Mark Huang @_markkh

Scientific SuperIntelligence @periodiclabs Joined October 2013

Tweets

600
Followers

917
Following

248
Likes

460

Rishabh Agarwal @agarwl_

a week ago

Not sure if there's any recording for my CoLM talk but you can find the slides here. agarwl.github.io/slides/context… Posting couple of slides which look directionally good (to me)

6 21 300 21K 347

0 0 1 127 0

View Details

Chris Barber @chrisbarber

3 weeks ago

I asked people which neolabs under $10B have the highest talent density: 32 votes: Core Automation (@MillionInt, @_arohan_) 26 votes: Periodic Labs (@LiamFedus, @ekindogus) 20 votes: Flapping Airplanes (@spectorb, @amspector100, @aidanmantine) 18 votes: Standard Intelligence (@G413N, @devanshpandey) 8 votes: Ricursive (@annadgoldie, @Azaliamirh) 5 votes each: - Ineffable (David Silver) - Mirendil (@bneyshabur, @HarshMeh1a, @shayan_, @tararezaeikh) - Recursive (@RichardSocher, @_rockt, @timshi_ai, @josh_tobin_, @CaimingXiong, @jeffclune, @tydsh, Alexey Dosovitskiy) 3 votes each: - Applied Compute (@ypatil125, @rhythmrg, @lindensli) - Isara (@ezhang7423, @hegasz) - Prime Intellect (@vincentweisser, @johannes_hage) 2 votes each: - Elorian (@AndrewDai, @yinfeiy, @SethInternet) - Goodfire (@eric_ho, @DanJBalsam, @banburismus_) - Inherent (@tantumscollins, @edwardfhughes, @LouisKirschAI, @kallyaleksiev) - Trajectory (@rronak_, @QuantumArjun, @MichaelElabd) - World Labs (@drfeifei, @jcjohnss, @BenMildenhall, @chlassner) People couldn't pick a company they work at or founded. Disclosure: I'm a small investor in Applied Compute, Factory, Standard Intelligence, Trajectory and Wafer. I didn't vote. Continued:

22 47 719 271K 1K

View Details

Mark Huang @_markkh

a month ago

The model just wants to learn! @periodiclabs is building the flywheel starting from the closed data loop.

The economics of a Neolab. A neolab is loosely defined as a startup of AI researchers who raises a lot of money pre-production to be able to finance GPU compute to take on a large AI problem. To buy 1000 GB300s or ~14 NVL72 racks will set you back $125-150M for 3yrs with 15-30%

177 157 2K 372K 2K

1 0 2 273 0

View Details

Mark Huang @_markkh

a month ago

@deedydas Did you mean @periodiclabs ?

0 0 2 299 0

View Details

Zijie Yan @zijie_y

a month ago

Neon’s scientific traces are long and vary widely in length. Training on them shouldn’t mean GPUs chewing through padding or waiting for each other. Here’s how we pack sequences, balance the work, and distribute long contexts across GPUs at @periodiclabs 👇 periodic.com/news/ai-infras…

2 6 45 13K 18

View Details

Rohan Pandey @khoomeik

a month ago

Another Neon infra tidbit: @periodiclabs accelerated checkpoint conversion ~30x down to just 1 minute, helping us deploy faster to get rapid feedback from our labs. Our models train in Megatron, but SGLang inference consumes weights in HuggingFace format. Traditionally, conversion ran serially: 1. rebuild the entire checkpoint from individual weight tensors 2. convert the checkpoint 3. reshard it to be inference-ready This could take 30 min for a trillion parameter model! Noticing that conversion shouldn't require materializing the full checkpoint, @hsu_byron introduced Fast Resharding: 1. parallelize conversion across Ray actors 2. convert expert chunks directly instead of assembling the full expert tensor in memory We upstreamed Fast Resharding to Miles in PR #1371. Now even you can deploy trained models to production in a matter of minutes.

5 10 116 8K 66

View Details

Mark Huang @_markkh

a month ago

At Periodic Labs we build labs that learn. Our team recognizes that the transition from digital environments to the physical world is highly complex and unpredictable. The next breakthroughs in science are deeply coupled with intelligence that is given the opportunity to explore through experimentation. I've been thrilled and honored to help contribute to the product surface areas that help theorists, experimentalists and industry practitioners in the physical sciences accelerate their goals. Here's a preview of the scientific and technological progress we've made in closing the experimental loop for XRD analysis using Neon our post-trained 1T model.

Periodic Labs @periodiclabs

a month ago

We’ve spent the past year building high-throughput labs and infrastructure for large-scale physics experiments in our search for better superconductors and magnets. We ran our first experiments by hand. Now, experiments run continuously in the labs. We’re building AI to make