Charith Mendis

@charith_mendis

PhD student - Compilers | Machine Learning for Program Optimization | High Performance Computing

Cambridge, MA
charithmendis.com
Vrijeme pridruživanja: lipanj 2019.
4 fotografije ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @charith_mendis

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @charith_mendis

  1. 21. sij

    We just sent out an RFC for implementing our timing framework BHive () inside LLVM-exegesis. Please provide feedback to improve our design. ()

    1 proslijeđeni tweet 6 korisnika označava da im se sviđa
    Poništi
  2. proslijedio/la je Tweet
    17. sij

    Heard about AVX-512 performance transitions party from the the cool kid down the block, but not sure if your parents will let you go? I'll go there so you don't have to: Includes run-on sentences weighed down by up to three footnotes each.

    8 replies 96 proslijeđenih tweetova 281 korisnik označava da mu se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  3. proslijedio/la je Tweet
    14. sij

    How Fast Does This Code Run? MIT researchers have trained a neural network to tell you how fast any code that you present it with will run. Sounds fun but why do we need it?

    2 proslijeđena tweeta 7 korisnika označava da im se sviđa
    Poništi
  4. proslijedio/la je Tweet
    3. sij

    A Metric-Guided Method for Discovering Impactful Features and Architectural Insights for Skylake-Based Processors Ahmad Yasin, Jawad Haj-Yahya, Yosi Ben-Asher, Avi Mendelson ACM Transactions on Architecture and Code Optimization () 16(4) 2019

    1 reply 22 proslijeđena tweeta 71 korisnik označava da mu se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    15. pro 2019.

    BHive: A Benchmark Suite and Measurement Framework for Validating x86-64 Basic Block Performance Models

    1 reply 5 proslijeđenih tweetova 48 korisnika označava da im se sviđa
    Poništi
  6. 12. pro 2019.

    Visit our poster (#165) today (Thursday 12th) during the evening poster session from 5 to 7 pm to learn about our work on "Compiler Auto-Vectorization with Imitation Learning".

    11 korisnika označava da im se sviđa
    Poništi
  7. 6. stu 2019.

    We have open-sourced our x86-64 basic block throughput profiling tool at . It allows you to profile arbitrary memory accessing basic blocks without crashing.

    1 reply 2 proslijeđena tweeta 30 korisnika označava da im se sviđa
    Poništi
  8. proslijedio/la je Tweet
    16. kol 2019.

    This tool has already blown my mind a couple of times. It's great for analysis of blocks that lie in that grey area, that are close to several different speed limits (see ), rather than one obvious limit.

    How fast will your code run on the latest Intel chip? Find out with Ithemal (). With only black-box access to a processor, we use machine learning to answer the question () with half the error of Intel's own tools.
    Prikaži ovu nit
    1 reply 9 proslijeđenih tweetova 26 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  9. 3. ruj 2019.

    Happy to announce that our paper on "A Benchmark For Validating x86-64 Performance Models" was accepted at IISWC 2019. Benchmark suite and timing tools coming soon! Work in collaboration with

    18 korisnika označava da im se sviđa
    Poništi
  10. proslijedio/la je Tweet
    29. kol 2019.

    "There are Turing Awards ready to be picked up by people who can make Python run faster" -David Patterson, 2018 ACM A.M. Turing Award Co-Recipient (with John Hennessy) "A New Golden Age for Computer Architecture"

    5 proslijeđenih tweetova 6 korisnika označava da im se sviđa
    Poništi
  11. 1. srp 2019.

    6 1/2 avenue at NY city!

    2 korisnika označavaju da im se sviđa
    Poništi
  12. 29. lip 2019.

    Nice work on data augmentation!

    Data augmentation is even more crucial for detection. We present AutoAugment for object detection, achieving SOTA on COCO validation set (50.7 mAP). Policy transfers to different models & datasets. Paper: , Code: , details in thread.
    Prikaži ovu nit
    Poništi
  13. 26. lip 2019.

    Stuck in Phoenix again. Second time this happened to me while coming back from on an American flight starting at Phoenix. In 2016 with Ben and Stephen, in 2019 with and Fredrik

    1 reply 5 korisnika označava da im se sviđa
    Poništi
  14. proslijedio/la je Tweet
    23. lip 2019.

    And our workshop concludes with a panel discussion. We are honored to have Dave Patterson, HT Kung, Cliff Young, and Farinaz Koushanfar as our panelists. See you next year!

    1 proslijeđeni tweet 14 korisnika označava da im se sviđa
    Poništi
  15. 23. lip 2019.

    Ithemal won the best paper award at the ML for systems workshop @isca 2019!

    2 proslijeđena tweeta 21 korisnik označava da mu se sviđa
    Poništi
  16. 20. lip 2019.

    Learn more about how Deep Learning can be used for performance modeling; I will be giving a talk about Ithemal () at the ML for systems workshop ()

    3 proslijeđena tweeta 6 korisnika označava da im se sviđa
    Poništi
  17. proslijedio/la je Tweet
    21. tra 2019.

    😀! I'm looking forward to this collaboration on Ithemal:

    Congratulations to the recipients of the 2018 Google Research Awards! This round we received over 900 proposals from universities across the globe, covering a diverse set of research areas such as HCI, machine perception, distributed computing and more.
    2 proslijeđena tweeta 5 korisnika označava da im se sviđa
    Poništi
  18. 12. lip 2019.

    METHOD: Ithemal just needs raw assembly instructions as input and uses a hierarchical recurrent neural network (RNN) to perform the final prediction. No featurization of instructions is required!

    5 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  19. 12. lip 2019.

    RESULTS: Using a corpus of over 1M basic blocks and just their measured performance, Ithemal learns to predict performance with half the error of LLVM's and Intel's tools.

    2 proslijeđena tweeta 5 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  20. 12. lip 2019.

    How fast will your code run on the latest Intel chip? Find out with Ithemal (). With only black-box access to a processor, we use machine learning to answer the question () with half the error of Intel's own tools.

    1 reply 18 proslijeđenih tweetova 36 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi

@charith_mendis još nije objavio nijedan Tweet.

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·