#dcay

[roguesenses]

❛ i won’t stop killing until you kill me. ❜

&. 𝐝𝐚𝐫𝐤 𝐚𝐧𝐝 𝐚𝐧𝐠𝐬𝐭𝐲 𝐬𝐞𝐧𝐭𝐞𝐧𝐜𝐞 𝐬𝐭𝐚𝐫𝐭𝐞𝐫𝐬. (accepting)

Collapsed  buildings,  ruptured  ground,  and  crimson-soaked  soil.  The  air  held  a  faint  metallic  scent  from  all  the  blood  shed  by  heroes  during  this ��battle.  The  people  that  he  had  admired,  cared  for,  and  perhaps  even  loved.  Every  time  he  drew  in  a  breath,  Izuku  tasted  their  agony  and  was  reminded  of  his  failure.  

Guilt  transformed  quickly  into  anguish,  and  from  there  gave  away  to  anger.  His  mind,  for  a  time,  begged  to  be  taken  over  by  pure  rage  because  in  that  state  he  wouldn't  have  to  acknowledge  or  experience  the  pain  of  his  losses.  

Except  he  could  not  let  himself  do  that...must  not  be  that  irresponsible  when  there  was  still  so  much  sitting  on  his  shoulders.  He  has  to  be  a  hero,  turn  the  situation  around  and  rescue  everyone.  

And  that  included  him.  

"Why?"  Izuku  demanded.  "Why  won't  you  just  stop?"  

Perhaps  they  were  past  the  point  of  communicating,  but  Izuku  could  not  stop  the  words  from  spilling  from  his  lips.  Emerald  eyes  trained  on  his  supposed  nemesis,  bright  light  spilling  from  the  corners  of  his  eyes  instead  of  tears.  "Isn't  there  any  part  of  you  that  wants  more  than  this  from  the  world?"  He  gestured  at  the  wreckage  that  surrounded  them,  the  mess  that  had  become  of  what  was  once  beautiful.

"I  won't  let  you  destroy  everything,"  he  told  the  villain.  "All  for  One's  greed  for  power  and  your  hatred  will  not  triumph  over  our  collective  will."  Izuku  was  not  in  this  fight  alone,  and  all  the  previous  users  of  One  for  All  remain  determined.  "I  will  defeat  you,  Shigaraki,  and  then  I  will  save  you."  

There  was  a  child  there,  buried  deep  inside,  thin  shoulders  shaking  as  he  cried.  No  amount  of  wrath  would  make  Izuku  forget  the  boy,  just as  no  amount  of  deterrence  should  turn  a  hero  away  from  his  core  ideal  of  trying  to  save  everyone.  

[kibouhero]

//continued from here.

Honestly, as much as Izuku wanted to defend the friends he had, he couldn't find it in him to do it. As much as he hated to agree with Tenko, he was sure his brother had a decent point. Though, just thinking about the fact that even his childhood friends were fickle made him sad.

"Yeah... maybe you're right..." he relented quietly.

His eyes widened in interest when Tenko gave the name of the game. He'd at least heard of it but obviously never played it. "...Oh, a beat -em up fighting game...!" he exclaimed as he watched his brother pick a character.

Yeah, he could see their mom taking issue with the violence in the game...

He gave a nod. "Okay, I won't mention anything!" he said more quietly before he focused on picking a character.

Eventually, all of the relevant choices were selected, and they waited for the stage to load. Izuku's hands tightened over the controller from anticipation.

@dcay

[explodcor]

//continued from here (new post cuz legacy editor).

How in the fuck--?!

It was like all Shigaraki had to do was wave his hand and they were immediately in another part of the city. He glanced around, looking for any sort of landmark that could indicate their location, but nothing looked familiar.

His eyes then locked onto the villain's back. They were alone--as far as he knew--It'd be so easy to just--

His palms were practically itching to attack the bastard. Hell, even just shouting for help didn't see all that crushing to his pride anymore. He was sure there were people nearby.

Katsuki's resistance died in his throat the moment Crusty said that he knew he wanted to attack him. Shit!

He knew he was fucked. Shigaraki wasn't bluffing. If Katsuki attacked, he'd likely be dead within seconds.

He sucked in a deep breath and let it out. He quickened his walking pace, so he could walk side by side with the villain rather than remain behind him. "...Okay, chat about what?" he asked quietly.

@dcay

[houseofmonsterz]

(spoilers: the third one is you)

[teensy-universe]

i've been meaning to make posts here that aren't rbs but haven't gotten around to it. urrmm have him

[m1shapanda]

if i’m being honest i feel like shit :(( like i just want to day to end already

nauruururur :(( i hope u feel better soon!! huhu sometimes shit just sucks huhu

[feralmoonlight]

The DCAI gave me a fullass plot for Vampire Sun actually working in the daycare and playing it off like he's not a vampire (but he does think biting is neat) and teasing his handler about asking about vampires and building a little relationship and one day at crafts time makes them a little red ribbon 'necklace' and something something oops he's actually a vampire, is reader gonna let him bite them for real?

[mod-python]

if it makes you feel any better anon, I only do digital art and I never claim anyone who says no traditional art in their form because it annoys the hell out of me. some of my oc’s most treasured artworks are traditional comms or dcay pieces. if you do manage to get someone who’s not up their own ass about traditional stuff it might end up being their favorite thing they’ve gotten.

.

[mfgentlegrowl]

My first DCAY post in a long time :)

This is Caramella for Fan_Endey

[invece-sto-sdraiato]

thank you @ne-spivay-ranenko for tagging me 🥰🥰

I've been listening to the same stuff on repeat so there's not much variety here...

I tag @anxious-witch @jerevision @arendellepeach @vikersund @vse-kar-vem @saallotar (no pressure, anyone who wishes to do can ofc 😘)

[godofvillains-arch]

Art by dcay!

Mon (or Mon-chan. Mon the II) Is Shigaraki's puppy and ALSO a emotional support animal. Shigaraki even sometimes brings him to stores and the sorts to help calm him down. Mon also wakes Shigaraki from nightmares and helps soothe him when he is angry or depressed or suffering from a PSTD episode. Mon does a LOT to help Shigs and the two have a very tight bond.

[h0useofw0lves]

best release of 2022 ?

fundatons of dcay (sorry boring answr i just don’t listen to much new music i’m sory i know it’s bad 😔)

[systemtek]

Denmark's new AI supercomputer Gefion ranked as 7th fastest storage systems in the world - Denmark's state-of-the-art AI supercomputer, Gefion, has achieved another global recognition by ranking 7th overall and 4th in the 10-node benchmark on the internationally recognised IO500 list. Gefion, Denmark's cutting-edge AI supercomputer, has secured a prominent position as the 7th fastest production storage system globally, while also ranking 4th in the 10-node benchmark on the IO500 list, which rates the world's fastest storage systems annually. This recognition highlights Gefion's exceptional data-handling capabilities, positioning it not only as a standout AI and high-performance computing resource but also as one of the most efficient data-handling systems worldwide. In addition to this achievement, Gefion recently ranked 21st globally on the TOP500 list of the world's most powerful supercomputers, further cementing its status as a high-performance powerhouse. The IO500 benchmarks evaluate the ability of supercomputers to read, write and transport data at lightning-fast speeds, a crucial factor in modern AI and high-performance computing. Gefion's performance on this list reinforces its reputation as a cutting-edge system, excelling both in computational power and data management efficiency. The significance of this 7th-place ranking lies in Gefion's outstanding input/output (IO) performance — its speed in accessing, retrieving, and storing data. This capability is vital for AI model training and complex large-scale simulations, where processing vast data sets — often measured in petabytes — can be completed in days or hours instead of weeks. "Gefion's ranking on the IO500 list is another testament to our commitment to pushing technological boundaries in advanced computing for AI and research," says Nadia Carlsten, CEO of the Danish Centre for AI Innovation (DCAI), which runs and operates Gefion.  "The powerful data storage and processing capabilities of Gefion mean that researchers can run larger, more intricate simulations and process datasets faster than ever," she continues. Strong IO performance can directly contribute to breakthroughs in areas such as climate modelling, drug discovery, and sustainable energy research. "Gefion is a monumental step forward for AI research and innovation, delivering the scale and precision needed to tackle the most complex challenges in fields like healthcare, sustainable energy, and quantum computing," says Alex Bouzari, Co-founder and CEO of Data Direct Networks (DDN), who is supplying storage for the Gefion AI supercomputer. "By combining DDN's industry-leading data platform with NVIDIA's groundbreaking technology, Gefion is not only advancing Denmark's leadership in AI but also enabling a new era of scientific discovery and global progress. We're proud to support this world-class initiative, which epitomizes the transformative power of AI and HPC when supported by the right infrastructure", he adds. Gefion's technological backbone includes NVIDIA DGX SuperPOD architecture, featuring 1,528 NVIDIA H100 Tensor Core GPUs interconnected with NVIDIA Quantum-2 InfiniBand networking. These components ensure not only unmatched processing power but also superior data access speeds, an essential factor for projects that rely on rapid results. DCAI designed Gefion with future scalability and flexibility in mind, allowing it to integrate upcoming generations of NVIDIA accelerated computing technologies, ensuring that Denmark remains at the forefront of AI and high-performance computing innovation. Danish Centre for AI Innovation (DCAI) Read the full article

[houseofmonsterz]

evil horrible creature with malicious intent

[carolynhyde]

Data Centric ai Data-centric AI (DCAI), a cutting-edge branch of AI era, is concerned with expertise, the use of, and drawing conclusions from facts. AI changed into once heavily depending on hints and heuristics in advance than becoming information-centric. These is probably beneficial in some occasions, but even as used on fresh data sets, they frequently produce much less-than-perfect outcomes or errors. By adding tool analyzing and massive facts analytics equipment, statistics-centric Artificial Intelligence modifies this through allowing it to investigate from statistics rather than relying on algorithms. It can for that reason make wiser options and deliver more particular effects. It moreover has the functionality to scale up a long way more than conventional AI techniques do. Data-centric AI will possibly keep growing in importance as statistics units get larger and more complicated. What is Data Centric Architecture in AI? The facts-centric approach includes meticulously optimising datasets to elevate the accuracy of AI structures. This strategy has capability, in step with machine analyzing professionals, due to the truth processed information produces better results than uncooked facts. A information-centric method locations awesome statistics input beforehand of converting the model's parameters. In gadget mastering, tagged photographs, textual content, audio documents, movies, and one of a kind information are used as training records. If the schooling information is awful, the created model and its optimisation will perform poorly. With AI-powered chatbots, this could bring about terrible consumer reports, but it may be disastrous for natural algorithms or unbiased automobiles. How Does Data Centric AI Work? Data augmentation, interpolation, and extrapolation are 3 strategies utilized by facts-centric AI to comply in your corporation's necessities. Using facts-centric AI, you do not need to educate a model on a particular dataset. Instead, the device makes a easy prediction from the education data provided thru your enterprise. This manner that a model evolved using facts out of your business agency will probable moreover artwork well with one-of-a-kind datasets. You may additionally additionally decorate the awesome of your fashions through manner of developing extra instances of an modern-day instance thru extrapolation or interpolation. Additionally, it entails growing new facts instances from older ones. Either extrapolation or interpolation may be used to finish this mission. The following steps make up a records-centric AI technique in famous: Labelling your datasets efficiently and solving any errors casting off noisy facts times from the analysis Feature Engineering Error Analysis for Data Augmentation Better consequences can be acquired with the resource of using area professionals to evaluate the accuracy or consistency of data factors. Data Centric AI Architecture Advantages Enhance Performance. A information-centric approach includes growing AI systems with exquisite statistics, ensuring that the records transmits the statistics the AI wants to learn. In addition to reducing wasteful trial and mistakes time spent growing the version without enhancing inconsistent data in a particular facts set, this aids teams in accomplishing the needed overall performance level. When first-rate control is statistics-centric, collaboration is superior amongst managers, professionals, and developers. They can collaborate whilst growing defects or tags with the intention to be regular via agreeing on them or with the resource of developing a model earlier than reading the consequences if you want to perform extra optimisations. Data-centric AI accelerates improvement considering agencies may go concurrently and right now impact the facts that the AI device makes use of. Data Centric AI Architecture Disadvantages It is probably tough to reveal and manipulate the great of statistics. If statistics units don't correctly replicate the populace, they may be prejudiced. This method can be steeply-priced because a number of facts is wanted to educate the models. Conclusion The results of fashions are extra accurate way to facts-centric AI, which additionally opens up new applications for this concept. As developers running with AI place a extra emphasis on models than information, it is gaining traction. The first-rate of the input records is now more regularly considered while enhancing effects than inside the beyond when engineers hired model-centric strategies to increase the outcomes and accuracy of version predictions.

[govindhtech]

DDR5 MRDIMM: The Future Of Data Transfer And Performance

Intel Data Center Chips are enhanced with new ultrafast memory.

DDR5 MRDIMM provide a novel, effective module design to improve system performance and data transfer speeds. In order to enable applications to surpass DDR5 RDIMM data speeds, multiplexing enables the combination and transmission of several data signals over a single channel, hence expanding the capacity without the need for more physical connections.

How by cleverly doubling the memory bandwidth of conventional DRAM modules, Intel and industry partners were able to release top-end Xeon CPUs with a plug-and-play solution. System memory, or DRAM, is essential to performance even though Intel’s main product emphasis is on the processors, or brains, that power computers. In servers, this is particularly true, since the number of processor cores has increased faster than the memory bandwidth (i.e., the memory bandwidth available per core has decreased).

This mismatch has the potential to be a bottleneck in heavy-duty computing tasks like weather modeling, computational fluid dynamics, and certain forms of artificial intelligence.

What Is MRDIMM

Intel experts have discovered a way to break through that constraint after years of work with industry partners. They have developed a unique method that has produced the quickest system memory ever and is expected to establish a new open industry standard. The first to take use of this new memory, known as MRDIMM, for improved performance in the most plug-and-play way possible are the newly released Intel Xeon 6 data center CPUs.

According to Intel’s Xeon product manager in the Data Center and AI (DCAI) division, “a significant percentage of high-performance computing workloads are memory-bandwidth bound,” which is the kind of task that MRDIMMs are most likely to help with.

This is the tale of the DDR5 (Multiplexed Rank Dual Inline Memory Module)MRDIMM for narrative efficiency. It sounds almost too wonderful to be true.

Bringing Parallelism to System Memory, with Friends

What Is RDIMMs

As it happens, the most widely utilized memory modules for data center tasks, called RDIMMs, actually feature parallel resources, much like contemporary computers. That’s just not how they’re utilized.

A senior principle engineer in memory pathfinding in DCAI is one of the two ranks for performance and capacity in the majority of DIMMs. It’s the ideal location.

One way to conceptualize ranks is as follows:

Banks: one set of memory chips on a module would belong to one rank, while the others would belong to the other. Data may be separately stored and retrieved across many ranks using RDIMMs, but not at the same time.

The electrical load of each MRDIMM is combined by the mux buffer, enabling the interface to function more quickly than RDIMMs. Additionally, the bandwidth of memory is increased because both ranks may now be read in tandem.

This jump, which would typically require many generations of memory technologies to accomplish, results in the fastest system memory ever constructed (in this example, peak bandwidth climbs by approximately 40%, from 6,400 mega transfers per second (MT/s) to 8,800 MT/s).

Same Standard Memory Module, Just Faster

You may now be asking yourself, “Wait a minute,” whether Intel is returning to the memory market. No. Throughout its history, Intel has abandoned its different memory product businesses, some of which are extremely well-known, despite having begun as a memory firm and inventing technologies like EPROM and DRAM.

The MRDIMM‘s simplicity of usage is what makes it unique. There is no need to modify the motherboard since it uses the same connection and form factor as a standard RDIMM (even the tiny mux chips fit in previously unoccupied spaces on the module).

The reliability, availability, and serviceability (RAS) and error-correcting capabilities of RDIMMs are also present in MRDIMMs. To Vergis, data integrity is preserved regardless of how distinct requests are multiplexed over the data buffer.

All of this implies that data center clients have the option to choose MRDIMMs when placing an order for a new server, or they may subsequently remove the server from the rack and replace the RDIMMs with new MRDIMMs. Enjoy the improved performance without changing a single line of code.

Xeon 6 + MRDIMM 

A CPU that is compatible with MRDIMMs is necessary, and the Intel Xeon 6 processor with Performance-cores, code-named Granite Rapids, which was released this year, is the first one on the market.

Two identical Xeon 6 systems one with MRDIMMs and the other with RDIMMs were evaluated in recent independent testing. Up to 33% more work was done by the MRDIMM-equipped machine.

Small language models(SLMs) and conventional deep learning and recommendation system workloads are among the AI workloads that can readily operate on Xeon and benefit from the bandwidth improvement that MRDIMM provides.

MRDIMMs have been released by leading memory suppliers, and more memory manufacturers are anticipated to follow suit. With assistance from OEMs like NEC, high-performance computing facilities like the National Institute for Fusion Science and the National Institute for Quantum Science and Technology are aggressively implementing Xeon 6 with P-cores due to MRDIMMs.

Read more on Govindhtech.com