dheevatsa @dheevatsa

AI systems at NVidia sites.google.com/site/dheevatsa California, USA Joined February 2009

Tweets

157
Followers

189
Following

454
Likes

450

dheevatsa @dheevatsa

5 months ago

Very excited to share the work that we've been doing on continuing to push the boundary for AI training by efficiently scaling across multiple data-centers that are over thousands of kilometers apart, using NeMo/M-Core. bit.ly/3StaYi9

0 0 0 57 0

dheevatsa @dheevatsa

a year ago

Awesome to see the Distributed Shampoo optimizer top AlgoPerf ! “28% faster training than baseline ... 19% faster than 2nd place " Kudos to the team's tenacity for persistently improving over many months, not only surpassing strong baselines but also making it practically viable!

MLCommons @MLCommons

a year ago

5 45 270 424K 141

0 0 1 193 0

dheevatsa @dheevatsa

3 years ago

Grand Teton - Meta’s next-gen compute platform for AI ! Embodies a lot of exciting things that we’ve co-designed over past couple of years, enabling pushing our AI workloads further and beyond #ai #ocpsummit22 #codesign lnkd.in/gFG9PTBF

0 1 1 0 0

dheevatsa @dheevatsa

4 years ago

Very happy share that our paper on “Software-Hardware Co-design for Fast and Scalable Training of Deep Learning Recommendation Models” has been accepted for the industry track at ISCA this year :) It’s really great to be able to showcase this work, that…lnkd.in/g9e5FJsj

0 1 4 0 0

Christy Pambianchi @ChristyP_CPO

4 years ago

Despite being smuggled out of his hometown as a teen, @intel’s @BharatBKaul says the “universe has been kind” to him. Inspiring. #IAmIntel intel.ly/3ANXJ29

4 11 71 0 1

dheevatsa @dheevatsa

4 years ago

It was great to be able present one of the exec talks at the OCP global summit last week along with whitney zhao, introducing our AI training cluster and talking about the general challenges/opportunities we are seeing with buildin…lnkd.in/gHgjrmEy lnkd.in/g6B4y9i9

0 0 1 0 0

Engineering at Meta @fb_engineering

4 years ago

💻@Meta’s Director of Engineering, Omar Baldonado, spoke on stage today at the 2021 OCP Global Summit, sharing the incredible work our Meta Infrastructure teams have done over the past 10 years through the Open Compute Project. Learn more about our work: bit.ly/3ojwSFk

5 4 18 0 1

Download Image

dheevatsa @dheevatsa

4 years ago

I’m looking forward to speaking at the AI Hardware summit this year! Perfect venue to talk about all of the interesting work on Co-designing AI HW/SW at scale at FB :) AI HW Summit 2021 - hubs.ly/H0NQMV40 #AIHWSummit #codesign lnkd.in/dzaXzz8

0 2 9 0 0

dheevatsa @dheevatsa

4 years ago

Our work on pushing the state-of-art for training DLRMs, enabling efficient training of models with upto 12 Trilion parameters ! lnkd.in/gwJr2f8

0 1 6 0 0

PyTorch @PyTorch

5 years ago

NVIDIA GTC 2021 is a week-long event, that shares breakthroughs in AI, data center, accelerated computing, healthcare, gaming technology and more. This year, GTC 2021 is hosting over 50 different #PyTorch-related sessions. Learn more below: pytorch.medium.com/4312e9e8f3ee?s…

0 34 124 0 6

Nityānanda Miśra (मिश्रोपाख्यो नित्यानन्दः) @MisraNityanand

6 years ago

At my talk on Samskrita names y’day, a lady asked me if what her name Vyusti (she pronounced व्यूस्ति) meant? I said it didn’t sound like a Samskrita word I later realized the word (and name) is Vyushti (वयुष्टि) = (1) dawn (2) fruit (3) prosperity (4) praise (5) special beauty

5 16 63 0 1

Download Image

AI at Meta @AIatMeta

5 years ago

Introducing TT-Rec, a new method to dramatically shrink memory-intensive Deep Learning Recommendation Models and make them easier to deploy at scale. arxiv.org/abs/2101.11714

4 135 581 0 74

Download Image

dheevatsa @dheevatsa

5 years ago

lnkd.in/gF7Z_wd

0 0 1 0 0

Rohan Saphal @RohanSaphal

5 years ago

Excited to share the acceptance of my first-author publication, "SEERL : Sample Efficient Ensemble Reinforcement Learning " at @aamas2021 !! Joint work with @ravi_iitm @dheevatsa Sasikanth Avancha and @BharatBKaul during my time at @iitmadras @rbc_dsai_iitm and @intel

Ulle Endriss @ulle_endriss

5 years ago

0 12 40 0 3

4 4 27 0 0

dheevatsa @dheevatsa

5 years ago

Our AI System Co-design team with the PyTorch team recently conducted a tutorial at KDD’2020, where we talk about all the awesome features the PyTorch supports for large scale DLRMs. Now available on the official PyTorch YouTube channel - youtube.com/playlist?list=…, check it out !

0 0 0 0 0

dheevatsa @dheevatsa

5 years ago

Sharing this really neat work from our team - on quantized collectives to improve distributed training of large-scale deep learning recommendation models, addressing the needs for both data- and model- parallelism. lnkd.in/g3AePjU