NVIDIA GeForce RTX 4070 AD103 vs Lisuan Tech LX MAX Comparison

Head-to-Head Benchmarks

The database does not contain any direct head-to-head benchmark results for the NVIDIA GeForce RTX 4070 AD103 versus the Lisuan Tech LX MAX. Neither GPU has recorded benchmark scores, and the winsA and winsB counters are both at zero. This absence of measured data is itself informative. It means the comparison must rely entirely on the architectural and specification differences recorded in the database, rather than on performance deltas from actual testing.

Both cards occupy the same percentile rank against all GPUs, sitting at the 50th percentile. Their average benchmark scores are identical at zero. Without recorded scores, the relative performance between these two cards cannot be stated as a percentage lead or deficit. The available data does, however, allow for a predictive analysis based on their respective compute resources, memory configurations, and clock behavior.

The most striking contrast lies in the compute architecture. The NVIDIA card uses 5,888 shading units, while the Lisuan card uses 6,144 shading units. That is a 256-unit advantage for the Lisuan Tech part, roughly 4.3 percent more shading processors. The Lisuan also holds a significant lead in render output units, with 96 ROPS against the NVIDIA card's 64 ROPS. That is a 50 percent advantage in pixel processing hardware. The pixel rate reflects this: the Lisuan delivers 192.0 GPixel/s, while the NVIDIA card delivers 158.4 GPixel/s. The Lisuan is ahead by 33.6 GPixel/s, a gap of approximately 21 percent.

The texture side tells a different story. NVIDIA's card has 184 texture mapping units, while the Lisuan has 192 TMUs, a smaller advantage of 8 units. Yet the NVIDIA card produces a higher texture rate: 455.4 GTexel/s versus 384.0 GTexel/s. The NVIDIA card is ahead by 71.4 GTexel/s, roughly 18.6 percent. This inversion comes from clock speeds. The NVIDIA card runs at a base clock of 1920 MHz and a boost clock of 2475 MHz. The Lisuan card has no recorded base or boost clock, which suggests its texture rate is derived from lower operating frequencies.

The floating-point compute comparison shows a similar inversion. The Lisuan card has more shading units but a lower FP32 throughput: 24.58 TFLOPS versus 29.15 TFLOPS for the NVIDIA card. The NVIDIA card leads by 4.57 TFLOPS, or about 18.6 percent, consistent with the texture rate gap. However, the Lisuan card holds a major advantage in FP16 compute. It delivers 49.15 TFLOPS with a 2:1 ratio to FP32, while the NVIDIA card delivers 29.15 TFLOPS with a 1:1 ratio. The Lisuan is ahead by 20.0 TFLOPS, roughly 68.6 percent, in half-precision work.

Architecture Differences

The two cards come from entirely different design lineages. The NVIDIA GeForce RTX 4070 AD103 is built on the Ada Lovelace architecture and uses the AD103 chip. The Lisuan Tech LX MAX uses the TrueGPU architecture with the 7G106 chip. The NVIDIA card belongs to the GeForce 40 generation, while the Lisuan card belongs to the 7G100 generation.

The manufacturing process differs by one nanometer. The NVIDIA card is fabricated on a 5 nm process, while the Lisuan card uses a 6 nm process. Both are produced at TSMC, so the foundry is the same, but the process node is not. The NVIDIA card has a known transistor count of 45,900 million and a die size of 379 mm², yielding a transistor density of 121.1 million transistors per square millimeter. The Lisuan card has no recorded transistor count, die size, or density, leaving a gap in the manufacturing comparison.

Memory architecture shows both similarities and differences. Both cards use 12 GB of memory on a 192-bit bus. The NVIDIA card uses GDDR6X memory, while the Lisuan card uses GDDR6. The memory clock differs as well. The NVIDIA card runs at 1313 MHz with 21 Gbps effective, while the Lisuan card runs at 2250 MHz with 18 Gbps effective. The bandwidth results favor the NVIDIA card: 504.2 GB/s versus 432.0 GB/s. The NVIDIA card leads by 72.2 GB/s, or approximately 16.7 percent.

The Lisuan card has no recorded RT cores or tensor cores. The NVIDIA card has 46 RT cores and 184 tensor cores. This suggests the Lisuan card may not have dedicated hardware for ray tracing or tensor operations, or the database lacks that information. The API support shows both cards support DirectX 12 Ultimate with the 12_2 feature level and OpenGL 4.6. The NVIDIA card supports Vulkan 1.4, while the Lisuan card supports Vulkan 1.3.

Power consumption differs as well. The NVIDIA card has a TDP of 200 W, while the Lisuan card has a TDP of 225 W. The Lisuan card draws 25 W more, an increase of 12.5 percent. Both cards use a single 16-pin power connector and recommend a 550 W power supply. Both are dual-slot cards, but the Lisuan is physically larger: 248 mm long, 118 mm high, and 48 mm wide, versus the NVIDIA card's 240 mm, 110 mm, and 40 mm dimensions.

The display outputs differ. The NVIDIA card has 1x HDMI 2.1 and 3x DisplayPort 1.4a. The Lisuan card has 4x DisplayPort 1.4a and no HDMI output. The production status also differs: the NVIDIA card is end-of-life, while the Lisuan card is active.

FAQ

Q: Which card has more shading units?

A: The Lisuan Tech LX MAX has 6,144 shading units, while the NVIDIA GeForce RTX 4070 AD103 has 5,888. The Lisuan card has 256 more, a difference of roughly 4.3 percent.

Q: Which card delivers higher FP32 compute performance?

A: The NVIDIA card delivers 29.15 TFLOPS, while the Lisuan card delivers 24.58 TFLOPS. The NVIDIA card is ahead by 4.57 TFLOPS, approximately 18.6 percent.

Q: How does memory bandwidth compare between the two?

A: The NVIDIA card has a bandwidth of 504.2 GB/s using GDDR6X memory. The Lisuan card has a bandwidth of 432.0 GB/s using GDDR6 memory. The NVIDIA card leads by 72.2 GB/s, about 16.7 percent.

Q: Does the Lisuan card have ray tracing cores?

A: The database records no RT cores for the Lisuan Tech LX MAX. The NVIDIA card has 46 RT cores. The Lisuan card also has no recorded tensor cores, while the NVIDIA card has 184.

Q: What is the difference in FP16 performance?

A: The Lisuan card delivers 49.15 TFLOPS in FP16 with a 2:1 ratio to FP32. The NVIDIA card delivers 29.15 TFLOPS with a 1:1 ratio. The Lisuan card leads by 20.0 TFLOPS, approximately 68.6 percent.

Q: Which card has a higher pixel rate?

A: The Lisuan card has a pixel rate of 192.0 GPixel/s, while the NVIDIA card has 158.4 GPixel/s. The Lisuan card leads by 33.6 GPixel/s, about 21 percent.

Specification Differences

The two cards differ in several key fields. The chip is different: AD103 for NVIDIA, 7G106 for Lisuan. The architecture is different: Ada Lovelace versus TrueGPU. The generation differs: GeForce 40 versus 7G100. The process node differs: 5 nm for NVIDIA, 6 nm for Lisuan. The transistor count is known only for NVIDIA at 45,900 million; the Lisuan count is unknown. The die size is 379 mm² for NVIDIA; the Lisuan die size is unknown.

The clock fields differ significantly. The NVIDIA card has a base clock of 1920 MHz and a boost clock of 2475 MHz. The Lisuan card has no base or boost clock recorded. The memory clock differs: 1313 MHz for NVIDIA versus 2250 MHz for Lisuan. The effective memory speed is 21 Gbps for NVIDIA and 18 Gbps for Lisuan.

The memory type differs: GDDR6X for NVIDIA, GDDR6 for Lisuan. The bus width is the same at 192 bit. The bandwidth differs: 504.2 GB/s versus 432.0 GB/s. The shading units, TMUs, and ROPS all differ: 5888 versus 6144, 184 versus 192, and 64 versus 96. The NVIDIA card has 46 RT cores and 184 tensor cores; the Lisuan card has none recorded.

The pixel rate differs: 158.4 GPixel/s versus 192.0 GPixel/s. The texture rate differs: 455.4 GTexel/s versus 384.0 GTexel/s. The FP32 throughput differs: 29.15 TFLOPS versus 24.58 TFLOPS. The FP16 throughput differs: 29.15 TFLOPS versus 49.15 TFLOPS. The TDP differs: 200 W versus 225 W.

The display outputs differ: NVIDIA has 1x HDMI 2.1 and 3x DisplayPort 1.4a, while Lisuan has 4x DisplayPort 1.4a. The Vulkan support differs: 1.4 for NVIDIA, 1.3 for Lisuan. The dimensions differ: NVIDIA is 240 mm by 110 mm by 40 mm, while Lisuan is 248 mm by 118 mm by 48 mm. The production status differs: NVIDIA is end-of-life, Lisuan is active. The release dates differ: NVIDIA was released on 2024-02-29, Lisuan on 2026-03-16. The NVIDIA card has a launch MSRP of 599 USD; the Lisuan card has no launch MSRP recorded.

The Verdict

The data points to a split decision based on workload type. The NVIDIA GeForce RTX 4070 AD103 holds the advantage in FP32 compute, texture rate, and memory bandwidth. Its 29.15 TFLOPS FP32 output is 18.6 percent higher than the Lisuan's 24.58 TFLOPS. Its 455.4 GTexel/s texture rate is 18.6 percent higher. Its 504.2 GB/s memory bandwidth is 16.7 percent higher. These metrics favor the NVIDIA card for traditional 3D rendering and games that rely on single-precision shading and texture-heavy scenes.

The Lisuan Tech LX MAX holds the advantage in pixel rate, ROPS count, and FP16 compute. Its 192.0 GPixel/s pixel rate is 21 percent higher. Its 96 ROPS are 50 percent more than the NVIDIA card's 64 ROPS. Its 49.15 TFLOPS FP16 output is 68.6 percent higher than the NVIDIA card's 29.15 TFLOPS. These metrics favor the Lisuan card for workloads that stress pixel fill and half-precision computation, such as certain compute tasks and machine learning inference.

The Lisuan card also has more shading units and TMUs, but those advantages do not translate into higher texture or FP32 throughput because of the clock differences. The NVIDIA card's higher boost clock of 2475 MHz appears to be the deciding factor in those categories. The Lisuan card's lack of recorded base and boost clocks makes its operating frequency unclear, but the recorded throughput figures imply a lower effective clock.

The NVIDIA card is end-of-life, while the Lisuan card is active. The NVIDIA card has a recorded launch MSRP of 599 USD. The Lisuan card has no launch MSRP in the database. The NVIDIA card includes RT cores and tensor cores, which the Lisuan card lacks in the recorded data. For ray tracing and tensor-based features, the NVIDIA card has the hardware support, while the Lisuan card does not show any.

Where Each One Wins

The NVIDIA GeForce RTX 4070 AD103 wins in FP32 compute, texture rate, memory bandwidth, ray tracing capability, tensor core availability, and Vulkan API version. The 18.6 percent lead in FP32 and texture rate suggests it is the stronger card for standard rasterization workloads where single-precision shaders dominate. The 16.7 percent lead in memory bandwidth helps with high-resolution textures and memory-intensive scenes. The presence of 46 RT cores gives it dedicated hardware for ray-traced effects, which the Lisuan card cannot match according to the database. The 184 tensor cores enable AI-accelerated features that have no counterpart on the Lisuan card.

The Lisuan Tech LX MAX wins in pixel rate, ROPS count, shading unit count, TMU count, FP16 compute, and physical size. The 21 percent lead in pixel rate and 50 percent lead in ROPS suggest it is the stronger card for fill-rate-bound scenarios, such as high-resolution rendering with simple shaders or multi-sample anti-aliasing. The 68.6 percent lead in FP16 compute indicates a clear advantage for half-precision workloads, including certain compute kernels and inference tasks that can use the 2:1 ratio. The 6,144 shading units give it more raw processing elements, even if the clock speed limits their utilization in FP32 tasks.

The Lisuan card also draws 25 W more power, at 225 W versus 200 W, and is physically larger in every dimension. It has four DisplayPort outputs, while the NVIDIA card has one HDMI and three DisplayPort. The Lisuan card has no HDMI output, which matters for users with HDMI-only displays. The NVIDIA card has a newer Vulkan version, 1.4 versus 1.3.

The production status difference is notable. The NVIDIA card is end-of-life, meaning it is no longer produced. The Lisuan card is active, meaning it is currently available. The release dates in the database place the Lisuan card's release at 2026-03-16, later than the NVIDIA card's 2024-02-29 release. The NVIDIA card has a predecessor in GeForce 30 and a successor in GeForce 50. The Lisuan card has neither predecessor nor successor recorded.