Technology

• 512GB DRAM sounds huge, but don’t hold your breath for consumer availability
• NEO’s 3D X-DRAM stacks layers sky-high, but price and practicality remain unclear
• AI and enterprise systems will get the speed, regular users probably won’t

NEO Semiconductor’s push into 3D X-DRAM memory marks an ambitious attempt to rethink DRAM design for the AI and high-performance computing era.

While the promises - stacked layers, enhanced bandwidth, and reduced power consumption - are impressive, the practicality and consumer accessibility of these technologies remain open to scrutiny.

With the company projecting that its most advanced modules could reach densities of up to 512GB, it’s hard not to ask: who is this memory really for?

Complex architectures with limited consumer impact

At the core of NEO’s approach is a vertically stacked architecture that mimics the structure of 3D NAND.

In NEO’s own words, the array is “segmented into multiple sectors by vertical slits,” with “word line layers connected through staircase structures.”

The company compares its 3D X-DRAM density to the current 0a-node planar DRAM’s 48GB and claims to reach 512GB, but the implication that such capacities will trickle down into mainstream consumer products seems tenuous at best.

The proof-of-concept chips are still in the early stages. NEO is currently developing a test version of the simpler 1T0C architecture, with the more complex, and more promising, 1T1C version planned for 2026.

The 1T1C variant utilizes IGZO transistors paired with a cylindrical high-k dielectric capacitor. It promises improved retention time, reportedly beyond 450 seconds, and supports stacking up to 128 layers.

With further refinements, including the addition of 5nm-thick spacers to reduce parasitic capacitance, NEO claims stacking could exceed 512 layers.

The 3T0C design, which incorporates dual IGZO layers, is geared toward in-memory computing and AI applications.

Still, NEO’s statements about eliminating the need for TSV and enabling up to 32K-bit bus widths raise eyebrows.

Such bandwidth sounds transformative, especially compared to the projected 2K-bit bus width of HBM4, but scaling this level of performance in real-world systems is a non-trivial task.

From a broader perspective, the DRAM market hasn’t shifted significantly in terms of cost-per-GB over the past decade. Despite some fluctuations, the downward trend slowed considerably after 2012.

One might expect the MacBook Pro, for instance, to ship with far more RAM by default today than it did a decade ago, but that hasn’t happened.

Even with some price drops - DDR3 vs. DDR5 comparisons show modest improvement - the advances haven’t been revolutionary.

Commodity pricing may fluctuate, but the overall curve has flattened. Forecasts suggest we may be near a low point before another upswing.

So while 3D X-DRAM may indeed deliver bigger, faster memory by 2026, it’s unlikely these 512GB modules will be available to consumers anytime soon.

More likely, such capacity and speed will be reserved for AI servers and enterprise systems, rather than everyday desktops or laptops.

You might also like

• These are the best mobile workstations you can buy right now
• We've also listed the best mini PCs for every budget
• Nearly 90% of businesses that switched from WordPress are content with their new CMS, survey finds

• Linus Tech Tips and Kioxia have smashed the Pi calculation world record
• 300 trillion digits of Pi were calculated using Kioxia NVMe SSD cluster
• The seven-month compute effort ended with Guinness recognition

After StorageReview previously claimed the Pi calculation world record with over 202 Trillion Digits, now Linus Media Group, the creators of the Linus Tech Tips YouTube channel, has taken it even further.

Working with Kioxia, LMG officially set a new Guinness World Record for the "Most Accurate Value of Pi", reaching an incredible 300 trillion digits.

This milestone was achieved using a high-performance storage setup featuring 2.2PB of Kioxia’s CM Series 30.72TB and CD Series 15.36TB PCIe NVMe SSDs.

(Image credit: Kioxia)Seven months and no SSD failures

The drives were organized in a NAS system connected to a dual-CPU compute server. The entire operation ran continuously for nearly seven and a half months.

“We knew breaking the Pi record with distributed network storage was going to be difficult - no one had really done it before due to the performance challenges associated with remote storage," said Jake Tivy, Writer & Host, Linus Media Group.

"Fortunately for us, the reliability and performance of Kioxia's NVMe SSDs enabled us to run continuous, intensive compute operations at speeds up to 100+ GB/s for nearly seven months straight, without a single SSD failure."

The project not only broke the previous record but did so by a wide margin. StorageReview’s 202 trillion digit milestone was huge, but it wasn’t officially verified by Guinness World Records.

The last recognized benchmark from Guinness was 62 trillion digits, so this new effort pushed that number nearly five times higher.

"Attaining a Guinness World Records title for the most accurate value of Pi is a tremendous achievement, emphasizing the courage of taking on a challenge and the power of great cooperation and teamwork," said Axel Stoermann, Vice President and CTO for Embedded Memory and SSD at Kioxia Europe.

"Kioxia America's successful collaboration with Linus Media Group enabled the demonstration of the robust capabilities of our NVMe SSDs under the most demanding of workloads. We will continue to advance the capabilities of our flash memory and SSD technology to support supercomputing applications," he added.

The Linus Tech Tips channel released a video documenting the effort which you can watch below. It also revealed the final digit of the record-breaking result. (Spoiler alert, the 300 trillionth digit of Pi is 5.)

You might also like

• These are the fastest SSDs you can buy right now
• And these are the largest HDDs and SSDs on the market at the moment
• Tech publisher breaks Pi calculation world record, hitting over 202 trillion digits

More games are coming to Xbox Cloud Gaming (Beta) soon, too.

Fast internet can be hard to come by in parts of Aurora, but our broadband experts found the fastest providers and the best values in town.

Take a trip through the realm(s) with these previews.

If you're reaching for a chatbot instead of a search engine or a trusted source, think twice before acting on the information it gives you.

I've never owned another smartphone apart from an iPhone up until this year. However, as AI makes its way onto every tech product on the planet, I needed to try Android to understand the differences between artificial intelligence in the two ecosystems.

After using a Samsung Galaxy S25 for a few weeks, I returned to my iPhone 16 Pro Max. Not because it was better, but because the ecosystem you've built your life in equates to the deciding factor when it comes to choosing between flagship smartphones.

Once back on iOS, I found myself missing one specific AI feature more than others, and without access on iPhone, I quickly defaulted back to living with an Android device.

That AI feature I'm talking about is Gemini Live, and while you could access it on iOS, the experience was dumbed down. That was until yesterday, at Google I/O 2025, when Google announced that all of Gemini Live's capabilities are rolling out on iPhone, and at no cost.

Here's why Gemini Live is the best AI tool I've ever used, and how adding all of its capabilities to iPhone means I'm ready to jump back to Apple.

What Visual Intelligence wanted to be

(Image credit: Apple)

Gemini Live already existed in the Gemini app on iOS, but it lacked two crucial elements that make the Android version that much better. Firstly, Gemini Live on iOS was unable to access your iPhone's camera, and secondly, it couldn't see what you were doing on your screen. I/O 2025 changed all that.

Now, iPhone users can give Gemini Live access to their camera and screen, allowing for new ways to interact with AI that we've not really seen on iOS before.

Gemini's camera ability is single-handedly one of, if not the, best AI tool I've used to date, and I'm thrilled iPhone users can now experience it.

What is Gemini Live's camera feature? Well, imagine a better version of what Apple wanted Visual Intelligence to be. You can simply show Gemini whatever you're looking at and ask questions without needing to describe the subject.

I've found Gemini Live's camera functionality thrives in situations like cooking. I used it last week to make Birria Tacos, and not only was I given advice every step of the way, but it was also able to see everything I was doing and help direct me towards a delicious dinner.

Not only did propping my S25 on a stand give Gemini Live the perfect angle, but because it can connect to Google apps, I could ask it to get information on a recipe directly from the content creator's video. No need to constantly touch your phone with dirty hands in the kitchen, and no need to even check a recipe anymore. Gemini Live can do it all.

An AI companion every step of the way

Screen sharing allows Gemini Live to see what's on your display at any time, allowing you to ask questions related to imagery, something you're working on, or even how to complete a puzzle in a game. It's seriously cool, similar to Apple Intelligence-powered Siri we were promised but never received back at WWDC 2024.

Gemini Live's full free rollout has just started, so we're yet to see how this functionality will work on iOS. That said, if it works half as well as it does on Android, this will be a feature I could see a lot of people falling in love with.

Gemini Live and its multiple ways of interacting with the world completely unlock AI on a smartphone, and now that iPhone users can access it too, I have no reason not to return to the Apple ecosystem.

@techradar

♬ original sound - TechRadar You might also like

• Google I/O will be a major event for Android XR glasses and headsets – here are 3 announcements I'm expecting to see
• Google I/O: 7 things we expect to see at the event
• I compared Google Gemini's image generation to ChatGPT's

• Google launches Python ADK v.1.0.0 and Java ADK v.0.1.0
• Its agent-to-agent communication protocol has also been improved
• Python SDK for A2A makes it even easier to manage inter-agent communications

Google has announced a series of updates to its AI agents in the hope that they become more streamlined as efficient colleagues, rather than tools that colleagues use.

"We envision a future where intelligent agents are not just tools, but collaborative partners in solving complex challenges, streamlining workflows, and unlocking new possibilities," Product Manager Polong Lin and Developer Advocate for Cloud AI Holt Skinner said in a blog post.

The latest updates focus on enhanced agent management and improved agent-to-agent communication – a sign of AI becoming even more autonomous.

Google wants to make AI agents more autonomous

The tech giant has enhanced its Agent Development Kit (ADK) to include a new production-ready and stable v.1.0.0 release of its Python ADK, which it says is already used by Renault Group, Box and Revionics. To coincide with this stable release, Google has also confirmed its first release of the Java ADK v0.1.0.

In terms of management, acknowledging that the Vertex AI Agent Engine already helps developers deploy, manage and scale agents in production, Google has also launched its Agent Engine UI for better agent lifecycle management.

Key features of the new Google Cloud console UI include a dashboard for managing deployed agents, metrics such as requests and CPU usage, session management and tracing, and deployment details with debugging tools.

Further improvements to the Agent2Agent (A2A) protocol have also been confirmed, including v0.2 which brings stateless interaction support for lightweight communications and standardized authentication using OpenAPI-like schema. A Python SDK for A2A has also been released to make it easier for developers to integration agent-to-agent communication within Python-based agents.

Keen to show off the power of A2A, Google boasted of support from Auth0, Box AI, Microsoft, SAP and Zoom.

"These advancements in our ADK, Agent Engine, and A2A protocol are designed to provide you with a comprehensive and flexible platform to bring your most ambitious agent-driven projects to life," Lin and Skinner commented.

You might also like

• We've listed the best AI tools and best AI writers
• Check out the best productivity tools for a handy boost at work
• Salesforce gives AI agents the power to be proactive and autonomous like never before

It's an all-Premier League showdown in Bilbao.

• 'Motors' allowed threat actors to take over admin accounts
• This enabled full website takeover
• The developers released a fix

Motors, a premium theme for WordPress, was carrying a critical-severity vulnerability that allowed malicious actors to fully take over compromised websites.

The privilege escalation flaw, due to the theme improperly validating user identities before updating passwords, is now tracked as CVE-2025-4322, and has a severity score of 9.8/10 (critical).

Security researchers Wordfence, who first spotted this bug, explained how threat actors could use it to “change arbitrary user passwords, including those of administrators, and leverage that to gain access to their account."

TechRadar Pro readers can get 60% off Premium Plans at RoboForm now!

New users can take advantage of RoboForm’s exclusive deal and get 60% off the Premium Plan. With this deal, you can get unlimited password storage, one-click login & autofill, password sharing, two-factor authentication for added protection, cloud backup, and emergency access for trusted contacts. To claim this deal, visit this link and sign up for the Premium Plan to lock in this huge discount.

Preferred partner (What does this mean?)View Deal

Premium themes

Obviously, having access to an admin account grants the malicious actors all kinds of privileges, including complete website takeover. All versions up to 5.6.68 are affected. The update that addresses the flaw was released on May 14, 2025. Since themes are not as simple to suspend, or swap, as plugins, users are advised to update their Motors as soon as possible.

Motors is a car dealer WordPress theme, designed for auto dealers, classified listing, auto rental, boats, repair services, and motorcycle dealers. It is developed by a company called StylemixThemes and, according to BleepingComputer, is one of the top-selling themes of its kind. On the Envato market, it is selling for $79 and has been sold more than 22,300 times.

WordPress is the world’s number one website builder platform, powering more than half of all websites on the internet. This also makes it a major target for cybercriminals but, since it’s mostly secure, hackers are looking for exploits in themes and add-ons, which are used as stepping stones for further compromise.

For example, in early March this year, news broke that malicious JavaScript code was deployed into more than 1,000 WordPress websites, following compromised extras. Users are advised to only keep the add-ons they are actually using, and to keep them updated at all times.

Via BleepingComputer

You might also like

• Thousands of WordPress sites targeted with malicious plugin backdoor attacks
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

There's a new airline safety rule for everyone's favorite travel tech because of the risk of fire.

• Wednesday season 2’s latest sneak peek teases more family-based lore
• This news comes after a big casting announcement, confirming additions including Joanna Lumley and Steve Buscemi
• The series arrives on Netflix in two parts in August and September

The first season of Wednesday was a huge success for Netflix, and as it returns for a second outing, the creators have teased that we should expect a lot more Addams Family lore.

Recently, I raved about the Netflix show's confirmed all-star cast, many of whom will be joining the family. Season one put a lot of focus on Wednesday Addams' time at Nevermore Academy, and while we will be returning there, we're also going to see more of Wednesday’s family and the dynamics between its various oddball members.

This is excellent news if, like me, you’ve been dying to see more from Morticia and Gomez especially, as their relationship is one of the most iconic on-screen partnerships out there.

Take a look at Netflix's latest teaser below for a fun, behind-the-scenes look at what's coming up in Wednesday season 2.

What do we know about Wednesday season 2?

Wednesday season 2 is the highly anticipated follow-up to one of the best Netflix shows, and the second installment will be split into two parts – an approach Netflix has taken with other shows recently, including Cobra Kai season 6, which was split into three parts.

The first four episodes will be released on August 6, followed by part two on September 3, when the remaining four will premier. That means there isn't too much of a wait between each edition, thankfully, but this won’t be a series you can binge-watch in a whole weekend unless you wait patiently until September.

Plenty of familiar faces will be returning for Wednesday season 2 including Tim Burton as creator, Jenna Ortega as the titular protagonist, and Catherine Zeta-Jones as Morticia.

Big-name newcomers include Steve Buscemi as the new Nevermore Academy principal, and Joanna Lumley as Wednesday's grandmother. The cast really does look excellent this season, and we should be in for a wild ride.

You might also like

• Netflix won't work on some Amazon Fire TV devices from June – here are the affected models
• 5 Netflix thriller movies with 100% on Rotten Tomatoes to watch right now
• Netflix is picking up Sesame Street and doing something HBO should've done all along

• Fujifilm X-E5 rumored to be launching in June
• News comes four years after the release of the X-E4
• X-E4 received mixed reviews for its minimalist approach

With retro looks and tactile controls, Fujifilm’s early X-E cameras gained an enthusiastic following. But when the X-E4 launched in 2021, it split opinion. Praised for its compact style, fans of the series felt that Fujifilm had stripped away too many features. Now, fresh rumors suggest its successor is set to land next month. Many will be hoping it rights those minimalist wrongs.

According to Fuji Rumors – a reliable industry source – the Fujifilm X-E5 is due to make its debut in June. While Fujifilm hasn’t officially confirmed the model’s existence, news of a summer launch doesn’t come as a huge surprise. Talk of the new model’s imminent arrival has been circulating since late last year.

The big question is what approach Fujifilm will take with the X-E5. Back in December, we touted it as one of the most exciting cameras of 2025. Since then, we’ve learnt nothing substantive about its specs. There have been no major online leaks, which is unusual for a camera that’s deep into development.

In the absence of anything concrete, we can only speculate about what the X-E5 will look like – and what we want to see from Fujifilm’s comeback camera.

What the X-E5 needs to get right

On paper, the X-E4 had the makings of a winner: it inherited the 26MP X-Trans sensor and X-Processor 4 from Fujifilm’s more expensive X-mount cameras. It also retained the compact proportions and attractive rangefinder styling of previous versions. But the physical execution proved divisive. In pursuit of minimalism, Fujifilm removed a lot of the physical controls found on the X-E3.

Gone were the rear control dial, AF lever and flash, plus a few buttons. Fujifilm ditched the grip, too. With no in-body image stabilization or weather-sealing either, many felt that the X-E4 prioritized style over substance. Critics pointed to the lack of dual SD card slots as evidence that the X-E4 wasn’t a tool for serious photographers. All of which was harder to justify with a body-only launch price of $850 / £799 / AU$1,399.

When the X-T30 II launched in November 2021, it looked like much better value. In a telling indictment of the X-E4’s shortcomings, Fujifilm stopped X-E4 production after just a couple of years. Ironically, the resulting rarity of the model actually drove its price up.

Against that background, the X-E5 has two jobs to do: win back favor among Fujifilm enthusiasts and re-establish the position of X-E series in the maker’s mirrorless line-up.

The former doesn’t have to be difficult. Looking at online comments, the formula for success needs to include the reintroduction of certain physical controls, including the rear dial and a built-in grip. Other common wishlist items are a high-res EVF, a bigger battery and a 40MP APS-C sensor. In-body image stabilization would really sweeten the deal.

That spec sheet would signal a clear shift back towards the enthusiast roots of the X-E series. In reality, it’s unlikely that we’ll see all of those features on the new camera, but it’s not a fools bet to suggest that Fujifilm will pitch the X-E5 as a proper photographer’s camera. Especially with the X-M5 now catering for videographers.

If Fujiiflm can inject the X-E5 with enough substance while keeping its analog charisma intact, it could be one of the sleeper hits of the year. Then it simply needs to make enough units to meet the inevitable demand. Rest assured, we’ll be keeping a close eye on Fujifilm announcements and rumors over the next few weeks.

You might also like...

• The rumored Fujifilm X-E5 is the camera I'm most excited about in 2025 – here's why
• Camera rumors for 2025: new gear we're expecting soon from DJI, Sony, Canon and more
• Best Fujifilm camera 2025: top mirrorless and compact cameras

Microsoft filed a brief in support of Epic Games and says Apple is in the way of it offering a full Xbox storefront on iOS.

Subscribers can access classic Activision games -- like Commando and Pitfall! -- on all four tiers.

• Lies of P is set to receive a huge free update
• It features two new difficulty modes in addition to the current standard
• Battle Memories and Death March modes will let players replay boss fights

If you thought Lies of P: Overture was the only piece of substantial content coming to one of the best soulslike games this year... well, you thought wrong. That's because Lies of P is also set to receive a free update that will be made available to all players - irrespective of whether you're buying the Overture expansion or not.

The update is bringing two new difficulty modes to the game. These are Butterfly's Guidance and Awakened Puppet, both of which provide a more story-focused, easier difficulty for those not used to the intensity of soulslikes. The default difficulty that's in the game right now is being renamed 'Legendary Stalker.'

There's no word yet on exactly what effects these difficulty modes will have on Lies of P. But it's easy to imagine reduced damage scaling for enemies and perhaps a decrease in enemy density depending on which difficulty you select.

The update is also bringing a dedicated boss rush mode to the game. Named Battle Memories, you'll be able to access this mode at the Hotel Krat Stargazer after clearing the game at least once.

Battle Memories will let you challenge previously defeated bosses from both the base game and the Overture DLC (downloadable content). Each boss has five difficulty tiers, with the harder Tiers 4 and 5 locked until you've cleared the previous ones (you'll get access to Tier 4 after beating Tier 3, for example).

Publisher Neowiz has also teased additional unlockables for those willing to brave the harder tiers: "Who knows what rewards await those who overcome these hardening challenges?"

But that's not all, as Battle Memories is being paired with another mode called Death March. This mode challenges players to defeat a minimum of three bosses consecutively with a limited pool of items and healing Pulse Cells. You're also able to set custom scenarios here, selecting the bosses you wish to face. The difficulty tiers of Battle Memories can also be applied here.

It's sounding like a very substantial update to go alongside the upcoming Overture expansion. With multiple difficulty options, boss rush modes, and the potential for unique rewards, there's really something for everyone here.

Finally, this Lies of P update will arrive the same day that the Overture DLC releases, sometime in the Summer of 2025.

You might also like...

• PNY's microSD Express cards for Nintendo Switch 2 are available to buy now in 128GB and 256GB options
• After five long years, Fortnite has finally returned to the iOS App Store - but it's not available everywhere yet
• The newly revealed Logitech G522 Lightspeed could become the new mid-range wireless gaming headset to beat

• Adobe is increasing the price of its Creative Cloud All Apps plan
• The new plan will include AI tools and credits
• The plan’s name is also changing to Creative Cloud Pro

Adobe’s Creative Cloud suite contains some fantastic apps for creatives, with Photoshop, Illustrator and more among its contents. But if you currently subscribe to Creative Cloud’s top tier, you might want to look away now, as Adobe says its price is about to jump to a pricey $770 a year.

As it is, Adobe’s All Apps plan for Creative Cloud will send you back $659.88 if you pay annually, meaning the changed pricing will cost you over $140 more than before. And that’s not the only change to the All Apps plan’s pricing – if you pay monthly, you’ll be asked for fork out $104.99 instead of the current $89.99 fee.

Corporate per-seat plans are going up to $99.99 per month (up from $89.99), while student and teacher pricing will increase from $34.99 to $39.99. And Adobe is also changing the name of the Creative Cloud All Apps plan to Creative Cloud Pro.

At a time of ongoing economic uncertainty, this will be an unwelcome change indeed for many of Adobe’s users. You’ll need to work out if the price rises are worth paying for if you should start looking for an alternative.

What else is changing?

(Image credit: Future)

In return for the increased cost, Adobe is putting a big focus on artificial intelligence (AI) tools. The company says that customers will get unlimited access to AI features like Photoshop’s Generative Fill and Generative Remove in Lightroom, as well as unlimited access to Firefly Boards, which are used for planning and brainstorming.

Elsewhere, users will be able to use the new Adobe Firefly generative AI app and integrate their own AI models into it. Adobe is offering 4,000 monthly credits for premium AI video, audio and image generation, in addition to the unlimited credits for standard tools.

If you’re not feeling enthusiastic about these new AI features – and the price tags that come with them – you don’t need to stay with Adobe. There are plenty of excellent replacements out there, from the best Photoshop alternatives to the best InDesign alternatives.

And if you do want to stick around, Adobe says the changes will come into effect on your first renewal after June 17, 2025.

You might also like

• Best Adobe Photoshop alternative of 2025: Avoid Creative Cloud subscriptions with these top apps
• The 5 biggest new Photoshop, Firefly and Premiere Pro tools that were announced at Adobe Max London 2025
• What is Firefly: everything you need to know about Adobe’s safe AI image generator

• Peter Green Chilled told its clients it suffered a ransomware attack
• It temporarily stopped delivering goods
• Markets are feeling the sting of the attack

Peter Green Chilled, a UK logistics company that distributes chilled and frozen food to UK supermarkets, suffered a ransomware attack recently that caused serious problems throughout the supply chain.

According to multiple media reports, the company mailed its customers on May 15, to notify them of the cybersecurity incident that occurred the day before. The BBC, citing the company’s managing director, Tom Binks, said the transport activities were operational, but new orders were not being processed.

No further details about the attack itself were given. Therefore, we don’t know who the attackers are, how they managed to infiltrate Peter Green Chilled’s IT infrastructure, or how much they’re demanding in ransom.

Save up to 68% on identity theft protection for Techradar readers

TechRadar editors praise Aura's upfront pricing and simplicity. Aura also includes a password manager, VPN, and antivirus to make its security solution an even more compelling deal.

Preferred partner (What does this mean?)View Deal

How the attack affects supermarkets

We also don’t know if the attackers stole any sensitive files, as is usual in ransomware attacks. The company is not responding to media inquiries at the moment, it seems.

Peter Green Chilled is not the largest organization in its industry, but it still plays an important role, as it supplies major UK supermarkets, including Tesco, Sainsbury’s, and Aldi. It also services Co-op and M&S, who are currently addressing cyberattacks of their own.

The attack at Peter Green Chilled sent ripples throughout the industry. Speaking at a morning radio show, the founder of The Black Farmer food brand explained how Peter Green not delivering the goods will probably cost the business around $133,000 a week. And that’s just one business, for one week’s work.

Cybercriminals love targeting critical infrastructure providers, since the pressure is enormous and forces the organizations to pay the ransom demand to keep the business going. For Jamie Moles, Senior Technical Manager at NDR provider ExtraHop, cybersecurity in retail and logistics needs to be treated with the same severity as it’s being treated in critical infrastructure.

“Cybersecurity in retail and supply chain logistics must be treated with the same severity as critical infrastructure,” he told TechRadar Pro in an emailed statement. “Protecting digital systems is no longer optional, and modernising how organisations can see into their networks will improve resilience against threats like ransomware, ensuring continuity and trust in the systems we all rely on every day."

Via The Register

You might also like

• Food retail giant behind several major US supermarket brands confirms data stolen in major ransomware breach
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

Sergey Brin, the man most responsible for Google Glass, is out of retirement and back at the company he cofounded, helping make AI happen. He's also touting a new kind of "Glass" or, more accurately, Android XR smart glasses while admitting to some big Google Glass missteps.

Brin unexpectedly took the stage at Google I/O 2025 for a sitdown with Big Technology Podcast’s Alex Kantrowitz and DeepMind's CEO Demis Hassabis on Tuesday. While the discussion mostly revolved around Gemini, Google's generative AI platform, Kantrowitz asked Brin what he learned from the Google Glass experience and how he might apply that to the modern Android XR Glasses project.

With bracing candor, Brin told Kantrowitz, "I definitely feel like I made a lot of mistakes with Google Glass, I'll be honest."

It's quite an admission from Brin, who more than a decade ago was Google Glass' biggest champion and memorably hosted a launch featuring wing-suit skydivers jumping from a plane while wearing Google Glass. It was a spectacular moment, but even there, Brin now sees a flaw and how he might avoid a similar misstep.

Talking about how much he missed that big launch moment, Brin turned his attention to the new XR glasses' own launch. "Maybe...we should probably polish the product this time, when it's ready and available, and then we'll do a really cool demo. So, that's probably a smart move."

Brin may see the blemishes and ultimate failure of Google Glass, the first wearable smart glasses that let you use gestures, head tilts, and even blinks to take a picture, but there was a moment between 2012 to 2014, when Google Glass was everywhere.

I wore my pair to a fashion show and CES, and on numerous network TV shows. Brin was spotted on the subway wearing his. They were, for a time, a cultural phenomenon, but also ridiculous, almost as quickly spawing the meme "Glassholes".

Would be cool if this guy showed up again at #GoogleIO. This is from a decade ago (at a different event) when Sergey showed me how to take a selfie with Google Glass pic.twitter.com/y0zlrMyR5cMay 14, 2024

Still, I appreciated Brin's enthusiasm, and when I ran into him more than a decade ago, while wearing Google Glass, naturally, he graciously took the time to show me how you could take a selfie with Google Glass (yes, it involved removing them from your face).

Glass suffered early on from availability and pricing issues ($1,500 / £1,000, or around AU$2,000) and Brin points to his naivete as the cause: "I just didn't know anything about consumer electronics supply chains really, and how hard it would be to build that and have it be at a reasonable price point....This time, we have great partners that are helping us build this."

The point here is that instead of Google trying to figure any of this out, it now has companies like Samsung, Gentle Monster, and Warby Parker, all experts in building consumer products, supply chains, and retail pricing, building the Android XR Glasses for them.

AI will make the clear difference

Google Glass faded into ignominy, but Brin sees Google's return to smart glasses in a fresh light, and it's mostly because of his new pet project at Google: AI and Gemini.

Aside from the form factor, which did not look like normal glasses, Google Glass might have been ahead of its time. They were "smart" without really having any true smarts of their own.

"Now, in the AI World," said Brin, "the things these glasses can do to help you out without constantly distracting you, that capability is much higher." And it potentially makes interacting with Android XR Glasses much more natural. It's also why Google and its partners are putting Gemini at the center of these Android XR wearables.

A Google presenter wearing Google Android XR Glasses at Google I/O 2025. (Image credit: Future)

Finally, Hassabis, who'd been watching the whole exchange, chimed in, "I feel like the universal assistant is the killer app for smart glasses, and I think that's what's gonna make it work."

This all makes sense; a powerful AI sitting resident on your face in a familiar form factor that asks for nothing more than voice commands to do your bidding, but can also watch and act on your behalf. That's the future of smart glasses, and one that Google Glass aspired to but never achieved.

Android XR Glasses will be Brin's second smart glasses act and, perhaps, the one he'll ultimately be best remembered for.

You might also like

• Google I/O 2025 as it happened: AI Search, Veo, Flow ...
• I finally tried Samsung's Project Moohan Android XR ...
• Google's Veo 3 marks the end of AI video's 'silent era'
• Google CEO: AI is not a 'zero-sum moment' for search

• Senua's Saga: Hellblade 2 officially confirmed to be coming to PS5 this summer
• The game was originally released as an Xbox Series X|S and PC exclusive
• PS5 Pro enhancements confirmed, along with other new features

Developer Ninja Theory has today, on the game's first anniversary, confirmed that Senua's Saga: Hellblade 2 will be coming to PlayStation 5 "this summer".

In an official announcement in a post on the developer's website, as well as a video that you can see below, Studio Head Dom Matthews confirmed the news and will also include some "new features" that will also be coming to PC and Xbox versions of the game via an update. No firm date has been given for the release right now, however.

The post goes on to say that Ninja Theory has been "working hard to fully optimise Hellblade 2 for PS5 and PS5 Pro, to give you the very best experience we can," which I'm hoping means some excellent enhanced visuals or graphics features - especially for PS5 Pro and to make use of PSSR. However, no specifics are mentioned, and it looks like we'll have to wait to hear more.

I really enjoyed Hellblade 2 and was longing for a reason to replay it, so this is music to my ears. The game is rather bleak and almost miserable, but it's technically brilliant, has an intensely gripping story, and some of the most richly atmospheric settings and locations I've experienced that make it absolutely mesmerizing and something that has long lived in my memory since I finished it for my review.

Ninja Theory games have long been on PlayStation, and it was somewhat strange to see the studio release a game and not have it launch on PlayStation, too. This is down to the studio being acquired by Xbox Game Studios back in 2018 and thus releasing the game exclusively on PC, Xbox Series X and Series S, and Xbox Game Pass last year when it launched on May 21, 2024.

But now, Senua's Saga: Hellblade 2 can join the ranks of Ninja Theory PlayStation games that include its predecessor, as well as Heavenly Sword, Enslaved: Odyssey to the West, and DmC: Devil May Cry when it arrives this summer.

You might also like...

• After five long years, Fortnite has finally returned to the iOS App Store - but it's not available everywhere yet
• The newly revealed Logitech G522 Lightspeed could become the new mid-range wireless gaming headset to beat
• PNY's microSD Express cards for Nintendo Switch 2 are available to buy now in 128GB and 256GB options