Intel Arc B370 vs NVIDIA RTX PRO 5000 72 GB Blackwell Comparison

Head-to-Head Benchmarks

The recorded database contains a single shared benchmark between these two GPUs: 3DMark Steel Nomad, a DirectX 12 workload. The results are not close. The Intel Arc B370 scores 1184 points, while the NVIDIA RTX PRO 5000 72 GB Blackwell scores 9579.5 points. This represents a delta of -87.6% for the Intel part, meaning the NVIDIA card delivers roughly 8.1 times the performance in this specific test. The NVIDIA GPU wins the only head-to-head comparison, and the Intel part records zero wins across the entire benchmark set.

The Intel Arc B370 sits at the 5th percentile of all GPUs in the database. Its average benchmark score is 1184, which places it among very old or low-end parts. Its nearest rivals in the database are the ATI Mobility Radeon HD 5570 with an average score of 1186, the ATI Radeon HD 5770 at 1190, the AMD Radeon HD 7650M at 1192, and the AMD FirePro M2000 at 1168. The Intel part trails the HD 5570 by -0.2%, trails the HD 5770 by -0.5%, trails the HD 7650M by -0.7%, and leads the FirePro M2000 by 1.4%. These delta values are all within roughly one and a half percentage points, indicating that the Arc B370 performs in the same tier as these older discrete and mobile graphics solutions.

The NVIDIA RTX PRO 5000 72 GB Blackwell sits at the 37th percentile of all GPUs in the database. Its average benchmark score across all recorded tests is 6407, which is substantially higher than the Intel part. However, the NVIDIA card's nearest rivals in the database are surprisingly modest: the NVIDIA GeForce GTX 460 SE at 6389, the NVIDIA GeForce GTX 580M at 6389, the AMD Radeon Vega 10 Mobile at 6476, and the NVIDIA Quadro M5000M at 6481. The RTX PRO 5000 leads the GTX 460 SE by 0.3%, leads the GTX 580M by 0.3%, trails the Vega 10 Mobile by -1.1%, and trails the Quadro M5000M by -1.1%. The average score is pulled down by several low PassMark results, which include 175 in DirectX 10, 219 in DirectX 11, 78 in DirectX 12, 311 in DirectX 9, 914 in G2D, 24310 in G3D, and 15667 in GPU compute. The G3D score of 24310 is by far the strongest result, while the legacy DirectX tests show much lower figures.

In the single shared benchmark, the NVIDIA card is 87.6% ahead of the Intel part. That is the only direct comparison available, and it dominates the overall picture. The Intel Arc B370's 1184 score in Steel Nomad is comparable to its own average of 1184, meaning the single test defines its entire database profile. The NVIDIA card's Steel Nomad score of 9579.5 is far above its average of 6407, indicating that this GPU performs much better in modern DirectX 12 workloads than in older API tests.

The Verdict

The data points to two completely different performance classes. The Intel Arc B370 delivers 6.144 TFLOPS of FP32 compute, while the NVIDIA RTX PRO 5000 delivers 66.94 TFLOPS. The NVIDIA card is 10.9 times higher in raw floating-point throughput. Pixel throughput tells a similar story: 48.00 GPixel/s for Intel versus 380.3 GPixel/s for NVIDIA, a factor of 7.9. Texture rate is 96.00 GTexel/s versus 1,045.9 GTexel/s, a factor of 10.9.

Memory separates the two even further. The Intel part uses system shared memory with system dependent bandwidth, so it has no dedicated VRAM figure. The NVIDIA card carries 72 GB of GDDR7 on a 384 bit bus, delivering 1.34 TB/s of bandwidth. The Intel part draws 25 W and requires no power connectors, while the NVIDIA card draws 300 W and uses a single 16-pin connector with a 700 W suggested power supply. The NVIDIA part is a dual-slot card measuring 267 mm in length, 111 mm in height, and 40 mm in width. The Intel part is an integrated graphics processor with no slot width, no power connectors, and no separate dimensions.

Anyone selecting between these two based purely on the database results should pick the NVIDIA RTX PRO 5000 for any task that benefits from high compute throughput, large memory capacity, or fast memory bandwidth. The Intel Arc B370 is the only option for a system that requires an IGP with no discrete power delivery, but its performance percentile of 5 and its closest rivals from the late 2000s and early 2010s make its capability level clear.

Where Each One Wins

The Intel Arc B370 wins in power efficiency and integration. It consumes 25 W, uses no external power connector, and fits as an IGP. It has no dedicated memory, so it cannot win on capacity or bandwidth. Its only benchmark win against a direct rival is a 1.4% lead over the AMD FirePro M2000, which is a narrow margin. It also leads 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% in terms of its own average score relative to those parts, but those are all cases where the Intel part is actually behind in raw score, with the delta values representing its deficit, not a win. The Intel part's positive delta only applies to the FirePro M2000.

The NVIDIA RTX PRO 5000 wins in every performance category that the database records. It holds the only head-to-head benchmark victory. It has higher shading units (14080 versus 1280), more texture mapping units (440 versus 40), more raster operation units (160 versus 20), more ray tracing cores (110 versus 10), and 440 tensor cores where the Intel part lists none. Its FP32 rate is 66.94 TFLOPS versus 6.144 TFLOPS. Its FP16 rate is 66.94 TFLOPS at a 1:1 ratio, while the Intel part achieves 12.29 TFLOPS at a 2:1 ratio. The NVIDIA part supports 72 GB of dedicated GDDR7 memory, while the Intel part relies on system shared memory.

FAQ

Q: Which GPU scores higher in 3DMark Steel Nomad?

A: The NVIDIA RTX PRO 5000 72 GB Blackwell scores 9579.5, while the Intel Arc B370 scores 1184. The NVIDIA card wins by 87.6%.

Q: How does the Intel Arc B370 compare to its nearest rivals?

A: 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 leads the AMD FirePro M2000 by 1.4%.

Q: What is the memory configuration of each GPU?

A: The Intel Arc B370 uses system shared memory with system dependent bandwidth. The NVIDIA RTX PRO 5000 has 72 GB of GDDR7 on a 384 bit bus with 1.34 TB/s bandwidth.

Q: What are the FP32 compute figures?

A: The Intel Arc B370 delivers 6.144 TFLOPS. The NVIDIA RTX PRO 5000 delivers 66.94 TFLOPS.

Q: What power delivery does each GPU require?

A: The Intel Arc B370 draws 25 W and uses no power connector. The NVIDIA RTX PRO 5000 draws 300 W and uses one 16-pin connector with a 700 W suggested power supply.

Q: Which GPU has tensor cores?

A: The NVIDIA RTX PRO 5000 has 440 tensor cores. The Intel Arc B370 lists none.

Architecture Differences

The Intel Arc B370 is built on Intel's 3 nm process at Intel foundry. It uses the Panther Lake chip with Xe3-LPG architecture and belongs to the Arc Graphics-M (Panther Lake) generation. The NVIDIA RTX PRO 5000 uses TSMC's 5 nm process, the GB202 chip, and Blackwell 2.0 architecture from the Blackwell PRO W (x000) generation. The NVIDIA die contains 92,200 million transistors on a 750 mm² die, giving a transistor density of 122.9M per mm². The Intel part's transistor count and die size are listed as unknown.

The NVIDIA GPU has 14080 shading units, 440 texture mapping units, 160 raster operation units, 110 ray tracing cores, and 440 tensor cores. The Intel GPU has 1280 shading units, 40 texture mapping units, 20 raster operation units, and 10 ray tracing cores, with no tensor core count listed. The NVIDIA part's FP16 performance runs at a 1:1 ratio with FP32, while the Intel part uses a 2:1 ratio, meaning the NVIDIA card maintains full throughput for both precision formats.

Both GPUs support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The NVIDIA card uses a PCIe 5.0 x16 bus interface, while the Intel part uses an IGP interface. Display outputs differ as well: the NVIDIA card provides 4x DisplayPort 2.1b, while the Intel part's outputs are portable device dependent. The NVIDIA card is a dual-slot design with a 267 mm length, 111 mm height, and 40 mm width. The Intel part is an IGP with no listed dimensions.

Specification Differences

The two GPUs differ in nearly every measurable specification. The Intel Arc B370 has a base clock of 300 MHz and a boost clock of 2400 MHz. The NVIDIA RTX PRO 5000 has a base clock of 1740 MHz and a boost clock of 2377 MHz. The NVIDIA card's memory clock is 1750 MHz with 28 Gbps effective data rate, while the Intel part's memory clock is listed as system shared.

Memory capacity, type, bus width, and bandwidth all differ. The Intel part uses system shared memory with no dedicated VRAM. The NVIDIA card uses 72 GB of GDDR7 on a 384 bit bus with 1.34 TB/s of bandwidth. The NVIDIA card's pixel rate is 380.3 GPixel/s versus 48.00 GPixel/s for Intel. Its texture rate is 1,045.9 GTexel/s versus 96.00 GTexel/s for Intel.

Power consumption differs by a factor of 12: 25 W for Intel versus 300 W for NVIDIA. The NVIDIA card requires a 16-pin power connector and a 700 W suggested power supply. The Intel part uses no power connector and has no suggested power supply. The NVIDIA card is dual-slot with physical dimensions of 267 mm by 111 mm by 40 mm. The Intel part is an IGP with no slot width, dimensions, or power connector requirements.

Release dates also differ. The NVIDIA RTX PRO 5000 was released on October 20, 2025, and its predecessor is listed as Workstation Ada. The Intel Arc B370 was released on January 26, 2026, with no predecessor listed. Both parts are marked as active in production. Neither part has a recorded launch MSRP in the database. The NVIDIA part's average benchmark score is 6407 across eight recorded tests, while the Intel part's average is 1184 from a single test. The percentile rankings place the Intel part at 5 and the NVIDIA part at 37.