amiranda_hpc's profile picture. HPC researcher, storage enthusiast, low-level programmer, C++ aficionado, part of the Storage Systems for Extreme Computing team @BSC_CNS
⬆️⬆️⬇️⬇️⬅️➡️⬅️➡️🅱️🅰️▶️

Alberto Miranda

@amiranda_hpc

HPC researcher, storage enthusiast, low-level programmer, C++ aficionado, part of the Storage Systems for Extreme Computing team @BSC_CNS ⬆️⬆️⬇️⬇️⬅️➡️⬅️➡️🅱️🅰️▶️

Alberto Miranda 님이 재게시함

👥✨Thanks to all the partners who attended online / in person and made our 𝔽𝕚𝕣𝕤𝕥 ℙ𝕝𝕖𝕟𝕒𝕣𝕪 𝕄𝕖𝕖𝕥𝕚𝕟𝕘 𝕠𝕗 𝕥𝕙𝕖 𝔸𝔻𝕄𝕀ℝ𝔼 𝕡𝕣𝕠𝕛𝕖𝕔𝕥 possible. Until next time!!!👋🔹

admire_eurohpc's tweet image. 👥✨Thanks to all the partners who attended online / in person and made our 𝔽𝕚𝕣𝕤𝕥 ℙ𝕝𝕖𝕟𝕒𝕣𝕪 𝕄𝕖𝕖𝕥𝕚𝕟𝕘 𝕠𝕗 𝕥𝕙𝕖 𝔸𝔻𝕄𝕀ℝ𝔼 𝕡𝕣𝕠𝕛𝕖𝕔𝕥 possible. Until next time!!!👋🔹
admire_eurohpc's tweet image. 👥✨Thanks to all the partners who attended online / in person and made our 𝔽𝕚𝕣𝕤𝕥 ℙ𝕝𝕖𝕟𝕒𝕣𝕪 𝕄𝕖𝕖𝕥𝕚𝕟𝕘 𝕠𝕗 𝕥𝕙𝕖 𝔸𝔻𝕄𝕀ℝ𝔼 𝕡𝕣𝕠𝕛𝕖𝕔𝕥 possible. Until next time!!!👋🔹

Alberto Miranda 님이 재게시함

📢We are pleased to inform you that the 𝗙𝗶𝗿𝘀𝘁 𝗣𝗹𝗲𝗻𝗮𝗿𝘆 𝗠𝗲𝗲𝘁𝗶𝗻𝗴 in Madrid will be held on 19 October✈️ The event will take place at Puerta de Toledo campus of the @uc3m. 👀Check out our website: admire-eurohpc.eu

admire_eurohpc's tweet image. 📢We are pleased to inform you that the 𝗙𝗶𝗿𝘀𝘁 𝗣𝗹𝗲𝗻𝗮𝗿𝘆 𝗠𝗲𝗲𝘁𝗶𝗻𝗴 in Madrid will be held on 19 October✈️

The event will take place at Puerta de Toledo campus of the @uc3m.

👀Check out our website: admire-eurohpc.eu

Checking the venue for the 1st plenary meeting of the @admire_eurohpc project!

amiranda_hpc's tweet image. Checking the venue for the 1st plenary meeting of the @admire_eurohpc project!

Alberto Miranda 님이 재게시함

WP4 Presentation “I/O Scheduler”. Leader: Alberto Miranda (BSC)

admire_eurohpc's tweet image. WP4 Presentation “I/O Scheduler”. Leader:  Alberto Miranda (BSC)

Alberto Miranda 님이 재게시함

Kick off Meeting of the ADMIRE project starts. Jesús Carretero presents the general lines of ADMIRE. #Admire-eurohpc

admire_eurohpc's tweet image. Kick off Meeting of the ADMIRE project  starts. Jesús Carretero presents the general lines of ADMIRE. #Admire-eurohpc

Alberto Miranda 님이 재게시함

Did we mention that 595 people joined us at the @VI4IO and #IO500 BOF? Thank you all! That was amazing! #SC20 #HPC


Alberto Miranda 님이 재게시함

Jean Luca Bez presenting user-space I/O forwarding at #PDSW20. Brilliant--divorce collective I/O optimizations (e.g., ROMIO) from compute nodes; instead roll that into dedicated I/O forwarding nodes (think DVS, IOFSL--or VAST CNODES!)

glennklockwood's tweet image. Jean Luca Bez presenting user-space I/O forwarding at #PDSW20.  Brilliant--divorce collective I/O optimizations (e.g., ROMIO) from compute nodes; instead roll that into dedicated I/O forwarding nodes (think DVS, IOFSL--or VAST CNODES!)

Alberto Miranda 님이 재게시함

Omni-Path HPC Interconnect Reemerges as Intel Spin-out, Receives $20M Investment Round from Intel Capital and Others insidehpc.com/2020/09/omni-p… @intelhpc #HPC #Omni-Path #cornelisnetworks


Alberto Miranda 님이 재게시함

Intel Data Parallel C++ 2020 P3HPC presentation by Jeff Hammond (@science_dot) youtube.com/watch?v=vfX-wp… #HPC #GPU #SYCL

HPC_Guru's tweet card. 2020 P3HPC Forum, Speaker: Jeff Hammond

youtube.com

YouTube

2020 P3HPC Forum, Speaker: Jeff Hammond


Alberto Miranda 님이 재게시함

The @RookieHPC website just reached its first quarter of a million pages read! Happy to see it helps :D


Alberto Miranda 님이 재게시함

Garbage Collection

System32comicsA's tweet image. Garbage Collection

Alberto Miranda 님이 재게시함

PSA: I keep reading papers saying containers are the only path to scalable python/shared lib loading. Large #hpc jobs can DDOS filesystems by loading many libs/.py files. Spindle solves this *generally*: github.com/hpc/Spindle It does *not* need containers or special staging.


Alberto Miranda 님이 재게시함

The HiPEAC 2020 Collaboration Grants are open! If you're a PhD or a junior post-doc student and want to find new ways to improve your research in other institutions, this call is for you! Go to hipeac.net/news/6931/hipe… and make a proposal before 15 of October! #research #grants


Alberto Miranda 님이 재게시함

Excited to learn about some of the details of the future MareNostrum5 in the talk of Javier Bartolomé @JURES20 @BSC_CNS @EuroHPC_JU

rosabadia's tweet image. Excited to learn about some of the details of the future MareNostrum5 in the talk of Javier Bartolomé @JURES20 @BSC_CNS @EuroHPC_JU

Alberto Miranda 님이 재게시함

C++ 20 Draft International Standard was approved unanimously! Congrats to everyone! And I'm already taking the new bits for granted and being cranky about adoption. 😂


Alberto Miranda 님이 재게시함

Introduction to Darshan - a scalable #HPC I/O characterization tool press3.mcs.anl.gov/atpesc/files/2… #ATPESC2020 #storage

HPC_Guru's tweet image. Introduction to Darshan - a scalable #HPC I/O characterization tool 

press3.mcs.anl.gov/atpesc/files/2…

#ATPESC2020 #storage

We have a new publication with Jean Bez from UFRGS on #HPC I/O forwarding on Future Generation Computer Systems! Check it or download it for free until Sept. 8th authors.elsevier.com/c/1bRDJ,3q5xZv…


Alberto Miranda 님이 재게시함

Agreed. I too am impressed with how DAOS did on the @IO500benchmark. Perhaps a silver lining in all the current negative news surrounding @intel .... #HPC #Storage #Exascale

I'm usually not impressed by benchmarks, but on IO500 (vi4io.org/io500/start) I discovered DAOS filesystem (github.com/daos-stack/daos) A full optane/nvme storage, with HDF5 optimization looks promising for #HPC and the bw/iops are amazing.



Alberto Miranda 님이 재게시함

As we transition #SC20 to fully virtual, we will keep as many elements of the in-person program as possible, though we have shuffled things around a bit to allow for maximum attendance. Learn about our new schedule & more in our FAQ's: okt.to/pVbdFW

Supercomputing's tweet image. As we transition #SC20 to fully virtual, we will keep as many elements of the in-person program as possible, though we have shuffled things around a bit to allow for maximum attendance. Learn about our new schedule & more in our FAQ's: okt.to/pVbdFW

Our file system #GekkoFS makes it to the Top 10 in the @IO500benchmark 10-node challenge, proving that ephemeral file systems using @intelhpc’s DCPMM have a lot in store for HPC applications! vi4io.org/io500/list/20-…


United States 트렌드

Loading...

Something went wrong.


Something went wrong.