Who contributed to Google on May 10, 2026?

2 developers shipped this update, including copybara-service[bot] and mattjj.

What were the notable Google updates?

Partially rolling forward the bits needed to compile Pallas Triton to PTX, add config option to disable auto-pvary insertion, and Fix typo in Pallas pipelining documentation.

Google

JAX, the GenAI SDK, and the Cloud libs - Google's open source layer

Pick a date

Topics: AI / ML Python Full archive →

The Wire · Showcase

JAX ROLLS BACK PALLAS TRITON CHANGES, ADDS CONFIG GUARDRAILS

By RepoJournal · Filed 06:02 UTC on May 10, 2026 · About Google

2 people shipped this

copybara-service[bot] @copybara-service[bot] 2 cited

mattjj @mattjj 1 cited

JAX is partially reverting a Pallas Triton PTX compilation change while shipping new configuration controls to prevent auto-pvary insertion from breaking downstream code.

The team partially rolled forward bits needed for Pallas Triton to compile to PTX [1], but this required reverting a prior commit — the change doesn't alter existing Pallas Triton behavior, just restores compilation infrastructure. Parallel to that, JAX is shipping a new config option to disable auto-pvary insertion [2], a necessary escape hatch for teams hitting unexpected behavior from automatic parallelization. The move suggests the auto-pvary feature is too aggressive in its current state and catching real production patterns it shouldn't touch. Minor documentation fixes landed in Pallas pipelining [3], cleaning up typos in the guides. These aren't flashy changes, but they're the friction-reduction work that keeps large compute workloads from breaking between releases.

Action items

→ Review if auto-pvary insertion is affecting your Pallas workloads — use new disable config if needed google/jax [monitor]
→ Update to latest JAX if using Pallas Triton; Pallas changes are isolated and safe google/jax [plan]

References

[1] Partially rolling forward the bits needed to compile Pallas Triton to PTX ↗ google/jax
[2] add config option to disable auto-pvary insertion ↗ google/jax
[3] Fix typo in Pallas pipelining documentation ↗ google/jax

Quick answers

What shipped in Google on May 10, 2026?: JAX is partially reverting a Pallas Triton PTX compilation change while shipping new configuration controls to prevent auto-pvary insertion from breaking downstream code. In total, 6 commits and 4 pull requests landed.
Who contributed to Google on May 10, 2026?: 2 developers shipped this update, including copybara-service[bot] and mattjj.
What were the notable Google updates?: Partially rolling forward the bits needed to compile Pallas Triton to PTX, add config option to disable auto-pvary insertion, and Fix typo in Pallas pipelining documentation.

JAX FIXES REMAT3 CONSTANT HANDLING CRASH, GOOGLE CLOUD PYTHON CENTRALIZES REST TRANSCODING

Matthew Johnson patched a critical remat3 regression where closed-over constants were silently dropped, causing immediate crashes in production rematerialization pipelines [ref:1].

ai 50 shipped 1-min read

Google 1 day ago

JAX TIGHTENS XLA OPTIMIZATION CONTROLS, GOOGLE CLOUD PYTHON SETS UP FEATURE GATES

JAX makes XLA optimization levels human-readable while Google Cloud Python client libraries prepare a unified feature gating system for observability.

ai 54 shipped 1-min read

Google 2 days ago

GENAI AUDIO TRANSCRIPTION SHIPS, PROTO-PLUS FIXES RACE CONDITIONS

Google's AI SDK gains native audio processing while the core Python cloud libraries ship critical stability fixes across the stack.

ai 60 shipped 1-min read

Google 3 days ago

JAX CUTS ASYNC COLLECTIVE PRIMITIVES, GENAI SDK ADDS LIVE API VOCABULARY

JAX is simplifying its async collective operations while the Python GenAI SDK ships custom vocabulary support for the Live API, marking a shift toward leaner abstractions and richer real-time capabilities.

ai 100 shipped 1-min read

Elsewhere on the wire

AI Agents about 9 hours ago

CLAUDE OPUS 5 LANDS ACROSS THE STACK

The newest Anthropic model is now live in langchain, Cline, and llama-index, with native support for extended reasoning and 1M context windows.

ai-agents 28 shipped 1-min read

Local LLMs about 9 hours ago

OLLAMA LANDS LAGUNA SUPPORT AND CRUSHES MEMORY LEAKS WHILE SGLANG HITS V0.5.16 WITH CONFIDENCE-DRIVEN SPECULATIVE DECODING

Ollama shipped three critical performance and reliability fixes for Metal residency and concurrent access patterns, while SGL-Lang released 0.5.16 with a new speculative algorithm hitting 383.7 tok/s on DeepSeek-V4.