NVIDIA Jetson T4000 vs Lisuan Tech LX ULTRA Comparison

Head-to-Head Benchmarks

The recorded database contains no head-to-head benchmark entries for the NVIDIA Jetson T4000 versus the Lisuan Tech LX ULTRA. The win tally stands at zero for both parts, and neither device has an average benchmark score above zero. This absence of direct measurement data means the comparison must rely entirely on the architectural specifications and computed throughput rates captured in the database.

The most striking numerical contrast appears in raw compute throughput. The Lisuan Tech LX ULTRA delivers 24.58 TFLOPS of FP32 performance, which is 5.23 times the 4.700 TFLOPS offered by the NVIDIA Jetson T4000. In FP16 workloads, the gap widens further: the LX ULTRA reaches 49.15 TFLOPS using a 2:1 ratio, while the Jetson T4000 maintains a 1:1 ratio at 4.700 TFLOPS. That puts the LX ULTRA at roughly 10.5 times the FP16 throughput of the Jetson T4000.

Texture and pixel fill rates follow the same pattern. The LX ULTRA processes 384.0 GTexel/s, compared to 73.44 GTexel/s for the Jetson T4000, a 5.23x advantage. Pixel throughput shows 192.0 GPixel/s versus 24.48 GPixel/s, an 8x difference in favor of the Lisuan part. These ratios derive directly from the shading unit, texture mapping unit, and ROP counts: 6144 shading units, 192 TMUs, and 96 ROPs on the LX ULTRA versus 1536 shading units, 48 TMUs, and 16 ROPs on the Jetson T4000.

Memory bandwidth tells a different story. The LX ULTRA uses 24 GB of GDDR6 on a 192-bit bus, achieving 432.0 GB/s. The Jetson T4000 pairs 64 GB of LPDDR5X with a 256-bit bus, reaching 273.2 GB/s. The Lisuan card holds a 1.58x bandwidth advantage despite the narrower bus, because its memory clock runs at 2250 MHz (18 Gbps effective) versus 1067 MHz (8.5 Gbps effective) on the NVIDIA part. Capacity, however, belongs to NVIDIA: 64 GB versus 24 GB, a 2.67x difference.

Power consumption and physical design diverge sharply. The Jetson T4000 draws 90 W and functions as an integrated graphics processor (IGP) with no power connectors and no display outputs. The LX ULTRA is a dual-slot card requiring one 16-pin connector, rated at 225 W, with four DisplayPort 1.4a outputs. System power supply recommendations in the database are 250 W for the Jetson T4000 and 550 W for the LX ULTRA.

The Jetson T4000 uses the GB10B chip on TSMC 5 nm process with a 391 mm² die. The LX ULTRA uses the 7G105 chip on TSMC 6 nm, with die size listed as unknown. Both are classified as active production parts, and both sit at the 50th percentile against all GPUs in the database, with zero recorded benchmark scores. The Jetson T4000 has a launch MSRP of 1,999 USD; the LX ULTRA has no launch MSRP recorded.

The Verdict

The data points to two devices built for fundamentally different deployment contexts, not direct competition. The Lisuan Tech LX ULTRA wins decisively on every raw throughput metric: FP32, FP16, pixel fill, texture fill, and memory bandwidth. It also provides standard display outputs, a conventional PCIe 4.0 x16 interface, and full DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.3 API support. The Jetson T4000 offers no display outputs, no graphics API support, and a lower PCIe 5.0 x8 connection.

The Jetson T4000 counters with 64 GB of memory, 2.67x the capacity of the LX ULTRA, and a 90 W power envelope that is 60% lower than the LX ULTRA's 225 W. It also uses a smaller physical footprint: 87 mm length, 100 mm height, 15 mm width versus 268 mm length, 112 mm height, 40 mm width. The NVIDIA part integrates as an IGP, requiring no separate power connector, while the Lisuan card needs a 16-pin connector and a 550 W recommended PSU.

Benchmark results indicate the LX ULTRA is the performance-oriented part, suited for workloads that demand maximum compute and graphics throughput. The Jetson T4000 is the capacity-and-efficiency-oriented part, suited for memory-bound or power-constrained environments where 64 GB of LPDDR5X and a 90 W budget matter more than raw TFLOPS. No recorded benchmark scores exist for either device, so the verdict rests on architectural specifications and the computed rates listed in the database.

Where Each One Wins

Lisuan Tech LX ULTRA wins on compute density. Its 24.58 TFLOPS FP32 and 49.15 TFLOPS FP16 are the highest numbers in this comparison. The 2:1 FP16 ratio indicates dedicated half-precision throughput, whereas the Jetson T4000's 1:1 ratio means FP16 runs at the same rate as FP32. For AI inference or scientific computing using half precision, the LX ULTRA offers 10.5x the throughput.

Lisuan Tech LX ULTRA wins on graphics and display. The database lists 4x DisplayPort 1.4a outputs, DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.3. The Jetson T4000 lists no outputs and N/A for all three APIs. Any workload requiring rasterization, ray tracing, or video output must use the LX ULTRA.

Lisuan Tech LX ULTRA wins on memory bandwidth. At 432.0 GB/s, it exceeds the Jetson T4000's 273.2 GB/s by 58%. The 18 Gbps effective memory speed on a 192-bit bus outperforms the 8.5 Gbps effective speed on a 256-bit bus.

NVIDIA Jetson T4000 wins on memory capacity. 64 GB versus 24 GB is a 2.67x advantage. Large language model weights, big datasets, or in-memory databases favor the NVIDIA part.

NVIDIA Jetson T4000 wins on power efficiency. At 90 W, it consumes 40% of the LX ULTRA's 225 W. The suggested PSU of 250 W versus 550 W reinforces this gap. The Jetson T4000 also requires no power connector and fits in an IGP slot, simplifying system integration.

NVIDIA Jetson T4000 wins on physical footprint and interface generation. The 87 mm length and 15 mm width versus 268 mm and 40 mm make it suitable for compact or embedded chassis. Its PCIe 5.0 x8 interface, while narrower than the LX ULTRA's PCIe 4.0 x16, offers a newer bus generation.

FAQ

Q: Which GPU has higher FP32 performance?

A: The Lisuan Tech LX ULTRA delivers 24.58 TFLOPS, which is 5.23 times the 4.700 TFLOPS of the NVIDIA Jetson T4000.

Q: How much memory does each card have?

A: The Jetson T4000 has 64 GB of LPDDR5X on a 256-bit bus. The LX ULTRA has 24 GB of GDDR6 on a 192-bit bus.

Q: Can the NVIDIA Jetson T4000 drive displays?

A: No. The database lists no display outputs for the Jetson T4000. The LX ULTRA provides 4x DisplayPort 1.4a outputs.

Q: What is the power consumption difference?

A: The Jetson T4000 is rated at 90 W with a suggested PSU of 250 W. The LX ULTRA is rated at 225 W with a suggested PSU of 550 W. The Jetson T4000 uses no power connectors, while the LX ULTRA requires one 16-pin connector.

Q: Which GPU supports modern graphics APIs?

A: Only the LX ULTRA supports graphics APIs, with DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.3. The Jetson T4000 lists N/A for all three.

Q: What are the physical dimensions?

A: The Jetson T4000 measures 87 mm by 100 mm by 15 mm and is an IGP. The LX ULTRA measures 268 mm by 112 mm by 40 mm and is a dual-slot card.

Architecture Differences

The NVIDIA Jetson T4000 uses the GB10B chip built on the Blackwell architecture, produced on TSMC 5 nm with a die size of 391 mm². Its generation is listed as Server Blackwell (Bxx), with a predecessor of Server Hopper and a successor of Server Rubin. The chip contains 1536 shading units, 48 TMUs, 16 ROPs, 12 RT cores, and 64 tensor cores. The FP32 and FP16 rates are identical at 4.700 TFLOPS, indicating a 1:1 ratio with no separate half-precision path.

The Lisuan Tech LX ULTRA uses the 7G105 chip based on the TrueGPU architecture, produced on TSMC 6 nm with die size unknown. Its generation is 7G100, with no predecessor or successor recorded. The chip contains 6144 shading units, 192 TMUs, and 96 ROPs. RT core and tensor core counts are listed as null. The FP32 rate is 24.58 TFLOPS, and the FP16 rate is 49.15 TFLOPS at a 2:1 ratio, meaning half-precision runs at twice the FP32 rate.

Memory subsystems differ in type and clock. The Jetson T4000 uses LPDDR5X at 1067 MHz with 8.5 Gbps effective data rate, achieving 273.2 GB/s over a 256-bit bus. The LX ULTRA uses GDDR6 at 2250 MHz with 18 Gbps effective data rate, achieving 432.0 GB/s over a 192-bit bus. The NVIDIA part carries 64 GB, the Lisuan part 24 GB.

Bus interfaces also differ. The Jetson T4000 connects via PCIe 5.0 x8, while the LX ULTRA uses PCIe 4.0 x16. Process nodes are close, 5 nm versus 6 nm, both from TSMC. Transistor counts are unknown for both devices, and transistor density is not recorded.

Power delivery and cooling reflect the architectural choices. The Jetson T4000 is an IGP with no power connectors, a 90 W TDP, and a 250 W suggested PSU. The LX ULTRA is a dual-slot card with one 16-pin power connector, a 225 W TDP, and a 550 W suggested PSU. Display outputs exist only on the LX ULTRA, with 4x DisplayPort 1.4a. The Jetson T4000 has none.

API support is exclusive to the LX ULTRA. DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.3 are listed for the Lisuan part. The Jetson T4000 lists N/A for all three, confirming its role as a compute-only accelerator rather than a graphics adapter. The release dates differ by roughly two months, with the Jetson T4000 dated 2026-01-04 and the LX ULTRA dated 2026-03-16. Both remain in active production per the database.