
A Chinese video card from Lisuan Technology has passed the first Geekbench test with results on par with modern mid-range solutions from competitors.
It seems that Lisuan Technology has released a new revision of its Lisuan G100 graphics accelerator. It has 48 main computing units with a maximum frequency of up to 2000 MHz and 12 GB of video memory. The declared characteristics differ significantly for the better from the initial version with 32 Compute Units, 300 MHz and 256 MB, which about a month ago «lit up» in Geekbench and demonstrated performance at the level of a very old GeForce GTX 660 Ti. Probably, an early prototype was tested then.

The updated specifications of the new product are more reminiscent of modern GPUs and, for example, in terms of the number of computing units, the card is not far behind the AMD Radeon RX 9070 GRE (48 Compute Units) presented in May and NVIDIA GeForce RTX 5070 (48 Streaming Multiprocessors), which was released in March. Of course, it is incorrect to directly compare the executive modules of different architectures, but you can get a general idea. The only thing we know about the TrueGPU architecture is that Lisuan Technology developed it from scratch, without using any third-party intellectual property.
So far, the card has only passed the performance test on the OpenCL platform in Geekbench. Result amounted to 111,290 points. If we compare it with the average results of other graphics cards in this benchmark, we see that the G100 is 2% faster than the Intel Arc A770, 10% faster than the NVIDIA RTX 4060, and 27% faster than the AMD Radeon RX 9060 XT. At the same time, the RTX 5060 is 9% ahead of the G100.

In the near future, Lisuan Technology is going to hold a presentation of the new product, the company announced on its official WeChat account. So we should soon learn even more information about the G100.
Source: Tom’s Hardware
Spelling error report
The following text will be sent to our editors: