Intel Arc B390 vs Intel Arc Graphics 32EU Comparison

Head-to-Head Benchmarks

The recorded data contains a single direct comparison between the Intel Arc B390 and the Intel Arc Graphics 32EU: the 3DMark Steel Nomad DX12 test. In this benchmark, the Intel Arc B390 scores 1482 points, while the Intel Arc Graphics 32EU scores 733 points. This represents a 102.2% advantage for the B390, meaning it delivers more than double the performance of the 32EU in this specific workload.

The B390's score places it in the 9th percentile of all GPUs in the database. Its nearest rivals are the NVIDIA GeForce GT 520MX (1463 points, 1.3% slower), the NVIDIA GeForce 800M (1460 points, 1.5% slower), the NVIDIA GeForce GT 625 OEM (1446 points, 2.5% slower), and the NVIDIA GeForce GT 710 (1443 points, 2.7% slower). This clustering shows the B390 sits in a narrow performance band, with all four rivals within a 2.7% margin.

The 32EU, by contrast, records a 3rd percentile placement. Its nearest rivals include the Intel Arc Graphics 24EU (733 points, 0% difference) and the Intel Arc Graphics 64EU (733 points, 0% difference), which indicates that the 24EU and 64EU variants produce identical scores in this test. The AMD Radeon HD 6470M trails by 1.4% with 723 points, while the NVIDIA GeForce GT 415M leads the 32EU by 2.4% with 751 points.

The delta between the two Intel parts is substantial. The B390 outperforms the 32EU by more than a factor of two, and the percentile gap (9th versus 3rd) reflects a meaningful tier separation. While both are integrated graphics solutions, the benchmark results indicate the B390 operates in a different performance class entirely.

Architecture Differences

The two processors share the Xe-LPG architecture family but belong to different generations and use different underlying chips. The Intel Arc B390 is built on the Panther Lake chip with the Xe3-LPG architecture, while the Intel Arc Graphics 32EU uses the Arrow Lake-S chip with the Xe-LPG architecture. Both are part of the Arc Graphics-M lineup, but the B390 belongs to the Panther Lake generation and the 32EU to the Arrow Lake generation.

The manufacturing details differ. Both use a 3 nm process node, but the foundry differs: the B390 is fabricated by Intel, while the 32EU is fabricated by TSMC. The 32EU has documented transistor and die specifications: 17,800 million transistors on a 243 mm² die, yielding a transistor density of 73.3M per mm². The B390's transistor count, die size, and density are not recorded in the database.

The compute resources diverge sharply. The B390 contains 1536 shading units, 48 texture mapping units, 24 raster output units, and 12 ray tracing cores. The 32EU contains 256 shading units, 16 TMUs, and 8 ROPs, with no ray tracing cores listed. This represents a 6x difference in shading units, a 3x difference in TMUs, and a 3x difference in ROPs. The B390's ray tracing capability is absent on the 32EU entirely.

Clock behavior also differs. Both have a 300 MHz base clock, but the boost clock reaches 2500 MHz on the B390 versus 1950 MHz on the 32EU. The B390's higher boost clock, combined with its larger execution resources, drives its performance advantage.

Both parts share the same API support: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Memory configuration is also identical in type: system shared memory with system dependent bandwidth and a system shared bus width. The B390 carries an 80 W TDP, while the 32EU carries a 65 W TDP.

Where Each One Wins

The head-to-head data records one win for the Intel Arc B390 and zero wins for the Intel Arc Graphics 32EU. The B390 wins the sole benchmark by a wide margin, so the use-case split follows directly from that result.

The B390 is the choice for any workload where 3D rendering performance matters. Its 1482 Steel Nomad score places it above the GeForce GT 710 and GeForce GT 625 OEM, which are older discrete cards, and it does so while operating as an integrated graphics solution with an 80 W TDP. The 12 ray tracing cores provide hardware acceleration for ray-traced effects, a feature entirely absent on the 32EU. For games or applications that use DirectX 12 Ultimate features, the B390 has both the shading throughput and the ray tracing hardware to handle them.

The 32EU still functions for basic graphics tasks. Its 733 score, while low, is identical to the Intel Arc Graphics 24EU and 64EU, which suggests that this test is not sensitive to the differences between those EU configurations. The 32EU's 65 W TDP is lower than the B390's 80 W, so in power-constrained scenarios where only minimal graphics capability is needed, the 32EU draws less. Its predecessor is listed as HD Graphics-M, which places it in a lineage of basic integrated graphics.

The B390's win is not marginal. A 102.2% delta means the B390 delivers double the frame throughput in the Steel Nomad workload. Any application that taxes the GPU will show a dramatic difference. The 32EU is suitable for desktop, productivity, and light media tasks; the B390 extends into gaming and 3D content at playable settings on modest hardware.

Specification Differences

The following table summarizes the recorded differences between the Intel Arc B390 and Intel Arc Graphics 32EU:

| Field | Intel Arc B390 | Intel Arc Graphics 32EU |

|-------|----------------|--------------------------|

| Chip | Panther Lake | Arrow Lake-S |

| Architecture | Xe3-LPG | Xe-LPG |

| Generation | Arc Graphics-M (Panther Lake) | Arc Graphics-M (Arrow Lake) |

| Foundry | Intel | TSMC |

| Transistors | unknown | 17,800 million |

| Die Size | unknown | 243 mm² |

| Transistor Density | null | 73.3M / mm² |

| Boost Clock | 2500 MHz | 1950 MHz |

| Shading Units | 1536 | 256 |

| TMUs | 48 | 16 |

| ROPs | 24 | 8 |

| Ray Tracing Cores | 12 | null |

| Pixel Rate | 60.00 GPixel/s | 15.60 GPixel/s |

| Texture Rate | 120.0 GTexel/s | 31.20 GTexel/s |

| FP32 | 7.680 TFLOPS | 998.4 GFLOPS |

| FP16 | 15.36 TFLOPS (2:1) | 1.997 TFLOPS (2:1) |

| TDP | 80 W | 65 W |

| Bus Interface | IGP | Ring Bus |

| Display Outputs | Portable Device Dependent | Motherboard Dependent |

| Release Date | 2026-01-26 | 2024-10-23 |

| Predecessor | null | HD Graphics-M |

Fields not listed here are identical between the two parts: base clock (300 MHz), memory type (System Shared), memory size (System Shared), bus width (System Shared), bandwidth (System Dependent), slot width (IGP), DirectX version (12 Ultimate), OpenGL version (4.6), Vulkan version (1.4), and production status (Active).

The B390 also lists power connectors as "None" and a TDP of 80 W, while the 32EU has no power connector data recorded. The B390 was released later, on 2026-01-26, versus 2024-10-23 for the 32EU.

FAQ

Q: Which GPU has a higher benchmark score?

A: The Intel Arc B390 scores 1482 in 3DMark Steel Nomad DX12, while the Intel Arc Graphics 32EU scores 733, giving the B390 a 102.2% advantage.

Q: Do the two GPUs use the same architecture?

A: No. The B390 uses Xe3-LPG on the Panther Lake chip, while the 32EU uses Xe-LPG on the Arrow Lake-S chip. Both are built on a 3 nm process, but the B390 is fabricated by Intel and the 32EU by TSMC.

Q: How do the shading units compare?

A: The B390 has 1536 shading units, 48 TMUs, 24 ROPs, and 12 ray tracing cores. The 32EU has 256 shading units, 16 TMUs, and 8 ROPs, with no ray tracing cores.

Q: What is the boost clock difference?

A: The B390 boosts to 2500 MHz, while the 32EU boosts to 1950 MHz. Both share a 300 MHz base clock.

Q: Which GPU has the lower TDP?

A: The Intel Arc Graphics 32EU has a 65 W TDP, compared to the B390's 80 W TDP.

Q: How does the B390 compare to its nearest rivals?

A: The B390's 1482 score is 1.3% ahead of the GeForce GT 520MX (1463), 1.5% ahead of the GeForce 800M (1460), 2.5% ahead of the GeForce GT 625 OEM (1446), and 2.7% ahead of the GeForce GT 710 (1443).

The Verdict

The benchmark data presents a clear decision between these two integrated graphics solutions. The Intel Arc B390 outperforms the Intel Arc Graphics 32EU by 102.2% in the sole recorded test, and it does so across every major compute metric. The B390's 1536 shading units versus 256, its 12 ray tracing cores versus none, its 2500 MHz boost versus 1950 MHz, and its 7.680 TFLOPS FP32 versus 998.4 GFLOPS all point in the same direction.

The B390's 1482 score places it in the 9th percentile of all GPUs, with nearest rivals including the GeForce GT 710 and GeForce GT 625 OEM. The 32EU's 733 score places it in the 3rd percentile, alongside the Arc Graphics 24EU and 64EU at identical scores. This means the B390 is not just faster than the 32EU; it sits in a different performance tier, one that approaches entry-level discrete graphics from a few generations back.

The 32EU's advantages are limited to power draw (65 W versus 80 W TDP) and earlier availability (released 2024-10-23 versus 2026-01-26). It also has documented transistor and die specifications, while the B390's are unknown. For users who need minimal graphics acceleration with the lowest power footprint, the 32EU serves that role.

For anyone who plans to run 3D applications, games, or GPU-accelerated workloads, the data supports choosing the B390. The 102.2% performance delta is not a marginal improvement; it is a doubling of throughput. The presence of ray tracing cores on the B390 adds a hardware capability that the 32EU lacks entirely. The B390 also carries a higher boost clock and significantly more texture and pixel throughput, with texture rate at 120.0 GTexel/s versus 31.20 GTexel/s and pixel rate at 60.00 GPixel/s versus 15.60 GPixel/s.

The percentile data reinforces the separation. A 9th percentile placement versus 3rd percentile means the B390 lands above 90% of all GPUs in the database, while the 32EU lands above only 3%. The nearest rivals for each part confirm the gap: the B390 competes with low-end discrete GeForce parts, while the 32EU competes with other integrated and legacy mobile parts.

The choice depends on the workload. The 32EU is a basic display and media accelerator, adequate for desktop use and light tasks. The B390 is a capable integrated GPU that can handle more demanding 3D workloads, with the ray tracing hardware and compute throughput to justify its higher power draw. For a system builder choosing between the two, the B390 delivers the performance, and the 32EU delivers the efficiency.