monica

wannabe writer.

[Google Cloud @googlecloud]

Meet TPU 8i Max and TPU 8tx, the third and fourth systems in our eighth-generation TPUs.

Together, these four systems give customers a purpose-built accelerator for every stage of the agentic lifecycle:

– TPU 8t for scale-up training
– TPU 8i for low-latency reasoning
– TPU 8tx for confidential computing
– TPU 8i Max for massive context windows and the highest compute density per rack

Read more in our announcement post: https://t.co/CqcqUtvSsW

[Amin Vahdat @AminVahdat]

Thrilled to share a big milestone from our AI & Infrastructure team at @Google.

Today we’re expanding our 8th generation TPU portfolio with:
• TPU 8i Max: Built for massive context windows and the highest compute density per rack
• TPU 8tx: Purpose-built for secure, confidential inference

This marks the first time we’ve introduced four distinct TPU systems in a single generation, in one year.

For developers, hardware is only as good as the software stack. And we want you to be able to build with TPUs as easily and efficiently as we do at Google. You can now:
↳ Run native PyTorch & Hugging Face models via TorchTPU (zero code rewrites)
↳ Cut KV transfer latency >90% with open-source TPU Sync
↳ Eliminate idle compute waste with Agent Compute Engine

Read more about our newest TPUs on our blog: https://t.co/8eaFhq3jhi

I look forward to connecting with builders, researchers, engineers and organizations pushing the boundaries of AI and computing!

[Irena Buzarewicz @IrenaBuzarewicz]

Thursday

[SemiAnalysis @SemiAnalysis_]

The Iron(wood) Throne:
Google's TPU Stakes Its Claim on Inference
Viable Alternative of CUDA Moat,
NVIDIA Working Nights & Weekends,
Jensen’s Counter Submission, Kimi K3,
Meta Muse Spark Open, GLM5.3,
TorchTPU, SGLang, vLLM, llm-d, Ray
https://t.co/mh16HIWmpJ

[Google Open Source @GoogleOSS]

Train in Python, serve in C++. ⚡

Meet XLA Serving Engine (XSE), our open source C++ runtime for serving JAX and StableHLO models across TPUs, GPUs, and CPUs with 2–5x the throughput of a Python server.

👇 Link in first reply

[Matthew Zeitlin @MattZeitlin]

peter thiel is the only one who understands that the whole point of getting rich is getting people to take your essays seriously

[monica @monicadsong]

unfortunately your agent will not read the bible for you!!

[amanda @jailedamanda]

parents please check your kids' halloween candy. just found hieronymus bosch’s garden of earthly delights inside of a peanut butter cup.

[Dwarkesh Patel @dwarkesh_sp]

Robert Caro on loneliness:

[monica @monicadsong]

"the bazaar of ideas is bustling... psyops in every direction"

source: https://t.co/Kqfp6raY2O

[Ben Pouladian @benitoz]

TPU Tuesday at The Information 😄

[monica @monicadsong]

the neurosurgeon who becomes a physical therapist. the lawyer who becomes a hair stylist. the engineer who becomes an instagram influencer.

[Google Cloud @googlecloud]

Native PyTorch on TPUs is here—and TorchTPU is officially open source.

Train and serve models using the tools you already know: @vllm_project, @sgl_project, @_llm_d_, TorchTitan, @huggingface, and @raydistributed.

Learn more from #PyTorchConf2026 → https://t.co/mISCD65vA4

[monica @monicadsong]

love this phrase: “iterated contact with reality”

[François Chollet @fchollet]

String theory is what happens when an intelligent system is starved of sensory input and starts hallucinating. Intelligence needs grounding.

For this reason, intelligence is practically never the bottleneck to solving impactful problems. The bottleneck is iterated contact with reality.

[Elon Musk @elonmusk]

Next year, I will give one of my SpaceX shares to every kid in America.

This will be the first of many.

[XDOF @xdofai]

We heard the hype around LLMs and robotics, so we put Astra and Fable to the test.

16 tasks testing 13 capabilities.

Our finding: both did surprisingly well on reasoning and planning and struggled with dexterity. But Astra did better overall.

Here’s what happened:

[Derek Thompson @DKThomp]

New newsletter: I KNOW WHY THE MATHEMATICIANS WEEP

I wrote about "technological grief"—the uncanny and ego-eviscerating feeling of watching a machine do an intellectual job on which you based your sense of self and worth. Also:

- what makes humans special
- what makes math special
- why science is both a craft (discovery) and a conversation (the relationships you build with fellow discoverers)
- how AI disrupting the former (intelligence) might increase the significance of the latter (relationships)
- why other fields should be prepared for emotional and occupational disorientation that mathematicians are going through

With many thanks to deeply wise words of @stevenstrogatz

https://t.co/P25n34a2lK

[Lennart Heim @ohlennart]

TIL that global spending on alcohol is still larger than AI infrastructure spending this year ($1.7T vs $1.5T)

[monica @monicadsong]

pachinko!

[François Chollet @fchollet]

I've been having a lot of fun trying a wide range of LLM post-training experiments using KerasHub + JAX + TPUs. Might open-source some of that