Repository Issue Activity (beta)

#29 [FEA]: Offset In load and store function

#71 [BUG]: Hang forever on B200 but not GB10

#86 [FEA]: Add a viewas function for type casting without modifying data

#88 [BUG]: DLPack tensors may be released before async launch work is finished

#83 Add support for python 3.14 and prebuilt wheel for python 3.14

#27 [Question] Support for other architectures (i.e., Hopper, Ada, etc.)

#47 [Question] Clarification on FP8 Micro-block Scaling and FP4 Support Timeline

#37 [FEA]: Require ct.barrier for multi stage kernels

#36 [FEA]: Require ct.count for statistical kernels

#69 [FEA]: CUDA C or PTX Injection

#84 [BUG]: Source install fails when cmake defaults to Ninja generator

#80 [BUG]: CuTile doesn't support graph mode

#81 cutile cpp

#24 [Question]: Regarding tuple iteration

#51 [BUG]: Memory leak in fp8 kernel launch

#58 [BUG]: Cannot build from source with custom CUDA location

#60 [BUG]: hardcoded linker flag -nostdlib prevents security hardening

#65 [FEA]: Does cutile-python have plans to integrate NCCL as the communication backend?

#67 [BUG]: Test suite assumes experimental/internal features are available

#66 [Question] : TileTypeError

#79 [BUG]: reductions crash on restricted floats with `TileInternalError: Internal error`

#78 [Question] What's the difference tiled_view then load and ct.load

#74 [Question] how to alloc tmem tensor

#49 [FEA]: Identity matrix

#23 [BUG]: PaddingMode.NEG_INF not working under float8_e4m3