Intel Arc B390 vs NVIDIA RTX A400 Comparison

Where Each One Wins

The recorded benchmark data splits cleanly along workload lines. The Intel Arc B390 has a single entry in the database, a 3DMark Steel Nomad DX12 score of 1482. That places it at the 9th percentile of all GPUs, a position that puts it in the company of legacy NVIDIA notebook parts. The RTX A400, by contrast, has nine separate benchmark entries spanning OpenCL, Vulkan, DirectX 9 through 12, and compute workloads. Its average benchmark score of 6078 puts it at the 35th percentile, a much stronger overall standing.

The RTX A400 wins every recorded comparison that exists. Its best single result is the Geekbench OpenCL score of 22844, followed closely by the Geekbench Vulkan score of 22237. The Passmark G3D score of 5983 and the Passmark G2D score of 899 round out its strongest showings. The Intel part has no comparable entries in those tests, so the database contains no head-to-head wins for the Arc B390. The wins column in the database shows zero for the Intel part and zero for the NVIDIA part, meaning the recorded tests do not overlap, but the available data still allows a use-case split. The Arc B390 is a mobile integrated graphics processor with a single DX12 test result, while the RTX A400 is a discrete workstation card with a broad multi-API benchmark profile. The RTX A400 clearly wins on compute-oriented workloads like OpenCL and Vulkan, and it also holds the higher overall percentile ranking.

For gaming-specific loads, the Passmark DirectX scores for the RTX A400 show modest numbers: 87 in DirectX 9, 37 in DirectX 11, 32 in DirectX 10, and 27 in DirectX 12. Those are low absolute scores, but they are the only DirectX data points in the database. The Arc B390's single 3DMark result of 1482 cannot be directly compared to the Passmark suite, but the percentile difference is stark. The Arc B390 sits at the 9th percentile, while the RTX A400 sits at the 35th percentile, a 26-point gap that indicates the NVIDIA part is substantially better positioned across the full GPU landscape.

Architecture Differences

The two parts come from different foundries and process nodes. The Intel Arc B390 uses the Panther Lake chip built on a 3 nm process at Intel's own foundry. The NVIDIA RTX A400 uses the GA107 chip built on an 8 nm process at Samsung. The transistor counts differ sharply: the NVIDIA part lists 8,700 million transistors on a 200 mm² die, giving a transistor density of 43.5 million per square millimeter. The Intel part lists unknown transistor count and die size, so no density figure exists for it.

The architectures themselves are from different generations. The Arc B390 uses Xe3-LPG architecture from the Arc Graphics-M (Panther Lake) generation. The RTX A400 uses Ampere architecture from the Workstation Ampere (Ax000) generation. The Intel part ships with 1536 shading units, 48 texture mapping units, 24 raster output units, and 12 ray tracing cores. The NVIDIA part has 768 shading units, 24 texture mapping units, 16 raster output units, 6 ray tracing cores, and 24 tensor cores. The Intel part has no tensor core count listed, while the NVIDIA part's tensor cores are a notable feature for AI workloads.

Clock behavior differs significantly. The Arc B390 has a base clock of 300 MHz and a boost clock of 2500 MHz, a wide frequency range that suggests aggressive power scaling. The RTX A400 has a base clock of 1417 MHz and a boost clock of 1762 MHz, a much narrower range. Memory configurations are entirely different. The Intel part uses system shared memory with system dependent bandwidth, while the NVIDIA part has dedicated 4 GB of GDDR6 on a 64-bit bus with 96.00 GB/s of bandwidth. The memory clock for the NVIDIA part is 1500 MHz with 12 Gbps effective data rate.

The Intel part is an integrated graphics processor (IGP) with no slot width, no power connectors, and a portable device dependent display output. The RTX A400 is a single-slot discrete card with four mini-DisplayPort 1.4a outputs, a PCIe 4.0 x8 interface, and a length of 163 mm. The TDP values also differ: the Arc B390 draws 80 W, while the RTX A400 draws 50 W with a suggested power supply of 250 W.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The RTX A400 has an average benchmark score of 6078, compared to the Arc B390's average of 1482, a difference of 4596 points.

Q: How do their percentile rankings compare?

A: The Arc B390 sits at the 9th percentile of all GPUs, while the RTX A400 sits at the 35th percentile, a 26-point gap in favor of the NVIDIA part.

Q: What memory configurations do the two use?

A: The Arc B390 uses system shared memory with system dependent bandwidth. The RTX A400 uses 4 GB of GDDR6 on a 64-bit bus with 96.00 GB/s bandwidth.

Q: Which part has more shading units?

A: The Arc B390 has 1536 shading units, double the 768 shading units of the RTX A400.

Q: What is the process node difference?

A: The Arc B390 is built on a 3 nm process at Intel, while the RTX A400 is built on an 8 nm process at Samsung.

Q: Does the RTX A400 have tensor cores?

A: Yes, the RTX A400 lists 24 tensor cores, while the Arc B390 has no tensor core count in the database.

Specification Differences

The two parts differ in nearly every measured specification. The process node is 3 nm for Intel and 8 nm for Samsung. The foundry is Intel for the Arc B390 and Samsung for the RTX A400. The chip is Panther Lake for the Intel part and GA107 for the NVIDIA part. The architectures differ: Xe3-LPG versus Ampere.

Clock speeds diverge widely. The Arc B390 runs at 300 MHz base and 2500 MHz boost. The RTX A400 runs at 1417 MHz base and 1762 MHz boost. The Intel part has a wider boost range, but the NVIDIA part starts from a much higher base.

Memory is a major differentiator. The Arc B390 uses system shared memory with system dependent bandwidth. The RTX A400 has 4 GB GDDR6, a 64-bit bus, 96.00 GB/s bandwidth, and a 1500 MHz memory clock with 12 Gbps effective data rate.

Compute unit counts favor the Intel part in raw quantity. The Arc B390 has 1536 shading units, 48 TMUs, 24 ROPs, and 12 RT cores. The RTX A400 has 768 shading units, 24 TMUs, 16 ROPs, 6 RT cores, and 24 tensor cores. The pixel rate for the Arc B390 is 60.00 GPixel/s versus 28.19 GPixel/s for the RTX A400. The texture rate is 120.0 GTexel/s versus 42.29 GTexel/s. FP32 throughput is 7.680 TFLOPS versus 2.706 TFLOPS. FP16 rates differ in ratio: the Arc B390 delivers 15.36 TFLOPS with a 2:1 ratio, while the RTX A400 delivers 2.706 TFLOPS with a 1:1 ratio.

Power and form factor differ. The Arc B390 has an 80 W TDP, is an IGP with no slot width, no power connectors, and a portable device dependent display output. The RTX A400 has a 50 W TDP, is single-slot, has no power connectors, a 250 W suggested PSU, a PCIe 4.0 x8 interface, and four mini-DisplayPort 1.4a outputs. The NVIDIA card is 163 mm long and 69 mm high.

Release dates differ by about 21 months. The Arc B390 released on 2026-01-26, while the RTX A400 released on 2024-04-15. The NVIDIA part has a defined predecessor and successor in the database (Quadro Turing and Workstation Ada), while the Intel part lists neither.

Head-to-Head Benchmarks

The database contains no direct head-to-head benchmark entries between the two cards, so the comparison must be built from the individual scores and rival deltas. The Arc B390's single 3DMark Steel Nomad DX12 score of 1482 places it just 1.3% ahead of the NVIDIA GeForce GT 520MX, 1.5% ahead of the GeForce 800M, 2.5% ahead of the GeForce GT 625 OEM, and 2.7% ahead of the GeForce GT 710. Those rivals all have average scores in the 1443 to 1463 range, so the Arc B390 is only marginally faster than a cluster of decade-old entry-level parts.

The RTX A400's nearest rivals tell a different story. Its average score of 6078 sits essentially even with the GeForce MX230 at 6077, a 0% delta. It is 0.5% ahead of the Quadro P2000 at 6049, 0.6% behind the Intel Iris Pro Graphics 6200 at 6117, and 1% ahead of the AMD Radeon 760M at 6019. The NVIDIA card sits in a cluster of modern and near-modern parts, none of which are more than 1% away from its score.

The compute scores for the RTX A400 are its strongest entries. The Geekbench OpenCL result of 22844 and the Vulkan result of 22237 are far above its Passmark scores. The Passmark GPU compute score of 2557 also indicates solid compute throughput. The Arc B390 has no compute scores in the database, so no comparison is possible there.

The FP32 throughput gap is large. The Arc B390 delivers 7.680 TFLOPS, which is 2.8 times the 2.706 TFLOPS of the RTX A400. The texture rate gap is even wider: 120.0 GTexel/s versus 42.29 GTexel/s, a factor of 2.8. The pixel rate gap is 60.00 GPixel/s versus 28.19 GPixel/s, a factor of 2.1. These raw throughput numbers favor the Intel part, but the benchmark results favor the NVIDIA part. The RTX A400 has a 26-point higher percentile ranking and an average score that is 4.1 times higher. The database shows that raw compute specifications do not translate into benchmark superiority for the Arc B390.

The RTX A400's DirectX scores are low in absolute terms: 87 in DX9, 37 in DX11, 32 in DX10, and 27 in DX12. These numbers suggest that the card's gaming performance is not its primary strength, but they are the only DirectX data points available. The Arc B390's 3DMark result of 1482 is a single data point from a different benchmark suite, so cross-suite comparison is not possible.

The Verdict

The data points to a clear split by use case. The Intel Arc B390 offers higher raw compute specifications: more shading units, more TMUs, more ROPs, more RT cores, higher pixel rate, higher texture rate, and higher FP32 throughput. Its 3 nm process node and 2500 MHz boost clock are modern specifications. But its only benchmark result is a 3DMark score of 1482, which lands it at the 9th percentile of all GPUs, alongside NVIDIA parts from over a decade ago.

The NVIDIA RTX A400 has lower raw specifications but substantially better benchmark results. Its average score of 6078 is 4.1 times higher than the Arc B390's average of 1482. Its 35th percentile ranking places it 26 points higher. Its nearest rivals are modern parts like the GeForce MX230 and the AMD Radeon 760M, not legacy GPUs. The RTX A400 also has dedicated 4 GB GDDR6 memory with 96.00 GB/s bandwidth, while the Arc B390 relies on system shared memory with system dependent bandwidth.

For workloads that stress compute throughput, the RTX A400 is the clear choice based on the recorded data. Its Geekbench OpenCL score of 22844 and Vulkan score of 22237 are far above anything the Arc B390 can show, and its Passmark GPU compute score of 2557 confirms compute capability. For users who need a discrete workstation card with four mini-DisplayPort outputs and a PCIe 4.0 x8 interface, the RTX A400 is the only option with those features in this comparison.

The Arc B390 is an integrated part with an 80 W TDP and no power connectors, designed for portable devices. Its raw throughput numbers are higher, but the database shows no benchmark evidence that those numbers translate into real-world performance. The single 3DMark result places it near the bottom of the GPU hierarchy. The RTX A400, despite lower FP32 and texture rates, delivers benchmark scores that are 4.1 times higher on average and sits in a much stronger competitive position. The data favors the RTX A400 for any workload where benchmark scores matter, while the Arc B390's appeal would have to rest on its integrated form factor and raw specification sheet, neither of which is validated by the recorded performance data.