Chinese MTT S80 GPU Available For Purchase, But There’s A Catch
Moore’s Threads, a Chinese semiconductor company has launched its new gaming GPU based on the MUSA architecture. This GPU is the successor to the MUSA-Sudi series and is fabricated on a 12nm process. The MTT S80 uses the ‘Chunxiao‘ GPU and is actually a capable Graphics Card. This GPU features 4096 Shading Units which amount to 14TFLOPs of computing power at 1.8GHz, or in-between the 3060 and the 3060 Ti.
Chinese MTTS80 Available For Purchase
Today was the launch day of these GPUs and sure they are available for purchase but there’s a catch. Let us explain. Over at JD.com, the MTT S80 paired with the ASUS TUF GAMING B660M-PLUS D4 motherboard costs 2999 RMB or $420. The motherboard, however, is not selective meaning that it is compulsory to purchase the motherboard as well, being part of the deal.
The aforementioned motherboard costs roughly $170 or 1200 RMB which equates to the GPU costing just 1800 RMB or $250. That’s quite the deal for what appears to be a possible RTX 3060 contender (based on the compute power).
This GPU may not fancy everyone and is probably not on par with Lovelace or RDNA3. Although it does support DP 1.4a and HDMI 2.1, which is equal to the RTX 4090, pun intended. Alongside that, support for DirectX is also available with no specified version. This GPU appears to be fine-tuned for games like League Of Legends, Diablo 3, Minecraft, NFS: Hot Pursuit 3 just to name a few.
It does beat Intel, NVIDIA and AMD on all fronts in one specific segment. This GPU has claimed the title of the world’s first GPU to support PCIe Gen 5 technologies. Packed with 16GB of G6 memory, expect it to deliver adequate performance at high resolutions. It will be really interesting to see if this GPU falls into the hands of international reviewers like Gamer’s Nexus.
Specifications
GPU Name | MTT S80 |
Shading Units | 4096 |
Memory | 16GB GDDR6 |
Ports | DP 1.4a, HDMI 2.1 |
PCIe Support/Lanes | PCIe Gen 5 x 16 (128GB/s) |
Compute Power | 14.4TFLOPs |
Core Frequency | 1.8GHz |
Process Node | 12nm |
Tensor Computing Units | 128 |
Supported Codecs | 8K AVI/H.265/H.264, 32 channels 1080P@30fps |
API Support | DX, Vulkan, OpenGL, Open GL ES |