Discover and read the best of Twitter Threads about #Top500

Most recents (9)

Finally reached my true rank and this is only the beginning ♥️ This is what my ( the best mercy EU ) stats look like if you wanna know. I feel like i have truly unlocked Mercys' full potential with her recent changes ♥️😇#heroesneverdie #mercy #GMsupport
Next stop 🔜 #Top500
I wanna thank my mercy influencers for helping me improve on this hard journey. Improvement takes time, 34 seasons and counting i will never stop trying. @skiesti_ @Animetic @LuxyCoulee @KhaleesiBB @AngelxoTv @HoshizoraOW @eleyzhau @Niandra_ @AquaBoostTV
Read 4 tweets
The #MI250 is misleadingly marketed as "one chiplet GPU" with 13312C, 90TFLOPs & 128GB @ 3.2TB/s.

But it is not. The 2 GCDs are 2 separate GPUs with 64GB each, like a K80 dual-GPU but in a socket. One #GPU can't directly access the other's memory.
🧵2/6

amd.com/en/products/se…
To use both GCDs, the software needs to be multi-GPU capable. For many algorithms this is very difficult and for some it is entirely infeasible. The desire for large unified memory is huge.
The #MI250 promises exactly that with "128GB", but delivers only half.
🧵3/6
Read 8 tweets
1/11 Ever wondered how Supercomputers multiply? Efficient multiplication is a critical part of traditional high performance computing as well as machine learning. Let's look at the #IBM Blue Gene/Q, which led the #Top500 and #Green500 in the late 2000s.
en.wikipedia.org/wiki/IBM_Blue_…
2/11 IBM Open Sourced the core to Blue Gene/Q (the @OpenPOWERorg A2I) so it offers a fantastic insight into a CPU that taped out at frequency and in volume. The #VHDL for the core is available at:
github.com/OpenPOWERFound…
3/11 We quickly see that it uses a few well established techniques for optimising multiplication. The first step uses Booth encoding:
en.wikipedia.org/wiki/Booth%27s…
Read 11 tweets
#SC21 #Top500:

ATIP's More Realistic China Top100?

Better pictures in this thread compared to the photo from Tiffany I tweeted earlier

#HPC #AI via @Cozy57 @glennklockwood
@Cozy57 @glennklockwood 2/

#Exascale in China: 2 EFlops systems today, 3rd delayed

#HPC #AI #SC21 #Top500
@Cozy57 @glennklockwood 3/

#HPC in China: Out of 10 #supercomputing centers planned, 8 are in operation

#AI #SC21
Read 6 tweets
#Top500 -Still waiting for #Exascale: Japan's #Fugaku outperforms all competition once again

@Azure system called Voyager-EUS2 was the only machine to shake up the top spots, claiming #10

top500.org/news/still-wai…

#HPC
#Top500: Only one change in the Top10 systems - Voyager-EUS2 @Azure using AMD EPYC CPU and Nvidia A100 GPU

Perlmutter @NERSC improved its performance to 70.9 Pflop/s but remained at #5 on the list

#SC21 #HPC
#Top500:

#Fugaku continues to hold the No. 1 position that it first earned in June 2020

It was co-developed by Riken and Fujitsu and is based on Fujitsu’s ARM A64FX processor

It uses Fujitsu’s Tofu D interconnect

#HPC #SC21
Read 7 tweets
#SC21: Press briefing starting now

Bronis R. de Supinski @Livermore_Lab mentions SC21 is the first hybrid @Supercomputing conference

#HPC #AI
@Supercomputing ~3800 in-person attendees, according to Bronis

More than I expected....

#SC21 #HPC
@Supercomputing #SC21: @mark_epcc from @EPCCed covers the finalist papers for the Gordon Bell Prize, which recognizes outstanding achievements in #HPC

Also the finalists for GB for #Covid19
Read 11 tweets
#SC21: @HPC_Hyperion #HPC Market briefing (without the breakfast this time 😉 )

Despite Covid, #Fugaku single-handedly saved the HPC revenue numbers for 2020 - a billion dollars makes a difference!
@HPC_Hyperion .@HPE is the top #HPC vendor, followed by @DellTech

@Fujitsu_Global jumped due #3 due to #Fugaku - so this is their 15 minutes of fame :-)

@InspurServer & @lenovo make up the Top 5

#SC21
@HPC_Hyperion The broader on-prem #HPC market

#HPC #storage remains #2 after compute, which is the biggest item as usual

#SC21
Read 28 tweets
People have asked me about my #Top1000 list and how it is calculated so I thought I would do a thread about it.

I have been doing this list for over 10 years since I joined Twitter in April 2009 as I wanted a way to grow my followers and also reward engagement.

1/
It started as a #Top500 list but a few years ago I changed it to a #Top1000 list.

I don't determine where people are in the list as it is based on people engaging with with my tweets over a 7 day period since my last list update.

2/
I award Engagement Points for different activities as follows:

Retweets 9 points.
Replies 6 points.
Likes 3 points.

I accumulate the totals for each person in a file and I have an Excel Spreadsheet and VB Macros to produce the list. It also builds the blog post and tweets.

3/
Read 12 tweets
1/ How should we talk about DP FLOPS for A100? A thread. #HPC #AI #Nvidia
2/ Nvidia announced A100 this morning and you can read about it from any number of outlets, e.g. nextplatform.com/2020/05/14/nvi…
3/ Nvidia, as usual, have themselves published a great set of detailed blogs about the architecture and specs, really diving into the approaches they've taken and how they might impact workloads.
devblogs.nvidia.com/nvidia-ampere-…
Read 13 tweets

Related hashtags

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3.00/month or $30.00/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!