Intel Arc Pro B65 vs Lisuan Tech LX ULTRA Comparison

Intel Arc Pro B65 and Lisuan Tech LX ULTRA are both active, dual-slot workstation graphics cards aimed at professional compute and rendering workloads, but the data shows they target different corners of that market. The Intel Arc Pro B65 uses a larger, denser 5 nm chip with a narrow but fast memory subsystem, while the Lisuan Tech LX ULTRA uses a roomier 6 nm design with a much wider shader array and higher raw FP32 throughput. Benchmark results, which currently show both cards sitting at the 50th percentile among all GPUs with no recorded head-to-head scores, do not yet separate them by measured performance. The specification differences, however, reveal a clear split: the LX ULTRA leads in arithmetic throughput and texture fill, while the Arc Pro B65 counters with higher memory bandwidth, a newer PCIe interface, and a more advanced display output standard.

Where Each One Wins

The Intel Arc Pro B65 wins in memory bandwidth and interface generation. It carries 32 GB of GDDR6 on a 256-bit bus, producing 608.0 GB/s of bandwidth, which is 40.7% higher than the Lisuan Tech LX ULTRA's 432.0 GB/s from 24 GB on a 192-bit bus. For workloads that stream large datasets, such as high-resolution texture sampling, large model inference, or multi-monitor 8K compositing, the Arc Pro B65's wider memory path provides a measurable advantage. The card also uses PCIe 5.0 x16, doubling the per-lane transfer rate of the LX ULTRA's PCIe 4.0 x16, which matters for GPU-to-CPU transfers and NVMe-style data ingestion. Display output is another clear win: the Arc Pro B65 offers four DisplayPort 2.1 connections, while the LX ULTRA uses four DisplayPort 1.4a, so the Intel card supports higher refresh rates and bandwidth per display.

The Lisuan Tech LX ULTRA wins in compute throughput and texture processing. Its 6144 shading units deliver 24.58 TFLOPS of FP32 performance, exactly double the Arc Pro B65's 12.29 TFLOPS. FP16 performance follows the same pattern: 49.15 TFLOPS versus 24.58 TFLOPS. That doubling extends to texture rate, with 384.0 GTexel/s on the LX ULTRA matching the Arc Pro B65's texture rate despite the LX ULTRA having 192 TMUs versus 160 on the Intel card (the Intel card achieves the same texture rate through a higher clock). The LX ULTRA also has more ROPs, 96 versus 80, though both cards produce the same pixel rate of 192.0 GPixel/s due to clock differences. For pure shader-bound tasks, such as scientific simulation, ray tracing BVH traversal, or AI preprocessing, the LX ULTRA's wider execution resources give it a theoretical 2x edge in FP32 and FP16 throughput.

Architecture Differences

The two cards use different manufacturing nodes and chip designs. The Intel Arc Pro B65 uses the BMG-G21 chip built on TSMC's 5 nm process, with 19,600 million transistors on a 272 mm² die, yielding a transistor density of 72.1 million per square millimeter. The Lisuan Tech LX ULTRA uses the 7G105 chip on TSMC's 6 nm process, but its transistor count and die size are recorded as unknown, so density cannot be calculated. The process node difference gives the Intel card a manufacturing advantage in terms of density and power efficiency per transistor, though the LX ULTRA compensates with a larger shader count.

The Intel card runs at a fixed 2400 MHz base and boost clock, with memory clocked at 2375 MHz (19 Gbps effective). The LX ULTRA's core clocks are not recorded, but its memory runs at 2250 MHz (18 Gbps effective). The absence of core clock data for the LX ULTRA means the FP32 and texture rate figures are the only reliable compute indicators, and they show that the LX ULTRA's 6144 shading units at an unspecified clock produce exactly 2x the FP32 throughput of the Intel card's 2560 shading units at 2400 MHz. This implies the LX ULTRA runs at a similar or somewhat lower clock, but the shader count difference dominates.

Memory architecture diverges sharply. The Arc Pro B65 uses 32 GB GDDR6 on a 256-bit bus, while the LX ULTRA uses 24 GB on a 192-bit bus. The Intel card's bus width is 33.3% wider, and its effective memory speed is 5.6% higher, resulting in the 40.7% bandwidth advantage. The LX ULTRA's 24 GB capacity is still substantial but trails by 25% in capacity. For rendering scenes that exceed 24 GB, the Arc Pro B65's larger memory pool is decisive, while the LX ULTRA's lower bandwidth may bottleneck shader-heavy workloads that repeatedly fetch textures.

The power delivery systems differ. The Arc Pro B65 has a 200 W TDP with a single 8-pin power connector, while the LX ULTRA has a 225 W TDP with a single 16-pin connector. Both cards recommend a 550 W power supply. The LX ULTRA draws 12.5% more power, which aligns with its higher shader count and FP32 throughput, but the Intel card achieves its memory bandwidth advantage within a lower power envelope. The LX ULTRA also has a specific physical footprint: 268 mm length, 112 mm height, and 40 mm width, while the Intel card's dimensions are not recorded.

API support shows partial overlap. Both cards support DirectX 12 Ultimate (12_2) and OpenGL 4.6. The Arc Pro B65 supports Vulkan 1.4, while the LX ULTRA supports Vulkan 1.3. The Intel card also has 20 ray tracing cores, while the LX ULTRA's RT core count is not recorded, so ray tracing capability on the Lisuan card cannot be quantified from the data.

Head-to-Head Benchmarks

The recorded head-to-head benchmark array is empty, and both cards have zero benchmark scores and zero wins in the comparison. The percentileVsAllGpus field places both at 50, meaning they sit exactly at the median of all GPUs in the database, but no specific benchmark numbers exist to compare. The absence of data forces the analysis to rely on the documented specification differences.

The most significant quantitative gap is FP32 throughput. The LX ULTRA's 24.58 TFLOPS is exactly 2.0 times the Arc Pro B65's 12.29 TFLOPS. In FP16, the LX ULTRA's 49.15 TFLOPS is also exactly 2.0 times the Intel card's 24.58 TFLOPS. These ratios are clean and repeatable across both precision formats, indicating that the LX ULTRA's shader array is the dominant compute resource. The texture rate is identical at 384.0 GTexel/s, meaning the LX ULTRA's 192 TMUs at a lower clock match the Arc Pro B65's 160 TMUs at 2400 MHz. The pixel rate is also identical at 192.0 GPixel/s, despite the LX ULTRA having 96 ROPs versus 80, because the Intel card's higher clock compensates.

Memory bandwidth is where the Arc Pro B65 reverses the trend. The Intel card's 608.0 GB/s is 176.0 GB/s higher than the LX ULTRA's 432.0 GB/s, a 40.7% lead. This bandwidth gap is larger in percentage terms than the FP32 gap (100% in favor of the LX ULTRA) but smaller in absolute magnitude. For a workload that is memory-bound, such as large matrix multiplications that repeatedly read the same weights, the Arc Pro B65's bandwidth advantage could narrow or overturn the LX ULTRA's compute lead.

The memory capacity difference is 8 GB, with the Arc Pro B65 holding 32 GB versus 24 GB on the LX ULTRA. That 33.3% capacity advantage means the Intel card can hold larger datasets entirely in VRAM, avoiding PCIe transfers. The PCIe 5.0 interface on the Arc Pro B65 further reduces transfer overhead, while the LX ULTRA's PCIe 4.0 interface halves the per-lane bandwidth. The display output difference is generation-based: DisplayPort 2.1 on the Intel card supports higher resolutions and refresh rates than DisplayPort 1.4a on the Lisuan card, though both provide four outputs.

Power consumption favors the Arc Pro B65 at 200 W versus 225 W, a 12.5% reduction. The Intel card achieves its memory bandwidth and interface advantages with lower power draw, while the LX ULTRA's higher shader count requires more energy. The power connector difference (8-pin versus 16-pin) reflects the higher power ceiling on the LX ULTRA, though both recommend the same 550 W PSU.

FAQ

Q: Which card has higher raw compute performance?

A: The Lisuan Tech LX ULTRA delivers 24.58 TFLOPS FP32 and 49.15 TFLOPS FP16, both exactly double the Intel Arc Pro B65's 12.29 TFLOPS FP32 and 24.58 TFLOPS FP16. The LX ULTRA's 6144 shading units versus 2560 on the Arc Pro B65 drive this 2x advantage.

Q: Which card has more memory bandwidth?

A: The Intel Arc Pro B65 has 608.0 GB/s from 32 GB GDDR6 on a 256-bit bus, which is 40.7% higher than the LX ULTRA's 432.0 GB/s from 24 GB on a 192-bit bus. The Intel card also runs its memory at 2375 MHz (19 Gbps effective) versus 2250 MHz (18 Gbps effective) on the LX ULTRA.

Q: Do both cards support the same graphics APIs?

A: Both support DirectX 12 Ultimate (12_2) and OpenGL 4.6. The Intel Arc Pro B65 supports Vulkan 1.4, while the Lisuan Tech LX ULTRA supports Vulkan 1.3. The Intel card also has 20 ray tracing cores, while the LX ULTRA's RT core count is not recorded.

Q: Which card has a newer PCIe interface?

A: The Intel Arc Pro B65 uses PCIe 5.0 x16, while the Lisuan Tech LX ULTRA uses PCIe 4.0 x16. The Intel card's interface provides double the per-lane transfer rate, which benefits direct memory access and data streaming.

Q: What are the power requirements for each card?

A: The Intel Arc Pro B65 has a 200 W TDP and uses a single 8-pin power connector. The Lisuan Tech LX ULTRA has a 225 W TDP and uses a single 16-pin connector. Both cards recommend a 550 W power supply.

Q: Which card has more display outputs and newer display technology?

A: Both cards have four DisplayPort outputs. The Intel Arc Pro B65 uses DisplayPort 2.1, while the Lisuan Tech LX ULTRA uses DisplayPort 1.4a. DisplayPort 2.1 supports higher bandwidth and refresh rates.

The Verdict

The data points to a clear use-case split. For compute-bound workloads that rely on raw shader throughput, the Lisuan Tech LX ULTRA is the stronger choice. Its 6144 shading units and 24.58 TFLOPS FP32 performance provide exactly double the arithmetic capacity of the Intel Arc Pro B65, and its FP16 throughput of 49.15 TFLOPS doubles the Intel card's 24.58 TFLOPS. Texture rate matches at 384.0 GTexel/s, but the LX ULTRA achieves this with more TMUs, suggesting headroom for higher clocks. The LX ULTRA's 225 W TDP with a 16-pin connector indicates it can sustain this compute density, though its memory bandwidth of 432.0 GB/s may bottleneck shader-heavy loops that repeatedly fetch data.

For memory-bound and interface-sensitive workloads, the Intel Arc Pro B65 is the better pick. Its 608.0 GB/s bandwidth is 40.7% higher, its 32 GB capacity is 33.3% larger, and its PCIe 5.0 x16 interface is two generations ahead of the LX ULTRA's PCIe 4.0 x16. The DisplayPort 2.1 outputs also future-proof multi-monitor setups. The Arc Pro B65 achieves all of this at a lower 200 W TDP, making it the more power-efficient card per unit of bandwidth. The identical pixel rate of 192.0 GPixel/s and texture rate of 384.0 GTexel/s mean neither card has an edge in basic rasterization output, leaving the decision to the specific ratio of compute to memory traffic in the target application.

The lack of recorded benchmark scores means no empirical performance ranking exists yet. The 50th percentile for both cards in the database is a neutral placement, not a comparison. Users who prioritize FP32/FP16 compute for simulation, AI training, or dense rendering should choose the LX ULTRA. Users who prioritize memory capacity, bandwidth, and modern connectivity for large datasets, video editing, or multi-display output should choose the Arc Pro B65.

Specification Differences

The following fields differ between the Intel Arc Pro B65 and the Lisuan Tech LX ULTRA:

• Chip: Intel uses BMG-G21; Lisuan uses 7G105.
• Architecture: Intel uses Xe2-HPG; Lisuan uses TrueGPU.
• Generation: Intel is Battlemage (Pro Series); Lisuan is 7G100.
• Process node: Intel is 5 nm; Lisuan is 6 nm.
• Transistors: Intel has 19,600 million; Lisuan is unknown.
• Die size: Intel is 272 mm²; Lisuan is unknown.
• Transistor density: Intel is 72.1M / mm²; Lisuan is null.
• Base clock: Intel is 2400 MHz; Lisuan is null.
• Boost clock: Intel is 2400 MHz; Lisuan is null.
• Memory clock: Intel is 2375 MHz (19 Gbps effective); Lisuan is 2250 MHz (18 Gbps effective).
• Memory size: Intel is 32 GB; Lisuan is 24 GB.
• Memory bus width: Intel is 256 bit; Lisuan is 192 bit.
• Memory bandwidth: Intel is 608.0 GB/s; Lisuan is 432.0 GB/s.
• Shading units: Intel has 2560; Lisuan has 6144.
• TMUs: Intel has 160; Lisuan has 192.
• ROPs: Intel has 80; Lisuan has 96.
• RT cores: Intel has 20; Lisuan is null.
• FP32: Intel is 12.29 TFLOPS; Lisuan is 24.58 TFLOPS.
• FP16: Intel is 24.58 TFLOPS (2:1); Lisuan is 49.15 TFLOPS (2:1).
• TDP: Intel is 200 W; Lisuan is 225 W.
• Power connectors: Intel uses 1x 8-pin; Lisuan uses 1x 16-pin.
• Bus interface: Intel is PCIe 5.0 x16; Lisuan is PCIe 4.0 x16.
• Display outputs: Intel has 4x DisplayPort 2.1; Lisuan has 4x DisplayPort 1.4a.
• Vulkan support: Intel is 1.4; Lisuan is 1.3.
• Dimensions: Intel is null; Lisuan is 268 mm length, 112 mm height, 40 mm width.
• Release date: Intel is 2026-03-31; Lisuan is 2026-03-16.