mirror of
https://github.com/paboyle/Grid.git
synced 2024-11-09 23:45:36 +00:00
TODO list update
This commit is contained in:
parent
cb336aa8f8
commit
f710d7bd45
28
TODO
28
TODO
@ -1,17 +1,27 @@
|
||||
- Lattice_arith - are the mult, mac etc.. still needed after ET engine?
|
||||
- LinalgUtils ssp loop not offloaded
|
||||
- Mobius/Domain EOFA cache header implementaiotn has thread_loop
|
||||
- ImprovedStaggered accelerate
|
||||
- Lattice_reduction - remnant thread_loops must offload. Audit thread_loop in main code for non-accelerated code
|
||||
Lattice_rng
|
||||
Lattice_transfer.h
|
||||
|
||||
- Stencil.h : Thread loops in exchange code. Need to offload these
|
||||
|
||||
- Lebesque order reintroduction. StencilView should have pointer
|
||||
|
||||
- accelerate A2Autils
|
||||
|
||||
GPU branch code item work list
|
||||
-----------------------------
|
||||
|
||||
7) Accelerate the cshift
|
||||
|
||||
* 0) Single GPU
|
||||
- 128 bit integer table load in GPU code.
|
||||
- coalescedRead <- threadIdx.x
|
||||
- Gianluca's changes to Cayley into gpu-port
|
||||
- GPU accelerate EOFA
|
||||
- Clean up PRAGMAS, and SIMT_loop
|
||||
thread_loop interface revisit.
|
||||
for_n
|
||||
for
|
||||
- Staggered kernels -> GPU coalesced loop
|
||||
- Staggered kernels inline for GPU -- DONE
|
||||
|
||||
@ -23,9 +33,12 @@ GPU branch code item work list
|
||||
|
||||
* 3) Comms/NVlink
|
||||
- OpenMP tasks to run comms threads.
|
||||
- Remove explicit openMP in staggered.
|
||||
- Single parallel region around both the Kernel call
|
||||
and the comms.
|
||||
- Fix the halo exchange SIMT loop
|
||||
- Stencil gather
|
||||
- SIMD dirs in stencil
|
||||
|
||||
* 4) ET enhancements
|
||||
- eval -> scalar ops in ET engine
|
||||
@ -35,9 +48,7 @@ GPU branch code item work list
|
||||
|
||||
- Conserved current clean up.
|
||||
- multLinkProp eliminate
|
||||
- SIMD dirs in stencil
|
||||
|
||||
7) Accelerate the cshift
|
||||
|
||||
8) Merge develop and test HMC
|
||||
|
||||
@ -50,6 +61,11 @@ GPU branch code item work list
|
||||
|
||||
|
||||
=============================================================================================
|
||||
- Clean up PRAGMAS, and SIMT_loop -- DONE
|
||||
thread_loop interface revisit.
|
||||
_foreach
|
||||
_for
|
||||
|
||||
-- Figure what to do about "multLinkGpu" etc.. in FermionOperatorImpl. -- DONE
|
||||
-- Gparity is the awkward one -- DONE
|
||||
-- Solve non-Gparity first. -- DONE
|
||||
|
Loading…
Reference in New Issue
Block a user