AMD Radeon AI PRO 9600D vs Intel Arc B390 Comparison

The Verdict

The data in this database places the AMD Radeon AI PRO 9600D and Intel Arc B390 in entirely different performance strata. The AMD part scores in the 50th percentile of all GPUs, while the Intel Arc B390 sits in the 9th percentile, based on its recorded 3DMark Steel Nomad DX12 score of 1482. The AMD Radeon AI PRO 9600D is a discrete, single-slot add-in card designed for professional workstations, whereas the Intel Arc B390 is an integrated graphics processor (IGP) for portable devices.

For any workload that relies on raw compute, rasterization, or ray tracing, the AMD Radeon AI PRO 9600D is the only viable choice. Its FP32 throughput of 24.82 TFLOPS is more than triple the Intel part's 7.680 TFLOPS. The AMD card also provides 32 GB of dedicated GDDR6 memory with 576.0 GB/s of bandwidth, while the Intel Arc B390 depends entirely on system shared memory, making its performance dependent on the host platform.

The Intel Arc B390 is positioned for basic graphics duties in a compact, low-power package. Its 80 W TDP and lack of any power connectors indicate a design for thin-and-light systems where the AMD card's 150 W TDP and single 16-pin connector would be impractical. The benchmark data shows the Intel part is only 1.3% ahead of the NVIDIA GeForce GT 520MX in 3DMark Steel Nomad DX12, a result that places it in the company of legacy entry-level discrete GPUs. The AMD Radeon AI PRO 9600D has no recorded benchmark scores in this database, but its hardware specifications, including 3072 shading units and a 256-bit memory bus, place it far above that performance class.

FAQ

Q: Which GPU has higher FP32 compute performance?

A: The AMD Radeon AI PRO 9600D delivers 24.82 TFLOPS of FP32 performance, while the Intel Arc B390 delivers 7.680 TFLOPS. This gives the AMD part more than three times the single-precision compute throughput.

Q: How do the memory subsystems differ?

A: The AMD Radeon AI PRO 9600D uses 32 GB of dedicated GDDR6 memory on a 256-bit bus, providing 576.0 GB/s of bandwidth. The Intel Arc B390 uses system shared memory with system-dependent bandwidth, which means its memory performance relies entirely on the host system's configuration.

Q: What is the performance ranking of the Intel Arc B390?

A: The Intel Arc B390 records a 3DMark Steel Nomad DX12 score of 1482, placing it in the 9th percentile of all GPUs. Its nearest rivals are the NVIDIA GeForce GT 520MX (1.3% slower), NVIDIA GeForce 800M (1.5% slower), NVIDIA GeForce GT 625 OEM (2.5% slower), and NVIDIA GeForce GT 710 (2.7% slower).

Q: What are the power requirements for each GPU?

A: The AMD Radeon AI PRO 9600D has a 150 W TDP and requires a 450 W suggested power supply, using a single 16-pin connector. The Intel Arc B390 has an 80 W TDP, uses no power connectors, and has no suggested PSU because it is an integrated graphics processor.

Q: What process nodes do these chips use?

A: The AMD Radeon AI PRO 9600D uses a 4 nm process at TSMC, with 53,900 million transistors on a 357 mm² die. The Intel Arc B390 uses a 3 nm process at Intel, but its transistor count and die size are recorded as unknown in the database.

Q: Which GPU supports which APIs?

A: Both GPUs support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The AMD Radeon AI PRO 9600D is based on RDNA 4.0 architecture with 48 ray tracing cores, while the Intel Arc B390 uses Xe3-LPG architecture with 12 ray tracing cores.

Architecture Differences

The AMD Radeon AI PRO 9600D is built on the RDNA 4.0 architecture, specifically the Navi 48 chip from the Radeon Pro Navi (Navi IV Series) generation. This is a discrete GPU design manufactured on a 4 nm process at TSMC. The chip integrates 53,900 million transistors across a 357 mm² die, yielding a transistor density of 151.0M per mm². The architecture uses a 1:1 ratio for FP16 and FP32 performance, with both rated at 24.82 TFLOPS.

The Intel Arc B390 uses the Xe3-LPG architecture, specifically the Panther Lake chip from the Arc Graphics-M (Panther Lake) generation. This is an integrated graphics processor manufactured on a 3 nm process at Intel's own foundry. The transistor count and die size are listed as unknown in the database. The FP16 performance is 15.36 TFLOPS at a 2:1 ratio relative to FP32, meaning the FP16 throughput is double the FP32 throughput of 7.680 TFLOPS.

The AMD part features 3072 shading units, 192 texture mapping units, 96 render output units, and 48 ray tracing cores. The Intel part has 1536 shading units, 48 TMUs, 24 ROPs, and 12 ray tracing cores. In every compute and graphics pipeline stage, the AMD Radeon AI PRO 9600D has substantially more hardware resources.

Pixel and texture throughput further separate the two architectures. The AMD card achieves 193.9 GPixel/s and 387.8 GTexel/s, while the Intel IGP achieves 60.00 GPixel/s and 120.0 GTexel/s. The AMD card's pixel rate is more than triple the Intel part's, and its texture rate is also more than triple.

The memory architecture is fundamentally different. The AMD card uses dedicated GDDR6 on a 256-bit interface with fixed bandwidth of 576.0 GB/s. The Intel part uses system shared memory, with a bus width and bandwidth both marked as "System Shared" and "System Dependent" respectively. This means the Intel Arc B390's memory performance varies with the host platform, a characteristic typical of integrated graphics.

Specification Differences

The specification sheet shows clear divergence in nearly every measurable field.

The AMD Radeon AI PRO 9600D has a base clock of 1080 MHz and a boost clock of 2020 MHz, with the game clock also listed at 1080 MHz. Memory runs at 2250 MHz with 18 Gbps effective speed. The Intel Arc B390 has a base clock of 300 MHz and a boost clock of 2500 MHz, but no separate game clock is listed. Its memory clock is marked as "System Shared."

Memory capacity differs dramatically: 32 GB of GDDR6 on the AMD card versus "System Shared" on the Intel IGP. The bus width is 256 bit versus "System Shared." Bandwidth is 576.0 GB/s versus "System Dependent."

The AMD card's TDP is 150 W, and it requires a 450 W suggested power supply. It is a single-slot card with a 16-pin power connector. The Intel Arc B390 has an 80 W TDP, uses no power connectors, has no suggested PSU, and is classified as an IGP form factor.

The bus interface also differs. The AMD card uses PCIe 5.0 x16, a standard discrete GPU interface. The Intel part uses IGP, meaning it is integrated into the host processor and has no separate bus connection.

Display outputs differ as well. The AMD card provides one DisplayPort 2.1a output. The Intel Arc B390's display outputs are listed as "Portable Device Dependent," reflecting its role in mobile and portable systems.

Physical dimensions are recorded only for the AMD card: 241 mm in length (9.5 inches), 111 mm in height (4.4 inches), and 19 mm in width (0.7 inches). No dimensions are listed for the Intel part.

The release dates are also different. The AMD Radeon AI PRO 9600D was released on 2025-12-10, and the Intel Arc B390 followed on 2026-01-26. The AMD card's predecessor is listed as Radeon Pro Vega, while the Intel part has no predecessor recorded.

Head-to-Head Benchmarks

The head-to-head benchmark data in this database contains no direct comparison entries, and neither GPU records wins against the other in that section. However, the available benchmark data for the Intel Arc B390 and the full specification set for the AMD Radeon AI PRO 9600D allow for a meaningful analysis.

The only recorded benchmark score belongs to the Intel Arc B390. In 3DMark Steel Nomad DX12, it scores 1482 points. This places the Intel part in the 9th percentile of all GPUs, a low ranking that indicates performance below the vast majority of discrete graphics solutions.

The nearest rivals to the Intel Arc B390 are all legacy NVIDIA discrete GPUs. The GeForce GT 520MX scores 1463, which is 1.3% lower. The GeForce 800M scores 1460, 1.5% lower. The GeForce GT 625 OEM scores 1446, 2.5% lower. The GeForce GT 710 scores 1443, 2.7% lower. These margins are small, meaning the Intel Arc B390 performs essentially on par with decade-old entry-level discrete graphics.

The AMD Radeon AI PRO 9600D has no recorded benchmark scores in this database, so no direct percentile comparison is available. Its percentile rank of 50 places it at the midpoint of all GPUs, which is consistent with a professional workstation card rather than a gaming or entry-level part.

Using the recorded hardware specifications as a proxy for performance, the AMD part holds decisive advantages. Its FP32 throughput of 24.82 TFLOPS is 3.2 times the Intel part's 7.680 TFLOPS. The pixel rate of 193.9 GPixel/s is 3.2 times the Intel part's 60.00 GPixel/s. The texture rate of 387.8 GTexel/s is 3.2 times the Intel part's 120.0 GTexel/s.

The memory bandwidth differential is even larger in practical terms. The AMD card provides a fixed 576.0 GB/s from dedicated GDDR6, while the Intel IGP's bandwidth is "System Dependent," meaning it shares the host memory bus with the CPU and other components. In any realistic workload, the AMD card's dedicated memory subsystem avoids the contention and latency that the Intel IGP must endure.

The ray tracing hardware also favors the AMD part by a factor of four: 48 ray tracing cores versus 12. Both support DirectX 12 Ultimate (12_2), so ray tracing workloads are possible on both, but the AMD card has substantially more dedicated hardware for that task.

The thermal and power envelopes reflect these performance differences. The AMD card consumes 150 W and is a single-slot discrete card, while the Intel IGP consumes 80 W with no additional power connectors. The AMD card's 150 W TDP is nearly double the Intel part's 80 W TDP, which is consistent with the AMD card's larger die and dedicated memory.

The FP16 comparison is the only area where the Intel part shows a relatively stronger specification. The Intel Arc B390 achieves 15.36 TFLOPS of FP16 performance, which is double its FP32 rate due to the 2:1 ratio. The AMD part achieves 24.82 TFLOPS of FP16 at a 1:1 ratio. In absolute terms, the AMD card still delivers 1.6 times the FP16 throughput of the Intel part, but the Intel architecture's 2:1 ratio indicates a different design priority for mixed-precision workloads.

Overall, the recorded data shows a clear hierarchy. The AMD Radeon AI PRO 9600D is a professional-grade discrete GPU with high compute throughput, large dedicated memory, and broad API support. The Intel Arc B390 is a low-power integrated GPU whose only recorded benchmark places it within 3% of legacy NVIDIA entry-level cards. Any workload that demands significant graphics or compute performance will favor the AMD part by a wide margin.