Intel Arc Pro B370 vs NVIDIA GeForce RTX 5090 Comparison

Intel Arc Pro B370 and NVIDIA GeForce RTX 5090 occupy opposite ends of the graphics hardware spectrum. The Arc Pro B370 is an integrated graphics processor within Intel's Panther Lake platform, while the RTX 5090 is a discrete flagship add-in board. The recorded data shows a 92nd percentile ranking for the RTX 5090 versus a 50th percentile for the Arc Pro B370. Their specifications, from process node to memory subsystem, define two entirely different product categories.

FAQ

Q: What is the core architectural difference between the two processors?

A: The Intel Arc Pro B370 uses the Xe3-LPG architecture on Intel's 3 nm process, while the NVIDIA GeForce RTX 5090 uses Blackwell 2.0 on a 5 nm TSMC process. The RTX 5090 integrates 92,200 million transistors on a 750 mm² die, whereas the Arc Pro B370's transistor count and die size are not recorded in the database.

Q: How do their memory configurations compare?

A: The RTX 5090 has 32 GB of GDDR7 memory across a 512-bit bus, delivering 1.79 TB/s of bandwidth. The Arc Pro B370 uses system shared memory, with its bandwidth listed as system dependent. This is a fundamental difference between a dedicated memory pool and a shared memory architecture.

Q: What are the power delivery requirements?

A: The RTX 5090 carries a 575 W TDP, requires a single 16-pin power connector, and the database recommends a 950 W power supply. The Arc Pro B370 has a 25 W TDP, uses no power connectors, and has no suggested PSU rating because it is an integrated processor.

Q: Which GPU has higher compute throughput?

A: The RTX 5090 delivers 104.8 TFLOPS of FP32 performance and 104.8 TFLOPS of FP16 (1:1). The Arc Pro B370 provides 6.144 TFLOPS FP32 and 12.29 TFLOPS FP16 (2:1). The NVIDIA part offers roughly 17 times the FP32 throughput.

Q: What is the physical size difference?

A: The RTX 5090 is a dual-slot card measuring 304 mm in length, 137 mm in height, and 40 mm in width. The Arc Pro B370 is an integrated graphics processor with no recorded dimensions, no slot width beyond IGP, and a bus interface of IGP rather than PCIe.

Q: What API support do both GPUs share?

A: Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The RTX 5090 adds 680 tensor cores and 170 RT cores, while the Arc Pro B370 has 10 RT cores and no tensor core count listed.

The Verdict

The RTX 5090 is the clear choice for high-end discrete graphics workloads. The database shows it at the 92nd percentile of all GPUs, with an average benchmark score of 79,842. It outperforms its nearest rival, the AMD Radeon Pro Vega 64X, by 1.4 percent, and leads the NVIDIA Tesla P100 PCIe 16 GB by 0.3 percent. Its 21760 shading units, 680 TMUs, and 176 ROPs provide the compute foundation for such results.

The Arc Pro B370 serves a completely different role. As an integrated processor with 1280 shading units, 40 TMUs, and 20 ROPs, it targets systems where discrete graphics are not an option. Its 25 W TDP and lack of power connectors make it suitable for compact or portable devices. The database records a 50th percentile ranking, placing it at the midpoint of all GPUs.

Users who require maximum performance, large memory capacity, or dedicated ray tracing hardware should select the RTX 5090. Users who need basic graphics capability integrated into a processor, with minimal power draw, should consider the Arc Pro B370. The data does not suggest any overlap in their intended use cases.

Head-to-Head Benchmarks

The RTX 5090 has recorded benchmark scores across multiple tests in the database. Its highest single score is 376,728 in Geekbench Vulkan, followed by 334,370 in Geekbench OpenCL. The Passmark G3D test shows 39,650, and the Passmark GPU Compute test shows 26,756. The Arc Pro B370 has no recorded benchmark scores in the database, so direct numerical comparison is limited to specification analysis.

The RTX 5090's average benchmark score of 79,842 places it just 0.3 percent above the NVIDIA Tesla P100 PCIe 16 GB and 0.6 percent above the Tesla P100 PCIe 12 GB. It trails the AMD Radeon Pro Vega 64X by 1.4 percent, which scores 80,959. These deltas indicate the RTX 5090 sits in a competitive band at the top of the performance stack.

The Arc Pro B370's FP32 throughput of 6.144 TFLOPS is dwarfed by the RTX 5090's 104.8 TFLOPS. Pixel fill rate also differs dramatically: the RTX 5090 achieves 423.6 GPixel/s versus 48.00 GPixel/s for the Arc Pro B370. Texture fill rate shows the RTX 5090 at 1,636.8 GTexel/s against 96.00 GTexel/s for the Intel part.

Memory bandwidth presents the largest gap. The RTX 5090's 1.79 TB/s GDDR7 implementation is an order of magnitude beyond the system dependent bandwidth of the Arc Pro B370. The RTX 5090 also has 32 GB of dedicated memory, while the Intel processor relies on shared system memory with no fixed capacity.

Specification Differences

The RTX 5090 uses a GB202 chip with 92,200 million transistors on a 750 mm² die, fabricated by TSMC on a 5 nm process. The Arc Pro B370 uses a Panther Lake chip on Intel's 3 nm process, with transistor count and die size marked as unknown. The RTX 5090's transistor density is recorded at 122.9M per mm².

Clock speeds differ substantially. The Arc Pro B370 has a base clock of 300 MHz and a boost clock of 2400 MHz. The RTX 5090 runs at a 2017 MHz base and 2407 MHz boost, with memory clocked at 1750 MHz or 28 Gbps effective.

The RTX 5090 has 21760 shading units, 680 TMUs, 176 ROPs, 170 RT cores, and 680 tensor cores. The Arc Pro B370 has 1280 shading units, 40 TMUs, 20 ROPs, and 10 RT cores, with no tensor core count recorded. These ratios reflect the RTX 5090's position as a full-scale discrete GPU.

Power and physical characteristics diverge completely. The RTX 5090 draws 575 W with a dual-slot cooler, a 16-pin connector, and a 950 W suggested PSU. It measures 304 mm by 137 mm by 40 mm. The Arc Pro B370 draws 25 W, has no slot width beyond IGP, no power connectors, and no recorded dimensions. The RTX 5090 uses PCIe 5.0 x16, while the Arc Pro B370 uses an IGP bus interface.

Display outputs also differ. The RTX 5090 provides 1x HDMI 2.1b and 3x DisplayPort 2.1b. The Arc Pro B370's display outputs are listed as portable device dependent.

Architecture Differences

The Intel Arc Pro B370 is built on Xe3-LPG architecture, part of the Arc Graphics-WM generation for Panther Lake. Its predecessor is HD Graphics-WM, and its release date is recorded as January 26, 2026. The NVIDIA GeForce RTX 5090 uses Blackwell 2.0 architecture in the GeForce 50 series, with GB202 as its chip. Its predecessor is the GeForce 40 series, and its successor is the GeForce 60 series.

Process technology separates the two designs. Intel fabricates the Arc Pro B370 on a 3 nm node at its own foundry. TSMC fabricates the RTX 5090 on a 5 nm node. The Intel part uses a shared memory architecture where memory size, type, and bus width are all system dependent. The NVIDIA part uses dedicated GDDR7 memory with fixed specifications.

The RTX 5090's tensor core count of 680 and RT core count of 170 enable AI and ray tracing workloads at scale. The Arc Pro B370's 10 RT cores provide basic ray tracing capability, but its lack of recorded tensor cores limits AI acceleration options. FP16 processing also differs: the RTX 5090 achieves 104.8 TFLOPS at 1:1 ratio, while the Arc Pro B370 reaches 12.29 TFLOPS at 2:1 ratio.

The RTX 5090's release date of January 29, 2025, and launch MSRP of 1,999 USD position it as a premium product. The Arc Pro B370 has no launch MSRP recorded. Both GPUs support the same DirectX, OpenGL, and Vulkan versions, indicating comparable software API compatibility despite their hardware differences.