Intel Arc Pro B370 vs NVIDIA GeForce RTX 5090 SE Comparison

Head-to-Head Benchmarks

The database contains no recorded head-to-head benchmark results for the Intel Arc Pro B370 versus the NVIDIA GeForce RTX 5090 SE. Neither part has an average benchmark score entered, and both sit at the 50th percentile among all GPUs in the database. The absence of direct measurements means the comparison must be built entirely from the architectural and specification data recorded for each product.

The raw compute figures show an enormous gulf. The RTX 5090 SE delivers 66.94 TFLOPS of FP32 throughput, while the Arc Pro B370 delivers 6.144 TFLOPS. That places the NVIDIA part at roughly 10.9 times the single-precision compute of the Intel part. The FP16 comparison is similarly lopsided: the RTX 5090 SE sustains 66.94 TFLOPS at a 1:1 ratio, whereas the Arc Pro B370 reaches 12.29 TFLOPS via a 2:1 rate. Even accounting for the Intel part's packed-rate advantage, the NVIDIA GPU still holds a 5.4x lead in half-precision throughput.

Texture and pixel throughput follow the same pattern. The RTX 5090 SE achieves 1,045.9 GTexel/s against the Arc Pro B370's 96.00 GTexel/s, an 10.9x margin. Pixel fill rate shows a 7.9x gap: 380.3 GPixel/s versus 48.00 GPixel/s. These figures derive directly from the shading unit, TMU, and ROP counts: 14,080 shading units, 440 TMUs, and 160 ROPs on the NVIDIA side, versus 1,280 shading units, 40 TMUs, and 20 ROPs on the Intel side.

The RTX 5090 SE also carries 110 ray tracing cores and 440 tensor cores. The Arc Pro B370 has 10 ray tracing cores and no recorded tensor core count. The architectural disparity in dedicated acceleration hardware is substantial, though no benchmark scores exist to quantify the real-world impact.

Where Each One Wins

The Intel Arc Pro B370 has clear advantages in power draw and physical integration. Its TDP is recorded at 25 W, compared to 500 W for the RTX 5090 SE. The Intel part requires no power connectors and occupies an integrated graphics processor slot width, meaning it is designed to be part of a Panther Lake platform rather than a discrete add-in card. The RTX 5090 SE is a dual-slot card using a single 16-pin connector and carrying a suggested power supply rating of 900 W.

Memory architecture also differentiates the two. The Arc Pro B370 uses system shared memory, with bandwidth described as system dependent. The RTX 5090 SE uses 24 GB of GDDR7 on a 384-bit bus, delivering 1.34 TB/s of dedicated bandwidth. For workloads that depend on dedicated VRAM capacity and consistent memory bandwidth, the NVIDIA card is the only one of the two with a fixed, quantified memory subsystem.

Clock behavior is another split. The Arc Pro B370 has a base clock of 300 MHz and a boost clock of 2400 MHz, a range that suggests power-sensitive operation and opportunistic boosting. The RTX 5090 SE starts at 1740 MHz base and boosts to 2377 MHz. The two parts operate in different power envelopes, so raw clock rates are not directly comparable.

The process node and foundry differ as well. Intel uses its own 3 nm process for the Panther Lake chip, while NVIDIA uses TSMC's 5 nm process for GB202. The RTX 5090 SE's die is recorded at 750 mm² with 92,200 million transistors, giving a transistor density of 122.9 million per square millimeter. The Arc Pro B370's transistor count and die size are listed as unknown, so no density comparison is possible.

Architecture Differences

The Intel Arc Pro B370 is built on the Xe3-LPG architecture, part of the Arc Graphics-WM generation for Panther Lake. It is classified as an integrated GPU with an IGP bus interface. Its predecessor is listed as HD Graphics-WM. The chip uses Intel's 3 nm foundry process.

The NVIDIA GeForce RTX 5090 SE uses the Blackwell 2.0 architecture on the GB202 chip, fabricated by TSMC on a 5 nm process. It belongs to the GeForce 50-series generation and lists GeForce 40 as its predecessor and GeForce 60 as its successor. The card connects through PCIe 5.0 x16 and presents three DisplayPort 2.1b outputs plus one HDMI 2.1b output. The Arc Pro B370's display outputs are listed as portable device dependent.

Both parts support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The API compatibility is identical in the database, so software-level feature support is not a differentiator.

The RTX 5090 SE has a recorded release date of December 31, 2025, and the Arc Pro B370 follows with a January 26, 2026 release date. Both are marked as active production status. The NVIDIA card has a recorded launch MSRP of 1,499 USD. No launch MSRP is recorded for the Intel part.

The RTX 5090 SE measures 267 mm in length, 111 mm in height, and 40 mm in width. The Arc Pro B370 has no recorded physical dimensions, consistent with its integrated design.

The memory clock for the RTX 5090 SE is recorded as 1750 MHz with 28 Gbps effective data rate. The Arc Pro B370's memory clock is listed as system shared, with no independent memory clock value.

The Verdict

The recorded data supports only one conclusion for raw performance: the RTX 5090 SE is in a different performance class entirely. Its FP32 compute is 10.9x higher, its texture rate is 10.9x higher, and its pixel rate is 7.9x higher. It has 11x the shading units, 11x the TMUs, 8x the ROPs, and 11x the ray tracing cores compared to the Arc Pro B370. It also has 24 GB of dedicated GDDR7 memory on a 384-bit bus with 1.34 TB/s bandwidth, whereas the Intel part depends entirely on shared system memory.

The Arc Pro B370's strengths are power and integration. At 25 W with no power connectors and an IGP form factor, it fits into platforms where a 500 W dual-slot card with a 900 W suggested power supply cannot. The RTX 5090 SE requires a PCIe 5.0 x16 slot, a 16-pin connector, and substantial chassis and cooling accommodations. The Intel part is a low-power integrated solution; the NVIDIA part is a high-end discrete accelerator.

Neither part has benchmark scores in the database, so the percentile ranking of 50 for both reflects the absence of data rather than measured parity. The verdict from the recorded specifications is straightforward: the RTX 5090 SE is the performance leader by every quantified compute metric, and the Arc Pro B370 is the appropriate choice only for systems constrained by power, space, or the need for integrated graphics.

The release dates place the Arc Pro B370 about four weeks after the RTX 5090 SE, so the Intel part is the newer product. The NVIDIA predecessor is listed as GeForce 40 and successor as GeForce 60, which frames the 5090 SE as a transition generation. The Intel part's predecessor is HD Graphics-WM, with no successor recorded.

FAQ

Q: How much faster is the RTX 5090 SE in FP32 compute than the Arc Pro B370?

A: The RTX 5090 SE delivers 66.94 TFLOPS of FP32 throughput, while the Arc Pro B370 delivers 6.144 TFLOPS. That is approximately 10.9 times higher.

Q: What is the memory configuration difference between the two GPUs?

A: The RTX 5090 SE has 24 GB of GDDR7 on a 384-bit bus with 1.34 TB/s bandwidth. The Arc Pro B370 uses system shared memory with system dependent bandwidth.

Q: Does the Arc Pro B370 have tensor cores?

A: No tensor core count is recorded for the Arc Pro B370. The RTX 5090 SE has 440 tensor cores.

Q: What is the power draw difference?

A: The Arc Pro B370 has a TDP of 25 W and requires no power connectors. The RTX 5090 SE has a TDP of 500 W, uses a single 16-pin connector, and has a suggested power supply rating of 900 W.

Q: Which GPU has more ray tracing cores?

A: The RTX 5090 SE has 110 ray tracing cores. The Arc Pro B370 has 10 ray tracing cores.

Q: What are the physical form factors of the two parts?

A: The Arc Pro B370 is an integrated GPU with an IGP slot width and no recorded dimensions. The RTX 5090 SE is a dual-slot card measuring 267 mm by 111 mm by 40 mm.