-
Hi, I added some lines in hopper to support the head dim of 192, but it didn't work. Could you plz help me? And I wonder if this relates to the reason that you currently experiment on the head_dim of …
-
Hello,
I am fixing code for my own functionality and have some questions about **wargroup_wait** called in **mma** function of **sm90_mma_tma_gmma_x_warpspecialized_x.hpp** files. Why do below files…
-
After upgrading my Nexus 7 to Android 4.2 today, the VNC server package stopped working.
Using the framebuffer method only yields a garbled image of the recovery system, while the surfaceflinger metho…
-
Can you please make jsoar to connect to the database through a single file API? This way one could connect a separate database or a different one if they don't want to use sqlite.
-
Currently, GDB + OpenOCD only supports debugging of RISC-V SoC. It should not be able to directly access trace components such as the Trace Encoder and SMEM. Furthermore, it cannot parse trace data. I…
-
# Context
Platform: macOS 14.1.1
Architecture: arm64
circt version: Firtool 1.58.0 release, also f124fb041
# Description
When generating a design from Chisel with arrays of SRAM instances, …
-
Hi,
I'd like to copy the framebuffer contents (/dev/fb0) to main memory on raspbian using `memcpy_dma_config`.
Could you please give me a hint on if this library is appropriate for the task and h…
-
On Ampere, we schedule split-K by first unswizzling to smem (usually) then each CTA in turn performs vectorized stores and loads to a temporary buffer in a deterministic order to perform the sum.
Hop…
-
Thanks a lot for the implementation of fmha!
Is it possible to extend the range of supported sequence lengths (currently its only up to 512) (or maybe different Head / Hidden size configurations) t…
-
### 🚀 The feature, motivation and pitch
Currently the handling of view in scheduler is sub-optimal.
For views inside the fusion group that connects fusion, it makes sense, since this usually gives…