zouyee

👨🏻‍💻 Kubernetes | CNCF Ambassador | Container Runtimes | Ziglang | Open Source Enthusiast | Speaker | Writer | He/Him

[zouyee @zouyee2]

I am Chinese!!!!!

[South China Morning Post @SCMPNews]

China’s Wuhan lockdown helped limit global spread of coronavirus, says World Health Organisation https://t.co/gQ4BRqv5T6

[zouyee @zouyee2]

You think we don't have that back home? What he loves might be this — but what he probably loves even more is the bubble-blowing dollar VCs.😄

[JUNDE WU @JundeMorsenWu]

I’m an AI PhD at Oxford wanted to move to the US for years, but it never worked out. Maybe I can share why.

To me, it feels like the US is making it harder and harder for Chinese talent to come or stay.

First, Trump signed PP10043, which meant I couldn’t go to the US for any graduate school. That’s why I had to turn down my Stanford offer and come to Oxford.

Later, I qualified for an EB-1A (the extraordinary ability green card). But because I was born in China, I’m stuck in a country-specific backlog that could take another 4–5 years.

Meanwhile, on the other side, I’m getting 3–4 emails from China almost every week. They offer high salaries, free housing, generous research funding, and even offered an astonishing amount of money just for me to come back and have a conversation.

I’ve always loved the American spirit. I truly believe my abilities could have a bigger impact there. But honestly, the whole process has been incredibly frustrating.

All of this makes me wonder whether I should just give up.

Why am I spending so much energy trying to go to a country that doesn’t seem to want me, when another country is doing everything it can to bring me back?

[Richard Feldman @rtfeldman]

Since the @bunjavascript folks recently wrote about their experience moving from Zig to Rust, it seemed like a good time to write about our experience rewriting @roc_lang from Rust to Zig!

https://t.co/JkPboqnGTn

[JetBrains @jetbrains]

Our full conversation with Andrew Kelley, creator of Zig, is out now.

No AI policy, leaving GitHub, a $670K nonprofit, and why there's no 1.0 (yet).

[bubble boi @bubbleboi]

People really need to stop hating on China man. Every time I’m in a pickle it’s China saving me.

Gas prices too high? I’m riding in the BYD. DRAM costs an arm and a leg? CXMT floods the market. Anthropic and OpenAI fucking me on token costs? Hello my friend Mr. Qwen.

The CCP has done more for my cost of living than my own government.

[zouyee @zouyee2]

Why is Bun in Rust better than the original?

⚽️Because I migrated from Zig to Rust.

Why fix all the bugs inherited from the original?

😗Because I can.

[Jarred Sumner @jarredsumner]

Bun in Rust is better than the original, and it’s going to keep getting better.

We fixed a lot of bugs inherited from the original. We’re fuzzing a lot more.

[Mitchell Hashimoto @mitchellh]

I strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out.

I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really).

It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely.

The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture.

We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying.

I worry.

[Joran Dirk Greef @jorandirkgreef]

I wrote these words 7 months ago.

I am more grateful today for Andrew's leadership of the ZSF, that the foundations of TigerBeetle, our compiler, should not be vibed out beneath us.

Standing up to “trillion dollar” big corp…

Zig is hard core quality.

https://t.co/xPyfOQLVDL

[zouyee @zouyee2]

dmlx update: serve-mode benchmark pipeline + expert cache tuning

Added: — Serve-mode benchmark pipeline with auto-generated perf reports — kqueue non-blocking accept (mitigates macOS VM pressure) — Expert cache warmup at startup (first-request misses -85%)

Changed: — Expert cache default 4GB → 10GB

Performance (DeepSeek-V4-Flash 4-bit, M4 Pro 48GB): — Prefill: 216ms (+12%) — Decode: 9.1 tok/s (serve mode, warm cache) — 100 tokens: 11s server-side — 7/7 correctness pass

Next: cache hit rate 42% → 95%, MLX compile fusion

https://t.co/kN8Gge4KAl

[zouyee @zouyee2]

It's insane that Bun rewrote in Rust, passes all tests, shrinks binary size, AND fixes memory leaks. Congrats, you built Deno with extra steps. I'm out. ☺️

[Jarred Sumner @jarredsumner]

Zig has been great for Bun. This branch is a fun experiment, and not a decision - here’s what I wrote on HN:

[zouyee @zouyee2]

Bun → Rust Deno → Rust

[Armin Ronacher ⇌ @mitsuhiko]

Say what you want: this is impressive. https://t.co/tAIAkuKtsZ

[zouyee @zouyee2]

dmlx perf update: +38% throughput from a single line change.

Removed a redundant GPU copy + forced sync in the expert streaming path. readExpertRows already returns contiguous data — the extra ops.copy + eval() was just a leftover from debugging.

Before: 95ms/tok (10.5 tok/s) After: 69ms/tok (14.5 tok/s)

DeepSeek-V4-Flash 284B running locally on M4 Pro 48GB. 247ms prefill, 7/7 correctness preserved.

Sometimes the biggest wins are removing code, not adding it.

https://t.co/kN8Gge5ipT

[zouyee @zouyee2]

@antirez ds4 is the real deal. I only had 48GB, so I built a toy that streams DeepSeek V4 experts from disk. Different path, same obsession — local inference that actually works.

https://t.co/dJssThJls6

[zouyee @zouyee2]

https://t.co/8OX0t0fTVY

[zouyee @zouyee2]

zig-learning has been updated with the latest from the Zig ecosystem!

New additions include:

• codedb – Zig code intelligence + MCP toolkit for AI agents
• libxev – Mitchel l Hashimoto's high-performance cross-platform event loop
• pixi – a pixel art editor built in Zig
• zm x – session persistence for terminal processes
• dvui – immediate mode GUI for apps & games

Plus:
Zig 0.16.0 spotlight, type resolution deep dive, and more.

#Zig #ZigLang #OpenSource #SystemsProgramming

[zouyee @zouyee2]

https://t.co/VOiZPMHkAA

[zouyee @zouyee2]

https://t.co/3mf4DiMgFb

[zouyee @zouyee2]

I just published How dmlx runs a ~150GB MoE model on a 48GB MacBook Pro https://t.co/1OZMTWybf1

[zouyee @zouyee2]

https://t.co/lrp5mACakO

[zouyee @zouyee2]

https://t.co/ou5uaql3El

[zouyee @zouyee2]

Now you can run 284B DeepSeek V4 on 48GB MacBook Pro via https://t.co/dJssThJls6 😅

[Basic Apple Guy @BasicAppleGuy]

Apple removed the 256GB memory option for the M3 Ultra Mac Studio. Only configuration is now 96GB.