The new Chinese Lisuan 7G106 graphics card runs Black Myth: Wukong in 4K |
|
 |
|
|
Chinese company Lisuan Technology has announced its first 6nm graphics card, the 7G106, based on the TrueGPU architecture. This is the first solution of its kind to be developed entirely in-house, without licenses from NVIDIA or AMD. The new product is aimed at the mid-range segment and demonstrates impressive performance, albeit with one important limitation. The Lisuan 7G106 has 48 computing units, 12 GB of GDDR6 video memory, and DirectX 12 support, with performance estimated at 24 TFLOPS. The graphics card interacts with the system via a PCIe 4.0 x16 interface and uses a single 8-pin power connector, indicating a possible power consumption of up to 225 W. In terms of multimedia, the 7G106 supports AV1 encoding in 4K @ 30 fps and HEVC  High Efficiency Video Coding (HEVC, or H.265) is the latest video compression standard developed by the Motion Picture Experts Group (MPEG), finalized in 2013. HEVC is able to compress video up to twice as efficiently as AVC, letting you encode a video in a file that is about half the size of AVC at the same quality level. The higher compression efficiency also means that for a given bit rate or file size, HEVC encodes video with significantly higher quality than AVC..in 8K, and can also decode video up to 8K @ 60 fps. Four DisplayPort 1.4 ports are available for connecting monitors (HDMI is not provided).At the presentation, the developers showed Black Myth: Wukong running in 4K with high graphics settings and a ?playable? frame rate. In synthetic tests, the Lisuan 7G106 scored 26,800 points in 3DMark Fire Strike and 111,290 points in Geekbench 6 OpenCL. This is about 10% higher than the RTX 4060, but 9% lower than the results of the new RTX 5060. However, the Lisuan 7G106 has one important limitation: the card cannot work with ray tracing. Lisuan also presented a professional version of the graphics card with 24 GB of memory and support for GPU GPU-accelerated computing is the use of a graphics processing unit (GPU) together with a CPU to accelerate deep learning, analytics, and engineering applications. virtualization for up to 16 containers. Mass production of the new products will start in September, but no information about prices is available yet.
|
