TFLOPS. That places the B390 at roughly 4.3 times the FP32 output of the mobile part. The FP16 figures follow the same pattern: 15.36 TFLOPS for the B390 versus 3.584 TFLOPS for the 64EU Mobile, again a 4.3x advantage. These numbers represent the peak theoretical throughput, and they set the ceiling for anything the shader array can do.

Texture and pixel processing show a similar hierarchy. The B390 reaches 120.0 GTexel/s and 60.00 GPixel/s, while the 64EU Mobile stops at 56.00 GTexel/s and 28.00 GPixel/s. The B390 doubles the pixel fill rate and more than doubles the texture fill rate. This directly affects rasterization-heavy workloads, where the B390 can push more geometry and fill more screen area per unit time.

Clock speeds also favor the B390. Both parts share a 300 MHz base clock, but the boost clock differs sharply: 2500 MHz for the B390 versus 1750 MHz for the 64EU Mobile. That 750 MHz gap contributes to the throughput advantage, though the larger shader count is the primary driver.

Shader resources are heavily skewed. The B390 packs 1536 shading units, 48 TMUs, and 24 ROPs. The 64EU Mobile carries 512 shading units, 32 TMUs, and 16 ROPs. The B390 has triple the shading units, 1.5x the texture mapping units, and 1.5x the render output units. These are not marginal differences; they define entirely different performance classes.

Neither part has recorded benchmark wins in the database. The winsA and winsB fields are both zero. This is an important caveat: the specification deltas are clear, but there are no measured application or game results to confirm how those specs translate into real-world performance. The data indicates the B390 should dominate in every measurable category, but the absence of benchmark scores leaves some uncertainty about driver efficiency, thermal behavior, and workload-specific scaling.

The percentile ranking is identical for both: 50th percentile among all GPUs. This is a coarse measure, and it does not reflect the large specification gap. The database may not yet have enough measured data to differentiate them, or the percentile field may be a placeholder pending benchmark submissions.

FAQ

Q: Which GPU has higher FP32 performance?

A: The Intel Arc Pro B390 records 7.680 TFLOPS, which is 4.3 times the 1.792 TFLOPS of the Intel Arc Graphics 64EU Mobile.

Q: Do both GPUs use the same architecture?

A: No. The 64EU Mobile uses Xe-LPG architecture on a 10 nm process, while the B390 uses Xe3-LPG architecture on a 3 nm process.

Q: What is the boost clock difference?

A: The B390 boosts to 2500 MHz, while the 64EU Mobile boosts to 1750 MHz. Both have a 300 MHz base clock.

Q: Which GPU has more shading units?

A: The B390 has 1536 shading units, triple the 512 shading units of the 64EU Mobile.

Q: Do either GPUs support ray tracing hardware?

A: The B390 includes 12 RT cores. The 64EU Mobile does not list any RT core count in the database.

Q: What is the memory configuration for both?

A: Both use system shared memory with system dependent bandwidth. Neither has dedicated VRAM.

Architecture Differences

The two GPUs represent different architectural generations and process nodes. The Intel Arc Graphics 64EU Mobile is built on Meteor Lake silicon using Xe-LPG architecture, fabricated on Intel's 10 nm process. The Intel Arc Pro B390 uses Panther Lake silicon with Xe3-LPG architecture, fabricated on a 3 nm process. The process node jump from 10 nm to 3 nm is a major generational shift, allowing the B390 to pack far more transistors into a similar power envelope.

The B390 adds ray tracing hardware, with 12 RT cores explicitly listed. The 64EU Mobile has no RT core count recorded, indicating it lacks dedicated ray tracing acceleration. This is a functional difference, not just a performance gap. Workloads that use ray tracing will either run on the B390's dedicated hardware or fall back to compute shaders on the 64EU Mobile, which is significantly slower.

DirectX support differs as well. The 64EU Mobile supports DirectX 12 (12_1), while the B390 supports DirectX 12 Ultimate (12_2). The 12_2 feature level includes additional capabilities such as mesh shaders, variable rate shading, and enhanced ray tracing support. Both support OpenGL 4.6 and Vulkan 1.4, so those API paths are identical.

The B390's larger shader array and higher clocks are enabled by the denser 3 nm process. The 64EU Mobile's 512 shading units and 1750 MHz boost represent a more conservative design aimed at lower power consumption. The TDP figures reflect this: 80 W for the B390 versus 65 W for the 64EU Mobile. The B390 delivers roughly 4.3x the FP32 throughput while drawing only 15 W more, a substantial efficiency improvement.

Both parts share the same memory architecture: system shared, with system dependent bandwidth. This means neither has dedicated VRAM, and performance will depend heavily on the host system's memory speed and capacity. The B390's higher compute throughput may be constrained by shared memory bandwidth in some workloads, but the database does not provide separate bandwidth figures for either.

The bus interface differs. The 64EU Mobile uses a Ring Bus, while the B390 uses IGP. Both are integrated graphics parts with a slot width of IGP, meaning they are designed to be embedded in a processor rather than installed as discrete cards. Display outputs are portable device dependent for both, confirming their mobile and integrated positioning.

Specification Differences

The following fields differ between the two parts:

• Chip: Meteor Lake versus Panther Lake
• Architecture: Xe-LPG versus Xe3-LPG
• Generation: Arc Graphics-M (Meteor Lake) versus Arc Graphics-WM (Panther Lake)
• Process Node: 10 nm versus 3 nm
• Boost Clock: 1750 MHz versus 2500 MHz
• Shading Units: 512 versus 1536
• TMUs: 32 versus 48
• ROPs: 16 versus 24
• RT Cores: None listed versus 12
• Pixel Rate: 28.00 GPixel/s versus 60.00 GPixel/s
• Texture Rate: 56.00 GTexel/s versus 120.0 GTexel/s
• FP32: 1.792 TFLOPS versus 7.680 TFLOPS
• FP16: 3.584 TFLOPS versus 15.36 TFLOPS
• TDP: 65 W versus 80 W
• Power Connectors: None listed versus None
• Bus Interface: Ring Bus versus IGP
• DirectX: 12 (12_1) versus 12 Ultimate (12_2)
• Release Date: 2023-12-13 versus 2026-01-26
• Predecessor: HD Graphics-M versus HD Graphics-WM

Fields that are identical: manufacturer (Intel), base clock (300 MHz), memory size (system shared), memory type (system shared), bus width (system shared), bandwidth (system dependent), OpenGL version (4.6), Vulkan version (1.4), production status (Active), slot width (IGP), and display outputs (portable device dependent). Neither part has a launch MSRP recorded.

The release date gap is notable: the 64EU Mobile launched in December 2023, while the B390 is dated January 2026. This places them in different product generations, with the B390 arriving more than two years later. The predecessor names also differ, with HD Graphics-M preceding the mobile part and HD Graphics-WM preceding the B390.

The Verdict

The recorded data shows a clear performance hierarchy. The Intel Arc Pro B390 is the stronger GPU in every measurable specification. It has triple the shading units, 1.5x the TMUs, 1.5x the ROPs, a 750 MHz higher boost clock, and a 4.3x advantage in both FP32 and FP16 throughput. It also adds 12 RT cores and supports DirectX 12 Ultimate, while the 64EU Mobile has no RT hardware and caps at DirectX 12 (12_1).

The B390's 3 nm process node gives it a significant efficiency advantage. It draws 80 W versus 65 W for the 64EU Mobile, yet delivers more than four times the compute throughput. The 15 W difference is small relative to the performance gap. This makes the B390 the better choice for any workload that can use the additional shader throughput, texture fill, or ray tracing capability.

The 64EU Mobile's case rests on its lower TDP and earlier release. At 65 W, it is better suited for thin-and-light portable devices where power budgets are tight. Its 512 shading units and 1750 MHz boost are sufficient for basic graphics acceleration, media playback, and light productivity. But there is no benchmark data in the database to suggest it competes with the B390 in compute-heavy tasks.

The absence of head-to-head benchmarks means this verdict is based entirely on specification analysis. The database shows no measured wins for either part, so the conclusions here rely on the recorded architectural and clock differences. Those differences are large enough that the B390 should win in almost any scenario that stresses the GPU.

For users who need maximum integrated graphics performance, ray tracing support, and modern API features, the B390 is the clear choice. For users who prioritize minimum power draw and are working with older, lighter software, the 64EU Mobile remains a viable option, but it is not in the same performance class.

Where Each One Wins

Intel Arc Pro B390 wins in: raw compute throughput, with 4.3x higher FP32 and FP16 performance. It wins in pixel fill rate at 60.00 GPixel/s versus 28.00 GPixel/s, and texture fill rate at 120.0 GTexel/s versus 56.00 GTexel/s. It wins in ray tracing workloads, as it is the only part with RT cores. It wins in modern DirectX 12 Ultimate applications, given its 12_2 feature level. It wins in any workload that scales with shading units, since it has triple the count. It wins in efficiency per watt, delivering 4.3x the performance at only 15 W higher TDP. It wins in clock speed headroom with a 2500 MHz boost versus 1750 MHz.

Intel Arc Graphics 64EU Mobile wins in: power draw, at 65 W versus 80 W. It wins in compatibility with older software that does not require DirectX 12 Ultimate features. It wins in scenarios where the host system's memory bandwidth is the limiting factor, as its lower compute throughput is less likely to saturate shared memory. It wins in release timing for systems designed before 2026, as it launched in December 2023. It wins in any workload that is not GPU-bound, where its lower power footprint leaves more thermal headroom for the CPU. It is also the only part with a Ring Bus interface, which may integrate differently with certain host systems.

The B390 wins in every performance category recorded in the database. The 64EU Mobile wins only in power efficiency and release timing. There is no benchmark data suggesting the mobile part outperforms the B390 in any application. The choice reduces to whether the 15 W power savings matter more than the 4.3x compute advantage, and for most users the B390's additional capability will be the deciding factor.