What's Changed
- UTILS/ARCH : Add FUJITSU-MONAKA CPU by @mentOS31 in #1257
- CI: improve ucc mpi tests - check ip exists by @dpressle in #1263
- TL/CUDA: enable NVLS build by default by @ikryukov in #1200
- TL/CUDA: fix NVLS arch by @ikryukov in #1264
- EC/CUDA: reduce compilation time by @ikryukov in #1265
- CI: update nvls tests timeout logic by @dpressle in #1266
- SCHEDULE: fix task restart in ppln schedule by @Sergei-Lebedev in #1262
- CI: run coverity in container by @dpressle in #1267
- TEST: disable hanging ctx team tests by @ikryukov in #1273
- CORE: move cuda topology to core by @Sergei-Lebedev in #1261
- BUILD: bump version to v1.8 by @Sergei-Lebedev in #1276
- CI: enhance NVLS tests by @ikryukov in #1269
- CI: fix SLURM job release in onfail/always by @ikryukov in #1282
- CI: move gtest builds to slurm by @dpressle in #1280
- CI: change OMPI by @ikryukov in #1286
- TL/CUDA: Add INT32, INT64, UINT32, UINT64 support for NVLS by @Juee14Desai in #1259
- CI: GHA workflow & build optimizations by @ikryukov in #1281
- CI: abort old builds by @dpressle in #1285
- CI: blossom ci hot fix by @ikryukov in #1289
- CI: convert slurm pipelines to use jenkins library by @dpressle in #1291
- coll/datatype: Add a collective checker function by @QiaoK in #1292
- TL/CUDA: NVLS and topo robustness improvements by @ikryukov in #1287
- TL/UCP: add allreduce ring algorithm by @wfaderhold21 in #1258
- TOOLS: matrix generator shuffle columns by @yaeliyac in #1251
- TL/CUDA: link libstdc++ for aarch64 (#1296) by @ikryukov in #1298
- TL/CUDA: multinode cleanup fixes (#1295) by @ikryukov in #1299
- TL/CUDA: NVLS allreduce small buffers (#1304) by @janjust in #1307
- TL/CUDA: fix nvls alignment (#1312) by @janjust in #1313
New Contributors
- @mentOS31 made their first contribution in #1257
- @Juee14Desai made their first contribution in #1259
- @QiaoK made their first contribution in #1292
Full Changelog: v1.7.0...v1.8.0