Amit Kumar Jha

Writing code for science at @openteamsinc (prev @Quansightai) and building @CirunHQ

my opinion doesn't matter, don't take it seriously.

[Amit Kumar Jha @iaktech]

That's the front page of @Github Marketplace! 🎉🥳

cc @CirunHQ 🤌

[Amit Kumar Jha @iaktech]

I wish more people put ethics over money.

[Mitchell Hashimoto @mitchellh]

@benhylak 50K is the typical offer for me. I get one per week or so. I haven’t done any either and never plan to.

[Amit Kumar Jha @iaktech]

You could mortgage a flat for that price in Manchester.

[Amit Kumar Jha @iaktech]

I do not understand, why would you have that as default.

[Amit Kumar Jha @iaktech]

@badlogicgames @threepointone

[Unsloth AI @UnslothAI]

2-bit Qwen3.6-27B GGUF made 26 tool calls, triaged 15 GitHub issues and fixed, tested + reproed our repo’s 3 latest issues. 🔥

Try this locally in Unsloth Studio with just 12GB RAM. Studio also has a new look!

GitHub: https://t.co/aZWYAtakBP

[Amit Kumar Jha @iaktech]

@badlogicgames @antirez @Kimi_Moonshot @mitsuhiko I have a little extension which adds support for cloudflare provider in pi, to be able to use any model from workers-ai including Kimi-K2.6

[Amit Kumar Jha @iaktech]

Did the math on a @CirunHQ customer, ~6.5M CI-minutes last month on 16–64 core self-hosted GitHub Actions runners.

At typical per-minute pricing (matched core count) by most providers, they'd have been billed $400K+. Their actual cloud + cirun bill was less than 1/10th of that.

That's over $4M saved per year.

[Amit Kumar Jha @iaktech]

@Kimi_Moonshot M5 Ultras are going to be more expensive than gold, if they make it to the market.

[clem 🤗 @ClementDelangue]

You can now visualize Pi traces that you upload on
@huggingface!

Let's make sharing agent traces 10x more common to make agent AI more open and collaborative! Also, because it's fun to analyze @badlogicgames's traces 😂😂😂

[Amit Kumar Jha @iaktech]

Now I just need half a dozen H100s

[RyanLee @RyanLeeMiniMax]

https://t.co/UhsQSbFXD2

[Amit Kumar Jha @iaktech]

@mitsuhiko @fatihxkarahan It would be great to know what model I’m running when swapping local models on the same endpoint.

[Amit Kumar Jha @iaktech]

@0xSero The delivery was pretty fast!

[Amit Kumar Jha @iaktech]

Tried Muse Spark (from @meta's msl) just now for image editing, asked to convert the 3090 on the left to 5090, it came up with the image on the right.

[Amit Kumar Jha @iaktech]

Been trying Gemma 4 26B-A4B (MoE) for coding and its been interesting...

[Amit Kumar Jha @iaktech]

@badlogicgames Nice! I did something similar recently: https://t.co/wamfPLUsBc

[Amit Kumar Jha @iaktech]

Turns out $25 for a PR review isn't such great value when the reviewer leaks its own source code.

[Chaofan Shou @Fried_rice]

Claude code source code has been leaked via a map file in their npm registry!

Code: https://t.co/jBiMoOzt8G

[Amit Kumar Jha @iaktech]

Ref: https://t.co/iBwBkuj0G4
https://t.co/78WExQBGy5

[Amit Kumar Jha @iaktech]

Everyone's been GPU-brained for 3 years. But agentic AI flips this. CPUs now account for 50-90% of total latency doing orchestration, tool calls, API handling, memory lookups. If you don't have enough CPUs, your GPUs just sit there waiting to be fed.

Intel literally can't keep up with demand. AMD's calling it a surge.

Meanwhile Arm just made its own chip for the first time ever. 136 cores, 300W, and Meta's already buying. The boring chip is suddenly the one everyone needs.

[Amit Kumar Jha @iaktech]

Step 1: share random shit with confidence
Step 2: Then ask grok to verify

[Wolf Vollprecht @wolfvollprecht]

Europe and China are both investing heavily in RISC-V chips. Thanks to @openKylin the @condaforge project is making progress as well. Today I ran Pixi on RISC-V for the first time.
If you're investing in RISC-V infrastructure and want the package tooling to match — let's talk.