Intel Arc Pro B65 vs NVIDIA GeForce RTX 5090 D Comparison

Head-to-Head Benchmarks

The NVIDIA GeForce RTX 5090 D holds a decisive performance advantage over the Intel Arc Pro B65 across every measurable category, though the Arc Pro B65 does not appear in the database's benchmark suite, leaving its recorded scores at zero. The RTX 5090 D achieves an average benchmark score of 77712, placing it in the 92nd percentile of all GPUs tracked by the database, while the Arc Pro B65 sits at the 50th percentile with no recorded benchmarks.

In compute-oriented workloads, the RTX 5090 D delivers 104.8 TFLOPS of FP32 performance versus 12.29 TFLOPS for the Arc Pro B65, a ratio of roughly 8.5 to 1. The FP16 comparison is even more lopsided: the RTX 5090 D sustains 104.8 TFLOPS at a 1:1 ratio, while the Arc Pro B65 reaches 24.58 TFLOPS at a 2:1 ratio, meaning the NVIDIA part maintains full-rate FP16 without the penalty the Intel architecture applies. Texture throughput shows a similar gap, with the RTX 5090 D producing 1,636.8 GTexel/s against 384.0 GTexel/s for the Intel card, and pixel fill rates stand at 423.6 GPixel/s versus 192.0 GPixel/s.

Memory bandwidth tells a comparable story. The RTX 5090 D moves data at 1.79 TB/s across a 512-bit GDDR7 interface, while the Arc Pro B65 manages 608.0 GB/s over a 256-bit GDDR6 bus. That is nearly three times the bandwidth, which directly influences high-resolution rendering and large dataset workloads. The RTX 5090 D's nearest rivals in the database, the AMD Radeon RX 6850M XT and NVIDIA Tesla P100 variants, sit within 1.6 to 2.4 percent of its average score, confirming that the 77712 figure represents a tightly clustered performance tier rather than an outlier.

The RTX 5090 D's raw compute resources dwarf the Intel part: 21,760 shading units versus 2,560, 680 texture mapping units versus 160, 176 ROPs versus 80, and 170 RT cores versus 20. The NVIDIA card also integrates 680 tensor cores, a feature entirely absent from the Arc Pro B65's specification sheet. These structural differences translate directly into the benchmark results, where the RTX 5090 D records a 3DMark Steel Nomad DX12 score of 14,326, a Geekbench OpenCL score of 310,674, and a Geekbench Vulkan score of 376,915.

Where Each One Wins

The RTX 5090 D wins in every workload category where the database has recorded measurements. In DirectX 12 rendering, the 3DMark Steel Nomad score of 14,326 indicates strong modern-game performance. PassMark results paint a consistent picture: the RTX 5090 D scores 434 in DirectX 9, 371 in DirectX 11, 219 in DirectX 12, and 231 in DirectX 10, with a G3D score of 44,065 and a GPU compute score of 28,396. The 2D score of 1,487 rounds out the suite.

For the Arc Pro B65, the database contains no benchmark entries, so its wins cannot be quantified. However, its specification sheet reveals areas where it may hold practical advantages. The Intel card draws 200 W of power against 575 W for the NVIDIA part, and its suggested power supply is 550 W versus 950 W. It uses a single 8-pin power connector rather than the 16-pin connector required by the RTX 5090 D. These factors make the Arc Pro B65 easier to integrate into systems with lower power budgets, though the database does not record performance data to confirm what that efficiency buys in real workloads.

The Arc Pro B65 also offers four DisplayPort 2.1 outputs, while the RTX 5090 D splits its outputs between one HDMI 2.1b and three DisplayPort 2.1b connectors. For multi-monitor setups that rely exclusively on DisplayPort, the Intel card provides one additional native DisplayPort connection. The RTX 5090 D's launch MSRP is 2,299 USD, while the Arc Pro B65 has no recorded launch price.

Architecture Differences

The two GPUs come from fundamentally different design philosophies. Intel's Arc Pro B65 uses the BMG-G21 chip built on the Xe2-HPG architecture, part of the Battlemage (Pro Series) generation. NVIDIA's RTX 5090 D uses the GB202 chip on the Blackwell 2.0 architecture, part of the GeForce 50-series. Both are fabricated by TSMC on a 5 nm process, but the similarities end there.

The transistor counts reveal the scale gap. The RTX 5090 D packs 92,200 million transistors on a 750 mm² die, achieving a density of 122.9 million transistors per square millimeter. The Arc Pro B65 contains 19,600 million transistors on a 272 mm² die, with a density of 72.1 million per square millimeter. The NVIDIA die is nearly three times larger in area and holds more than four times as many transistors.

Memory technology differs as well. The RTX 5090 D uses GDDR7 at 1750 MHz with 28 Gbps effective speed, while the Arc Pro B65 uses GDDR6 at 2375 MHz with 19 Gbps effective speed. Both cards offer 32 GB of memory, but the NVIDIA part's 512-bit bus versus the Intel part's 256-bit bus accounts for the bandwidth disparity. The RTX 5090 D's FP16 throughput matches its FP32 at 104.8 TFLOPS, indicating a 1:1 ratio architecture, whereas the Arc Pro B65's FP16 runs at 24.58 TFLOPS against 12.29 TFLOPS FP32, a 2:1 ratio that halves throughput for half-precision work.

Ray tracing hardware differs sharply: 170 RT cores in the NVIDIA part versus 20 in the Intel part. Tensor cores exist only on the RTX 5090 D, with 680 units, and the NVIDIA card's 21,760 shading units outnumber Intel's 2,560 by a factor of 8.5. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, and both use a PCIe 5.0 x16 bus interface.

Specification Differences

The two cards diverge on nearly every measurable specification except memory capacity and bus interface. Both offer 32 GB of VRAM and PCIe 5.0 x16 connectivity, but the memory type, bus width, and bandwidth all favor the NVIDIA part. Clock speeds differ in opposite directions: the Arc Pro B65 runs at 2400 MHz base and boost, while the RTX 5090 D runs at 2017 MHz base and 2407 MHz boost. The Intel card maintains a flat clock, while the NVIDIA card boosts 390 MHz above its base.

The Arc Pro B65 uses GDDR6 memory clocked at 2375 MHz (19 Gbps effective) on a 256-bit bus for 608.0 GB/s bandwidth. The RTX 5090 D uses GDDR7 at 1750 MHz (28 Gbps effective) on a 512-bit bus for 1.79 TB/s bandwidth. Memory clock measurements are not directly comparable across generations because GDDR7 achieves higher effective data rates at lower physical clocks.

Power requirements show the largest practical gap. The Arc Pro B65 has a TDP of 200 W and a suggested PSU of 550 W, while the RTX 5090 D has a TDP of 575 W and a suggested PSU of 950 W. The Intel card uses one 8-pin power connector; the NVIDIA card uses one 16-pin connector. Both are dual-slot designs. Physical dimensions are recorded only for the RTX 5090 D: 304 mm length, 137 mm height, and 48 mm width.

Release dates place the RTX 5090 D first at 2025-01-29, with the Arc Pro B65 following on 2026-03-31. The NVIDIA card lists a predecessor (GeForce 40) and successor (GeForce 60), while the Intel card lists neither. Display outputs favor Intel by count (four DisplayPort 2.1 versus one HDMI 2.1b plus three DisplayPort 2.1b on NVIDIA), though the NVIDIA card adds an HDMI option.

FAQ

Q: Which GPU has more shader cores?

A: The NVIDIA GeForce RTX 5090 D has 21,760 shading units, while the Intel Arc Pro B65 has 2,560. The NVIDIA card also has 680 texture mapping units, 176 ROPs, 170 RT cores, and 680 tensor cores, compared to 160 TMUs, 80 ROPs, and 20 RT cores on the Intel part.

Q: Do both cards have 32 GB of memory?

A: Yes, both the Intel Arc Pro B65 and NVIDIA GeForce RTX 5090 D have 32 GB of VRAM. However, the RTX 5090 D uses GDDR7 on a 512-bit bus for 1.79 TB/s bandwidth, while the Arc Pro B65 uses GDDR6 on a 256-bit bus for 608.0 GB/s.

Q: What is the power draw difference?

A: The Intel Arc Pro B65 has a TDP of 200 W with a suggested power supply of 550 W, while the NVIDIA GeForce RTX 5090 D has a TDP of 575 W with a suggested power supply of 950 W. The Intel card uses one 8-pin connector; the NVIDIA card uses one 16-pin connector.

Q: How does the RTX 5090 D compare to its nearest rivals in the database?

A: The RTX 5090 D's average benchmark score is 77712, putting it 1.1 percent ahead of the AMD Radeon RX 6650M XT (76904), 1.6 percent behind the AMD Radeon RX 6850M XT (78940), 2.1 percent behind the NVIDIA Tesla P100 PCIe 12 GB (79396), and 2.4 percent behind the NVIDIA Tesla P100 PCIe 16 GB (79605). It sits in the 92nd percentile of all GPUs.

Q: What are the recorded benchmark scores for the Arc Pro B65?

A: The database lists no benchmark entries for the Intel Arc Pro B65. Its average benchmark score is 0, and its percentile versus all GPUs is 50. The RTX 5090 D, by contrast, has ten recorded benchmark scores across 3DMark, Geekbench, and PassMark tests.

Q: Which card supports ray tracing and tensor operations?

A: Both cards support ray tracing, but with different hardware counts: the RTX 5090 D has 170 RT cores versus 20 on the Arc Pro B65. Tensor cores exist only on the NVIDIA card (680 units); the Intel card lists no tensor core count.

The Verdict

The data points to a clear separation in performance class. The NVIDIA GeForce RTX 5090 D records an average benchmark score of 77712 and sits in the 92nd percentile of all GPUs, with verified results across DirectX 9 through DirectX 12, OpenCL, Vulkan, and compute workloads. The Intel Arc Pro B65 has no recorded benchmarks and a 50th percentile standing, so the database provides no evidence of competitive performance.

For workloads that demand maximum throughput, the RTX 5090 D is the only choice supported by measurements. Its FP32 output of 104.8 TFLOPS, FP16 output of 104.8 TFLOPS at 1:1 ratio, 1.79 TB/s memory bandwidth, and 1,636.8 GTexel/s texture rate are all recorded figures. The Arc Pro B65's 12.29 TFLOPS FP32 and 608.0 GB/s bandwidth place it in a different performance tier, even before considering the missing tensor core capability.

The Arc Pro B65 does offer a lower power profile at 200 W TDP with a 550 W suggested PSU, and it provides four DisplayPort 2.1 outputs versus three on the NVIDIA card. For systems constrained by power delivery or those requiring maximum DisplayPort connectivity, the Intel part has structural advantages. But the database records no benchmark results to quantify what those advantages deliver in practice, whereas the RTX 5090 D's performance is fully documented across ten separate tests.

The RTX 5090 D's nearest rivals in the database, the AMD Radeon RX 6850M XT and Tesla P100 variants, sit within 2.4 percent of its average score, indicating tight competition at the top tier. The Arc Pro B65, with no scores and no rival comparisons, cannot be assessed on the same terms. Users seeking measured performance should select the RTX 5090 D; users prioritizing lower power draw and DisplayPort count may consider the Arc Pro B65, but do so without benchmark validation.