Technology

Here are some hints — and the answers — for Connections No. 594 for Saturday, Jan. 25.

Here are some hints and the answer for Wordle No. 1,316 for Saturday, Jan. 25.

Commentary: "AI" has been a constant buzzword at tech events, but some of Samsung's latest offerings could prove to be quite helpful.

High cholesterol is considered 240 mg/dL. The man’s was over 1,000 mg/dL.

The FCC's press release on ending DEI efforts could have an immediate impact on broadband efforts to bridge the digital divide.

• Asus recently released its Q-Release Slim feature for GPUs
• Said feature is reportedly damaging GPUs
• There's no current fix for this issue from Asus

Asus’s Q-Release Slim was first revealed in August 2024 during Gamescom. It is an eject feature on 800-series AMD and Intel motherboards that makes removing graphics cards much easier. However, the recently released feature is reportedly causing major headaches for buyers.

According to HXL and Andreas Schilling from HardwareLuxx (reported on by Tom’s Hardware), Q-Release Slim has been causing damage to graphics cards by scraping off the sides of the connector. This could potentially ruin your expensive graphics cards, including the recent RTX 5090.

The Q-Release Slim is supposed to be an improved version of the Q-Release launched back in 2021 for high-end Asus ROG motherboards. It lets you quickly eject cards by titling them while pulling upwards. However, the fault lies in the peg side of the PCIe slot on the motherboard, which has a metallic piece embedded in it that scrapes against the connector. If the card is ejected frequently, this can result in significant damage — something reviews and overclockers tend to do.

Uniko’s Hardware offers more insight into the matter, including several photos from the bilibili channel. The first is of the design flaw reportedly causing this issue, while two more are before and after the original card’s condition versus the damaging results of using the Q-Release Slim eject 60 times.

We’ve also reached out to Asus concerning this issue and will update this story with any statements from them.

This isn’t the only issue plaguing Asus either

This isn’t the only recent report concerning Asus motherboards either. Patch notes slipped out early, which revealed a BETA bios fix for a “microcode signature verification vulnerability,” apparently affecting Asus gaming motherboards.

Tavis Ormandy, the security researcher who spotted the leak, stated, "It looks like an OEM leaked the patch for a major upcoming CPU vulnerability, ie: 'AMD Microcode Signature Verification Vulnerability.’” He then continues, “I'm not thrilled about this. The patch is not currently in Linux firmware, so this is the only publicly available patch."

By the time Asus edited the patch notes to remove mentioning AMD’s microcode issue, AMD had already confirmed the vulnerability to The Register. "AMD is aware of a newly reported processor vulnerability. Execution of the attack requires both local administrator-level access to the system, and development and execution of malicious microcode," the company stated.

It seems that Asus motherboards have been embroiled in some serious trouble, which could spell disaster for both processors and graphics cards. Hopefully, the manufacturer can correct the issue soon, as the timing couldn’t be worse — with the release of the RTX 5000-series and new processors on the horizon.

• AMD confirms processor security flaws after Asus patch slips out early
• Painful GPU extraction woes could be a thing of the past with new motherboards
• Some of the most iconic AMD chips have a serious security flaw — which the company says it probably won't patch now

Lexington residents have multiple providers to choose from, including Spectrum, Comporium and Kinetic. Here’s how the best broadband providers in Lexington stack up.

Back in August 2023, Meta revealed an ‘all-in-one’ AI translation model capable of understanding close to 100 different languages.

Dubbed SeamlessM4T (Massively Multilingual and Multimodal Machine Translation), this is Meta’s attempt at creating a ‘universal translator’ akin to the Babel Fish in Douglas Adams’ classic sci-fi series The Hitchhiker’s Guide to the Galaxy.

The team behind the SeamlessM4T tool has now detailed its work in a piece in the journal Nature, revealing the advanced system delivers an all-in-one solution for text-to-text, speech-to-text, speech-to-speech, and text-to-speech translations across an impressive, and growing, array of languages.

Over 400 years of raw audio

SeamlessM4T, which, among other things, is being used to automatically dub videos on Facebook and Instagram, currently supports speech-to-speech translation from 101 to 36 languages, speech-to-text translation for from 101 to 96 languages, text-to-text translation for 96 languages, text-to-speech translation from 96 to 36 languages, and automatic speech recognition for 96 languages. This unified approach overcomes the limitations of traditional cascaded systems, which often require separate subsystems for speech recognition, translation, and text-to-speech synthesis.

By streamlining these processes, Meta says SeamlessM4T outperforms existing models, achieving up to 23% higher BLEU (Bilingual Evaluation Understudy) scores in translation accuracy and demonstrating impressive resilience to background noise and speaker variations.

To create SeamlessM4T, Meta started with 4 million hours (over 400 years) of multilingual raw audio originating from a publicly available repository of crawled web data. The team developed SeamlessAlign, a multimodal corpus containing over 470,000 hours of aligned speech and combined the dataset with cutting-edge machine learning techniques, including SONAR (Sentence-level Multimodal and Language-Agnostic Representations) embeddings, which enable multilingual and modality-agnostic encoding for text and speech.

Meta says that by addressing social and ethical challenges through the use of safeguards, SeamlessM4T can be a valuable tool for global communication. These safeguards reduce gender bias - errors in grammatical gender determination - and mitigate the problem of added toxicity - where offensive words appear in translations but not in the original source.

You might also like

• This is the best translation software you can use right now
• Meta reveals bot that can translate almost 100 languages
• Alibaba releases new translation-focused AI model

I had a chance to see Dreame's newest robot vacuum, the Dreame X50 Ultra Complete, in action. With its cutting-edge features, this brand is the one to watch. In fact, it might even make me switch allegiance from my current favorite bot brand, Roborock.

We saw plenty of impressive robot vacuum innovations at CES 2025 in January, but the X50 Ultra Complete has the most... well, complete feature set I've seen.

The standout headline feature is that it has tiny little mechanical legs. These lift the bot and help it hoist itself over tall thresholds in the home. This 'ProLeap System' means the X50 Ultra can tackle steps up to 4.2cm tall in a single bound or up to 6cm tall in two. In action, it looks kind of like the wooing ritual of an exotic bird. It's really quite majestic.

Slightly less graceful is the bot's descent down steps. There is a "shock-absorption system" designed to soften collisions and absorb sound, as well as a little wheel in the front that means the X50 is not quite face-planting, but it still comes down with a bit more of a bump than I was expecting or am wholly comfortable with.

What's afoot?

Beyond those little feet, there's more to show off, though. It also has a navigation puck that can pop up and down (a feature appearing on one of the newest Roborock bots). That enables it to use LiDAR navigation without permanently raising the height of the bot.

With the puck retracted, the height of the robot vacuum is just 8.9cm tall, shallow enough that it can sneak under low-sitting furniture, where dust and hair might otherwise collect. It'll then pop back up again when space allows.

Those are two 2025 innovations, but there are further features that have appeared on previous Dreames – including 2024's highly-rated Dreame L40 Ultra robot vacuum and Dreame X40 Ultra Complete robovac – and help round out what is an extremely capable setup.

For example, the spinning mop pads attach using magnets, which means the X50 Ultra Complete can drop them off in its base when they're not required and then return to pick them up again as needed, with no manual intervention required from you.

That means there's absolutely no danger of wet mop pads dragging over your carpets (the mop pads can also lift up to avoid this, but it's not such a failsafe approach if you have thick rugs, for example).

(Image credit: Future)

To ensure the edges of rooms aren't missed, the Dreame X50 not only has an extending side brush to sweep debris into the robot's suction path (fairly common on premium bots), but also a mop pad that can kick out to the side.

I saw the Dreame X50 Ultra, alongside the brand's other products, at the grand opening of its first flagship store in Birmingham, UK – a "significant milestone" in the brand's expansion into the UK.

We're in the process of testing the Dreame X50 Ultra and will report back if it's as good as it looks. It will go on sale in the UK from February 7 at a list price of £1,299. It'll be available to purchase directly from Dreame (in-store and online) as well as via Amazon.

US release information and pricing has been requested, and we'll update this when we hear back.

You might also like...

• 5 overdue robot vacuum innovations I want to see this year
• Narwal’s new robovac will moonwalk your floors to get even the thickest carpets clean
• Browse the best robot vacuums for pet hair

Commentary: We want to see Android phones with built-in magnets, but official cases are a step towards that.

Exclusive: Check out this reveal of two cards from the next Magic: The Gathering set, leaning into Aetherdrift's interplanar race theme.

• The FBI’s missive follows three previous ones in as many years
• Statement is aimed at educating businesses and warding off domestic collaborators
• Suggested remedies include employing endpoint protection on computer systems and checking applications for “typos and unusual nomenclature”

The FBI has claimed North Korean IT workers are extorting US companies which have hired them by leveraging their access to steal source code.

In a statement, the agency warned domestic and international firms employees turned threat actors, “facilitate cyber-criminal activities and conduct revenue-generating activity” using stolen data “on behalf of the regime.”

It recommended endpoint protection, and monitoring network logs to identify where data has been compromised across “easily accessible means” like shared internal drives and cloud storage drives.

FBI guidance on remote hiring processes

The FBI also recommended a litany of actions that all amount to taking care to know who you’re hiring, which sounds like good practice even if you’re not especially worried about unwittingly hiring a threat actor.

It recommended stringent identity verification processes throughout the recruitment process and cross-checking applicants’ details against that of others in the pile, and across different HR systems.

It also claimed these applicants are using AI tools to obfuscate their identities, but, if true, offered little advice to counter them beyond conducting recruitment processes in person; which isn’t always possible.

The agency also suggested recruiters ask applicants “soft questions” about their whereabouts and identity, but we’d suggest that this is good practice all round too.

North Korean IT workers have been a target of the FBI for some time, having released separate guidance in 2022, 2023, and 2024. In the latter, it expressed concern that US-based individuals were, knowingly or unknowingly, helping facilitate state-sponsored threat actors by setting up US-based infrastructure such as front addresses and businesses.

You may also like

• North Korean hackers target job seekers with fake interviews
• FBI warns of North Korean hackers using VPNs to infiltrate businesses
• We’ve also listed the best malware removal tools right now

Check out our picks for the best AV receivers from Onkyo, Sony, Yamaha, Denon and Marantz.

The company is pushing back on the claims, stating Apple Watch bands are "safe for users to wear."

The AI agent is part of a greater effort to automate more aspects of our daily lives.

• QNAP said it addressed six flaws in its Hybrid Backup Sync tool
• The flaws stemmed from rsync, an open-source file syncing tool
• Users are advised to update their HBS immediately

QNAP has addressed half a dozen vulnerabilities affecting its Hybrid Backup Sync (HBS) software.

In a security advisory, the company noted the vulnerabilities were discovered in rsync, an open source file synchronization tool used to transfer and sync files between systems. It supports local and remote operations via SSH, and minimizes data transfer with incremental updates. Many backup solutions use rsync, including Duplicity, Bacula, Rclone, and others.

HBS is a data backup and disaster recovery solution that supports local, remote, and cloud storage services.

Arbitrary code execution

The bugs are tracked as CVE-2024-12084, CVE-2024-12085, CVE-2024-12086, CVE-2024-12087, and CVE-2024-12088, and affect HBS 3 Hybrid Backup Sync 25.1.x. QNAP said they could have been used to run malicious code remotely against unpatched Network Attached Storage (NAS) endpoints. Apparently, threat actors would only need anonymous read access to vulnerable servers, in order to exploit the flaws.

"When combined, the first two vulnerabilities (heap buffer overflow and information leak) allow a client to execute arbitrary code on a device that has an Rsync server running," CERT/CC said when rsync 3.4.0 was released. "The client requires only anonymous read-access to the server, such as public mirrors. Additionally, attackers can take control of a malicious server and read/write arbitrary files of any connected client."

To secure their systems, administrators are advised to update their HBS 3 Hybrid Backup Sync to version 25.1.4.952, by logging into QTS or QuTS hero as an admin, opening App Center and searching for HBS 3 Hybrid Backup Sync, and clicking on the Update button.

According to BleepingComputer, there are currently more than 700,000 IP addresses with exposed rsync servers, but it’s difficult to determine how many can be exploited.

Via BleepingComputer

You might also like

• Huge data breach exposes over 600,000 records, including background checks, vehicle, and property records
• Here's a list of the best antivirus tools on offer
• These are the best endpoint protection tools right now

• Pre-orders should go live for AMD’s RX 9070 GPUs on March 23, according to one US retailer
• There are a couple of reasons why this is likely wrong, though
• We’ve also seen more strangeness around the previously rumored January 23 launch for RDNA 4, as an ad campaign appears to have misfired

As if enough weirdness hasn’t already surrounded AMD’s RDNA 4 GPU launch – it’s been an odd affair indeed – something at Team Red appears to have misfired on the marketing side again, and we’ve also caught a mention of a possible new release date for the RX 9070 graphics cards.

However, I think the date shared by a US retailer is off the mark, and just a placeholder, so let’s deal with that first.

The story here is that B&H Photo previously posted what it believed to be the date that gamers would be able to pre-order RX 9070 models, namely January 23 – but of course that didn’t happen.

Instead, AMD made an announcement to let us know that the RX 9070 graphics cards won’t arrive until March (still within the official Q1 2025 timeframe that was previously promised, it’s worth noting).

Following that, VideoCardz spotted that B&H Photo now has a date of March 23 for pre-orders on RDNA 4 graphics cards – and the retailer still has this date on its site (it hasn’t been taken down yet).

What’s strange about this is that March 23 is a Sunday, so it seems highly unlikely that a sales milestone (pre-orders, or indeed the actual on-sale date) would be on a weekend. Surely, AMD would wait until the Monday (March 24) if this was the plan.

Also, bearing in mind that the previous date B&H posted was January 23, it feels like what might have happened here is that a staff member working on the product listings just replaced the name of the month in the date, and left the day the same (though why an update was posted at all to the RDNA 4 listings, when nothing’s official, is a good question).

Hence my conclusion that this looks very much like a placeholder date, given the way this has happened, and the additional fact that it’s a Sunday.

(Image credit: Gorodenkoff / Shutterstock) Reddit campaign wasn’t canceled?

The second oddity that backs up some previous rumors – and the initial January date B&H attached to RX 9070 pre-orders – is that some AMD ads have popped up, going by reports, claiming that you can ‘play now’ with the new RDNA 4 GPUs.

This was again noticed by VideoCardz, and it happened yesterday, on January 23, so one theory is that this was a date AMD had indeed planned for something big to happen, either a launch or reveal. (If you can ‘play now’ though, then the GPUs would surely be on sale). Team Red had a bunch of marketing stuff scheduled for Reddit (as flagged up here) around this, which accidentally still went ahead somehow (with these ads appearing in a few different countries in Europe, seemingly being published in German, Italian, and Polish).

The not-unreasonable conclusion folks are reaching is that AMD did have something planned for January 23, especially as there were a lot of rumors this was the case (albeit some mentioned January 24, though it was all in the same ballpark of the week we find ourselves in now). But all this has now been put off to March, although officially, as mentioned, AMD has never said anything concrete for the RX 9070 launch outside of it being Q1 2025.

The rumors around RDNA 4 have taken some very strange turns indeed, then, but the good news is that it seems these next-gen graphics cards will be worth waiting for. Even if we do have to hang on until later in March, rather than seeing the RX 9070 GPUs emerge earlier in the month – and really, I wouldn’t be swayed at all by the new B&H pre-order date for all the reasons mentioned.

You might also like...

• Still worried RTX 5000 GPUs don’t have enough VRAM? Nvidia’s secret weapon is powerful AI texture compression, and it’s calmed some of my fears
• AMD vs Nvidia: who is the graphics card champion?
• Top gaming PCs: great rigs for serious PC gaming

Now that Netflix has unveiled which titles will be leaving in February 2025, take this as your sign to get a head start on catching the best Netflix movies before it's too late. But while Netflix doesn't tend to remove a huge amount of titles each month (which is a good thing), there are always one or two gems thrown into the mix that I know I'll miss dearly. As for this month, those movies are Pearl (2022) and Stand By Me (1986), but there are plenty new additions on the way to make up for it, looking at everything coming to Netflix in February.

When it comes to Netflix series, I find that these are less prone to being axed from Netflix's library. This month, only two TV shows (Brooklyn Nine-Nine and The Mindy Project) are being removed compared to over 30 movies, so with that said, you can't say that Netflix isn't one of the best streaming services when it comes to TV content.

Everything leaving Netflix in February 2025

Leaving on February 1

Cocaine Cowboys 2 (movie)
Plus One (movie)
Run All Night (movie)

Leaving on February 11

The Fast and the Furious (movie)
2 Fast 2 Furious (movie)
The Fast and the Furious: Tokyo Drift (movie)
Fast Five (movie)
Fast & Furious 6 (movie)
The Pope's Exorcist (movie)

Leaving on February 14

The Catcher Was a Spy (movie)
White Boy (movie)

Leaving on February 15

47 Meters Down: Uncaged (movie)
Blackhat (movie)
Pearl (movie)

Leaving on February 20

Book Club (movie)
Southpaw (movie)

Leaving on February 21

All Good Things (movie)

Leaving on February 24

U Turn (movie)

Leaving on February 25

Brooklyn Nine-Nine seasons 1-2 (TV show)

Leaving on February 28

21 Bridges (movie)
A Haunted House (movie)
A Haunted House 2 (movie)
Aloha (movie)
The Angry Birds Movie (movie)
Blended (movie)
Cinderella Man (movie)
Due Date (movie)
Green Lantern (movie)
Inception (movie)
Legends of the Fall (movie)
Little (movie)
The Mindy Project seasons 1-6 (TV show)
Oblivion (movie)
The Other Guys (movie)
Scooby-Doo (movie)
Scooby-Doo 2: Monsters Unleashed (movie)
Sixteen Candles (movie)
Stand by Me (movie)
Without a Paddle (movie)

You might also like

• Squid Game season 2 is Netflix's reigning champion – here are 3 more K-Dramas with perfect 100% Rotten Tomatoes scores
• The Night Agent season 2 stars tease where season 3 of the hit Netflix spy show will go next and it leaves Peter in a 'tight spot'
• Back in Action is Netflix's new #1 movie but its bad reviews are no secret – watch these 3 with over 90% on Rotten Tomatoes instead

Frank Scheufens, Product Manager at PNY Technologies explains the benefits and uses of Digital Twins in this second installment.

Operating Digital Twins: virtually manage, monitor and optimize physical assets

A digital twin operates through a cycle of data collection, transmission, analysis, and user interaction. The process begins with data collection, where physical assets are fitted with sensors that monitor various parameters—such as temperature, pressure, or movement. This data is then transmitted to the digital twin via cloud or edge computing, allowing the digital model to update in real-time. Advanced algorithms and analytics are applied to the incoming data, helping the twin predict future states, identify anomalies, and provide actionable insights.

The real power of a digital twin lies in its interactivity. Operators can interact with the model to test hypothetical scenarios, diagnose problems, and adjust operations without affecting the actual physical asset. This simulation capability, coupled with real-time data, provides a dynamic environment for experimentation and optimization.

The role of AI in Digital Twin projects

Artificial intelligence (AI) plays a crucial role in maximizing the potential of digital twins. AI’s ability to handle massive datasets, identify patterns, and make predictions allows digital twins to go beyond basic monitoring. In many cases, machine learning algorithms are integrated into the twin to enable predictive analytics—forecasting potential failures or maintenance needs before they occur. This predictive capability helps organizations shift from reactive to proactive maintenance, reducing costs and avoiding unexpected downtime.

AI can also help with anomaly detection, identifying unusual behavior patterns in the data. In industries like manufacturing or aviation, these patterns might indicate a need for maintenance or signal a potential issue. Additionally, AI algorithms can simulate various scenarios within the digital twin, allowing organizations to optimize processes and test changes without risking real-world consequences. In some advanced applications, AI even enables the digital twin to make autonomous decisions, adjusting operations based on the analyzed data.

Challenges in Digital Twin projects

While digital twins offer immense benefits, they also come with challenges that need to be managed carefully. One of the most significant is data management. The vast quantities of data generated by sensors can be overwhelming, requiring robust storage, processing, and analysis capabilities. Integrating a digital twin with existing, sometimes outdated, systems can also be challenging. Legacy infrastructure isn’t always compatible with modern digital twins, necessitating extensive adjustments and integrations. Another challenge is the high initial cost. Developing a digital twin, especially for complex systems, involves significant investment. This includes installing sensors, building models, integrating AI capabilities, and ensuring cybersecurity. The initial expense can be a barrier for smaller organizations or those with tight budgets.

Cybersecurity is another critical concern. Digital twins rely on a constant flow of data, often transmitted through the internet or other networks. This data transmission can become a vulnerability, exposing the twin to potential cyber threats if not adequately secured. Keeping a digital twin accurate over time also requires continuous maintenance. Models must be regularly calibrated and adjusted to reflect changes in the physical object they represent, which can be a demanding and time-consuming process.

Why Digital Twins require special platforms

Developing and managing a digital twin demands a specialized platform due to the unique requirements of real-time data processing, integration, and visualization.

A dedicated platform offers several benefits:

• Data integration: These platforms can aggregate data from various sources—IoT devices, cloud databases, and other systems—seamlessly. This allows the digital twin to have a holistic view of the physical system.
• Advanced analytics: Many platforms include built-in analytics and machine learning tools, making it easier to apply AI to data analysis and predictions. • Scalability: A specialized platform can grow as the digital twin project expands, whether it’s adding more assets, increasing data points, or integrating additional processes.
• Security: Given the importance of data protection, dedicated platforms often include robust security features to safeguard sensitive information.
• Visualization & interaction: Digital twin platforms provide tools for visualizing the model and interacting with it in a meaningful way, such as through 3D interfaces or augmented reality.

These platforms streamline the complex process of building and maintaining a digital twin, ensuring the project is efficient, scalable, and secure.

Types of Digital Twins: from components to processes

Digital twins come in various forms, each catering to different aspects of a physical system:

1. Component Twins: These focus on individual parts or components of a system, like an engine or a pump. Component twins are often used to monitor the health and performance of critical parts, predicting failures before they occur.
2. Asset Twins: These model entire assets, such as a vehicle or a machine. Asset twins are used to track the overall health, efficiency, and lifecycle of the asset, providing valuable insights for maintenance and upgrades.
3. System Twins: System twins represent groups of assets working together, like a factory production line. They help optimize workflows, identify bottlenecks, and improve overall efficiency.
4. Process Twins: These are the most comprehensive, simulating entire processes or operations. Process twins allow organizations to see how different components and systems interact over time, providing insights that can streamline complex workflows and improve productivity. Each type of digital twin offers a unique perspective, from the granular level of individual parts to the broad view of entire processes.

Security considerations for Digital Twins

Security is a top priority when it comes to digital twins, as they involve constant data exchanges between the physical and digital realms. The level of security varies based on implementation, but several challenges need to be addressed:

• Data transmission: With data continuously flowing from the physical object to the digital twin, encryption is crucial. End-to-end encryption ensures that data remains secure during transmission, protecting it from interception.
• Integration points: Digital twins are often connected to other systems, like cloud platforms or analytics tools. Each connection point can be a potential vulnerability, so it’s vital to secure them with proper authentication and encryption measures.
• IoT device security: Many digital twins rely on IoT devices to gather real-time data, and these devices can be targets for cyberattacks. Ensuring they have up-to-date firmware, secure communication channels, and strong access controls is essential.

A multi-layered security approach, regular updates, and proactive monitoring are key to ensuring that digital twins remain secure while handling sensitive data.

Conclusion

Digital twins are revolutionizing industries by creating a bridge between the physical and digital worlds. They provide a dynamic, real-time view of physical systems, offering unprecedented opportunities for optimization, prediction, and innovation. However, they also come with challenges, from high initial costs to cybersecurity concerns. With the right platform, tools, and security measures in place, digital twins can offer significant advantages, helping organizations navigate the complexities of modern technology and stay ahead of the competition. As AI and IoT technologies continue to advance, the capabilities and applications of digital twins are only expected to grow, making them an integral part of the future of smart systems and connected industries.

We've featured the best IT management tool.

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

A former USDS employee called the repurposing of the Digital Service into DOGE an “A+ bureaucratic jiu-jitsu move.” It will give Musk and his associates access to unclassified data in every government agency.