AMD Radeon PRO W7400 vs Intel Arc B390 Comparison

Where Each One Wins

The benchmark data splits these two GPUs into entirely different performance strata. The AMD Radeon PRO W7400 occupies the 50th percentile among all GPUs in the database, placing it squarely in the mid-range of recorded hardware. The Intel Arc B390, by contrast, sits at the 9th percentile, a position that places it among the lowest-performing accelerators the database tracks.

The Intel Arc B390 has one recorded benchmark result: a 3DMark Steel Nomad DX12 score of 1482. That figure positions it within 1.3% of the NVIDIA GeForce GT 520MX (1463), 1.5% of the GeForce 800M (1460), 2.5% of the GeForce GT 625 OEM (1446), and 2.7% of the GeForce GT 710 (1443). These are narrow margins, all within three percentage points, indicating that the Arc B390 trades blows with a cluster of legacy entry-level discrete GPUs.

The AMD Radeon PRO W7400 has no recorded benchmark entries in the database. Its wins must therefore be assessed from its architectural specifications and computed throughput rates rather than direct test scores. The absence of benchmark data for the Radeon PRO W7400 means the database cannot confirm a measured win in any synthetic test for that part.

Where the Intel part wins is in its measured 3DMark Steel Nomad score, which outperforms all four nearest rivals listed in the database. The 1.3% lead over the GT 520MX is marginal, but it is a recorded win nonetheless. The Arc B390 also shows a clear advantage in FP16 compute: 15.36 TFLOPS versus the Radeon PRO W7400's 7.885 TFLOPS, a 2:1 ratio that reflects the Intel architecture's dedicated FP16 throughput path.

The Radeon PRO W7400 wins on raw FP32 compute, delivering 7.885 TFLOPS against the Arc B390's 7.680 TFLOPS. That is a slim 0.27% edge, but it is a lead. The AMD part also wins on pixel fill rate (70.40 GPixel/s versus 60.00 GPixel/s) and texture fill rate (123.2 GTexel/s versus 120.0 GTexel/s). These are small margins, but they point to the Radeon's advantage in rasterization-bound workloads.

The practical split is clear. The Intel Arc B390 is designed for integrated graphics duty in a mobile or portable context, with its performance profile clustering near decade-old discrete GPUs. The AMD Radeon PRO W7400 is a professional workstation card with a 55 W TDP and a single-slot form factor, aimed at tasks where sustained FP32 throughput and DisplayPort 2.1 outputs matter more than a single synthetic benchmark score.

Architecture Differences

The two GPUs come from different foundries and process nodes. The AMD Radeon PRO W7400 uses the Navi 33 chip built on TSMC's 6 nm process, housing 13,300 million transistors on a 204 mm² die. That works out to a transistor density of 65.2M per mm². The Intel Arc B390 uses the Panther Lake chip on Intel's 3 nm process; the database records no transistor count, die size, or density for this part.

Architecturally, the Radeon PRO W7400 is built on RDNA 3.0 under the codename Hotpink Bonefish, part of the Radeon Pro Navi (Navi III Series) generation. The Arc B390 uses Xe3-LPG architecture and belongs to the Arc Graphics-M (Panther Lake) generation. These are fundamentally different design philosophies: RDNA 3.0 is a discrete GPU architecture optimized for rasterization throughput, while Xe3-LPG is an integrated graphics architecture intended to share system resources.

The Radeon PRO W7400 has 1792 shading units, 112 texture mapping units, 64 ROPs, and 28 ray tracing cores. The Arc B390 has 1536 shading units, 48 TMUs, 24 ROPs, and 12 ray tracing cores. The AMD part fields more of every execution unit type, with a 1.17x lead in shading units, a 2.33x lead in TMUs, a 2.67x lead in ROPs, and a 2.33x lead in ray tracing cores.

Neither part lists tensor cores in the database. The Radeon PRO W7400 delivers FP16 at a 1:1 ratio with FP32, meaning both peak at 7.885 TFLOPS. The Arc B390 delivers FP16 at a 2:1 ratio, reaching 15.36 TFLOPS while its FP32 peak is 7.680 TFLOPS. This indicates the Intel architecture uses a packed FP16 path that doubles throughput on half-precision workloads, a common approach in integrated GPUs targeting AI inference on shared memory.

Memory architecture differs fundamentally. The Radeon PRO W7400 has 8 GB of dedicated GDDR6 memory on a 128-bit bus, delivering 172.8 GB/s of bandwidth. The Arc B390 uses system shared memory with a system-dependent bus width and bandwidth; the database records no fixed figures for these fields. The clock speeds reflect this difference: the Radeon runs its memory at 1350 MHz (10.8 Gbps effective), while the Arc B390 has no dedicated memory clock.

Both parts support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The Radeon PRO W7400 outputs through 4x DisplayPort 2.1, while the Arc B390's display outputs are listed as portable device dependent. The Radeon uses a PCIe 4.0 x8 bus interface; the Arc B390 is an IGP with no external bus interface.

Head-to-Head Benchmarks

The database contains no direct head-to-head benchmark entries between these two GPUs. The only recorded score belongs to the Intel Arc B390, which logged 1482 in 3DMark Steel Nomad DX12. That score places it 1.3% above the nearest rival in its comparison set, the NVIDIA GeForce GT 520MX at 1463.

In the absence of a shared test, the comparison must draw on computed throughput rates. The Radeon PRO W7400 leads FP32 compute by 0.27%, delivering 7.885 TFLOPS versus 7.680 TFLOPS. This is a narrow but consistent advantage in single-precision workloads, which dominate traditional graphics rendering.

The Arc B390 leads FP16 compute by 94.9%, delivering 15.36 TFLOPS versus 7.885 TFLOPS. This is a substantial margin, but it only matters for workloads that can exploit half-precision arithmetic. The Radeon's 1:1 FP16 ratio means it cannot accelerate half-precision tasks beyond its FP32 rate, while the Arc's 2:1 ratio doubles its throughput for such workloads.

Pixel fill rate favors the Radeon PRO W7400 at 70.40 GPixel/s versus 60.00 GPixel/s, a 17.3% lead. Texture fill rate favors the Radeon at 123.2 GTexel/s versus 120.0 GTexel/s, a 2.7% lead. These rates derive from the Radeon's higher ROP and TMU counts combined with its boost clock of 1100 MHz versus the Arc's 2500 MHz. The Arc's much higher clock speed (2500 MHz boost versus 1100 MHz) compensates partially for its fewer execution units, but not enough to overcome the Radeon's larger resource pool.

The Arc B390's 3DMark score of 1482, when set against its nearest rivals, shows a performance envelope that the Radeon PRO W7400's 50th percentile position suggests it would exceed substantially. The Radeon's percentile places it above half of all GPUs in the database, while the Arc's 9th percentile places it below 91% of recorded parts. The delta between these percentile positions implies a large gap in measured performance, even though no direct benchmark links the two.

Specification Differences

| Specification | AMD Radeon PRO W7400 | Intel Arc B390 |

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

| Manufacturer | AMD | Intel |

| Chip | Navi 33 | Panther Lake |

| Architecture | RDNA 3.0 | Xe3-LPG |

| Codename | Hotpink Bonefish | None listed |

| Generation | Radeon Pro Navi (Navi III Series) | Arc Graphics-M (Panther Lake) |

| Process Node | 6 nm | 3 nm |

| Foundry | TSMC | Intel |

| Transistors | 13,300 million | Unknown |

| Die Size | 204 mm² | Unknown |

| Transistor Density | 65.2M / mm² | Not listed |

| Base Clock | 330 MHz | 300 MHz |

| Boost Clock | 1100 MHz | 2500 MHz |

| Memory Clock | 1350 MHz 10.8 Gbps effective | System Shared |

| Memory Size | 8 GB | System Shared |

| Memory Type | GDDR6 | System Shared |

| Memory Bus Width | 128 bit | System Shared |

| Memory Bandwidth | 172.8 GB/s | System Dependent |

| Shading Units | 1792 | 1536 |

| TMUs | 112 | 48 |

| ROPs | 64 | 24 |

| Ray Tracing Cores | 28 | 12 |

| Pixel Rate | 70.40 GPixel/s | 60.00 GPixel/s |

| Texture Rate | 123.2 GTexel/s | 120.0 GTexel/s |

| FP32 | 7.885 TFLOPS | 7.680 TFLOPS |

| FP16 | 7.885 TFLOPS (1:1) | 15.36 TFLOPS (2:1) |

| TDP | 55 W | 80 W |

| Slot Width | Single-slot | IGP |

| Power Connectors | None | None |

| Suggested PSU | 250 W | Not listed |

| Bus Interface | PCIe 4.0 x8 | IGP |

| Display Outputs | 4x DisplayPort 2.1 | Portable Device Dependent |

| Length | 168 mm 6.6 inches | Not listed |

| Height | 69 mm 2.7 inches | Not listed |

| Width | 20 mm 0.8 inches | Not listed |

| Release Date | 2025-08-02 | 2026-01-26 |

| Predecessor | Radeon Pro Vega | None listed |

FAQ

Q: Which GPU has more shading units?

A: The AMD Radeon PRO W7400 has 1792 shading units, while the Intel Arc B390 has 1536 shading units. The Radeon leads by 256 units, a 1.17x advantage.

Q: How do the FP32 compute figures compare?

A: The Radeon PRO W7400 delivers 7.885 TFLOPS of FP32 compute, slightly ahead of the Arc B390's 7.680 TFLOPS. This is a 0.27% lead for the AMD part.

Q: What does the Arc B390's 3DMark Steel Nomad score indicate?

A: The Arc B390 scored 1482 in 3DMark Steel Nomad DX12. That score sits 1.3% above the NVIDIA GeForce GT 520MX (1463), 1.5% above the GeForce 800M (1460), 2.5% above the GeForce GT 625 OEM (1446), and 2.7% above the GeForce GT 710 (1443).

Q: Why does the Arc B390 have higher FP16 throughput?

A: The Arc B390 uses a 2:1 FP16 ratio, reaching 15.36 TFLOPS, while the Radeon PRO W7400 uses a 1:1 ratio, capping FP16 at 7.885 TFLOPS. The Intel architecture processes two half-precision operations per clock per unit, doubling its throughput for such workloads.

Q: What are the memory configurations?

A: The Radeon PRO W7400 has 8 GB of GDDR6 memory on a 128-bit bus with 172.8 GB/s of bandwidth. The Arc B390 uses system shared memory with system-dependent bus width and bandwidth; the database records no fixed values for these fields.

Q: Which GPU has a higher boost clock?

A: The Intel Arc B390 boosts to 2500 MHz, more than double the Radeon PRO W7400's 1100 MHz boost clock. The Radeon's base clock of 330 MHz is slightly higher than the Arc's 300 MHz base clock.