Intel Arc B370 vs NVIDIA GeForce RTX 5090 Comparison

Head-to-Head Benchmarks

The recorded data shows a decisive outcome in the only directly comparable benchmark. In 3DMark Steel Nomad DX12, the NVIDIA GeForce RTX 5090 scores 18,355 points, while the Intel Arc B370 scores 1,184 points. This yields a delta of 93.5% in favor of the RTX 5090, meaning the Arc B370 trails by that percentage in this specific test. The margin is enormous, and the result places the two products at opposite ends of the performance spectrum.

The RTX 5090 also carries a much broader benchmark record in the database. Its average benchmark score across all recorded tests is 79,842, which places it in the 92nd percentile of all GPUs. The Arc B370, by contrast, has an average score of 1,184 and sits in the 5th percentile. The RTX 5090’s nearest rivals in the database include the NVIDIA Tesla P100 PCIe 16 GB (average score 79,605, a 0.3% difference), the NVIDIA Tesla P100 PCIe 12 GB (79,396, 0.6% difference), and the AMD Radeon RX 6850M XT (78,940, 1.1% difference). The Arc B370’s closest competitors are far less powerful: the ATI Mobility Radeon HD 5570 (1,186, 0.2% difference), the ATI Radeon HD 5770 (1,190, 0.5% difference), and the AMD Radeon HD 7650M (1,192, 0.7% difference).

In the head-to-head comparison, the RTX 5090 wins 1 out of 1 recorded benchmarks, while the Arc B370 wins none. The single test result is the only direct measurement, but it is consistent with the broader database averages. The gap between the two averages is roughly 78,658 points, which reflects the fundamental difference in their target roles.

Where Each One Wins

The RTX 5090 wins in every recorded category where both have data. Its 3DMark Steel Nomad score is 15.5 times higher than the Arc B370’s score (18,355 divided by 1,184). This is a pure rasterization and compute workload, and the RTX 5090’s advantage is overwhelming. The RTX 5090 also shows strong results in compute-oriented tests: its Geekbench OpenCL score is 334,370, its Geekbench Vulkan score is 376,728, and its Passmark GPU Compute score is 26,756. These are not directly compared to the Arc B370, but they indicate the RTX 5090’s capability in general-purpose and API-agnostic workloads.

The Arc B370 does not have any benchmark wins in the database. Its single recorded score of 1,184 in 3DMark Steel Nomad places it below all of its nearest rivals in that test, though the differences are small: it trails the ATI Mobility Radeon HD 5570 by 0.2%, the ATI Radeon HD 5770 by 0.5%, and the AMD Radeon HD 7650M by 0.7%. It does outperform the AMD FirePro M2000 by 1.4% (1,184 versus 1,168). This suggests the Arc B370 is positioned at the entry level, where small percentage differences separate competing products.

For use cases, the data indicates the RTX 5090 is the only choice for demanding 3D workloads, high-resolution rendering, or compute-heavy tasks. The Arc B370, with its low scores, is suited only for basic display output or very light 3D tasks, and even then it sits at the bottom of the database’s percentile ranking.

Architecture Differences

The two GPUs are built on fundamentally different architectures and process nodes. The Intel Arc B370 uses the Xe3-LPG architecture on a 3 nm process node, fabricated by Intel. Its chip is codenamed Panther Lake, and it belongs to the Arc Graphics-M (Panther Lake) generation. The NVIDIA GeForce RTX 5090 uses the Blackwell 2.0 architecture on a 5 nm process node, fabricated by TSMC. Its chip is GB202, and it belongs to the GeForce 50 generation.

The RTX 5090 is a discrete, full-scale GPU with 92,200 million transistors on a 750 mm² die, giving a transistor density of 122.9 million per square millimeter. The Arc B370’s transistor count and die size are not recorded in the database, but its role as an integrated graphics processor (IGP) suggests a much smaller implementation. The RTX 5090’s memory subsystem is substantial: 32 GB of GDDR7 on a 512-bit bus with 1.79 TB/s bandwidth. The Arc B370 uses system shared memory, with type, bus width, and bandwidth all designated as system dependent. This means the Arc B370 relies on the host system’s memory, while the RTX 5090 has dedicated high-speed VRAM.

Compute resources differ massively. The RTX 5090 has 21,760 shading units, 680 texture mapping units, 176 raster output units, 170 ray tracing cores, and 680 tensor cores. The Arc B370 has 1,280 shading units, 40 TMUs, 20 ROPs, and 10 ray tracing cores. The RTX 5090’s FP32 compute is 104.8 TFLOPS, while the Arc B370 delivers 6.144 TFLOPS. The FP16 figures also diverge: the RTX 5090 achieves 104.8 TFLOPS at a 1:1 ratio, while the Arc B370 reaches 12.29 TFLOPS at a 2:1 ratio.

Clock speeds tell a similar story. The RTX 5090 has a base clock of 2017 MHz and a boost clock of 2407 MHz, with memory running at 1750 MHz (28 Gbps effective). The Arc B370 has a base clock of 300 MHz and a boost clock of 2400 MHz, with memory speed designated as system shared. The RTX 5090’s pixel rate is 423.6 GPixel/s and texture rate is 1,636.8 GTexel/s, versus 48.00 GPixel/s and 96.00 GTexel/s for the Arc B370.

Power and physical design also separate them. The RTX 5090 has a TDP of 575 W, requires a dual-slot cooler, uses a single 16-pin power connector, and suggests a 950 W power supply. It connects via PCIe 5.0 x16 and measures 304 mm in length, 137 mm in height, and 40 mm in width. The Arc B370 has a TDP of 25 W, is an integrated processor with no slot width (IGP), uses no power connectors, and has no dedicated power supply suggestion. Its bus interface is IGP, and its display outputs are dependent on the portable device it is embedded in. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

FAQ

Q: Which GPU has the higher 3DMark Steel Nomad DX12 score?

A: The NVIDIA GeForce RTX 5090 scores 18,355, which is 93.5% higher than the Intel Arc B370’s 1,184.

Q: What is the performance percentile ranking for each GPU?

A: The RTX 5090 is in the 92nd percentile of all GPUs, while the Arc B370 is in the 5th percentile.

Q: How do the memory configurations differ?

A: The RTX 5090 has 32 GB of GDDR7 memory on a 512-bit bus with 1.79 TB/s bandwidth. The Arc B370 uses system shared memory, with its type, bus width, and bandwidth dependent on the host system.

Q: What are the TDP requirements for each card?

A: The RTX 5090 has a TDP of 575 W and suggests a 950 W power supply. The Arc B370 has a TDP of 25 W and requires no power supply recommendation because it is an integrated processor.

Q: Do both GPUs support the same APIs?

A: Yes, both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: What are the nearest rivals for each in the database?

A: The RTX 5090’s nearest rivals are the NVIDIA Tesla P100 PCIe 16 GB (0.3% difference), the NVIDIA Tesla P100 PCIe 12 GB (0.6% difference), and the AMD Radeon RX 6850M XT (1.1% difference). The Arc B370’s nearest rivals are the ATI Mobility Radeon HD 5570 (0.2% difference), the ATI Radeon HD 5770 (0.5% difference), and the AMD Radeon HD 7650M (0.7% difference).

The Verdict

The data is unambiguous. The NVIDIA GeForce RTX 5090 is a high-end discrete GPU designed for maximum performance, as evidenced by its 92nd percentile ranking and an average benchmark score of 79,842. It wins the only direct comparison by 93.5%, and its architectural resources (21,760 shading units, 32 GB GDDR7, 104.8 TFLOPS FP32) are in a different class entirely. The RTX 5090 is the appropriate choice for any workload requiring substantial 3D rendering, compute, or ray tracing capability.

The Intel Arc B370 is an integrated graphics processor with a 25 W TDP, designed for low-power portable devices. Its 5th percentile ranking and average score of 1,184 place it at the entry level, comparable to much older discrete GPUs like the ATI Radeon HD 5770 or the AMD Radeon HD 7650M. It can handle basic display output and light 3D tasks, but it is not intended for demanding applications. Its launch MSRP is not recorded, so no pricing comparison is possible from the data.

For users selecting between these two, the decision depends entirely on the intended use case. The RTX 5090 is required for high-end gaming, professional visualization, or compute workloads. The Arc B370 is suited only for ultra-low-power integrated graphics scenarios where discrete GPU performance is not needed. There is no middle ground in the recorded data.

Specification Differences

The following specifications differ between the Intel Arc B370 and the NVIDIA GeForce RTX 5090:

• Process Node: 3 nm (Intel) versus 5 nm (TSMC)
• Architecture: Xe3-LPG versus Blackwell 2.0
• Chip: Panther Lake versus GB202
• Generation: Arc Graphics-M (Panther Lake) versus GeForce 50
• Transistors: not recorded versus 92,200 million
• Die Size: not recorded versus 750 mm²
• Transistor Density: not recorded versus 122.9M / mm²
• Base Clock: 300 MHz versus 2017 MHz
• Boost Clock: 2400 MHz versus 2407 MHz
• Memory Size: System Shared versus 32 GB
• Memory Type: System Shared versus GDDR7
• Memory Bus Width: System Shared versus 512 bit
• Memory Bandwidth: System Dependent versus 1.79 TB/s
• Shading Units: 1,280 versus 21,760
• TMUs: 40 versus 680
• ROPs: 20 versus 176
• RT Cores: 10 versus 170
• Tensor Cores: not recorded versus 680
• Pixel Rate: 48.00 GPixel/s versus 423.6 GPixel/s
• Texture Rate: 96.00 GTexel/s versus 1,636.8 GTexel/s
• FP32: 6.144 TFLOPS versus 104.8 TFLOPS
• FP16: 12.29 TFLOPS (2:1) versus 104.8 TFLOPS (1:1)
• TDP: 25 W versus 575 W
• Slot Width: IGP versus Dual-slot
• Power Connectors: None versus 1x 16-pin
• Suggested PSU: not recorded versus 950 W
• Bus Interface: IGP versus PCIe 5.0 x16
• Display Outputs: Portable Device Dependent versus 1x HDMI 2.1b, 3x DisplayPort 2.1b
• Dimensions: not recorded versus 304 mm (length), 137 mm (height), 40 mm (width)
• Release Date: 2026-01-26 versus 2025-01-29
• Predecessor: not recorded versus GeForce 40
• Successor: not recorded versus GeForce 60
• Launch MSRP: not recorded versus 1,999 USD
• Average Benchmark Score: 1,184 versus 79,842
• Percentile: 5 versus 92