Technology

The Turtle Beach Stealth Ultra controller is a solid pro-level controller, but it makes some big compromises.

• Some Pixel 6a units will soon have much worse battery performance
• Affected users can claim a free battery replacement, or may be able to get cash
• Other Pixel models could be next in line

Usually, software updates bring desirable new features or bug fixes, but one upcoming Google Pixel 6a update could actually make your phone worse.

Google has announced that it will be rolling out a mandatory update to some Pixel 6a models, which “will reduce battery capacity and charging performance” after the phone reaches 400 charging cycles. So if your phone isn’t yet at 400 charging cycles then these changes won’t kick in yet, even once you download the update.

There are good reasons for these changes though, with Google stating that this will “reduce the risk of potential battery overheating which could pose a risk to users” – and we know how that can play out, as there have been multiple reports of Pixel 6a units catching fire.

Still, a battery that charges slower and needs charging more often than before isn’t exactly going to be desirable to users.

The Google Pixel 6a (Image credit: Future)

This update will start rolling out on July 8 as part of Android 16 – so at least you’ll get some Android 16 goodies to go with your battery downgrade.

Interestingly though, only some Pixel 6a units will get these new battery management ‘features’. It’s unclear why that is – perhaps Google used different batteries in some versions of the phone, but to see whether yours is affected, you can head to this support page.

If it is, then you will have some options other than just putting up with a battery downgrade.

For one thing, if you’re in the UK, US, or select other locations then Google will let you get a battery replacement at no charge. Alternatively, you may be able to claim either a discount off a new Pixel or a cash payment – though the exact amount of cash/discount and availability of these options hasn’t yet been disclosed.

Not the first and maybe not the last

Worryingly, the Google Pixel 6a might not be the last device to get this battery weakening update. It wasn’t the first, with Google applying a similar update to some Pixel 4a models back in January, so that’s reason enough to think that more models might endure this in future.

But that’s not the only reason to think this could apply to other models too, as Android Authority has also found evidence in a teardown of Google’s Settings Services app that devices with poor battery health will be forced to use the company's Battery Health Assistance feature, which reduces capacity and charging speeds as the battery ages.

This feature is so far optional, other than in the case of the Google Pixel 9a, which has it permanently enabled.

The only silver lining to all this – other than these changes minimizing the risk of your phone catching fire – is that at least in the case of the Pixel 6a and Pixel 4a, these mandatory battery updates have only applied to select units.

So it’s possible your phone has a less troublesome battery that won’t ever need these changes, but this does rather make it sound like Google might want to change some of its battery suppliers.

You might also like

• I tested the Google Pixel 9a, and it would be the perfect bargain phone if it wasn’t so dang ugly
• Best Pixel phones: our reviews help you choose the Google phone that's right for you
• Google Pixel 9 review: a fantastic phone* (*added fees may apply)

• The Galaxy tri-fold could go on sale by the end of 2025
• A new report tips a 10-inch screen and a Snapdragon processor
• We can expect a high price and limited availability

We've been patiently waiting for the Samsung tri-fold phone to launch since it was briefly teased back in January, but a new report suggests the waiting isn't over yet – while also hinting at some of the key specs of the foldable.

This information comes from South Korean outlet The Bell (via well-known tipster @Jukanlosreve), and perhaps the most interesting detail here is that mass production of the device will begin in September. That means it should be available to buy sometime in the fourth quarter of 2025 (October, November, or December).

The report matches up with what we've previously heard about the tri-fold arriving later than the Samsung Galaxy Z Fold 7 and the Samsung Galaxy Z Flip 7 – which have a single fold to split the main display into two panels. These foldables should be unveiled at the next Unpacked event on Wednesday, July 9, and go on sale shortly after.

From this latest report, it sounds as though Samsung may announce some details of the Galaxy tri-fold at the July 9 event, even if it won't be on sale for a while. Previous rumors have suggested it could be called the Samsung Galaxy G Fold.

It's going to cost you

The Huawei Mate XT is already in the wild (Image credit: Huawei)

The article from The Bell goes on to say that Samsung is only manufacturing a relatively small number of these tri-fold phones – it's apparently more of a tech demonstration showcase than a handset that's expected to shift a significant number of units.

There's also a mention of a high price for the tri-fold, which is to be expected given the technical complexity here. That will affect sales too, and earlier rumors pointed to an asking price of $3,000-$3,500 (about £2,195-£2,560 or AU$4,560-AU$5,320), as well as suggesting that the Galaxy G Fold is only going on sale in South Korea and China.

We've previously seen the Huawei Mate XT tri-fold go on sale for around the price of four PS5 Pros, so it's no surprise that Samsung's equivalent is going to cost a significant amount of cash. The Huawei model is currently only available in a limited number of Asian countries, which is perhaps another indicator of what Samsung might do.

We should get a few more official details next Wednesday, and of course we'll be covering the launch live. The Bell says the Samsung tri-fold will feature a 10-inch screen and run on the Snapdragon 8 Elite chipset, matching up with previous leaks.

You might also like

• The Samsung Galaxy tri-fold could match the Z Fold 7 with these 3 features
• Samsung drops another big hint about the tri-folding 'G Fold'
• The upcoming Samsung tri-fold phone may take a while to charge up

• HPE's $14 billion acquisition of Juniper Networks has now completed
• EU and UK approval came quickly, US approval not so much
• The networking business could account for 50% of revenue

More than 18 months after HPE initially disclosed plans to acquire Juniper Networks, and several regulatory hurdles later, the merger has now been completed.

With Juniper's shares ceasing trading on the NYSE and HPE's networking business set to double, HPE hopes the $14 billion acquisition will allow it to focus on higher-margin, higher-growth areas, creating "long-term profitable revenue growth."

The news comes shortly after the US Department of Justice approved the merger, subject to two key conditions which require HPE to divest its Instant On business and auction a license for Juniper’s AI Ops for Mist source code.

HPE-Juniper merger is now complete

"In addition to positioning HPE to offer our customers a modern network architecture alternative and an even more differentiated and complete portfolio across hybrid cloud, AI, and networking, this combination accelerates our profitable growth strategy as we deepen our customer relevance and expand our total addressable market into attractive adjacent areas," HPE CEO Antonio Neri commented.

Juniper Networks CEO Rami Rahim described the merger as an, "opportunity to disrupt the networking industry at the most important and relevant time."

HPE noted that the merger aligns with its vision for customers to access silicon, hardware, operating system, security, software and services all in one place.

Looking ahead, HPE envisions the combined networking business to contribute to more than half of the company's entire operating income, which stood at $7.6 billion, up 6%, in its most recent quarter.

Despite some conditions imposed by the DOJ, HPE was quick to gain approval from the European Commission, which found "no competition concerns in the European Economic Area" nearly a year ago. Just a couple of weeks later, the UK's Competition and Markets Authority also granted approval to the deal.

You might also like

• These are the best cloud hosting providers around today
• Check out our roundup of the best small business servers
• HPE set to cut thousands of employees despite results rise

• Apple has added more child protection features to FaceTime in iOS 26
• The latest one blurs videos when it detects nudity is present
• It currently affects adult accounts too, but that might be a bug

Apple has been adding parental control features that are designed to protect minors for years now, and it looks like a new one has just been found in the iOS 26 beta. Yet it’s turning out to be pretty controversial, as there are concerns that it could be something of an overreach on Apple’s part.

Specifically, the new feature has been added to the FaceTime video-calling app. When FaceTime detects that someone is undressing on the call, it pauses the call and instead displays a warning message that reads: “Audio and video are paused because you may be showing something sensitive. If you feel uncomfortable, you should end the call.” There are then buttons labeled “Resume Audio and Video,” and “End Call.”

At its WWDC 2025 in June, Apple published a press release covering new ways its systems will protect children and young people online. The release included a feature that lines up the new FaceTime behavior: “Communication Safety expands to intervene when nudity is detected in FaceTime video calls, and to blur out nudity in Shared Albums in Photos.”

The actual implementation was noted by iDeviceHelp on X. Below the post, @user_101524 added that the feature can be found in the Settings app in iOS 26 by going to Apps > FaceTime > Sensitive Content Warning.

By default, the feature is disabled, so it needs to be switched on by the user, but that hasn't stopped it from stirring up online debate...

Generating controversy

(Image credit: iDeviceHelp / Future)

While this new feature might seem sensible, it has actually generated a degree of controversy. That’s because right now, it seems to affect all users of iOS 26, not just those who are using a child account. This has ruffled some feathers among people who feel that Apple is potentially censoring the behavior of consenting adults.

As well as that, some users have questioned how Apple knows what is being displayed on-screen and whether the company has access to customer video calls. On this point, Apple has said the following:

“Communication Safety uses on-device machine learning to analyze photo and video attachments and determine if a photo or video appears to contain nudity. Because the photos and videos are analyzed on your child’s device, Apple doesn’t receive an indication that nudity was detected and doesn’t get access to the photos or videos as a result.”

Like many of Apple’s features, the on-device processing means that content is not sent to Apple’s servers and is not accessible by the company. Rather, it is using artificial intelligence (AI) to flag video content that likely contains nudity, then censoring it.

The fact that Apple’s Communication Safety features are aimed at protecting minors suggests that this latest FaceTime feature might not be intended to cover adults as well as children. Its inclusion on all accounts, therefore, might be an oversight or bug. While we don’t know for sure, we should find out by September when iOS 26 comes out of beta and releases in full to the public.

You might also like

• Apple just expanded its child safety features with age ratings that could lessen the chance of an inappropriate download
• Best parental control app of 2025: tested, ranked and reviewed by the experts
• iOS 26: new features, a new design, and everything you need to know

• More layoffs are underway at Xbox
• Perfect Dark developer The Initiative has been shuttered, on top of other game cancelations and cuts
• Microsoft Gaming CEO Phil Spencer said Xbox has "never looked stronger" in an email to staff

Microsoft Gaming CEO Phil Spencer claimed that the Xbox "platform, hardware, and game roadmap have never looked stronger" amid a studio closures, game cancelations, and widespread layoffs.

As reported by Bloomberg, Microsoft's gaming division has been hit by significant layoffs. This includes job cuts at mobile game maker King, ZeniMax, Rare, Halo Studios, and Turn 10 Studios.

The upcoming Rare game Everwild has seemingly been culled, as has an unannounced game from The Elder Scrolls Online developer ZeniMax Online Studios. The planned Perfect Dark reboot has been canned too, with the studio behind it, The Initiative, shut down.

In an email sent to staff, Spencer said that these decisions were intended "to position gaming for enduring success and allow us to focus on strategic growth areas" plus "increase agility and effectiveness" overall.

"I recognize that these changes come at a time when we have more players, games, and gaming hours than ever before. Our platform, hardware, and game roadmap have never looked stronger. The success we're seeing currently is based on tough decisions we've made previously," he continued.

These "tough decisions" presumably include previous studio closures and layoffs, such as the shuttering Arkane Austin, Alpha Dog Games, Roundhouse Games, and Tango Gameworks last year.

"We must make choices now for continued success in future years and a key part of that strategy is the discipline to prioritize the strongest opportunities. We will protect what is thriving and concentrate effort on areas with the greatest potential, while delivering on the expectations the company has for our business. This focused approach means we can deliver exceptional games and experiences for players for generations to come."

You might also like...

• Xbox cancels Perfect Dark, Everwild, and an unannounced online game amidst yet another round of mass layoffs
• Donkey Kong Bananza offers a no-holds-barred open world that encourages you to go bananas
• Nintendo Switch 2 owners, beware – some owners are reporting worrisome overheating problems

When it comes to cleaning habits, opinions vary massively. I manage all the floorcare articles on TechRadar, and while I have some of the very best vacuum cleaners in my cupboard, and spend a lot more time thinking about vacuuming than the average person, that doesn't mean I'm actually doing it that regularly.

To find out if my cleaning habits are up to scratch, I decided to find out – from a cleaning expert – exactly how often we should all be vacuuming our floors. I spoke to Kiril Natov, a carpet and upholstery cleaning technician and CEO of cleaning company Premium Clean, with 18 years of experience in the industry. Here's what he had to say…

How often should you vacuum carpets?

"Aim to vacuum carpets at least twice a week, even if they don't appear dirty," suggests Natov. If you have pets or children, he recommends upping that frequency.

Does it matter what kind of vacuum you use? "For cut pile carpets, use a vacuum with a beater bar or brush to enhance cleaning," Natov advises. A beater bar is just a rotating roller within the cleaner head, often with bristles, which helps agitate the carpet fibers. When the TechRadar team is recommending the best vacuums for carpet, we also look for models that have enough suction to pull up dust and debris that's wedged deeper into the carpet.

Natov also suggests you could use a robovac. My experience is that even today's best robot vacuums don't have enough suction to really deep clean carpet, but they are great for staying on top of regular, light cleans.

A beater bar will help agitate the carpet fibers (Image credit: Future)How often should you vacuum hard floor?

"We recommend vacuuming hardwood floors once a week," says Natov. He suggests using your vacuum's hard floor setting if it has one.

What kind of vacuum should you use for hard floors? "Any vacuum can suck dust, hair, and crumbs off your hardwood, tile, or vinyl floors, but some models do it better than others," says Natov. "To avoid scattering debris or possibly damaging delicate flooring, look for a vacuum that either lets you switch off the spinning brush roll or has a special cleaner head with soft bristles."

My own top picks for the best vacuum cleaners for hardwood floor are models that include a dedicated fluffy floor head – as Natov says, these are perfect for cleaning dust and debris from the surface of even delicate hard floors. Examples include the Dyson V15 Detect and the Dreame R20.

For hard floors, look for a vacuum with a dedicated soft roller – Dyson's version comes with a laser to highlight dust (Image credit: Future)How often should you vacuum your sofa?

"We recommend vacuuming the furniture and upholstery fabric once a month," says Natov, noting that pet owners might want to increase that to fortnightly. He suggests paying special attention to your sofa – it's one of the most high-traffic areas in your home, so it's bound to pick up dirt and dust.

When it comes to cleaning furniture and upholstery, it's all about the attachments. If you're buying a cordless vacuum, look for one with a mini motorized cleaner head – kind of like a shrunk-down version of the main floor head. This will help you cover larger, uneven surfaces like sofa cushions.

A mini motorized tool is ideal for cleaning sofas and other soft furnishings (Image credit: Future)

Natov's final hot tip for ensuring a thorough clean is to stay on top of vacuum cleaner maintenance. Clean your filters regularly (and make sure they're completely dry before putting them back into the machine – or your vacuum will start to smell), replace filters in line with manufacturer guidelines, stay on top of emptying the dust cup, and don't forget to clean the attachments, too.

You might also like...

• Corded vs cordless vacuums: which should I buy?
• How long should a vacuum cleaner last?
• Are Dyson vacuums worth the premium price? After testing most of the leading brands, here's my take

Here are the best TVs I have tested, including OLEDs and mini-LEDs, from top brands such as LG, Samsung and TCL.

• Kaspersky recently analyzed FunkSec, a new ransomware group
• This group uses AI to generate code in the encryptors and other tools
• Ransomware is steadily growing as a threat

The future of ransomware threats lies in Generative Artificial Intelligence (GenAI), as hackers are increasingly using the nascent technology to improve and streamline their coding processes, experts have warned.

The latest State of Ransomware report from Kaspersky’s Global Research and Analysis Team (GReAT) analyzed FunkSec, a relatively new ransomware group, first spotted in late 2024.

Despite its junior status, FunkSec already made a name for itself, “quickly surpassing many established actors by targeting government, technology, finance and education sectors across Europe and Asia,” Kaspersky said.

Get 55% off Incogni's Data Removal service with code TECHRADAR

Wipe your personal data off the internet with the Incogni data removal service. Stop identity thieves
and protect your privacy from unwanted spam and scam calls.View Deal

Lowering the barrier for entry

Analyzing the code in its products, the researchers determined that the group is actively using GenAI.

Telltale signs include generic placeholder comments (for example “placeholder for actual check”) and technical inconsistencies (commands for different operating systems that don’t align), they said.

Furthermore, they observed declared but unused functions such as modules included upfront but never utilized, which is something large language models are apparently used to doing.

“More and more, we see cybercriminals leveraging AI to develop malicious tools. Generative AI lowers barriers and accelerates malware creation, enabling cybercriminals to adapt their tactics faster. By reducing the entry threshold, AI allows even less experienced attackers to quickly develop sophisticated malware at scale,” said Marc Rivero, Lead Security Researcher at Kaspersky’s GReAT.

AI-powered attacks will probably require AI-powered defenses, as well. Today, many of the best antivirus and endpoint protection services use AI and machine learning, mostly to detect threats that traditional signature-based methods would miss.

Companies like CrowdStrike, SentinelOne, Sophos, Microsoft Defender for Endpoint, Palo Alto Networks, and many others, are vocal about their AI/ML capabilities, often emphasizing speed, accuracy, and lower false positives compared to legacy solutions.

In this report, Kaspersky recommended users enable ransomware protection for all endpoints, keep everything updated, and focus defense strategies on detecting lateral movements and data exfiltration, among other things.

You might also like

• One of the most powerful ransomware hacks around has been cracked using some serious GPU power
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

• Xbox has canceled Everwild, Perfect Dark, and an unannounced game from ZeniMax
• Perfect Dark developer, The Initiative, has also been shuttered
• Layoffs have occurred at several studios

Microsoft has once again made significant cuts to its gaming division, as Xbox cancels a handful of games, closes a studio, and performs yet another round of mass layoffs.

As reported by Bloomberg, Xbox has closed down studio The Initiative and subsequently canceled the upcoming Perfect Dark reboot. Rare's Everwild has also been canceled, as well as an unannounced online title from The Elder Scrolls Online developer Zenimax Online Studios.

Matt Booty, head of Xbox Game Studios, has said in an email to Xbox staff that the cuts “reflect a broader effort to adjust priorities and focus resources to set up our teams for greater success within a changing industry landscape.”

To facilitate these adjustments, Xbox has reportedly laid off staff across a number of its studios. These include ZeniMax, Candy Crush maker King, Call of Duty support studios Raven Software and Sledgehammer Games, Forza Motorsport developer Turn 10 Studios, and Halo Studios which is responsible for the wider Halo franchise.

In total, it seems that a harrowing 9,000 staff members have been affected by the cuts. Back in May 2025, Microsoft also cut an estimated 6,000 to 7,000 employees from its workforce, meaning the company has potentially let around 15,000 staff go in this year alone - a terrifyingly high number.

It's not the first time Xbox has elected to close entire studios, either. Last year, the company shuttered Arkane Austin, as well as Alpha Dog Games, Roundhouse Games, and perhaps most controversially, Tango Gameworks. Thankfully, the Hi-Fi Rush and The Evil Within developer was able to be rehoused under South Korean publisher Krafton, but not every studio was so lucky.

You can also check out...

• Nintendo Switch 2 owners, beware – some owners are reporting worrisome overheating problems
• Dying Light: The Beast won't have branching narrative choices, but the developer says it will pave the way for the series' future
• Steam's new performance overlay feature will help you monitor how your games are running

• Leaked photos show the Samsung Galaxy Z Fold 7 from various angles
• You can see it in a stylish blue shade, and it looks very thin
• Another leak suggests the phone will only have an IP48 rating

At this point, we’ve seen numerous leaked renders of the Samsung Galaxy Z Fold 7, but so far we haven’t seen much in the way of photos. One leaked photo showed the back of the phone, but that was about it – until now.

Today, @Jukanlosreve – who has a great track record for leaks – has shared photos showing the front, back, and sides of the Samsung Galaxy Z Fold 7, and while the details are no different to what we’ve already seen in renders, the phone looks much better in the flesh.

You can see that it’s in a blue shade here, which could be the ‘Blue Shadow’ that we’ve previously heard might be one of the Samsung Galaxy Z Fold 7 colors – though from the name we’d expect that to be darker.

Z Fold 7 pic.twitter.com/h8EhC7LbTPJuly 3, 2025

In any case, you can also see a triple-lens camera, and one of the photos provides a good look at just how slim this phone might be.

Previous leaks disagree on exactly how thin it will be, with sources pointing to anything from 3.9mm to 4.5mm when unfolded, but anywhere in that range would make it a lot thinner than the 5.6mm thick Samsung Galaxy Z Fold 6.

A dust disappointment

Fold7 IP48 pic.twitter.com/o9icyqDTmfJuly 2, 2025

In other Samsung Galaxy Z Fold 7 news a leaked energy rating label shared by @MysteryLupin lists it as having an IP48 rating. That’s the same as the current model, and is at odds with some earlier leaks that pointed to better dust resistance.

If the Samsung Galaxy Z Fold 7 does have this rating, then while it will be able to survive submersion in water to depths of 1.5 meters for up to 30 minutes, it will only have minimal dust resistance, so that would be disappointing.

We suspect this leaked label is correct though, since foldable phones always struggle with dust resistance, and since leaker @PandaFlashPro also recently claimed it has an IP48 rating.

We should find out for sure soon, as the Samsung Galaxy Z Fold 7 is set to launch on July 9, alongside the Samsung Galaxy Z Flip 7.

You might also like

• How to watch Samsung Galaxy Unpacked on July 9: get ready for new foldable phones and more
• A massive new Samsung Galaxy Z Fold 7 and Z Flip 7 leak suggests we could see some serious upgrades at next week's Unpacked
• Samsung Galaxy Z Fold 7 could launch with a big free bonus – and it's good news for AI fans

• Login credentials for an account with root access was found in Cisco's Unified Communications Manager
• There are no workarounds, just a patch, so users should update now
• Different versions of the tool are affected

Another hardcoded credential for admin access has been discovered in a major software application - this time around it’s Cisco, who discovered the slip-up in its Unified Communications Manager (Unified CM) solution.

Cisco Unified CM is an enterprise-grade IP telephony call control platform providing voice, video, messaging, mobility, and presence services. It manages voice-over-IP (VoIP) calls, and allows for the management of tasks such as user/device provisioning, voicemail integration, conferencing, and more.

Recently, Cisco found login credentials coded into the program, allowing for access with root privileges. The bug is now tracked as CVE-2025-20309, and was given a maximum severity score - 10/10 (critical). The credentials were apparently used during development and testing, and should have been removed before the product was shipped to the market.

Get 55% off Incogni's Data Removal service with code TECHRADAR

Wipe your personal data off the internet with the Incogni data removal service. Stop identity thieves
and protect your privacy from unwanted spam and scam calls.View Deal

No evidence of abuse

Cisco Unified CM and Unified CM SME Engineering Special (ES) releases 15.0.1.13010-1 through 15.0.1.13017-1 were said to be affected, regardless of the device configuration. There are no workarounds or mitigations, and the only way to address it is to upgrade the program to version 15SU3 (July 2025).

“A vulnerability in Cisco Unified Communications Manager (Unified CM) and Cisco Unified Communications Manager Session Management Edition (Unified CM SME) could allow an unauthenticated, remote attacker to log in to an affected device using the root account, which has default, static credentials that cannot be changed or deleted," Cisco said.

At press time, there was no evidence of abuse in the wild.

Hardcoded credentials are one of the more common causes of system infiltrations. Just recently Sitecore Experience Platform, an enterprise-level content management system (CMS), held a hardcoded password for an internal user. It was just one letter - ‘b’ - which was super easy to guess.

Roughly a year ago, security researchers from Horizon3.ai found hardcoded credentials in SolarWinds’ Web Help Desk.

Via BleepingComputer

You might also like

• New Chrome flaw leaks sensitive information across websites - your data could already be in the wrong hands
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

As AI models grow larger and more capable, the supporting infrastructure must evolve in tandem. AI’s insatiable appetite has Big Tech going as far as restarting nuclear power plants to support massive new datacenters, which today account for as much as 2% of global electricity consumption, or more than the entire country of Germany.

But the humble power grid is where we need to start.

Constructing the computing superstructure to support AI tools will significantly alter the demand curve for energy and put increasing strain on electrical grids. As AI embraces more complex workloads across both training and inference, compute needs – and thereby power consumption – are expected to increase exponentially. Some forecasts suggest that datacenter electricity consumption could increase to as much as 12% of the global total by 2030.

Semiconductors form the cornerstone of AI computing infrastructure. The chipmaking industry has focused primarily on expanding renewable energy sources and delivering improvements in energy-efficient computing technologies. These are necessary but not sufficient – they cannot sustainably support the enormous energy requirements demanded by the growth of AI. We need to build a more resilient power grid.

Moving from Sustainability to Sustainable Abundance

In a new report, we call for a different paradigm – sustainable energy abundance – which will be achieved not by sacrificing growth, but by constructing a holistic energy strategy to power the next generation of computing. The report represents the work of major companies across the AI technology stack, from chip design and manufacturing to cloud service providers, as well as thought leaders from the energy and finance sectors.

The foundational pillar of this new strategy is grid decarbonization. Although not a new concept, in the AI era it requires an approach that integrates decarbonization with energy abundance, ensuring AI’s productivity gains are not sidelined by grid constraints. In practical terms, this entails embracing traditional energy sources like oil and gas, while gradually transitioning toward cleaner sources such as nuclear, hydro, geothermal, solar and wind. Doing this effectively requires understanding of the upgrades needed for the electricity grid to enable rapid integration of existing and new energy sources.

Consuming electricity from the grid naturally assumes the emissions profile of the grid itself. It should come as no surprise that emissions related to the grid represent the single biggest component of the emissions bill facing any given company. In the conventional approach to sustainability, companies focused more on offsetting emissions derived from the grid rather than sourcing the grid with cleaner (or carbon-free) energy. To support the coming scale-out of AI infrastructure, access to a clean grid will be one of the most important aspects in reducing carbon footprint.

Strategically selecting locations for datacenters and semiconductor fabs will be critical. Countries and regions have a varying mix of clean energy in the power grid, which impacts their carbon emission profile. For example, the United States and France generate a similar percentage of their overall electricity from renewable sources. However, the United States has a significantly higher country emission factor, which represents the direct carbon emission per kilowatt-hour of electricity generated.

This is because most of the electricity in France is generated through nuclear power, while the United States still gets a significant percentage of electricity supplied through coal and natural gas. Likewise, there could be significant differences within a country such as the United States, with states like California having a higher mix of renewables compared to some other states.

Driving Innovation in Semiconductor Technology

A truly resilient grid strategy must start with expanded capacity for nuclear, wind, solar, and traditional forms of energy, while driving a mix shift to cleaner sources over time. However, to achieve this enhanced capacity, it will be necessary to invest in disruptive innovations. Transmission infrastructure must be modernized, including upgraded lines, substations and control systems. Likewise, the industry must take advantage of smart distribution technologies, deploying digital sensors and AI-driven load management techniques.

Semiconductors have an important role to play. Continued growth of GPUs and other accelerators will drive corresponding growth in datacenter power semiconductors, along with increasing semiconductor content in other components such as the motherboard and the power supply.

We forecast that the datacenter power semiconductor market could reach $9 billion by 2030, driven by an increase in servers as well as the number of accelerators per server. Approximately $7 billion of the opportunity is driven by accelerators, with the rest coming from the power supply and other areas. As the technology matures, we believe gallium nitride will play an important role in this market, given its high efficiency.

As the grid incorporates increasing levels of renewables, more semiconductors will be needed for energy generation. Silicon carbide will be important for solar generation and potentially wind as well. We estimate that renewable energy generation could grow to more than a $20 billion market for semiconductors by 2030. A similar opportunity exists for smart infrastructure such as meters, sensors and heat pumps.

Shifting Incentives for Sustainable Growth

Restructuring the power grid offers the single biggest opportunity to deliver sustainable, abundant energy for AI. Modernizing the power grid will require complex industry partnerships and buy-in from company leadership. In the past, sustainability initiatives were largely regarded as a compliance checkbox item, with unclear ties to business results. A new playbook is needed to enable the growth of AI while shifting business incentives toward generation, transmission, distribution and storage of clean energy and modernization of the power grid.

To truly harness the transformative productivity and prosperity potential of AI, we need a comprehensive sustainability strategy that expands clean energy capacity, modernizes energy infrastructure, and maintains diverse energy generation sources to ensure stable, abundant power for continued technological innovation. When combined with progress in energy-efficient computing and abatement measures, this holistic approach can realistically accelerate the pursuit of sustainability while mitigating the risk of curtailing growth due to insufficient energy resources.

We list the best IT infrastructure management service.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

Where we are today is not hybrid cloud rebranded. Hybrid was a transition strategy. Distributed is an entirely new operating environment, where cloud infrastructure and services are physically located in multiple, dispersed environments: on-premise data centers, multiple public clouds, edge locations, and sovereign zones. Yet they are managed as a single, cohesive system. Unlike centralized or hybrid approaches, distributed cloud treats geographic and architectural diversity as a feature, not a compromise.

This shift happened gradually. Organizations reacted to new regulatory frameworks like GDPR and FedRAMP, which enforce data locality and privacy standards that centralized architectures can’t always support. Meanwhile, latency-sensitive applications, like real-time analytics, pulled compute closer to the user, pushing cloud computing infrastructure to the edge. And cost became a concern: 66% of engineers report disruptions in their workflows due to lack of visibility into cloud spend, with 22% saying the impact is equivalent to losing a full sprint.

Distributed cloud addresses all of these challenges, enabling businesses to comply with regulations, improve performance, localize deployments, and maintain operational continuity in one architectural shift. But managing it to ensure that a distributed framework actually reaches its full potential requires serious rethinking. Infrastructure has to be modular and versioned by design, not patched together.

Dependencies need to be explicit, so changes don’t cascade unpredictably. Visibility should extend beyond individual cloud providers, and governance has to follow workloads wherever they run. Yet most organizations today operate without these principles, leaving them struggling with fragmentation, limiting their scalability, opening the door to security and competitive threats, and slowing innovation.

Old Tools, New Problems

There’s growing evidence to show just how widespread the shift toward distributed cloud has become: 89% of organizations now use a multi-cloud strategy, with 80% relying on multiple public clouds and 60% using more than one private cloud. The reasons are strategic: reducing vendor lock-in, complying with data localization laws, and improving performance at the edge.

But the consequences are operational. Fragmentation creates chaos. Teams struggle with version control, lifecycle inconsistencies, and even potential security lapses. Infrastructure teams become gatekeepers, and developers lose confidence in the systems they rely on.

Most organizations are still applying traditional centralized cloud management principles to a distributed world. They rely on infrastructure as code (IaC), stitched together with pipelines and scripts that require constant babysitting. This approach doesn’t scale across teams and regions. IaC also introduces new dependencies between layers that are invisible until they break.

All in all, the approach is problematic: 97% of IaC users experience difficulties, with developers often viewing IaC as a “necessary evil” that slows down application deployment. The result is a kind of paralysis: any change carries too much risk, so nothing changes at all.

A New Operating Model for a Fragmented World

Solving this requires more than another tool. It requires a new operating model and mindset. Infrastructure should be broken into modular, composable units with clear boundaries and pre-defined dependencies. Teams should be able to own their layer of the stack without impacting others. Changes should be trackable, auditable, and safe to automate.

Platforms that offer a single control plane across environments can make this possible. They turn complexity from a liability into a strategic asset: one that offers flexibility without sacrificing control. This is where emerging approaches like blueprint-based infrastructure management offer a compelling path forward. Instead of expecting AI or DevOps teams to connect workflows, infrastructure can be transformed into modular components. Think Lego bricks, except it’s a chunk of code that’s versioned, pre-approved, and reusable.

In this model, automation doesn’t mean giving up control. It means enabling teams to move faster within guardrails that are defined by architecture. The result is a system that scales, even across regulatory zones, business units, and tech stacks. This kind of shift doesn’t eliminate complexity, but it makes it manageable, legible, and strategic.

And we’re already seeing the rise of blueprint-based and modular infrastructure strategies across the industry—from Cisco’s UCS X-Series, which decouples hardware lifecycles for flexible upgrades, to Microsoft Azure Local’s unified control plane for hybrid deployments, and the growing adoption of platform engineering as a discipline for building reusable, scalable systems. It’s an evolution that just makes sense.

The Strategic Advantage of Managing Well

Distributed cloud isn’t something organizations opt into. It's the state they're already operating in. The real differentiator now is how well it's managed. Scaling infrastructure has always been achievable; distributing infrastructure, by contrast, demands a different kind of investment: in architecture, workflows, and operational discipline.

Without a system built specifically for elasticity, decoupling, and visibility across environments, complexity quietly erodes both speed and trust. Infrastructure becomes harder to change, risks accumulate invisibly, and innovation slows to a crawl.

The right foundation turns that story around. Distributed infrastructure, when managed deliberately, doesn’t have to become a barrier. It becomes a catalyst.

Elastic systems allow teams to localize deployments without fragmenting control. Decoupled architectures enable parallel innovation across business units, cost centers, and regions without introducing instability. Unified visibility makes governance a continuous function, not an afterthought. Managing complexity isn’t about eliminating it; it’s about structuring it so that distributed systems can scale sustainably, adapt safely, and operate predictably.

In that context, infrastructure becomes a lever for scale instead of a source of drag. Managing it well isn’t just an operational need. It’s a strategic advantage.

We list the best cloud database.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

When it comes to fixed broadband, Irvine is the fourth-fastest city in the US. These top ISPs offer the best internet plans.

Here are the answers for The New York Times Mini Crossword for July 3.

New troubling research found that entertainment is the most common reason teens use their phones behind the wheel, followed by texting and navigation.

• Crunchyroll aired anime with obviously AI-generated subtitles that included typos, clunky phrasing, and lines like “ChatGPT said.”
• Fans quickly noticed and criticized the lack of human oversight
• The incident highlights growing concerns about AI replacing creative roles without proper review, particularly in localization, where context and tone are crucial

There are mistranslations, and then there are ChatGPT subtitles that appear to have been deliberately written to upset people. That's what appeared to happen with some of the translated Japanese shown on screen during episodes of anime recently spotted and shared online.

The first example to gain attention online made it clear that ChatGPT was the culprit of awkward and outright wrong translations during an episode of Necronomico and the Cosmic Horror Show, Crunchyroll’s new anime series about occult weirdness and internet brain rot. It literally included the line "ChatGPT said" in both the German and English subtitles.

Fans started posting screenshots of bizarre sentence structures and dialogue that they had spotted, and now had an explanation and a source of blame for. Misspelled character names, inconsistent phrasing, and just outright made-up words and phrases were spotted everywhere.

(Image credit: Pixel/Bluesky)

I only watched about two minutes, and was so frustrated at the subs having errors that even a normal machine translation wouldn't have given.

— @hilene.bsky.social (@hilene.bsky.social.bsky.social) 2025-07-03T02:47:11.136Z

In case that wasn't enough, Crunchyroll’s president, Rahul Purini, had told Forbes in an interview only a few months ago that the company had no plans to use AI in the “creative process.” They weren’t going to mess with voice acting or story generation, he said. AI would be restricted to helping people find shows to watch and to recommending new shows based on what viewers had previously enjoyed.

Apparently, ChatGPT translations don't count under that rubric, but localization isn't a mechanical process, as any human translator could explain.

Localization art

Hey now, show some respect for the most storied of all anime subbers: Translator's name

— @viridianjcm.bsky.social (@viridianjcm.bsky.social.bsky.social) 2025-07-03T02:47:11.132Z

Localizing is a big deal among anime fans. Debates over whether certain subtitles are too literal, too loose, or too limited in their references to be understood outside Japan have raged for decades. But no one on any side of those debates is likely to claim these massive errors by ChatGPT are okay.

Crunchyroll hasn’t officially clarified how this happened, but reports suggest the subtitles came from the company's Japanese production partner. The generated subtitles may have been given to Crunchyroll to air without Crunchyroll being responsible for making them.

As several people pointed out, when you pay to stream anime from a major platform like Crunchyroll, you're expecting a certain baseline of quality. Even if you disagree with a localizer's choices, you can at least understand where they are coming from. The fact that apparently no one read the ChatGPT subtitles before they were uploaded to a global audience is harder to justify.

Translation is an art. Localization isn’t just about replacing Japanese with English. It’s about tone, cadence, subtext, and making a character sound like themselves across a language barrier. AI can guess what words go where, but it doesn’t know the characters or the show. It's like a little translation dictionary, which is fine as far as it goes, but it can't make a conversation make sense without a human piecing together the words. A few fans are outraged enough to call for unsubscribing and going back to sharing fansubs, the homebrewed subtitles unofficially written and circulated back in the days of VHS. In other words, the very thing Crunchyroll once helped make obsolete by offering higher-quality, licensed versions of shows.

At a time when more people are watching anime than ever before, Crunchyroll is apparently willing to gamble that most of us won’t notice or care whether the words characters say make any sense. If Crunchyroll wants to keep its credibility, it has to treat localization not as a tech problem to optimize, but as a storytelling component that requires human nuance and judgment. Otherwise, it might just be "gameorver" for Crunchyroll's reputation.

(Image credit: @pi8you/Bluesky)You might also like

• ChatGPT crosses a new AI threshold by beating the Turing test
• AI took a huge leap in IQ, and now a quarter of Gen Z thinks AI is conscious
• ChatGPT model matchup - I pitted OpenAI's o3, o4-mini, GPT-4o, and GPT-4.5 AI models against each other and the results surprised me

Make sure to use the Switch 2 cable included with the new gaming console.

• Geopolitics, not just technology, is quietly rewriting who can sell cameras in Western markets
• Hikvision says it’s unfair, but the shutdown shows trust is no longer automatic for Chinese firms
• Company denies wrongdoing, but surveillance fears are now enough to end entire business operations

Canada has ordered Chinese surveillance giant Hikvision to cease its operations in the country, citing national security concerns.

The ban follows a formal review conducted under the Investment Canada Act and marks a move against foreign technology firms.

"The government has determined that Hikvision Canada Inc.’s continued operations in Canada would be injurious to Canada’s national security," said Industry Minister Mélanie Joly.

International pressure and rising suspicion

Hikvision, one of the world’s largest producers of surveillance cameras, has operated in Canada since 2014.

However its expansive global reach and ties to state-linked projects in China have long drawn concern from Western countries.

Although the government has not made public the specific reasons behind its decision, it has stated intelligence and security assessments played a central role.

This silence is likely to fuel speculation, much like in previous crackdowns on Huawei, where classified intelligence was used to justify broad commercial restrictions.

The comparison to Huawei is not unwarranted. Hikvision now finds itself under the same kind of scrutiny that led to Huawei’s ejection from 5G infrastructure projects across the Five Eyes nations.

The US, UK, and Australia have all already taken measures against Hikvision, particularly over claims its cameras have been used to surveil Uyghur Muslims in China’s Xinjiang region, allegations that Beijing denies.

The FBI has also warned about malware targeting webcams, and the Western world often believes that Chinese IoT is arguably more dangerous than TikTok, which is considered spyware.

Unsurprisingly, Hikvision “strongly disagrees” with Canada's decision, saying, “We believe it lacks a factual basis, procedural fairness, and transparency,” the company claims the move appears “to be driven by the parent company’s country of origin.”

With geopolitical tensions continuing to define much of the West’s approach to Chinese firms, decisions like Canada’s risk being seen less as technology-based judgments and more as political posturing.

Hikvision claimed it cooperated fully with authorities and submitted all requested documents, but this did not alter the outcome.

It’s unclear how many public buildings in Canada still use Hikvision devices, but Joly has committed to reviewing and phasing out any remaining equipment.

“I strongly urge Canadians to take note of this decision and make their own decisions accordingly,” she warned.

The Canadian government appears to be focusing on surveillance risks, and this questions the trustworthiness of smart devices, like the webcams or parental control solutions.

As more homes and workplaces adopt smart cameras and monitoring tools, the line between convenience and intrusion becomes thinner.

If bans become more commonplace, vendors may need to prove more than just feature strength to remain competitive.

Whether you're selecting a home monitoring system or seeking the best antivirus software, the politics of hardware and software are becoming harder to ignore.

Via Economic Times

You might also like

• These are the best business smartphones right now
• We've also listed the best rugged laptops available
• Amazon has a million robots on its floors - close to outnumbering human workers