-
Notifications
You must be signed in to change notification settings - Fork 32
Pull requests: google/tpu-sync
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Add P2P transfer latency and PCIe byte metrics to TPU Sync telemetry schema.
#935
opened Sep 16, 2026 by
copybara-service
Bot
Loading…
Back off instead of spinning when accept() fails
#934
opened Sep 16, 2026 by
copybara-service
Bot
Loading…
Refactor KVCacheManagerWithTransfer to delegate control plane operations to ControlPlaneBackend.
#933
opened Sep 16, 2026 by
copybara-service
Bot
Loading…
Implement TcpControlPlaneBackend for KV cache control plane over POSIX TCP sockets.
#932
opened Sep 16, 2026 by
copybara-service
Bot
Loading…
Implements end-to-end single-node weight synchronization performance test for
#927
opened Sep 15, 2026 by
copybara-service
Bot
Loading…
Add PyTorch support to TPU Raiden H2D/D2H presubmit performance gating.
#925
opened Sep 14, 2026 by
copybara-service
Bot
Loading…
Add
torch_d2d_read_benchmark_runner.py under examples/microbenchmarks/ providing PyTorch parity with jax_d2d_read_benchmark_runner.py:
#924
opened Sep 14, 2026 by
copybara-service
Bot
Loading…
Improve multi-NIC discovery and NUMA resolution
#923
opened Sep 14, 2026 by
copybara-service
Bot
Loading…
ci: build the release matrix from the frameworks input
#917
opened Sep 13, 2026 by
yinlin09
Collaborator
Loading…
Test pull acceptance before acknowledgement
pull ready
#914
opened Sep 11, 2026 by
madongfly
Collaborator
Loading…
Test receive settlement across asynchronous work
pull ready
#910
opened Sep 11, 2026 by
madongfly
Collaborator
Loading…
[peregrine] handle socket send EPIPE error
#894
opened Sep 10, 2026 by
copybara-service
Bot
Loading…
[tpu-raiden] Add PollEvictedHashes API skeleton for KV cache eviction tracking.
#893
opened Sep 9, 2026 by
copybara-service
Bot
Loading…
Support uploading wheel to PyPI via OSS Exit Gate and build tpu_sync_torch
#892
opened Sep 9, 2026 by
copybara-service
Bot
Loading…
[TPU Raiden] Backend Tier Extension (Part 2/4): Core C++ Engine Implementation, NUMA Coordination, POSIX Driver & Benchmarks
#890
opened Sep 9, 2026 by
copybara-service
Bot
Loading…
Implement exponential backoff circuit breaker for failing producer peers in tpu_sync.
#887
opened Sep 8, 2026 by
copybara-service
Bot
Loading…
Chunk vectored socket I/O at
IOV_MAX instead of rejecting large iovec spans
#885
opened Sep 8, 2026 by
majunze2001
Loading…
Implements raw PJRT device<->host DMA transfers for PyTorch tensors, mirroring
#884
opened Sep 8, 2026 by
copybara-service
Bot
Loading…
Implement gRPC-based transfer control plane for KVCacheManagerWithTransfer.
#881
opened Sep 8, 2026 by
copybara-service
Bot
Loading…
[TPU Raiden] Backend Tier Extension (Part 1/4): Protocols, Interfaces & Abstract Driver Architecture
#880
opened Sep 8, 2026 by
copybara-service
Bot
Loading…
Add
torch_dma_kv_cache_benchmark.py under examples/microbenchmarks/ providing PyTorch parity with jax_dma_kv_cache_benchmark.py:
#871
opened Sep 6, 2026 by
copybara-service
Bot
Loading…
Instrument weight synchronization components with coarse-grained
RAIDEN_TRACE annotations.
#860
opened Sep 5, 2026 by
copybara-service
Bot
Loading…
Previous Next
ProTip!
Follow long discussions with comments:>50.