Technology

• Directive 8020 will feature "real-time threats" that will ramp up players' fear
• Creative director, Will Doyle, says the game's Turning Point narrative system will allow players to explore the story's branching outcomes
• Doyle explains that these two changes to the Dark Pictures formula are "critical" to how players interact with the story

Supermassive Games has introduced two "critical" changes to Directive 8020 that will affect how players interact with the game's branching narrative.

Speaking in an interview with TechRadar Gaming during Summer Game Fest, Directive 8020's creative director, Will Doyle, explained that the team has switched up the survival horror aspect for the next Dark Pictures anthology entry by implementing "real-time threats".

"There are two big things that have fundamentally changed what this game is," Doyle said. "One of them is the real-time threats, the effect that that's had on changing the control system, and the fear that it puts into you when you're playing it really, really gets your heart pounding."

The Until Dawn creators have also changed how players make important story choices by introducing something called the Turning Point system. This mechanic allows players to rewind at certain points in the narrative, which Doyle says will allow players to explore different outcomes, unlike in any other Dark Pictures entry.

"The other is the turning point system and the ability to explore our branching story. Because for me, that's kind of two critical aspects of what we've done," he said.

"We're a narrative game, and you've got a new system that lets you explore the narrative in a fun way, and we are a game that lets you move around and have action in it, and we've pushed that forward as well."

Directive 8020 is the fifth entry in the eight-planned Dark Pictures series and arrives on October 2 for PlayStation 5, Xbox Series X, Xbox Series S, and PC.

In the game, a colony ship called the Cassiopeia has crash-landed on the planet Tau Ceti f in the middle of an expedition to save Earth. The crew soon discovers they aren't alone and must survive being hunted by an alien organism capable of mimicking its prey.

You might also like...

• Best horror games: the scariest titles to play in 2025
• Xbox Games Showcase 2025: everything revealed from Call of Duty: Black Ops 7 and Persona 4 remake to a new handheld
• After seeing Resident Evil Requiem in action, I'm convinced it'll be another groundbreaking entry in the horror series

• Interpol and international partners conducted Operation Secure
• In four months, the police arrested dozens of people and seized vital data
• Thousands of IP addresses hosting infostealers were taken down, as well

Dozens of people have been arrested, and thousands of IP addresses seized, in an Interpol-led international law enforcement operation aimed at disrupting a network of infostealers and other malware.

The international law enforcement agency said Operation Secure took place between January and April 2025, and saw police agencies in 26 countries work together to locate servers, map physical networks, and move in to disrupt cybercriminal campaigns.

As a result, 32 people were arrested: 18 in Vietnam, 12 in Sri Lanka, and two in Nauru, including the individual suspected to have been running the entire operation, who was found with around $11,500 in cash, several SIM cards, and business registration documents which, according to the Interpol, point to a scheme “to open and sell corporate accounts”.

Servers, IP addresses, and gigs of data

In Hong Kong, the police analyzed more than 1,700 pieces of intelligence shared by Interpol, which helped them identify 117 command-and-control servers hosted on 89 internet service providers.

These servers were used as central hubs for different cybercriminal campaigns, including phishing, fraud, and social media scams.

Aside from the arrests, the police also seized 41 servers and obtained more than 100 GB of data.

Furthermore, it took down more than 20,000 malicious IP addresses linked to information stealers, and identified 216,000 victims and potential victims, some of whom have already been notified.

A few private cybersecurity companies also participated in the campaign, by providing valuable intelligence: Group-IB, Kaspersky, and Trend Micro.

“Interpol continues to support practical, collaborative action against global cyber threats,” commented Neal Jetton, the agency’s Director of Cybercrime.

“Operation Secure has once again shown the power of intelligence sharing in disrupting malicious infrastructure and preventing large-scale harm to both individuals and businesses.”

You might also like

• Top US dental firm spills over 8 million user files online
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

Writing an effective AI image prompt isn't hard, but it may take a little more work than you expect.

• Sam Altman says a ChatGPT prompt uses "0.34 watt-hours" of electricity, roughly one second of an oven
• He also says a single ChatGPT prompt uses "0.000085 gallons of water; roughly one-fifteenth of a teaspoon"
• While that's not a lot in isolation, ChatGPT has over 400 million weekly users, with multiple prompts per day

OpenAI CEO, Sam Altman has revealed ChatGPT's energy usage for a single prompt, and while it's lower than you might expect, on a global scale, it could have a significant impact on the planet.

Writing on his blog, Altman said, "The average query uses about 0.34 watt-hours, about what an oven would use in a little over one second, or a high-efficiency lightbulb would use in a couple of minutes. It also uses about 0.000085 gallons of water; roughly one-fifteenth of a teaspoon."

While that might not sound like a lot as an isolated prompt, ChatGPT has approximately 400 million active weekly users, and that number is growing at a rapid rate. Bear in mind there's a growing amount of AI tools and chatbots on the market, including Google Gemini and Anthropic's Claude, so general AI energy usage will be even higher.

Last month, we reported on a study from MIT Technology Review which found that a five-second AI video uses as much energy as a microwave running for an hour or more. While Altman's ChatGPT prompt energy usage reveal is nowhere near as high as that, there are still concerns considering how much people interact with AI.

We rely on AI, so is this energy consumption a concern?

There's a constant concern about ChatGPT's energy consumption, and it is becoming increasingly vocal as AI usage continues to rise. While Altman's blog post will put some minds at ease, considering the relatively low energy and water usage in isolation, it could also spark more uproar.

Earlier this week, a mass ChatGPT outage led to millions of people unable to interact with the chatbot. Over the 10 hour plus period, I received emails from thousands of readers who gave me a new perspective on AI.

While I'd be lying if I said AI's energy consumption doesn't concern me, it would be unfair to overlook the positives of the technology and how it is improving the lives of millions.

The climate crisis is not limited to me and you, but unfortunately, it's the working class that ultimately pays the price. ChatGPT's energy consumption at a mass scale may be a severe problem in the future, but then again, so are the private jets flying 10-minute flights.

The AI climate concerns are not black and white, and those who criticise the impact of the technology on the planet are equally vocal about the impact of other technologies. That said, we're only at the beginning of the AI revolution, and energy consumption will continue to rise. At what point should we be worried?

You might also like

• This is what really happened with Siri and Apple Intelligence, according to Apple
• Why don’t AI companies talk about energy usage?
• Sam Altman thinks superintelligence is within our grasp and makes 3 bold predictions for the future of AI and robotics

If the AI Action Summit in Paris earlier this year is the sign of things to come, establishing a coordinated approach to regulation and governance will be no easy task in the short or long term. To an extent, this is understandable – these processes rarely operate at pace, particularly when stakeholders are still trying to understand the impact of important trends, such as the emergence of advanced AI.

The problem this creates, however, is that without consensus, organizations must work against the backdrop of a more complex and unpredictable threat landscape, with the tools used by threat actors more advanced and accessible than ever before. In fact, the pace of change around AI tools is so rapid that it’s difficult to properly predict exactly what new cyber risks will emerge.

Not all one-way traffic

Thankfully, it’s not all one-way traffic. On one hand, threat actors are using generative tools to automate phishing campaigns, identify system vulnerabilities and write malicious code. On the other, forward-thinking organizations are using the same techniques to stay one step ahead. Nevertheless, security teams are in a difficult position – expected to respond to a growing range of threats with the same or fewer resources, while also managing a greater range of more serious risks.

The result is a situation where many organizations are constantly playing catch-up. Threat intelligence may be available, but without the right tools and frameworks in place to both distil and make use of it, much of that insight stays hidden or goes underutilized.

This plays out in different ways. In some organizations, core security functions – from threat intelligence and automation to incident response – remain siloed, with limited coordination or shared visibility. In others, strategies are developed in isolation, missing the opportunity to tap into the wealth of experience and insight already available across the broader security community.

The result? Individual businesses are left to fend off highly organized, fast-moving threat groups that thrive on shared intelligence and agile tactics, and are often several steps ahead.

The power of collaboration

To address these essential issues, organizations are relying more heavily on security collaboration and collective defense, with Information Sharing and Analysis Centers (ISACs) among the most established and effective approaches.

Operating across sectors, these groups are designed to collect, analyze and distribute actionable threat intelligence, while also equipping members with the tools and resources needed to strengthen resilience. Today, the National Council of ISACs, for example, includes nearly 30 sector-specific organizations – a clear sign of how far this model has evolved.

The industry clearly sees the value. According to recent research, more than 90% of respondents say collaboration and information sharing are either very important or critical to their cybersecurity strategy. The problem? Almost three-quarters (70%) believe they could be doing more, with nearly one in five admitting they could share significantly more intelligence than they currently do.

Worryingly, more than half of those surveyed (53%) said their organization doesn’t engage with an ISAC at all. Perhaps even more concerning, 28% weren’t even aware that ISACs exist, underlining how much ground there is still to cover in building a truly collaborative cyber defense ecosystem.

But, for any effective approach to collective defense to succeed, the goal is to establish workflows that allow for the rapid, structured exchange of indicators of compromise (IoCs), tactics, techniques and procedures (TTPs) and real-world incident reports. Communities that get this right create a multiplier effect – the more each participant shares, the stronger the whole becomes.

Proactive security

All of this supports a shift from reactive to proactive security. Today’s teams must be able to identify risks before they escalate and take preventive action in near real time.

But that’s easier said than done. Security operations are often flooded with data from multiple sources, making it hard to separate the signal from the noise. That’s why threat intelligence platforms (TIPs) are becoming vital in helping ingest and operationalize threat data and in the process, reducing manual overhead and improving decision-making.

The best TIPs also enable automated intelligence sharing with external communities. In doing so, they act as a nerve center, connecting internal teams with trusted partners outside the organization.

This can transform the levels of sophistication and speed that can be applied to threat intelligence, empowering security teams to replace their reliance on manual processes while boosting efficiency, time savings and improved accuracy.

TIPs can also broaden the different types of data used in the threat intelligence process, including integrating structured and unstructured data, which can then be delivered as standardized output.

Looking at the overall security landscape at present, the challenges posed by AI-powered cybercrime are already prompting regulators to push for higher standards. Across multiple industries, new rules demand that organizations go beyond point solutions and build resilience into their day-to-day security strategies.

In practical terms, that means engaging with trusted partners and building response frameworks that are ready for action. If and when international standards emerge, organizations that embrace the collective defense approach will be strongly placed to ensure their networks and data remain safe.

Train up with the best online cybersecurity courses.

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

• DC Studios has released one final trailer for Superman
• The James Gunn-directed superhero movie releases on July 11
• Its latest teaser contains some crucial details about its plot and the odd Easter egg

James Gunn's Superman movie is officially less than a month away (at the time of publication, anyway). And, as the countdown towards its July 11 release gathers pace, DC Studios has debuted one final trailer for the David Corenswet-led comic book movie.

The trailer, which also confirms tickets are now on sale for one of 2025's most anticipated new movies, doesn't just contain footage we've seen in the DC Universe (DCU) film's other teasers.

Indeed, there are numerous new clips that hint at the direction of its plot and contain the odd DC Comics Easter egg. So, here are four crucial details and/or references to previous Superman projects you might have missed upon first viewing. Potentially big spoilers follow, so proceed at your own risk.

1. A glimpse at Mister Handsome

Who's the mysterious individual standing behind Lex Luthor? (Image credit: DC Studios/Warner Bros. Pictures)

If your initial reaction to this individual is "...who?", you're not alone. Mister Handsome isn't an established DC Comics character – in fact, he appears to be a brand-new creation for this DCU Chapter One movie.

Okay, so who is he? We don't actually know, but that hasn't stopped eagle-eyed fans from thinking they've spotted him in Superman's latest trailer.

Look behind Nicholas Hoult's Lex Luthor in the above image, and you'll see what appears to be a pale skinned individual standing on some form of podium. Viewers think this human-looking creature is the aforementioned Mister Handsome.

Superman's behind-the-scenes vignette gave us a better look at Mister Handsome (Image credit: DC Studios/Warner Bros. Pictures)

There's further evidence to suggest this is the case. In Superman's behind-the-scenes featurette, we catch another glimpse of a being who not only resembles the character in Superman's final trailer, but also has a visibly different posture and appearance to most humans.

Additionally, peaking as part of a recent Fandango interview, Hoult and Gunn revealed that Hoult's son, who occasionally visited the Superman set, had a fondness for Mister Handsome. Elaborating on who this character is, Gunn teased: "Mister Handsome is Lex’s creature that he created in a petri dish who drives around on this flying platform that is the ugliest, grossest creature in the world."

This doesn't confirm that the pale skinned individual we briefly see is Mister Handsome, but Gunn's description lines up with what little we've seen of them. Throw in the incredibly ironic name for someone who's not attractive to look at and this has to be Mister Handsome.

2. Ultraman unmasked

Things are heating up between Ultraman and Superman (Image credit: DC Studios/Warner Bros. Pictures)

There's been plenty of speculation about Ultraman's actual identity. I've extensively covered the biggest fan theory about this secondary villain, so I won't do so again here. If you want more details on said hypothesis, read my pieces on Ultraman's supposed ties to another villain called the Hammer of Boravia, my breakdown of Superman 's second trailer, and the biggest Ultraman fan theory seemingly being confirmed by some Funko Pop figures.

Superman's third trailer doesn't deliver much in the way of new Ultraman footage, but there are a couple of missable shots of this mysterious individual without his mask. The first shows Ultraman blasting the Man of Steel with his own heat vision ability. Later, we see Supes whacking a demasked Ultraman with a large metallic object.

Neither clip is slow or long enough to give us a good look at Ultraman's face. Nonetheless, we know he'll lose his face covering at some point, which suggests we'll find out who he really is. My money is still on him being a Superman clone.

3. Green Lantern's 'Big Blue' call back

A superpowered humanoid alien is enough to make anyone green with envy (Image credit: DC Studios/Warner Bros. Pictures)

At the trailer's 1:20 mark, we see Nathan Fillion's Guy Gardner/Green Lantern squaring up to Kal-El. Clearly, something's happened between the pair and it seems Gardner is trying to goad Supes to attack him. You don't mock someone by saying "make a move, Big Blue" if you're not trying to antagonize them.

"Big Blue" might sound like a form of derision on Gardner's part, but it's actually a call back to one of Superman's oldest nicknames. Indeed, it's a reference to 'The Big Blue Boy Scout' moniker that the Son of Krypton also went by in the 1950s. It's also an alias that's been shortened to 'Big Blue' in countless other comic books.

This, then, appears to be one of many homages that'll be paid to the Man of Tomorrow's literary roots in his latest big-screen adventure.

4. Who is baby Joey in Superman?

One Kryptonian and a baby (Image credit: DC Studios/Warner Bros. Pictures)

The final big reveal in Superman's latest trailer concerns a character called Joey. He's the alien baby we see Clark Kent's superpowered alter-ego holding/rescuing as some form of cosmic explosion erupts around them.

We already know Joey's father is, too. Superman's Funko Pop figurine collection has already ruined that surprise, with Joey being the progeny of Anthony Carrigan's Metamorpho. For the uninitiated: essentially, he's a metahuman who can create any type of element out of thin air.

The prevailing fan theory – one strengthened by footage shown in Superman's second trailer – is Luthor is exploiting Metamorpho in order to keep Joey safe. And, given Metamorpho's unique powers, he's tailor-made to craft kryptonite, aka the only substance in the known galaxy that can harm/weaken someone like Superman. It seems, then, that Metamorpho will be coerced to create some kryptonite, which Luthor can use against Supes, so Luthor doesn't harm Joey.

You might also like

• James Gunn confirms new Wonder Woman movie is on the way, and DC fans say a Star Wars actor is the clear choice to play her
• Get the complete lowdown on Supergirl: Woman of Tomorrow
• Or find out everything we know about Peacemaker season 2

There are spin-offs to the popular game, like Quordle, and more puzzles that everyone can try now.

• Some Pixel 6a phones have been overheating while charging
• Google says it will address the issue with a software update
• The update will limit battery capacity and charging speed

You may have spotted several reports of Google Pixel 6a phones catching fire and burning up in recent weeks, and Google is now taking steps to stop any future incidents by limiting the battery capacity and charging speed on the 2022 handsets.

In a statement to Android Authority, a Google spokesperson said that a "subset" of Pixel 6a phones will soon get a "mandatory" software update, reducing battery capacity and charging performance once 400 cycles have been reached.

This should "reduce the risk of potential battery overheating" according to Google, though it will leave you with a phone that charges up more slowly and doesn't last as long between charges – not great for a handset that's only been out three years.

Google says users with affected phones will be contacted next month with details of what they need to do. Meanwhile, the Android Authority team has spotted a warning about a potential battery overheating issue on the Pixel 6a in the latest Android 16 beta.

Keep an eye on your phone

Evidence of a burned out Pixel 6a (Image credit: Ariella / Android Authority)

It's difficult to gauge just how widespread this problem is, but Android Authority has already recorded two separate incidents, and there are pictures as proof. It's scary to think that your phone could suddenly burst into flames while it's been left charging.

Given that Google only mentions a "subset" of Pixel 6a phones, it's possible that only certain handsets are affected. We may not see this update roll out for every single Pixel 6a, but right now it's not clear either way.

If you own a Google Pixel 6a, it's a good idea to keep an eye on it while it's charging: look out for any kind of deformation, and check that the handset isn't getting too hot to the touch. Very short battery life can also indicate a battery health problem.

Swelling and overheating can potentially happen to any lithium-ion battery, if it gets physically damaged or somehow malfunctions, but it's an issue that keeps happening with Pixel phones – most recently with the Google Pixel 7a.

You might also like

• The Pixel 10 could be launching earlier than we thought
• These are the best Pixel phones you can buy
• Google teases a refresh for the Pixel's always-on display

• The Google Pixel 10 series could have built-in magnets for magnetic wireless charging
• A new line of 'Pixelsnap' accessories might launch to take advantage of this feature
• The Pixel 10 series could also have improved speakers

Android phones have technically been capable of supporting MagSafe-like charging for a while now, but so far, no big-name handsets have fully embraced the technology. The Samsung Galaxy S25 series, for instance, offers a half-measure solution that requires a magnetic case – but the Google Pixel 10 series might actually have magnets built in.

This rumor comes from Android Authority, and if true, would mean you’d be able to magnetically attach wireless chargers and other accessories to the Google Pixel 10, with no special cases required.

Android Authority claims to have seen “credible marketing materials intended for retailers” that show Google is working on magnetic accessories for the Pixel 10. These accessories would use the Qi2 standard (which enables magnetic wireless charging), and they include a ‘Pixelsnap Charger’, a ‘Pixelsnap Charger with Stand’, and a ‘Pixelsnap Ring Stand’.

The first two are self-explanatory, while the ‘Pixelsnap Ring Stand’ might be a stand for your phone that doesn’t include a charger.

Given that Google has previously been rumored to be working on a ‘Hub Mode’ for its phones, these stands would make sense, as Hub Mode would work like a smart display, where a stand would come in handy.

Faster charging and better speakers

The Google Pixel 9 (Image credit: Philip Berne / Future)

In any case, Android Authority also found evidence of an upcoming Google wireless charger in a trade database, and its specs allow, in theory, for 60W wireless charging – though since current Pixel models can’t even charge that fast with wires, it’s likely that Google would restrict the speeds.

The database also mentions that this upcoming charger will come in ‘Rock Candy’ and ‘Mist’ shades, which Android Authority speculates will translate to black and off-white, respectively.

And as well as a charging upgrade, the Google Pixel 10 series could also get improved speakers, with Android Headlines claiming that they will be the best speakers ever on a Pixel phone, albeit ones lacking Dolby Atmos.

As ever with leaks, we’d take all of this with a pinch of salt for now, but it’s surprising that Android device makers aren’t yet embracing magnetic charging, so we’re hopeful that the Google Pixel 10 series will. The phones will probably launch in August, so we should find out whether they do then.

You might also like

• Android 16 is out now, but its big visual redesign might not arrive until September
• Google Pixel 10 rumors – 5 things we think we know and 3 we don't
• Best Google Pixel phones: our reviews help you choose the Google phone that's right for you

• Tens of thousands of Nvidia chips will be deployed across many European countries
• Nvidia is partnering with key telecom firms across the continent
• AI centers will further research and offer training opportunities

Nvidia has announced significant new plans to support European customers as the region looks to bolster sovereignty requirements amid ongoing global trade concern

The chipmaker has revealed France, Italy, Spain and the UK are all deploying Nvidia Blackwell systems to build their own sovereign AI infrastructure, and it has also pledged to build an AI factory in Germany for industrial manufacturing applications, together with further AI technology centers in Germany, Sweden, Italy, Spain, the UK and Finland.

Nvidia's expansion across Europe comes amid partnerships with a number of European firms, including Mistral AI, Orange, Swisscom, Telefónica and Telenor.

Nvidia European expansion

Though confirmed in an Nvidia press release, many of the infrastructure expansions rely on local partnerships.

Mistral AI will deploy 18,000 Grace Blackwell systems in France, Nebius and Nscale will roll out 14,000 Blackwell GPUs via new data centers in the UK, Germany will launch the world's first industrial AI cloud with 10,000 Blackwell GPUs and Italy is to develop the "Large Colosseum" reasoning model on Grace Blackwell Superchips.

Further partnerships with European telecommunication companies will see Orange, Fastweb, Telenor, Swisscom and Telefónica launch their own AI models and tools.

"Every industrial revolution begins with infrastructure. AI is the essential infrastructure of our time, just as electricity and the internet once were," CEO Jensen Huang explained.

UK Tech Secretary Peter Kyle expanded on Huang's notion: "Just as coal and electricity once defined our past, AI is defining our future."

A series of AI centers are also to be established across many European countries, including Germany, Sweden, Italy, Spain, the UK and Finland, with the aim of accelerating AI research and offering local upskilling options.

"With bold leadership from Europe’s governments and industries, AI will drive transformative innovation and prosperity for generations to come," Huang added.

The news follows Huang's recent appearance at London Tech Week 2025, where he joined UK Prime Minister Sir Keir Starmer on stage to announce new expansions in the country, and hail the impact of AI investment.

You might also like

• Check out the best AI tools and best AI writers
• We've listed our pick of the best cloud computing services
• The UK Government wants to save billions per year with new tech procurement site

• iOS 26 adds the new Call Screening feature to iPhone
• The new feature automatically asks callers for a name and reason for calling, and provides a real-time transcript
• New data suggests more than 1 billion hoax calls could be intercepted

A new feature in iOS 26 could block more than a billion spam calls each year, according to new data.

Apple is adding a new Call Screening feature to iPhone with iOS 26, and with Call Screening enabled, your iPhone will ask for a name and reason for calling before sending the call through, building on the Live Voicemail feature added with iOS 17.

Similarly to Live Voicemail, Call Screening provides a real-time transcription of the caller’s response to those initial questions, and then gives the user the choice to pick up or ignore the call.

The feature will also be available in the Wi-Fi calling-enabled Phone apps coming with iPadOS 26 and macOS Tahoe 26.

A billion blocked calls

The iPhone 16 family will all get Call Screening, whether its powered by Apple Intelligence or not. (Image credit: Future)

Analysis by second-hand phone marketplace Compare and Recycle suggests that this could block more than a billion scam calls each year in the UK alone.

Though the report provided by Compare and Recycle is UK-specific, it tracks that this figure could increase quite substantially with other countries factored in.

The report estimates that the average person in the UK gets four spam calls per month. The report also estimates that just shy of 24 million people in the UK will get access to Call Screening, working out to more than 1.1 billion intercepted calls each year.

Personally, I can’t stand spam calls, and there have been days – and even whole months – where I seem to get much more than the estimated four calls.

Additionally, I get plenty of calls that drop as soon as I pick up, or as soon as I say something. It’s never nice to imagine that my number’s been scraped or marked as ‘active’ in a database for scammers.

To be honest, the issue is so prevalent that I sometimes don’t pick up the phone at all, instead waiting for a voicemail or follow-up email to know exactly who's trying to reach me.

The addition of Call Screening could see me pop the SIM card out of my trusty Oppo Find X8 Pro and back into one of the best iPhones, for the sake of using my phone as an actual phone again.

With that said, the official Apple press release for iOS 26 doesn't make it clear whether Call Screening is an Apple Intelligence feature, in which case my iPhone 15 would be out of luck.

WWDC saw the announcement of plenty more features for iPhone, iPad, and Mac – head over to our WWDC2025 recap for a full rundown.

And be sure to let us know whether you’re looking forward to using Call Screening on your iPhone in the comments below.

You might also like

• Ooh, shiny! I hate myself for being excited about Liquid Glass in iOS 26
• Android 16 is out now, but its big visual redesign might not arrive until September
• Apple is out of surprises and focuses on making the future as clear as glass

Internet service providers in Los Angeles offer fiber plans as well as DSL. To help you narrow down the right option, here are CNET's top picks for home internet.

Here are hints and the answers for the NYT Connections: Sports Edition puzzle, No. 262, for June 12.

Here are the answers for The New York Times Mini Crossword for June 12.

A new setting in iOS 26 can extend battery life without resorting to Low Power Mode.

The third major tournament of the year sees Bryson DeChambeau aiming to defend his title at Oakmont.

• Dual Thunderbolt 5 ports and OCuLink elevate this eGPU beyond typical external GPU standards
• Nvidia's Ada Lovelace cards shine in this unit
• Compact design and bold specs make FEVM FNGT5 Pro a tempting power upgrade for PCs

External GPUs have long served as a way to upgrade a laptop’s graphical capabilities, particularly for users whose machines lack discrete GPUs.

The FNGT5 Pro from Chinese manufacturer FEVM is the latest entrant in this niche category, bringing an ambitious mix of high-end GPU options and modern connectivity features.

The FNGT5 Pro supports three RTX 40-series laptop GPUs, specifically the RTX 4060, 4080, and 4090. This might raise eyebrows, but it appears to be a calculated decision to balance power and heat management within such a compact enclosure.

RTX eGPU aims high

Measuring 142 x 100 x 60 mm and with a total volume of 0.86 liters, the FNGT5 Pro is compact and travel-friendly, though not quite pocket-sized.

Despite its portability, the device features dual Thunderbolt 5 ports (100W upstream and 30W downstream), a high-speed USB-A port, and an OCuLink interface.

Offering both Thunderbolt 5 and OCuLink sets it apart from most rivals, which typically offer just one of the two.

Display connectivity is handled by HDMI 2.1 and DisplayPort 1.4a outputs.

If you're part of the Apple ecosystem, however, don’t get too excited, you likely can’t use this eGPU with a MacBook Pro.

Apple has not supported external GPUs since its transition to Apple Silicon, and even earlier Intel-based Macs were only compatible with Thunderbolt 3 eGPUs using officially supported AMD GPUs.

Despite Thunderbolt 5 being theoretically backward-compatible and extremely fast, macOS lacks the driver-level support needed for Nvidia cards, especially those housed in non-certified enclosures.

So, while you could physically connect the FNGT5 Pro to a MacBook via Thunderbolt, it’s highly unlikely to function as intended.

As for pricing, the top-tier RTX 4090 Laptop GPU, with 16GB of memory and 9,728 CUDA cores, costs $1,374, steep, but in line with desktop equivalents.

The RTX 4080, featuring 7,424 CUDA cores and 12GB of memory, is priced at $1,040, while the entry-level RTX 4060, with 3,072 CUDA cores and 8GB of RAM, comes in at $555.

For users seeking the best laptop for video editing or for Photoshop, pairing a compatible system with a powerful eGPU like the FNGT5 Pro can help close the performance gap without committing to a full desktop setup.

Via TomsHardware

You might also like

• Check out the best 3D modeling software for 3D printing and more
• We've rounded up the best portable monitors available now
• This free Chrome extension makes Google way better and faster

• InnoGrit N3X SSD delivers 50 DWPD endurance but costs more than typical enterprise drives
• Built for caching, inference, and workloads that punish ordinary SSDs
• Runs entirely in SLC mode, sacrificing capacity for serious performance gains

The InnoGrit N3X SSD introduces a high-endurance storage solution aimed at enterprise workloads with extreme write demands.

Unveiled at Computex 2025, and featuring Kioxia’s second-generation XL-Flash operating in SLC mode, the drive is engineered to deliver 50 drive writes per day (DWPD) over five years, far exceeding the endurance of typical enterprise SSDs.

This level of durability is impressive, but it also raises questions about the cost of the device and whether its performance will justify the expected premium.

SCM roots and a specialized architecture

At the heart of the N3X is storage class memory (SCM), a memory tier designed to bridge the performance gap between DRAM and traditional NAND flash.

When used in SLC mode, Kioxia’s XL-Flash functions as a type of SCM, promising ultra-low latency and high endurance.

Unlike standard NAND, which stores multiple bits per cell, operating XL-Flash in SLC mode prioritizes speed and reliability over capacity.

This design choice closely mirrors the original goals of Intel’s now-discontinued Optane memory, positioning the N3X as a potential successor in that specialized niche.

While SCM technologies like XL-Flash are not new, they remain relatively rare due to their high cost and specialized applications.

InnoGrit’s use of the IG5669 PCIe 5.0 controller, with NVMe 2.0 support, allows for impressive performance claims: up to 14 GB/s read and 12 GB/s write speeds, along with 3.5 million random read IOPS.

Latency is where the N3X particularly stands out - read latency under 13 microseconds and write latency as low as 4 microseconds.

If consistently achieved, these figures would place the N3X among the fastest SSDs in development.

The drive is marketed for workloads involving sustained writes, in-memory computing, and real-time inference, areas where traditional NAND SSDs often struggle with latency and wear.

However, the decision to operate entirely in SLC mode significantly reduces the available capacity per die, resulting in smaller drive sizes and a higher cost per gigabyte.

While the drive is offered in capacities ranging from 400GB to 3.2TB, these fall short of what is expected from today’s largest SSDs.

Although the N3X possesses many of the technical qualities of the best portable SSDs, it is not intended for mainstream use.

Its reliance on SCM architecture, while enabling exceptional performance, places it firmly in the domain of niche enterprise deployments.

You might also like

• These are the best budget laptops available
• Apple Mac Mini M4 gets a helmet-style retro Macintosh docking station
• We've also listed the best workstations right now

Deep Cover brings an injection of comedy to the genre of the moment, making it a top contender to stream this weekend.

The latest tariff announcement doesn't change much, but you should still expect prices to increase. Here's how to save if you're shopping for one.