Intel Arc Pro B370 vs NVIDIA GeForce RTX 5090 D Comparison

FAQ

Q: What is the average benchmark score difference between the Intel Arc Pro B370 and the NVIDIA GeForce RTX 5090 D?

A: The Intel Arc Pro B370 has an average benchmark score of 0, while the NVIDIA GeForce RTX 5090 D scores 77,712. The RTX 5090 D holds the 92nd percentile among all GPUs in the database, whereas the Arc Pro B370 sits at the 50th percentile.

Q: How do the FP32 compute performances compare between these two GPUs?

A: The Intel Arc Pro B370 delivers 6.144 TFLOPS of FP32 performance, while the NVIDIA GeForce RTX 5090 D provides 104.8 TFLOPS. This makes the RTX 5090 D approximately 17 times faster in raw single-precision floating-point throughput.

Q: What are the memory specifications for each card?

A: The Intel Arc Pro B370 uses System Shared memory with System Dependent bandwidth, meaning it has no dedicated VRAM. The NVIDIA GeForce RTX 5090 D features 32 GB of GDDR7 memory on a 512-bit bus with 1.79 TB/s of bandwidth.

Q: Which GPU has more shading units and ray tracing cores?

A: The NVIDIA GeForce RTX 5090 D has 21,760 shading units and 170 ray tracing cores. The Intel Arc Pro B370 has 1,280 shading units and 10 ray tracing cores. The RTX 5090 D also includes 680 tensor cores, while the Arc Pro B370 lists no tensor core count.

Q: What is the thermal design power difference between the two?

A: The Intel Arc Pro B370 has a TDP of 25 W and is an integrated graphics processor (IGP) with no power connectors. The NVIDIA GeForce RTX 5090 D has a TDP of 575 W, requires a single 16-pin power connector, and the database suggests a 950 W power supply.

Q: When were these GPUs released relative to each other?

A: The NVIDIA GeForce RTX 5090 D was released on 2025-01-29, while the Intel Arc Pro B370 was released on 2026-01-26. The RTX 5090 D came to market nearly a full year earlier.

The Verdict

The benchmark data presents an extreme asymmetry. The NVIDIA GeForce RTX 5090 D records an average benchmark score of 77,712, placing it in the 92nd percentile of all GPUs in the database. The Intel Arc Pro B370 has no recorded benchmark scores, an average score of 0, and sits at the 50th percentile. For any workload that depends on dedicated graphics processing, the RTX 5090 D is the only viable choice based on recorded performance data.

The RTX 5090 D also holds a clear architectural advantage. It uses a dedicated GB202 chip with 92,200 million transistors on a 750 mm² die, fabricated by TSMC on a 5 nm process. The Arc Pro B370 is an integrated processor using Intel's Panther Lake chip, built in-house on a 3 nm node with unknown transistor and die dimensions. For users who need a discrete GPU with 32 GB of GDDR7 memory, the RTX 5090 D is the data-backed pick.

The Arc Pro B370 is suited for systems where integrated graphics are a requirement, not a choice. Its 25 W TDP and system-shared memory indicate it is designed for compact, low-power platforms. It offers no dedicated VRAM, no power connectors, and uses an IGP bus interface. The RTX 5090 D, by contrast, is a dual-slot, 304 mm long card with a PCIe 5.0 x16 interface and a 16-pin power connector.

The verdict is unambiguous: for users with discrete GPU workloads, the RTX 5090 D is the only option with measurable performance. The Arc Pro B370 is a capable integrated solution for power-constrained environments, but it cannot be compared on equal footing in any recorded benchmark category.

Head-to-Head Benchmarks

The head-to-head benchmark table shows zero wins for the Intel Arc Pro B370 and zero wins for the NVIDIA GeForce RTX 5090 D. This is because the Arc Pro B370 has no recorded benchmark entries in the database. The RTX 5090 D, however, has ten individual benchmark scores that demonstrate its performance range.

In 3DMark Steel Nomad DX12, the RTX 5090 D scores 14,326. This test is a demanding modern rasterization workload, and the score places it well above the database average. In Geekbench OpenCL, the RTX 5090 D scores 310,674, and in Geekbench Vulkan it reaches 376,915. These compute-oriented tests highlight the GPU's strength in general-purpose and graphics API workloads.

PassMark results for the RTX 5090 D show a spread across different DirectX versions. DirectX 9 scores 434, DirectX 10 scores 231, DirectX 11 scores 371, and DirectX 12 scores 219. The PassMark G3D score is 44,065, and the GPU compute score is 28,396. The 2D graphics score is 1,487. These numbers show that the RTX 5090 D is strongest in 3D rendering and compute tasks, while its legacy DirectX 9 performance is relatively modest compared to its modern API results.

The nearest rivals in the database for the RTX 5090 D are the AMD Radeon RX 6650M XT with an average score of 76,904 and a delta of +1.1%, the AMD Radeon RX 6850M XT with 78,940 and a delta of -1.6%, and two NVIDIA Tesla P100 variants: the PCIe 12 GB at 79,396 (-2.1%) and the PCIe 16 GB at 79,605 (-2.4%). The RTX 5090 D is nearly tied with these mobile and data center GPUs, sitting just 1.1% above the RX 6650M XT and 2.4% below the Tesla P100 16 GB. This suggests that while the RTX 5090 D is a high-percentile GPU, its average score is closely matched by several other products in the database.

Since the Arc Pro B370 has no benchmark scores, there are no wins to attribute to it. The recorded data only allows for a one-sided comparison, with all measurable performance belonging to the RTX 5090 D.

Specification Differences

The two GPUs differ in nearly every specification field. The Intel Arc Pro B370 has a base clock of 300 MHz and a boost clock of 2400 MHz. The NVIDIA GeForce RTX 5090 D has a base clock of 2017 MHz and a boost clock of 2407 MHz. Memory clocks differ as well: the Arc Pro B370 uses System Shared memory, while the RTX 5090 D runs at 1750 MHz with 28 Gbps effective speed.

Memory configuration is a major divergence. The Arc Pro B370 has System Shared memory size, type, and bus width, with bandwidth listed as System Dependent. The RTX 5090 D has 32 GB of GDDR7 on a 512-bit bus with 1.79 TB/s bandwidth. This is the most significant hardware difference for real-world workloads.

Compute unit counts differ drastically. The Arc Pro B370 has 1,280 shading units, 40 texture mapping units, 20 ROPs, and 10 ray tracing cores. The RTX 5090 D has 21,760 shading units, 680 TMUs, 176 ROPs, and 170 ray tracing cores. The RTX 5090 D also includes 680 tensor cores, whereas the Arc Pro B370 has no listed tensor core count.

Pixel and texture rates reflect these unit differences. The Arc Pro B370 delivers 48.00 GPixel/s and 96.00 GTexel/s. The RTX 5090 D delivers 423.6 GPixel/s and 1,636.8 GTexel/s. FP16 performance also differs: the Arc Pro B370 provides 12.29 TFLOPS with a 2:1 ratio, while the RTX 5090 D provides 104.8 TFLOPS with a 1:1 ratio.

Power and physical specifications are equally divergent. The Arc Pro B370 has a 25 W TDP, an IGP slot width, no power connectors, and an IGP bus interface. The RTX 5090 D has a 575 W TDP, a dual-slot width, one 16-pin power connector, a 950 W suggested PSU, and a PCIe 5.0 x16 bus interface. The RTX 5090 D measures 304 mm in length, 137 mm in height, and 48 mm in width.

Display outputs also differ. The Arc Pro B370 lists Portable Device Dependent outputs, while the RTX 5090 D has 1x HDMI 2.1b and 3x DisplayPort 2.1b. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Architecture Differences

The Intel Arc Pro B370 is built on the Xe3-LPG architecture using the Panther Lake chip, part of the Arc Graphics-WM (Panther Lake) generation. It is fabricated on a 3 nm process at Intel's own foundry. Transistor count and die size are unknown, and no transistor density is recorded. The predecessor is listed as HD Graphics-WM. This is an integrated graphics solution, meaning the GPU shares the system's memory and is not a discrete add-in card.

The NVIDIA GeForce RTX 5090 D uses the Blackwell 2.0 architecture with the GB202 chip, part of the GeForce 50-series generation. It is fabricated by TSMC on a 5 nm process. The chip contains 92,200 million transistors on a 750 mm² die, giving a transistor density of 122.9 million per square millimeter. Its predecessor is the GeForce 40 series, and its successor is the GeForce 60 series.

The architectural differences extend to core design. The RTX 5090 D has 680 tensor cores dedicated to AI and deep learning workloads, which the Arc Pro B370 does not list. The RTX 5090 D also has 170 ray tracing cores versus 10 on the Arc Pro B370, a 17-fold difference. The shading unit count difference is even larger, with 21,760 versus 1,280, a ratio of 17 to 1 as well.

Memory architecture is fundamentally different. The Arc Pro B370 uses system-shared memory, meaning its performance is dependent on the host system's RAM and memory controller. The RTX 5090 D has dedicated 32 GB GDDR7 memory with a 512-bit bus, eliminating competition with the CPU for memory bandwidth. The 1.79 TB/s bandwidth on the RTX 5090 D is a dedicated resource, whereas the Arc Pro B370's bandwidth is listed as System Dependent.

The process node difference is notable: Intel uses 3 nm for the Arc Pro B370, while TSMC uses 5 nm for the RTX 5090 D. This means the Intel part is built on a newer fabrication process, but the RTX 5090 D compensates with a vastly larger chip and higher power envelope. The RTX 5090 D also has a much higher transistor count, at 92,200 million versus unknown for the Arc Pro B370.

Where Each One Wins

The Intel Arc Pro B370 wins in power efficiency and integration. Its 25 W TDP is dramatically lower than the RTX 5090 D's 575 W, making it suitable for fanless or passively cooled systems. It requires no power connectors, no external power supply upgrade, and occupies no expansion slot since it is an IGP. For portable devices or compact platforms where space and thermal budget are the primary constraints, the Arc Pro B370 is the functional choice. Its 3 nm process node also indicates a modern manufacturing technology, though the database records no benchmark scores to quantify its real-world performance.

The NVIDIA GeForce RTX 5090 D wins in every measurable performance category. It has recorded scores in 3DMark, Geekbench OpenCL, Geekbench Vulkan, and multiple PassMark tests. Its 92nd percentile ranking places it among the top GPUs in the database. The 32 GB of GDDR7 memory with 1.79 TB/s bandwidth provides a clear advantage for large datasets and high-resolution textures. The 104.8 TFLOPS FP32 and FP16 performance, coupled with 680 tensor cores, makes it suited for compute-heavy workloads such as rendering, simulation, and AI inference.

For gaming and 3D rendering, the RTX 5090 D holds all recorded wins. The 3DMark Steel Nomad score of 14,326 and PassMark G3D score of 44,065 indicate strong rasterization performance. The Geekbench Vulkan score of 376,915 suggests robust API-level graphics performance. The Arc Pro B370 has no comparable data in these categories.

For legacy API compatibility, the RTX 5090 D also takes the lead. Its PassMark DirectX 9 score of 434, DirectX 10 score of 231, DirectX 11 score of 371, and DirectX 12 score of 219 cover a broad range of older and newer workloads. The Arc Pro B370 supports DirectX 12 Ultimate and Vulkan 1.4 per its API list, but no benchmark scores exist to verify its implementation.

The use-case split is therefore clear. The Arc Pro B370 is for low-power integrated systems where discrete graphics are not an option. The RTX 5090 D is for any workload that requires measurable graphics or compute performance. The database contains no evidence that the Arc Pro B370 can compete with the RTX 5090 D in any benchmarked scenario.