-
Notifications
You must be signed in to change notification settings - Fork 33
Pull requests: google/tpu-sync
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Move InitTransportServer and RegisterActivePlanDirect into ReshardSendSession::Create
#982
opened Sep 20, 2026 by
copybara-service
Bot
Loading…
Add AcquireWithTimeout and remove AcquireDynamicBlocks in StagingBlockAllocator
#981
opened Sep 20, 2026 by
copybara-service
Bot
Loading…
Add peregrine TransportMetrics and TransportAdapter::GetTransportMetrics().
#960
opened Sep 18, 2026 by
copybara-service
Bot
Loading…
[gmon] Add Monarch retention rules for P2P transfer latency and PCIe byte metrics.
#959
opened Sep 18, 2026 by
copybara-service
Bot
Loading…
Declare and instrument PCIe H2D and D2H byte counter and per-device label telemetry metrics:
#950
opened Sep 18, 2026 by
copybara-service
Bot
Loading…
Add TdsKVBackend storage driver with zero-copy O_DIRECT and 4KB bounce-buffering.
#948
opened Sep 18, 2026 by
copybara-service
Bot
Loading…
[tpu-raiden] Support shared memory DMA mapping and external tensor H2D/D2H transfers in KVCacheManager
#944
opened Sep 17, 2026 by
copybara-service
Bot
Loading…
[TPU Raiden] Backend Tier Extension (Part 3/4): Python Framework Bindings & Integration Tests (JAX & PyTorch)
#942
opened Sep 17, 2026 by
copybara-service
Bot
Loading…
Standardize MetricFamilyBuffer and BufferedMetricsExporter on PrometheusLabelView and remove duplicate label formatting logic
#939
opened Sep 17, 2026 by
copybara-service
Bot
Loading…
Internal build configuration cleanup for torch_tpu src/ layout
#938
opened Sep 17, 2026 by
copybara-service
Bot
Loading…
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…
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…
ci: build the release matrix from the frameworks input
#917
opened Sep 13, 2026 by
yinlin09
Collaborator
Loading…
[tpu-raiden] Add PollEvictedHashes API skeleton for KV cache eviction tracking.
#893
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…
Implement gRPC-based transfer control plane for KVCacheManagerWithTransfer.
#881
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!
Updated in the last three days: updated:>2026-09-17.