Discover and read the best of Twitter Threads about #GPU
Most recents (18)
DeepSquare integrates the neural network #text2image model #StableDiffusion.
This thread presents how we #designed and implemented the #workflow to execute Stable Diffusion on the DeepSquare Grid.
This thread presents how we #designed and implemented the #workflow to execute Stable Diffusion on the DeepSquare Grid.
We designed the workflow as follows:
The #prompt is sent over an environment variable.
Most of the parameters are sent via environment varibles.
Images are sent to transfer.deepsquare.run
The #prompt is sent over an environment variable.
Most of the parameters are sent via environment varibles.
Images are sent to transfer.deepsquare.run
The docker image is already compiled from github.com/Stability-AI/s… and is exported as registry-1.deepsquare.run/library/stable…
The steps are as follows:
1-Generate the images. Rows are distributed into task per #GPU.
2-Combine all the images into one grid.
The steps are as follows:
1-Generate the images. Rows are distributed into task per #GPU.
2-Combine all the images into one grid.
1/ #Nvidia is world’s eighth largest companies in terms of Market Cap (~$582 bn).
It is not as large in terms of revenue ($27 bn, FY23 Jan YE), but perhaps its business and its opportunities make the market value this high. Perhaps. Difficult to fathom.
A thread.
It is not as large in terms of revenue ($27 bn, FY23 Jan YE), but perhaps its business and its opportunities make the market value this high. Perhaps. Difficult to fathom.
A thread.
2/ Set up in 1993, Nvidia makes chips, software and systems. Or platforms.
It was set up in 1993.
In 1999, it invented the #GPU which forms its core business even today.
From early focus on PC Graphics, they have expanded to several computationally intensive fields.
It was set up in 1993.
In 1999, it invented the #GPU which forms its core business even today.
From early focus on PC Graphics, they have expanded to several computationally intensive fields.
3/ What is a GPU?
GPU is Graphics Processing Unit. Initially launched for graphics, it has now evolved into an important computational tech.
In 2006, Nvidia introduced CUDA programming model which opened parallel processing capabilities of GPU for general purpose computing.
GPU is Graphics Processing Unit. Initially launched for graphics, it has now evolved into an important computational tech.
In 2006, Nvidia introduced CUDA programming model which opened parallel processing capabilities of GPU for general purpose computing.
Yli kymmenen vuotta @Jimms_fi asiakkaana ja tuhansia euroja sinne käyttäneenä, nyt ensimmäinen ongelma heiltä ostetussa tuottessa.
@Jimms_fi toimii LAINVASTAISESTI takuu/RMA-tilanteissa, ja täyttää toiminnassaan jopa petoksen tunnusmerkit.
Ketju /1
#petos #RMA #Takuu #GPU
@Jimms_fi toimii LAINVASTAISESTI takuu/RMA-tilanteissa, ja täyttää toiminnassaan jopa petoksen tunnusmerkit.
Ketju /1
#petos #RMA #Takuu #GPU
Ongelma lyhyesti:
Näytönohjaimen keskimmäinen tuuletin alkoi elää omaa elämäänsä (0-100% välillä), ja sitten lakkasi yhtäkkiä toimimasta ilman minkäänlaisia käyttäjän toimia. GPU toimi ennen tätä virheettömästi, ja tämän jälkeenkin kahdella tuulettimella normaalisti.
/2
Näytönohjaimen keskimmäinen tuuletin alkoi elää omaa elämäänsä (0-100% välillä), ja sitten lakkasi yhtäkkiä toimimasta ilman minkäänlaisia käyttäjän toimia. GPU toimi ennen tätä virheettömästi, ja tämän jälkeenkin kahdella tuulettimella normaalisti.
/2
14.2 - Jimm's PC Store pyytää "lisätietoja valmistajalta" (ASUS) näytönohjaimen takuukäsittelyä varten.
/3
/3
Some initial results from the #CASP15 competition and we see that #Alphafold2 has become a fertile ground for experimentation by several research groups around the world. Facebook's ESM protein language models (pLMs) are the top non-MSA based methods.
Another slide showing some of the experimentation taking place: increase the number of models, recycles, get more diversity using dropout (how would that work?), subsample MSAs.
The #metaverse is inevitable and it's going to be worth TRILLIONS
BUT there’s one huge problem – our current infrastructure can’t support it
Protocols such as @RenderToken aim to fix this and, as such, have the potential to increase in value by 30x – 80x
Here’s how
👇
🧵
BUT there’s one huge problem – our current infrastructure can’t support it
Protocols such as @RenderToken aim to fix this and, as such, have the potential to increase in value by 30x – 80x
Here’s how
👇
🧵
2/
Render is a decentralized “marketplace” that allows users to rent out unused #GPU processing power to #render 3D objects (we’ll explain this in a second)
It has a MC of $123M, FDV of $260M, and its $RNDR token trades at $0.48
In 2021, the token price hit $8.76
Render is a decentralized “marketplace” that allows users to rent out unused #GPU processing power to #render 3D objects (we’ll explain this in a second)
It has a MC of $123M, FDV of $260M, and its $RNDR token trades at $0.48
In 2021, the token price hit $8.76
3/
This thread will cover the following:
• Industry overview
• What problem does Render solve?
• How does it work?
• What are its plans for the future?
• Who are the key players in the ecosystem?
• What are its #tokenomics?
• What’s the potential value of #RNDR?
This thread will cover the following:
• Industry overview
• What problem does Render solve?
• How does it work?
• What are its plans for the future?
• Who are the key players in the ecosystem?
• What are its #tokenomics?
• What’s the potential value of #RNDR?
1/ We've been working on Exactly that, for some time now. While the general idea is nice and correct...
Leveraging existing GPU mining infra to do ZK is in fact FAR from straight forward.
Time for a thread!🧵👇
Leveraging existing GPU mining infra to do ZK is in fact FAR from straight forward.
Time for a thread!🧵👇
2/ GPU mining was dominated by #Ethereum mining which is now, post merge, no longer possible.
So we came up with the idea of helping the community repurpose their GPUs for ZK compute a few weeks ago...
So we came up with the idea of helping the community repurpose their GPUs for ZK compute a few weeks ago...
3/ ...and we got CRAZY demand.
Which took us completely by surprise! (surely, we thought, miners with tens of thousands of GPUs had a plan B for after the merge, right? well no.. but that's another story to tell)
Which took us completely by surprise! (surely, we thought, miners with tens of thousands of GPUs had a plan B for after the merge, right? well no.. but that's another story to tell)
I did @officialBinotto's @ScuderiaFerrari SF71H in @FluidX3D #CFD on a supercomputer.
- 1s in real life @ 100km/h
- 20s 4K60 video (3x)
- 14h compute on 8x @AMDInstinct #MI200 64GB #GPU
- 144TB data visualized
What I found is absolutely wild. A #SimulationFriday #F1 thread: 🧵1/5
- 1s in real life @ 100km/h
- 20s 4K60 video (3x)
- 14h compute on 8x @AMDInstinct #MI200 64GB #GPU
- 144TB data visualized
What I found is absolutely wild. A #SimulationFriday #F1 thread: 🧵1/5
The @FluidX3D simulation was done at 10 billion voxel grid resolution (2152×4304×1076), over 217k time steps (1 second), at Re=3.75M (100km/h).
The fins on the front spoiler create a turbulent boundary layer and kick up it up onto the front wheels to reduce drag. 🧵2/5
The fins on the front spoiler create a turbulent boundary layer and kick up it up onto the front wheels to reduce drag. 🧵2/5
The streamlined chassis guides airflow under the spoiler to create down force. The halo - one of the best additions to the sport in terms of safety - is rather aerodynamic.
Each frame of the video is 120GB, 144TB for 1201 frames. @FluidX3D renders the data directly in VRAM. 🧵3/5
Each frame of the video is 120GB, 144TB for 1201 frames. @FluidX3D renders the data directly in VRAM. 🧵3/5
What is Ergo?
Ergo is a programmable PoW #blockchain like #Ethereum, without gas fees and builds upon #Bitcoin's UTXO model.
#Ergo serves as an efficient, decentralized and secure, smart contract platform, with optional privacy.
🧵👇 Let's Explore
Ergo is a programmable PoW #blockchain like #Ethereum, without gas fees and builds upon #Bitcoin's UTXO model.
#Ergo serves as an efficient, decentralized and secure, smart contract platform, with optional privacy.
🧵👇 Let's Explore
Fair launch, no pre-mine no #VCs
"#Cryptocurrency should provide tools to enrich ordinary people. Small businesses that are struggling to make ends meet, not big depersonalized financial capital. This is what inspired me. This is my dream"
@chepurnoy
ergoplatform.org/en/blog/2021-0…
"#Cryptocurrency should provide tools to enrich ordinary people. Small businesses that are struggling to make ends meet, not big depersonalized financial capital. This is what inspired me. This is my dream"
@chepurnoy
ergoplatform.org/en/blog/2021-0…
Q5K: Quake level viewer in 5K LUTs on a low cost, low power ice40 up5k #fpga! Custom #GPU, @risc_v CPU and SOC, capable of rendering #Quake's level with lightmaps.
How? Thread 👇
(Written in #Silice, here running on the #mch2022 badge fpga)
How? Thread 👇
(Written in #Silice, here running on the #mch2022 badge fpga)
The (tiny) GPU is my DMC-1 (Doom-Meets-Comanche) GPU, which also powers the Doomchip-onice demos (remember? Doom with a terrain!!).
It targets the ice40 UP5K, an entry-level fpga with great support from the Open Source toolchain #yosys/#nextpnr.
2/n
It targets the ice40 UP5K, an entry-level fpga with great support from the Open Source toolchain #yosys/#nextpnr.
2/n
There were four main hardware changes to enable Quake level rendering:
1) 32-bits per-column depth,
2) streaming of level data from QPI memory (SPIflash on icebreaker, PSRAM on mch2022 badge),
3) multi-texturing for lightmaps (!!)
3/n
1) 32-bits per-column depth,
2) streaming of level data from QPI memory (SPIflash on icebreaker, PSRAM on mch2022 badge),
3) multi-texturing for lightmaps (!!)
3/n
@boblautenbach @Alan_L_Lovejoy @Ryu_Zoku Yes Chris Lattner is very interesting, been following his #LLVM and #SWIFT work.
1/
1/
@boblautenbach @Alan_L_Lovejoy @Ryu_Zoku Actually #ZokuTalk adopts the portion of SWIFT #EBNF for #Unicode as they seems have done good job on that, as well as parts of the EBNF for numerics such as "1_234.00" to expand privative data types entry.
\2
\2
@boblautenbach @Alan_L_Lovejoy @Ryu_Zoku Borrowing C & Assembly data types: 0b0011, 0xFF, keeping 16r0A, plus many others.
Actually the numeric formats are one of more complicated part of the #EBNF of #ZokuTalk and the only non-#MOBS (#Messages #Objects #Block #Syntax) part of the evolved syntax.
\3
Actually the numeric formats are one of more complicated part of the #EBNF of #ZokuTalk and the only non-#MOBS (#Messages #Objects #Block #Syntax) part of the evolved syntax.
\3
I have benchmarked the new @AMDInstinct MI250 #GPU at @fzj_jsc, and it is disappointing but also impressive. Let me explain.
🧵1/6
cc @HPC_Guru @hpcprogrammer @IanCutress @ProfMatsuoka @sunitachandra29 @aschilling @wkmyrhang @VideoCardz @AMDGPU_
#HPC #Top500 #MI200 #Exascale
🧵1/6
cc @HPC_Guru @hpcprogrammer @IanCutress @ProfMatsuoka @sunitachandra29 @aschilling @wkmyrhang @VideoCardz @AMDGPU_
#HPC #Top500 #MI200 #Exascale
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…
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
The #MI250 promises exactly that with "128GB", but delivers only half.
🧵3/6
(1/20)@MessariCrypto 关于ETH2.0的到来将给矿工带来的巨大冲击作了分析，译了一下，饭后谈资。
重点见解：
1. 2.0将迫使以太坊价值 190 亿美元的采矿业寻找新家。
2.对于大多数现有的以太坊矿工来说，开采替代的工作量证明代币在经济上是不可持续的。
#Ethereum #eth2 #Miner #以太坊 #挖矿
重点见解：
1. 2.0将迫使以太坊价值 190 亿美元的采矿业寻找新家。
2.对于大多数现有的以太坊矿工来说，开采替代的工作量证明代币在经济上是不可持续的。
#Ethereum #eth2 #Miner #以太坊 #挖矿
(2/20)不包括 ETH 的 GPU 可开采代币的总市值为 41 亿美元，约占 ETH 市值的 2%。 ETH 还占 GPU 可开采硬币的每日矿工总收入的 97%。
3.许多大型矿工计划转向以数据中心为导向的业务，重点是提供高性能计算服务。
矿工可以将他们的 GPU 贡献给 @RenderToken 、@Liverpeer 和 @akashnet_ 等 Web3 协议
3.许多大型矿工计划转向以数据中心为导向的业务，重点是提供高性能计算服务。
矿工可以将他们的 GPU 贡献给 @RenderToken 、@Liverpeer 和 @akashnet_ 等 Web3 协议
(3/20)基于 ETH 核心开发者的论调和最近在 Ropsten 测试网上的成功，假设没有挫折，主网Merge似乎在 8 、9 月步入正轨。
2022 年 5 月，以太坊的哈希率或用于保护网络的总算力达到顶峰。从那以后，它的哈希率一直呈下降趋势，这表明矿工们预计Merge将很快发生，关停并开始售卖矿机(图为以太坊的哈希率)
2022 年 5 月，以太坊的哈希率或用于保护网络的总算力达到顶峰。从那以后，它的哈希率一直呈下降趋势，这表明矿工们预计Merge将很快发生，关停并开始售卖矿机(图为以太坊的哈希率)
A short thread on accelerated computing. How $TSLA's #Dojo, $GOOG's #TPU #Cerebras's #WSE-2, #Graphcore's #IPU, and others compare to $NVDA 's #GPU. This is an extract of work we published a few weeks ago. 👇1/6
It is difficult to compare chips. Designing a chip is all about managing trade-offs between multiple dimensions. A chip has limited potential resources and the architect allocates them over these multiple dimensions. 2/6
With this framework in mind, one sees alternatives to #GPU have fundamentally different architectures, favoring the flow of data across the chip vs. the flow of data between the memory and the chip. 3/6
1.Well You guys asked and we couldn't handle the hype of delivering our @GuildOfGuardian report 🧵
Have a GOG read⬇️
#GOG is a mobile #blockchain multiplayer, fantasy, action RPG where players build their #NFT dream team of Guardians and compete in a #guild to earn epic rewards
Have a GOG read⬇️
#GOG is a mobile #blockchain multiplayer, fantasy, action RPG where players build their #NFT dream team of Guardians and compete in a #guild to earn epic rewards
2.The game is developed by @StepicoGame and built on @Immutable.
#GOG ambitious goal started in early 2020 and it is set to take off thanks to these 4 main areas where the GOG team wants to focus the most when making the #blockchain game.
#GOG ambitious goal started in early 2020 and it is set to take off thanks to these 4 main areas where the GOG team wants to focus the most when making the #blockchain game.
[#Thread] Mesdames et messieurs, devant vos yeux ébahis (musique de cirque), nous allons aborder le thème :
Comment gagner de l'#argent avec les #cryptomonnaies
Thread à déguster très frais accompagné d'une coupe de Dom Pérignon bien assis dans votre fauteuil préféré. ⬇️
Comment gagner de l'#argent avec les #cryptomonnaies
Thread à déguster très frais accompagné d'une coupe de Dom Pérignon bien assis dans votre fauteuil préféré. ⬇️
Avant de commencer, sachez qu'il n'existe pas de méthode miracle et chaque #investissement vous demandera du #temps ou de l'#argent.
Certaines méthodes vous permettront de gagner de l'argent de façon passive, d’autres demanderont un peu de travail et certaines compétences.
Certaines méthodes vous permettront de gagner de l'argent de façon passive, d’autres demanderont un peu de travail et certaines compétences.
Ensuite, mon but est de simplement "donner des pistes" au travers de ce #thread.
Les caractères étant limités, des exemples ainsi que d'avantages d'#explications seront disponibles sur l'#article complet en fin d'article sur @AuCoinduBloc.
(L'article est un pavé)
Les caractères étant limités, des exemples ainsi que d'avantages d'#explications seront disponibles sur l'#article complet en fin d'article sur @AuCoinduBloc.
(L'article est un pavé)
(1) Possible buyout of #ARM by #NVIDIA does have an effect on the #Bioinformatics field: many applications now are deployable on CPU/GPUs with #ARM and/or #NVIDIA chips on them. Some recent examples are:
(a) the Oxford @nanopore MinION Mk1c device, which originally was specced at Jetson TX2 ARM+Pascal GPU accelerators (ARM processor 6 cores, 256 Core GPU), 8 GB RAM (may have changed since then.
#EdgeAI for everyone! — @NVIDIA Jetson Nano developer kit = low-cost #AI computer delivers compute performance to run AI workloads at unprecedented size, power, & cost, plus it connects to a diverse set of sensors: amzn.to/2ZrT9Cj
—————
#MachineLearning #IoT #IoTPL #IIoT
—————
#MachineLearning #IoT #IoTPL #IIoT
.@NVIDIA Jetson Nano 2GB — The Ultimate #AI and #Robotics #Edge platform with an entry-level price: dy.si/MByxJ
————
#BigData #DataScience #EdgeAI #IoT #IIoT #IoTPL #EdgeComputing #MachineLearning #DeepLearning @NVIDIAAI #GPU
————
#BigData #DataScience #EdgeAI #IoT #IIoT #IoTPL #EdgeComputing #MachineLearning #DeepLearning @NVIDIAAI #GPU
The #Nvidia Jetson Nano is boosting the growing industrial #IoT revolution: bit.ly/3koHOP2
——————
#abdsc #EdgeAnalytics #EdgeComputing #IIoT #IoTPL #StreamingAnalytics #BigData #DataScience #AI #EdgeAI #MachineLearning #GPU
———
See it here: amzn.to/3ahOie5
——————
#abdsc #EdgeAnalytics #EdgeComputing #IIoT #IoTPL #StreamingAnalytics #BigData #DataScience #AI #EdgeAI #MachineLearning #GPU
———
See it here: amzn.to/3ahOie5
I’m excited to announce the preprint of our paper “Bringing #UMAP Closer to the Speed of Light with #GPU Acceleration” is now available on Arxiv! Special thanks to all the co-authors @viclafargue1, @teju85, @EdwardRaffML, @oatesbag, @zstats @datametrician for making it possible.
This paper describes many design tricks, as well as core principals we followed, to accelerate #UMAP in @rapidsai #cuML and achieve SOA performance. Many of these tricks have been used to accelerate other algorithms in the #GPU computing ecosystem.
This paper was a team effort, and would not have been possible without such a wonderful group of co-authors. I’m very happy with how it turned out.
