Worth installing
650
fguzman82/gateGPT
Full Transformer into a custom chip. microGPT in RTL, generating names on a Virtex-5 FPGA at ~56k tokens/second.
—
—
Explore 3 GitHub repositories written in Verilog. Discover top-starred projects and those trending this week.
Full Transformer into a custom chip. microGPT in RTL, generating names on a Virtex-5 FPGA at ~56k tokens/second.
An open-source AI accelerator, developed by AI: RTL, ISA, simulator, compiler and profiler in one repo. Runs Qwen3, LFM2.5 and Qwen3.5 on a Kintex-7 PCIe card.
A GPU with transformation & lighting, rasterization, texture mapping, flat shading, double buffer, z-buffer
Full Transformer into a custom chip. microGPT in RTL, generating names on a Virtex-5 FPGA at ~56k tokens/second.
An open-source AI accelerator, developed by AI: RTL, ISA, simulator, compiler and profiler in one repo. Runs Qwen3, LFM2.5 and Qwen3.5 on a Kintex-7 PCIe card.
A GPU with transformation & lighting, rasterization, texture mapping, flat shading, double buffer, z-buffer