Intel Arc B770 vs NVIDIA GeForce RTX 5090 Comparison

FAQ

Q: How do the two GPUs differ in raw compute performance?

A: The NVIDIA GeForce RTX 5090 delivers 104.8 TFLOPS of FP32 compute, while the Intel Arc B770 delivers 19.66 TFLOPS. The RTX 5090 also provides 104.8 TFLOPS of FP16 (1:1 ratio), whereas the Arc B770 offers 39.32 TFLOPS of FP16 (2:1 ratio).

Q: What are the memory configurations of each card?

A: The Intel Arc B770 uses 16 GB of GDDR6 memory on a 256-bit bus with 512.0 GB/s bandwidth. The NVIDIA GeForce RTX 5090 uses 32 GB of GDDR7 memory on a 512-bit bus with 1.79 TB/s bandwidth.

Q: Which card has a higher transistor count and die size?

A: The RTX 5090 contains 92,200 million transistors on a 750 mm² die, while the Arc B770’s transistor count is unknown but its die size is 368 mm². The RTX 5090 also has a recorded transistor density of 122.9M per mm².

Q: What are the power requirements for each GPU?

A: The Arc B770 has a 225 W TDP and a suggested 550 W power supply. The RTX 5090 has a 575 W TDP and a suggested 950 W power supply.

Q: Which card ranks higher in the database percentile?

A: The RTX 5090 sits at the 92nd percentile of all GPUs, while the Arc B770 sits at the 50th percentile. The RTX 5090’s average benchmark score is 79,842, whereas the Arc B770 has no recorded benchmark entries.

Q: What are the release dates for these cards?

A: The RTX 5090 was released on 2025-01-29, and the Arc B770 was released on 2025-12-31.

Architecture Differences

The two cards are built on entirely different design philosophies. The Intel Arc B770 uses the BMG-G31 chip, based on the Xe2-HPG architecture, and belongs to the Battlemage (Arc 7) generation. The NVIDIA GeForce RTX 5090 uses the GB202 chip, based on the Blackwell 2.0 architecture, and belongs to the GeForce 50 series. Both are fabricated on a 5 nm process at TSMC, but the similarities end there.

The RTX 5090 packs 21,760 shading units, 680 TMUs, 176 ROPs, 170 RT cores, and 680 tensor cores. The Arc B770 has 4,096 shading units, 256 TMUs, 128 ROPs, and 32 RT cores. The RTX 5090 also includes dedicated tensor cores, a feature not listed for the Arc B770. These structural differences explain the large gap in compute throughput: the RTX 5090’s FP32 output is 104.8 TFLOPS versus 19.66 TFLOPS for the Arc B770.

Memory architecture also diverges sharply. The RTX 5090 uses 32 GB of GDDR7 on a 512-bit bus, reaching 1.79 TB/s of bandwidth. The Arc B770 uses 16 GB of GDDR6 on a 256-bit bus, with 512.0 GB/s of bandwidth. The RTX 5090’s memory clock is rated at 1750 MHz with 28 Gbps effective speed, while the Arc B770 runs at 2000 MHz with 16 Gbps effective speed.

Clock behavior differs as well. The Arc B770 has a base clock of 2100 MHz and a boost clock of 2400 MHz. The RTX 5090 has a lower base clock of 2017 MHz but a slightly higher boost clock of 2407 MHz. Despite the similar boost figures, the RTX 5090’s massive parallel resource count drives far higher fill rates: 423.6 GPixel/s and 1,636.8 GTexel/s versus 307.2 GPixel/s and 614.4 GTexel/s for the Arc B770.

Interface and connectivity also differ. The RTX 5090 uses PCIe 5.0 x16, while the Arc B770 uses PCIe 4.0 x16. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Display outputs are similar, with the Arc B770 offering 1x HDMI 2.1a and 3x DisplayPort 2.1, while the RTX 5090 offers 1x HDMI 2.1b and 3x DisplayPort 2.1b.

Head-to-Head Benchmarks

The RTX 5090 has ten recorded benchmark entries in the database; the Arc B770 has none. This means all quantitative comparisons rely on the RTX 5090’s recorded scores and its nearest rival data. The RTX 5090’s average benchmark score is 79,842, placing it at the 92nd percentile of all GPUs. Its nearest rival, the NVIDIA Tesla P100 PCIe 16 GB, scores 79,605, a delta of 0.3%. The Tesla P100 PCIe 12 GB scores 79,396, a 0.6% delta. The AMD Radeon RX 6850M XT scores 78,940, a 1.1% delta. The AMD Radeon Pro Vega 64X scores 80,959, putting it 1.4% ahead of the RTX 5090.

In individual benchmark tests, the RTX 5090 shows a wide variance across workloads. Its strongest synthetic results appear in Geekbench Vulkan with a score of 376,728 and Geekbench OpenCL with 334,370. In 3DMark Steel Nomad DX12, it scores 18,355. Passmark results are mixed: G3D scores 39,650, GPU Compute scores 26,756, G2D scores 1,413, DirectX 9 scores 395, DirectX 11 scores 341, DirectX 10 scores 226, and DirectX 12 scores 185.

The Arc B770 has no benchmark scores, no wins, and no nearest rival data in the database. Its percentile of 50 indicates an average position relative to all GPUs, but there is no recorded performance data to substantiate any workload-specific strength. The head-to-head comparison is therefore one-sided: the RTX 5090 is the only card with measurable results.

Specification Differences

The two cards differ in nearly every measurable specification. The RTX 5090 uses the GB202 chip with 92,200 million transistors, while the Arc B770 uses the BMG-G31 chip with an unknown transistor count. Die sizes are 750 mm² for the RTX 5090 and 368 mm² for the Arc B770. The RTX 5090 has a recorded transistor density of 122.9M per mm²; the Arc B770 has none listed.

Shading resources differ by a factor of roughly five: 21,760 shading units on the RTX 5090 versus 4,096 on the Arc B770. TMUs are 680 versus 256, ROPs are 176 versus 128, RT cores are 170 versus 32, and tensor cores are 680 versus none listed. Pixel rates are 423.6 GPixel/s versus 307.2 GPixel/s, and texture rates are 1,636.8 GTexel/s versus 614.4 GTexel/s.

Memory capacity is 32 GB versus 16 GB, with GDDR7 versus GDDR6 type, 512-bit versus 256-bit bus width, and 1.79 TB/s versus 512.0 GB/s bandwidth. Power consumption is 575 W versus 225 W TDP, with suggested power supplies of 950 W versus 550 W. Power connectors are 1x 16-pin for the RTX 5090 and 1x 6-pin plus 1x 8-pin for the Arc B770.

Physical dimensions are listed only for the RTX 5090: 304 mm length, 137 mm height, and 40 mm width. Both are dual-slot cards. Bus interfaces are PCIe 5.0 x16 for the RTX 5090 and PCIe 4.0 x16 for the Arc B770. Release dates are 2025-01-29 for the RTX 5090 and 2025-12-31 for the Arc B770. The RTX 5090 has a stated production status of Active and a launch MSRP of 1,999 USD.

The Verdict

The recorded data points in one direction. The RTX 5090 is a fully measured product with a 92nd percentile ranking, an average benchmark score of 79,842, and ten individual test results. The Arc B770 has no benchmark entries, no average score, and no rival comparisons, leaving its 50th percentile as the only positional signal. Given that the RTX 5090 outperforms its nearest recorded rival, the AMD Radeon Pro Vega 64X, by 1.4%, and that its average score of 79,842 sits well above the Arc B770’s zero recorded score, the RTX 5090 is the only card with demonstrated performance in the database.

The Arc B770’s specifications, however, describe a lower-power part: 225 W TDP, 16 GB of GDDR6, and 19.66 TFLOPS of FP32. Its smaller die, 368 mm², and reduced resource counts point to a different market segment. The RTX 5090’s 575 W TDP, 32 GB of GDDR7, and 104.8 TFLOPS place it in a much higher performance tier. For any workload requiring maximum compute throughput, memory bandwidth, or ray tracing resources, the RTX 5090 is the clear choice based on available measurements.

Where Each One Wins

The RTX 5090 wins in every category with recorded data. Its 104.8 TFLOPS of FP32 and FP16 (1:1) performance exceeds the Arc B770’s 19.66 TFLOPS FP32 and 39.32 TFLOPS FP16 (2:1) by a wide margin. Its 1.79 TB/s memory bandwidth is more than three times the Arc B770’s 512.0 GB/s. Its 170 RT cores and 680 tensor cores provide dedicated hardware that the Arc B770 does not list. Its pixel rate of 423.6 GPixel/s and texture rate of 1,636.8 GTexel/s both surpass the Arc B770’s 307.2 GPixel/s and 614.4 GTexel/s.

The Arc B770 does hold advantages in a few narrow areas. Its 225 W TDP is lower than the RTX 5090’s 575 W, and its suggested power supply of 550 W is far below the 950 W required for the RTX 5090. Its 5 nm process and 368 mm² die suggest a more compact implementation. Its dual power connectors, 1x 6-pin and 1x 8-pin, are more flexible than the RTX 5090’s single 16-pin connector. The Arc B770 also has a higher base clock, 2100 MHz versus 2017 MHz, though its boost clock of 2400 MHz is marginally lower than the RTX 5090’s 2407 MHz.

For users prioritizing low power draw, smaller die size, or a dual-connector power layout, the Arc B770 has a specification-level appeal. For any performance-based decision, the RTX 5090’s recorded benchmark results, 92nd percentile standing, and average score of 79,842 make it the only card with demonstrated capability in this comparison.