Connect with us

Tech

What are the storage requirements for AI training and inference? | Computer Weekly

Published

on

What are the storage requirements for AI training and inference? | Computer Weekly


Despite ongoing speculation around an investment bubble that may be set to burst, artificial intelligence (AI) technology is here to stay. And while an over-inflated market may exist at the level of the suppliers, AI is well-developed and has a firm foothold among organisations of all sizes.

But AI workloads place specific demands on IT infrastructure and on storage in particular. Data volumes can start big and then balloon, in particular during training phases as data is vectorised and checkpoints are created. Meanwhile, data must be curated, gathered and managed throughout its lifecycle.

In this article, we look at the key characteristics of AI workloads, the particular demands of training and inference on storage I/O, throughput and capacity, whether to choose object or file storage, and the storage requirements of agentic AI.

What are the key characteristics of AI workloads?

AI workloads can be broadly categorised into two key stages – training and inference.

During training, processing focuses on what is effectively pattern recognition. Large volumes of data are examined by an algorithm – likely part of a deep learning framework like TensorFlow or PyTorch – that aims to recognise features within the data.

This could be visual elements in an image or particular words or patterns of words within documents. These features, which might fall under the broad categories of “a cat” or “litigation”, for example, are given values and stored in a vector database.

The assigned values provide for further detail. So, for example “a tortoiseshell cat”, would comprise discrete values for “cat” and “tortoiseshell”, that make up the whole concept and allow comparison and calculation between images.

Once the AI system is trained on its data, it can then be used for inference – literally, to infer a result from production data that can be put to use for the organisation.

So, for example, we may have an animal tracking camera and we want it to alert us when a tortoiseshell cat crosses our garden. To do that it would infer the presence or not of a cat and whether it is tortoiseshell by reference to the dataset built during the training described above.

But, while AI processing falls into these two broad categories, it is not necessarily so clear cut in real life. It will always be the case that training will be done on an initial dataset. But after that it is likely that while inference is an ongoing process, training also becomes perpetual as new data is ingested and new inference results from it.

So, to labour the example, our cat-garden-camera system may record new cats of unknown types and begin to categorise their features and add them to the model.

What are the key impacts on data storage of AI processing?

At the heart of AI hardware are specialised chips called graphics processing units (GPUs). These do the grunt processing work of training and are incredibly powerful, costly and often difficult to procure. For these reasons their utilisation rates are a major operational IT consideration – storage must be able to handle their I/O demands so they are optimally used.

Therefore, data storage that feeds GPUs during training must be fast, so it’s almost certainly going to be built with flash storage arrays.

Another key consideration is capacity. That’s because AI datasets can start big and get much bigger. As datasets undergo training, the conversion of raw information into vector data can see data volumes expand by up to 10 times.

Also, during training, checkpointing is carried out at regular intervals, often after every “epoch” or pass through the training data, or after changes are made to parameters.

Checkpoints are similar to snapshots, and allow training to be rolled back to a point in time if something goes wrong so that existing processing does not go to waste. Checkpointing can add significant data volume to storage requirements.

So, sufficient storage capacity must be available, and will often need to scale rapidly.

What are the key impacts of AI processing on I/O and capacity in data storage?

The I/O demands of AI processing on storage are huge. It is often the case that model data in use will just not fit into a single GPU memory and so is parallelised across many of them.

Also, AI workloads and I/O differ significantly between training and inference. As we’ve seen, the massive parallel processing involved in training requires low latency and high throughput.

While low latency is a universal requirement during training, throughput demands may differ depending on the deep learning framework used. PyTorch, for example, stores model data as a large number of small files while TensorFlow uses a smaller number of large model files.

The model used can also impact capacity requirements. TensorFlow checkpointing tends towards larger file sizes, plus dependent data states and metadata, while PyTorch checkpointing can be more lightweight. TensorFlow deployments tend to have a larger storage footprint generally.

If the model is parallelised across numerous GPUs this has an effect on checkpoint writes and restores that mean storage I/O must be up to the job.

Does AI processing prefer file or object storage?

While AI infrastructure isn’t necessarily tied to one or other storage access method, object storage has a lot going for it.

Most enterprise data is unstructured data and exists at scale, and it is often what AI has to work with. Object storage is supremely well suited to unstructured data because of its ability to scale. It also comes with rich metadata capabilities that can help data discovery and classification before AI processing begins in earnest.

File storage stores data in a tree-like hierarchy of files and folders. That can become unwieldy to access at scale. Object storage, by contrast, stores data in a “flat” structure, by unique identifier, with rich metadata. It can mimic file and folder-like structures by addition of metadata labels, which many will be familiar with in cloud-based systems such as Google Drive, Microsoft OneDrive and so on.

Object storage can, however, be slow to access and lacks file-locking capability, though this is likely to be of less concern for AI workloads.

What impact will agentic AI have on storage infrastructure?

Agentic AI uses autonomous AI agents that can carry out specific tasks without human oversight. They are tasked with autonomous decision-making within specific, predetermined boundaries.

Examples would include the use of agents in IT security to scan for threats and take action without human involvement, to spot and initiate actions in a supply chain, or in a call centre to analyse customer sentiment, review order history and respond to customer needs.

Agentic AI is largely an inference phase phenomenon so compute infrastructure will not need to be up to training-type workloads. Having said that, agentic AI agents will potentially access multiple data sources across on-premises systems and the cloud. That will cover the range of potential types of storage in terms of performance.

But, to work at its best, agentic AI will need high-performance, enterprise-class storage that can handle a wide variety of data types with low latency and with the ability to scale rapidly. That’s not to say datasets in less performant storage cannot form part of the agentic infrastructure. But if you want your agents to work at their best you’ll need to provide the best storage you can.



Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Tech

A Humanoid Robot Set a Half-Marathon Record in China

Published

on

A Humanoid Robot Set a Half-Marathon Record in China


Over the weekend in China, a humanoid robot shattered world half-marathon record—the human record—by seven minutes.

The star performer was a robot developed by the Chinese company Honor (the smartphone maker), which finished the 13.1-mile race in 50 minutes, 26 seconds. The human record, set by Ugandan Olympic medalist Jacob Kiplimo, is 57 minutes, 20 seconds. The result marks an impressive milestone especially considering that, just a year earlier, the fastest robot at this half-marathon event took two and a half hours to complete the same distance.

But Honor’s robot was not the only participant. The event consisted of more than 100 humanoid robots from 76 institutions across China. The robots lined up alongside 12,000 human runners in Beijing’s E-Town, albeit on separate courses to avoid accidents. The contrast in performance between humans and robots was more than evident.

Run, Robot, Run

A humanoid robot is designed to mimic the structure and movement of the human body, with legs, arms, and sensors that allow it to interact with its environment. In this case, the winning robot incorporated features inspired by elite runners: long legs (almost a meter), advanced balance systems, and a liquid cooling mechanism, similar to that of smartphones, to prevent overheating during the race.

In addition, many of the participating robots operated autonomously, meaning without direct human control. Thanks to artificial intelligence algorithms, they could adjust their pace, maintain balance, and adapt to the terrain in real time. Notably, the Honor robot that achieved the 50-minute mark operated autonomously. The Chinese manufacturer presented another robot, operated by remote control, that ran the same stretch in even less time: 48 minutes, 19 seconds.

As expected, there were some accidents in the race. Some robots fell down, others veered off the path, and several needed technical assistance along the way. While the physical performance of humanoid robots has advanced rapidly, their reliability is still developing. Of course, the laughter and jeers are no longer as frequent as they used to be, replaced by applause and exclamations of surprise.

The winning robot, “Blitz,” from smartphone manufacturer Honor was on display at the awards ceremony after the Beijing E-Town Robot Half Marathon.

Photograph: Lintao Zhang/Getty Images

Robot Superiority

Just like the robots that went viral for their impressive martial arts display a few weeks ago, this long-distance race is part of a broader strategy by China to show off its leadership in the development of advanced robots.

You don’t need to be a robotics expert to see that this achievement demonstrates that machines can outperform humans at specific physical tasks under controlled conditions. (It’s hard to imagine that the winning robot could achieve the same result, for example, if it started to rain during the race.) But humans still have a few tricks up their sleeve: Running in a straight line is very different from performing complex real-world activities, such as manipulating delicate objects or interacting socially.

However, it’s understandable that the image of a robot crossing the finish line in record time, ahead of human athletes, raises several questions. Is this the beginning of a new era in which machines redefine physical limits?

One could argue that a car is a machine, and those have always been faster than humans. But a humanoid robot is designed to mimic humans. It’s more alarming to see one beat humanity at its own game—even if so many of them are still tripping over themselves.

This story originally appeared in WIRED en Español and has been translated from Spanish.





Source link

Continue Reading

Tech

War Memes Are Turning Conflict Into Content

Published

on

War Memes Are Turning Conflict Into Content


As ceasefire announcements between the US and Iran—and separately between Israel and Lebanon—dominated headlines over the past two weeks, they also prompted a look back at how war spread online: through memes.

There were jokes about conscription. Captions about getting drafted, but at least with a Bluetooth device. The song “Bazooka” went viral, with users lip-syncing to: “Rest in peace my granny, she got hit by a bazooka.” Military filters followed. So did posts about Americans wanting to be sent to Dubai “to save all the IG models.”

Across the Gulf, the tone was different but the instinct was the same. Memes joked that Iran was replying to Israel faster than the person you’re thinking about. Delivery drivers were shown “dodging missiles.” “Eid fits” became hazmat suits and tactical vests.

Dark humor is one of the oldest responses to fear, a way of reclaiming control, however briefly, over events that offer none. Variations of that idea appear across psychology and philosophy, including Freud’s relief theory, which frames humor as a release of tension.

But social media changes the scale and speed of that instinct.

A joke once shared within a small community can become a global template in minutes. Algorithms do not reward depth or accuracy; they reward engagement. The memes that travel fastest are usually stripped of context, easy to recognize and simple to remix.

Middle East scholar and media analyst Adel Iskandar traces political satire back centuries, from banned satirical papyri in ancient Egypt to cartoons during revolutions and gallows humor in modern wars. “Where there is hardship, there is satire,” he says. “Where there is loss of hope, there is hope in comedy.”

That tradition still exists online. But today it is fused with recommendation systems designed to keep attention moving.

Memes Spread Faster Than Facts

The word “meme” was coined by Richard Dawkins in his 1976 book The Selfish Gene, where he described how ideas replicate like genes. On today’s internet, replication follows platform logic.

Fitness means generality. A meme does not need to be accurate. It needs to feel familiar. It needs the right format, paired with trending audio and the right emotional shorthand.

“A meme is like a virus,” Iskandar says. “If it doesn’t travel, it’ll die.”

The most visible response online is not always the truest one. It is often just the easiest to spread. And once context disappears, one crisis can start to resemble any other.

Geography shapes humor too, and adds another level of tension. “If you live far away from the threat, you’re capable of producing content that ridicules it with an element of safety,” says Iskandar. “Whereas if you happen to be within close proximity, it is more of a fatalism.”

That divide matters. For some users, war exists mainly as mediated spectacle: clips, edits, graphics, headlines, and reaction posts. For others, it is sirens, uncertainty, disrupted flights, rising prices, and messages checking who is safe.

The same meme can function as entertainment in one country and emotional survival in another. Take the American experience of violence, which Sut Jhally, professor of communication at the University of Massachusetts Amherst, says “is very mediated.”

What much of the Western world has consumed instead is what cultural critic George Gerbner called “happy violence”: spectacular, consequence-free, and detached from the aftermath.

Jhally argues that the September 11 attacks remain the defining modern American experience of war-adjacent political violence. Much else has been cinematic: distant invasions, blockbuster destruction, video-game logic, apocalypse franchises.

The teenager from the Midwest joking about being drafted is drawing from zombie films and superhero apocalypses. “There is almost no discussion about what an actual Third World War would look like,” he says. “People do not have a perception of what that really looks like.”





Source link

Continue Reading

Tech

Hyundai’s New Ioniq 3 Has Hot-Hatch Looks, but Can It Beat BYD?

Published

on

Hyundai’s New Ioniq 3 Has Hot-Hatch Looks, but Can It Beat BYD?


Hyundai has unveiled its Ioniq 3, a fully electric compact hatchback for urban driving designed to be as aerodynamically efficient as possible yet still offer up a surprisingly spacious interior—a trick the carmaker is loftily calling Aero Hatch. The 3 is intended to fill the gap between Hyundai’s Inster supermini and Ioniq 5 crossover.

In profile, the Ioniq 3 has a sleek front end that transitions into a roofline that stays straight over both front and rear occupants before dropping to merge with the rear spoiler. It’s this roofline that maximizes interior headroom for the rear passengers, but it also offers a supposed class-leading drag coefficient of 0.263.

The Ioniq 3’s impressive aerodynamics will supposedly help it get more than 300 miles on a single charge.

Photograph: Courtesy of Hyundai

The car has the same underpinnings as its sibling brand, Kia’s EV2. Two battery options will deliver a projected WLTP distance of 344 km (around 214 miles) for the Standard Range Ioniq 3; the Long Range version is supposedly good for a competitive 308-mile range. Built on the group’s Electric-Global Modular Platform (E-GMP), the car has a 400-volt architecture to lower costs rather than the 800-volt system of the Ioniq 5 N, 6, or 9 SUV. Still, this means that if you can find sufficiently fast DC charging, you can, in theory, top up from 10 to 80 percent in approximately 29 minutes (AC charging capability is up to 22 kW).

This is fine, but it is not a match for BYD’s new Blade 2.0 battery tech that WIRED tried, astonishingly allowing the Denza Z9 GT to charge its battery in just over nine minutes from 10 percent. True, that battery tech was in a $100,000 “premium” EV, but it’s coming to BYD’s wider models. And if BYD makes good on its plans to deliver a charging network to rival Tesla’s Supercharger, then very soon buyers will be expecting comparable charge times, and 30 minutes will quickly feel awfully long.

I asked José Muñoz, Hyundai Motor Company president and CEO, whether this new battery technology from BYD concerns him, whether Hyundai—leading the EV pack with 800-volt architectures for so long—needs to match the Blade 2.0’s performance. “We welcome the challenge,” Muñoz tells me. “Every challenge is an opportunity to do better. And I can tell you that, lately, we have a lot of opportunities to do better.”

“We are also working on fast charging,” Muñoz says, adding that Hyundai’s success will be built on not merely one leading technology but many. “There are not more elements that may be offered by the Chinese that we can offer. It’s only a matter of how you mix them. A lot of times, you get stuck into one indicator. I’m an engineer. And we always have the example of the airplanes: What is more important in an airplane, altitude or speed? There is only one answer. You need to achieve both.”



Source link

Continue Reading

Trending