Latest News - Page 652

The current Apple Watch situation is a bit of a mess if you are planning on buying one in the United States, and it shows little sign of improving any time soon. But there does appear to be scope for that to happen in the future, according to a new report.

As a reminder, the Apple Watch Series 9 and Apple Watch Ultra 2 are currently prohibited from being sold with working blood oxygen sensors as a result of a ruling that confirmed Apple infringed upon a Masimo patent in relation to the technology those wearables use. Apple was forced to disable the feature at the turn of the year, although an appeal has been lodged. Now, it appears that those who buy an Apple Watch with the blood oxygen sensor technology disabled may well be able to get it reenabled at a later date should Apple's appeal be successful.

That's according to an IP Fray report which notes that new documentation has been published which explains exactly how Apple removed the feature from the Apple Watches being sold today. According to that report, Apple disables the blood oxygen sensor at the point when the wearable is paired with an iPhone but after Masimo tested the software lock it found that it was able to bypass it on a jailbroken device. And while not everyone has a jailbroken iPhone, the fact that the software lock could be bypassed suggests that there could be hope for the future.

A new paper penned by researchers from Google DeepMind, ETH Zurich, University of Washington, OpenAI, and McGill University reveals that OpenAI and Google's AI models have been cracked open.

The new paper reveals that thirteen computer scientists from the aforementioned locations were able to launch an attack on OpenAI and Google's closed AI services, and this attack resulted in the revealing of a significant hidden portion of the underlying transformer models. More specifically, the attack revealed the embedding projection layer of a transformer model through API queries. Notably, the attack technique was originally proposed back in 2016 and has since been built upon to achieve the breaking of OpenAI and Google's AI models.

The team made Google and OpenAI aware of their infiltration, which both companies responded to by implementing mitigation techniques for that specific type of attack. Furthermore, the team decided not to publish their findings online, which would have been the exact size of OpenAI's GPT-3.5-turbo model, as this information was deemed harmful to the product since bad actors would learn aspects of the model, such as total parameter count, weight, size, etc.

Apple released the macOS Sonoma 14.4 software update recently and while it doesn't bring any huge new features, it does seem to be causing some new problems for owners of USB hubs. Many people use hubs to allow them to connect multiple devices to computers like the MacBook Air thanks to its lack of ports, while plugging HDMI monitors in can require such a hub thanks to the lack of HDMI connectivity on some laptops. That means that such hubs can be vital and having them not work as they should can be hugely problematic.

A new MacRumors report says that people have started to complain on its forums and social media like X and Reddit, reporting that their hubs aren't always working as intended. They say that accessories and devices that are plugged into those hubs are sometimes not detected and that the issue only started after they installed the macOS Sonoma 14.4 update.

Monitors seem to be particularly problematic, the report notes, with people reporting that those who have monitors connected to Macs via hubs are no longer able to make them work. Mice and other input peripherals are similarly impacted.

Google has taken to its blog to share a 2023 Year in Review for its Vulnerability Reward Program, a community-driven security effort that Google pays decent money for.

Google has rewarded 632 security researchers from 68 different countries with $10 million for all of the bugs discovered in the company's products such as Android and Google-powered devices. Notably, the maximum payout per issue was $15,000, and the biggest payout for a vulnerability report throughout 2023 was $113,337. During 2023 Google added generative AI platforms such as Gemini were added to the Vulnerability Reward Program, and throughout the year 35 reports were paid out for a total of $87,000.

More specifically, Google writes that for Android and Google's own devices it paid out $3.4 million for bugs discovered, with bugs found in its Wear OS and Android Automotive operating systems generating $70,000 across 20 critical bug discoveries. Google Chrome security researchers earned a hefty chunk of money pulling in $2.1 million for 359 vulnerability reports. If you are interested in reading more about this story, check out Google's blog post here.

If you're an iPhone owner who hasn't yet updated to iOS 17.3.1, you never will. That's because Apple has now decided to stop signing the update which means that it can no longer be installed, even if you have the IPSW file required to do so.

Whenever an iPhone has its iOS operating system updated or downgraded, the phone phones home to Apple's servers to ensure that the software is valid. If Apple says that it is, the installation can continue. But if Apple says otherwise, the update will not install and there is no way to get around that. That's what has happened with iOS 17.3.1 which means that the only version of iOS currently available to iPhone owners is the latest iOS 17.4 release.

Apple regularly stops signing older versions of iOS in an attempt to ensure that as many people as possible are using the latest and greatest version of its software. There are a few reasons that it might want to do that.

VR gaming is here to stay, but the technology still feels like it needs a few elements to become a 'must-have.' That said, playing in a true virtual space can be transformative even while using the most affordable solutions. As fans of the technology, we've followed the development and progress of Virtuix's Omni One, which delivers a "full-body VR gaming system."

Per the name, the Omni One is an omnidirectional treadmill that gives you 360-degree movement so you can walk or even run in a virtual environment without having to worry about what's going on in the real-world. Freedom of movement for VR gaming is one of the reasons why thousands have pre-ordered the complete Omni One system (priced at $2,595), which includes a customized Pico 4 Enterprise headset.

With all pre-orders to be fulfilled during Q1 and Q2 of this year, the Omni One system is set to launch in late Q2 2024. With that, Virtuix has announced the full launch lineup for the Omni One, 35 games in total, including the well-received Ghostbusters: Rise of the Ghost Lord and Breachers.

For many people around the world home internet connection speeds are a problem, but scientists now believe they can solve this issue and it all has to do with a new chip design.

It's likely that people have put together that Wi-Fi connections slow down when there is more devices connected to that specific network, and this same principle applies for even larger networks such as ones provided by cell towers. An example of this is terrible connection speeds, or a lack of connection at sporting events. These rudimentarily called wireless traffic jams, but now researchers have an idea of solving it.

Typical Wi-Fi networks are powered by a "planar" processor, which produces signals at a flat or two-dimensional range. Due to these signals being emitted in a flat range there is a limited number of frequencies, or connections that can pair with the network at any given moment. To solve this researchers have proposed creating a 3D Wi-Fi chip that would enable the hardware to handle multiple frequencies at the same time, significantly reducing the congestion on the network, thus increasing speed.

A team of researchers from the University of California, Riverside, say they have stumbled onto a groundbreaking PC breakthrough technology that may be able to give up to twice as much performance to all existing PCs.

The team of researchers presented their findings, which have been published online, detailing a new system called simultaneous and heterogeneous multithreading, or SHMT. So, how does it work? Modern computers are currently made up of multiple processors, such as a CPU, GPU, and TPU. Due to different parts of the PC being dedicated to carrying out different workloads through instructions, there are bottlenecks and latency between the data moving between each of the different processors.

However, SHMT can solve this issue, as the technology enables a single task to be carried out across multiple processes simultaneously. Not only does this reduce the workload per processor, but it also speeds up the efficiency of the task and costs less energy to run. According to the recently published paper the team found that SHMT implementation in tests resulted in processing speeds increasing up to 1.95x, with energy usage dropping by close to 51%.

In just over 10 years the Moon will be home to a nuclear reactor that is completely build without humans, and it will be used to power a lunar base.

Russia's space agency Roscosmos has announced its considering sending an autonomous nuclear reactor to the surface of the Moon in a joint effort with China National Space Administration (CNSA). Both space agencies plan on constructing a lunar base together called the International Research Station (ILRS), which they claim will be open to "all interested countries and international partners".

However, it isn't likely that US astronauts will be able to visit the ILRS due to growing tensions between the US and China and the recent sanctions placed on Russia following its invasion of Ukraine. Roscosmos announced on March 5 that it may build a nuclear reactor with the CNSA and it will be delivered to the lunar surface "somewhere at the turn of 2033 - 2035," according to Roscosmos director general Yury Borisov.

Nintendo Switch 2 rumors, speculation, and inside information have been popping up everywhere lately. The latest confirms what many have been saying about the upcoming console: it will support NVIDIA's AI-powered DLSS rendering, boosting performance while maintaining next-gen image fidelity.

This comes from YouTuber Moore's Law is Dead, who states in a recent video that "most" games in development for the Nintendo Switch 2 will use DLSS. However, it won't be mandatory. That last bit is what we would have expected to see, as specific titles like 2D games would probably not have any need for leveraging upscaling technology.

Where it gets interesting is the news that Nintendo passed on using NVIDIA's cutting-edge Ada Lovelace architecture for the Switch 2 in favor of a more affordable and cost-efficient Ampere chip. According to Moore's Law is Dead, some Lovelace features, like efficiency tweaks, were added to the custom chip.

Late last year, first-party PlayStation developer Insomniac Games was the victim of a significant hack and data breach that saw details on upcoming games (including a playable section of the studio's Wolverine game) and internal Sony information. The hack revealed that the studio was also working on an unannounced multiplayer Spider-Man game called Spider-Man: The Great Web - which was canceled.

Alongside the cancelation of Naughty Dog's multiplayer The Last of Us project, studio closures, and skyrocketing development costs, Sony is seemingly moving away from focusing on multiplayer always-online live service games. A trailer for Insomniac's Spider-Man: The Great Web leaked overnight, showcasing a game that looks pretty far along development-wise.

The trailer shows co-op gameplay reminiscent of the Into the Spider-Verse animated movies, where various web slingers team up to tackle crime and villains in the game's open-world New York City. The trailer also shows the franchise's female web-slinger, Gwen, as a playable character. There even looks to be a Venom-style playable character.

NASA has shared an incredible image on its social channels as a reminder of the awesome power of the Hubble Space Telescope, a space telescope that is more than 30 years old.

The Hubble Space Telescope captured a star cluster officially called NGC 6611, and the above image is from the space agency's "Hubble Classic" catalog, with it originally uploaded to the database on June 26, 2018. However, NASA has taken to its social channels to repost the image and explain via a link to its blog that NGC 6611 is an open star cluster that formed approximately 5.5 million years ago within the Eagle Nebula, or Messier 16.

In astronomical age, NGC 6611 is considered to be a "very young cluster" as it contains many hot blue stars that cause the Eagle Nebula to glow brightly. NASA writes on its blog that the Eagle Nebula is predominately made of hydrogen gas that will eventually disperse and will be assisted by shockwaves caused by supernova explosions as the young stars end their lives.

It appears as though there are some major issues in the US with South Korean companies in Samsung, LG, and SK hynix concerned over US operations.

The Korea Times has a new report that said "concerned about their investments in the United States amid a spike in construction costs and lingering subsidy uncertainties ahead of the upcoming presidential election in the world's largest economy," according to officials and analysts on Tuesday.

Samsung Electronics invested a rather chunky $17 billion to construct a new chip factory based in Taylor, Texas, which will begin mass production of semiconductors usef for fifth-generation network systems or AI solutions in the second half of the year. But, because of soaring material and labor costs in the United States, Samsung is expected to pony up as much as $8 billion for the construction of the plant.

Micron was the first to announce mass production of its new ultra-fast HBM3E memory in February 2024, seeing the company ahead of HBM rivals in SK hynix and Samsung... leaving its HBM competitors shocked.

The US memory company announced it would provide HBM3E memory chips for NVIDIA's upcoming beefed-up H200 AI GPU, which will feature HBM3E memory, unlike its predecessor with the H100 AI GPU, which featured HBM3 memory.

Micron will make its new HBM3E memory chips on its 1b nanometer DRAM chips, comparable to 12nm nodes, which HBM leader SK Hynix is using on its HBM. According to Korea JoongAng Daily, Micron is "technically ahead" of HBM competitor Samsung, which is still using 1a nanometer technology, which is the equivalent of 14nm technology.

Crucial has just announced its new 12GB DDR5 memory DIMMs for both desktop PCs and laptops, which sits between 8GB and 16GB DIMMs.

We already have the usual 8GB, 16GB, and 32GB memory sticks, but also 24GB, 48GB, and 64GB DIMMs if you want oodles of RAM. Crucial's introduction of 12GB DDR5 memory DIMMs offers a 50% capacity boost over 8GB, offering 24GB in dual-channel DDR5 configurations (versus 16GB if you used 2 x 8GB sticks).

Crucial's new 12GB DDR5 memory is already listed on the Amazon UK website, spotted by industry insider "Momomo_US" on X. Crucial's new 12GB DDR5 sticks have a part number of "CT12G56C46S5" with speeds of DDR5-5600, but it can be downclocked to 5200 and 4800 if required.

It took a few years for Cyberpunk 2077 to live up to its initial hype and potential. In addition to fixing a litany of technical issues with the game, developer CD Projekt RED has spent the time since the game's 2020 debut adding new features, overhauling systems, and creating a more worthy follow-up to the iconic The Witcher 3.

Culminating with the release of the critically acclaimed Phantom Liberty expansion in 2023, the team seems ready to stop development and shift its focus toward the sequel.

Cyberpunk 2077 game director Gabe Amatangelo spoke with veteran journalist Stephen Totilo, telling him, "We're done." With the surprise addition of a functioning metro system late last year and things like custom paint jobs for vehicles in 2024, players shouldn't expect more new content.

Oh great, if Helldivers 2 wasn't already hard enough... flying bugs are coming, it seems. In a new video shared by "Projectpatdc" on the Helldivers 2 subreddit, we're seeing flying bugs... and I'm scared.

But it's the response from Helldivers 2 developer Arrowhead and its CEO that has me in stitches, Arrowhead CEO Johan Pilestedt tweeted: "I've heard rumors of flying bugs in Helldivers 2. I want to officially refute such preposterous claims. Everyone knows that 'Bugs can't fly.' And I'm not alone in thinking this. The Ministry of Truth agrees that this is propaganda from bug sympathizers that want to brainwash good people".

The evidence is there: flying bugs are taking to the skies in Helldivers 2, but the developer is being coy about it and I love it. You can clearly see in the official Helldivers 2 box art that there are flying bugs, but the developer is keeping hush-hush about it, referring to the bugs as propaganda. So good.

The Qualcomm Snapdragon X Elite could present a significant shake-up for efficient Windows 11 notebooks and laptops. The Arm-based SoC (CPU, GPU, and NPU) is closer to Apple's M3 chip powering MacBooks. Still, as it will power Windows 11 devices, it will directly compete with Intel's new Core Ultra 'Meteor Lake' processors and AMD's Ryzen 'HawkPoint' - two x86 chipsets - especially now, in the year of the AI PC.

Although a single example, one of Turkey's most well-known tech journalists has put the Snapdragon X Elite X1E80100 28 Watt SoC to the test, comparing it with the x86 'Meteor Lake' Intel Core Ultra 7 155H processor. How does it fare? Surprisingly well.

When testing the chip's NPU or AI capabilities, the Snapdragon X Elite leaves the Intel Core Ultra in its wake, scoring 1720 points in the UL Procyon test compared to 476 points for the Intel Core Ultra 7 155H. AI testing is never like-for-like, as the Intel chip in this test used OpenVINO while the Snapdragon X Elite used the Qualcomm SNPE SDK.

AIDA64 is a powerful tool for enthusiasts. It offers detailed system information covering hardware and software, as well as 64-bit benchmarks and stress tests to push everything from memory, storage, CPUs, and GPUs to their limit. For those who want to be in complete control or monitor their PC, AIDA64 can be invaluable.

Developer FinalWare has just dropped AIDA64 v7.20, a notable update that brings some welcome additions and updates to the app. First up, on the purely cosmetic front, there's a new Dark Mode with three different themes designed to match Windows Dark Mode settings. If you're a fan of Dark Mode for apps and browsing, like me, this is excellent news.

AIDA64's massive library has been expanded on the hardware front to support the new GeForce RTX 40 SUPER GPUs from NVIDIA and the Radeon RX 7600 XT from AMD.

A team of NHS surgeons at private Cromwell Hospital in London used Apple's new Vision Pro headset during an operation, a first for the UK.

The team at Cromwell Hospital used the new Apple Vision Pro during an operation on a male patient and his spine, navigating the surgery using augmented reality. A scrub nurse used the Apple Vision Pro headset to keep track of the process and choose the right tools.

The AR headset allowed the nurse to see the real-world with virtual screens displaying important information, with Suvi Verho, the lead scrub nurse at London Independent Hospital, told the Daily Mail that Apple's Vision Pro headset was a "gamechanger" and that "It gives you confidence in surgery," adding "It eliminates human error. It eliminates the guesswork".