Info-Direkt

🔒
❌ About FreshRSS
There are new articles available, click to refresh the page.
Before yesterdayTeknik

NASA is planning a huge computing upgrade for its next space missions

NASA has selected Microchip Technology to develop a High-Performance Spaceflight Computing (HPSC)  processor it says will provide at least 100 times the computational capacity of current spaceflight computers. 

NASA's spaceflight computing may be in dire need of an upgrade, as current systems "were developed almost 30 years ago" according to Wesley Powell, NASA’s principal technologist for advanced avionics, and "future NASA missions demand significantly increased onboard computing capabilities and reliability".

Microchip Technology, based in Chandler, Arizona, specializes in the manufacture of microcontroller, mixed-signal, analog, and Flash-IP integrated circuits and has been in operation for over 30 years.

How will the project work?

Microchip will work to design and deliver the HPSC processor over the next three years, with the goal of employing the processor on future lunar and planetary exploration missions.

The work will take place under a $50 million firm-fixed-price contract, with Microchip contributing significant research and development costs to complete the project.

Microchip’s new processor architecture reportedly offers the flexibility for the processing power to "ebb and flow" depending on current operational requirements, meaning certain processing functions can also be turned off when not in use, reducing power consumption. 

NASA says the capability will save a large amount of energy and improve overall computing efficiency for space missions. 

The new processor could have its own applications down here on earth as well.

NASA says the new tech may also be useful for commercial systems on Earth that require similar mission-critical edge computing needs as space missions, and need to be able to safely continue operations if one component of the system fails. 

These potential applications apparently include industrial automation, time-sensitive ethernet data transmission, artificial intelligence, and even Internet of Things gateways, which bridge various communication technologies.

High-Performance computing is obviously something that's on NASA's radar, as it recently deployed Azure Quantum as part of a project to reduce the time it takes to transmit instructions between spacecraft from hours to just minutes.

  • Looking to soup up for own computing power? Check out our guide to the best workstations

14th-gen Intel chips could support ray tracing – and truly bring it to the masses

Intel Meteor Lake processors, the next-next-generation CPU from Intel due out in 2023, are expected to feature ray tracing support on a tiled-GPU architecture, marking a major step forward for the technology.

Ray tracing is a rendering process that lights up a rendered scene by mimicking the way we actually see light. This sounds almost silly (don't we see all light?), but making an artificial scene look natural requires calculating the intricate physics involved in how light behaves, and this is a genuinely daunting task from a computational perspective.

As it stands, ray tracing technology is currently only viable in gaming PCs with the best graphics cards or the best gaming laptops with discrete GPUs, making it a revolutionary technology that remains locked away in more "elite" applications on high-end hardware.

Intel's bringing the technology to its "integrated" Intel Xe graphics platform is a major advance in the technology that could make it far more accessible, however. We put integrated in quotes here since – as Wccftech points out – it's not quite the same thing as current-gen integrated graphics; it's actually a tiled architecture that's more like a system-on-a-chip than traditional integration into a single CPU die.

What matters though is that this is the kind of chip that will be going into standard laptops like ultrabooks or even the best Chromebooks that feature beefier specs like the one in our HP Elite Dragonfly Chromebook review.

Intel is already working on its own graphics upscaling tech to pair with its somewhat-beleagured Intel Arc desktop graphics cards, and that tech could easily make its way into Intel's 14th-gen processors. If so, it could make some of the best cheap laptops very capable 1080p gaming machines.

Could Intel Meteor Lake bring next-gen graphics to everyday users?

Ray Tracing in Shadowlands

(Image credit: Blizzard Entertainment)

While we haven't even gotten our hands on Intel Raptor Lake chips yet – those are expected to launch as soon as October of this year – the first Intel Meteor Lake chips could land as soon as late 2023. If that's the case, you could be playing something like Cyberpunk 2077 with modest settings and some ray tracing on a base-spec Dell XPS 13 at a playable frame rate.

This isn't to say that a single processor is going to turn even the best Ultrabooks into true gaming laptops. But with the advances in GPU technology that comes with something like ray tracing, it will have ancillary benefits for overall performance and – with proper settings tweaks (and upscaling technology) – you won't be restricted to playing light-duty or casual games like Civilization VI on a basic laptop.

As for desktops, this will be less of an advance, since most desktops have a free PCIe slot to slap in one of the best cheap graphics cards around capable of ray tracing, and these will run circles around the integrated graphics of even the best processors in Intel's 14th-gen lineup.

So this is mostly an issue for laptops, which is the fastest growing segment of the computer market by far. Traditional laptops have tended to be fairly light on gaming performance since this dramatically improves energy efficiency (i.e., battery life) and keeps costs down.

By integrating a ray-tracing GPU into the standard processor package for these laptops, we're going to see much more powerful performance across a whole range of graphics-intensive applications from video streaming to gaming at a much more affordable price point.

We've gotten a taste of this kind of performance already, and it is truly impressive

An Asus Zenbook S 13 OLED on a laptop stand

(Image credit: Future)

We already saw something similar in our Asus ZenBook S 13 OLED review, which is one of the first laptops to feature an AMD processor with integrated RDNA 2 graphics. 

While the gaming performance wasn't mind-blowing in absolue terms, the fact that you could even run Port Royal on an Ultrabook is a major advance. Unfortunately, finding the ZenBook S 13 OLED is a frustrating challenge, something that wouldn't be the case with a laptop running Intel Meteor Lake.

While it will be some time yet before we see these laptops, it's good to know that they are coming, and it could offer consumers far more performance without having to make total sacrifices for affordability like they currently have to do. This can only be good for everyday consumers, whether you're a serious gamer or not.

Tesla's in-house supercomputer is something special - but the next will be even better

Tesla's in-house supercomputer has received an additional 1,600 GPUs, a 28% increase on the figure quoted a year ago.

Tesla Engineering Manager Tim Zaman claims this would place the machine 7th in the world by GPU count.

The machine now features a total of 7,360 Nvidia A100 GPUs, which are built specifically for data center servers, but utilize the same architecture as the company's top-of-the-line GeForce RTX 30-series cards. 

Telsa supercomputer upgrade

It's likely Tesla needs all the processing power it can get right now. The company is currently working on 'neural nets', which are used to process the vast quantities of video data that the company's cars collect.

The latest upgrade may be just the start of Tesla's high-performance computing (HPC) ambitions.

In June 2020, Elon Musk said "Tesla is developing a neural net training computer called Dojo to process truly vast amounts of video data", explaining the planned machine would achieve a performance of over 1 exaFLOPs, which represents one quintillion floating-point operations per second, or 1,000 petaFLOPs.

Performance of over 1 exaFLOPs would place the machine among the most powerful supercomputers worldwide, as only a few current supercomputers have officially exceeded the exascale barrier, including The Frontier supercomputer at the Oak Ridge National Laboratory in Tennessee, United States. 

You might even be able to get a job building the new computer. Musk asked his Twitter followers to "consider joining our AI or computer/chip teams if this sounds interesting".

Dojo won't be reliant on Nvidia hardware, however. The planned machine is set to be powered by Tesla's new D1 Dojo Chip, which the carmaker said could have specifications of up to 362 TFLOPs at its AI Day event.

  • Want to run your own AI research in the cloud? Check out our guide to the best cloud hosting

Via Tom's Hardware

Entire 13th-gen Intel Raptor Lake starting lineup specs leak online

With Intel Raptor Lake set to release in the coming months, it looks like the entire launch lineup for the next-gen processor series from Team Blue have leaked ahead of their announcement.

A user going by the name Extreme Player on the Chinese social media site Bili Bili posted the specs for 14 SKUs in total, covering Core i3 to Core i9 processors (first flagged by Wccftech).

The topline numbers show a 24-core i9 processor (eight performance cores and 16 efficiency cores) with 32 threads,which doubles the number of efficiency cores as the flagship Alder Lake i9-12900K. The rest of the specs are below, as posted on Bili Bili.

A table showing the different specs for the initial Intel Raptor lake processor lineup

(Image credit: Extreme Player / Bili Bili)

Obviously, these specs haven't been announced by Intel, so they need to be taken with a grain of salt, but the specs are in line with other rumors we've heard.


Analysis: Can Intel rein in some of its power?

One of the things about the big.LITTLE architecture pioneered by ARM is that it is supposed to be substantially more energy efficient than traditional pocessor architecture.

This is definitely the case with Apple silicon powering the best MacBook and Macs, as well as the best phones and best tablets on the market utilizing these kinds of processors.

It hasn't turned out to be the case with Intel Alder Lake though, which uses a lot of power compared to other chips released in recent years, which has been especially true of its mobile chips.

I'll be especially interested to see if Intel can make their chips more power-conscious with this generation, but given some of the benchmark scores we've seen lately, I think it's likely to be full steam ahead.

The rise of SSDs is pushing the hard drive market closer to the brink

The combined effect of a drop-off in PC demand and the pace of SSD adoption led to a significant fall in hard drive shipments last quarter, a new report suggests.

Data from analyst firm Trendfocus shows HDD shipments plummeted 15.4% in Q2 on a quarter-to-quarter basis, with unit sales among vendors like Toshiba dropping by as much as 17.3%.

Although the picture looks a little less bleak when it comes to performance-focused enterprise drives and nearline units (which bridge the gap between online and offline storage), demand for most HDD segments fell by doubt-digit percentages across the period.

The death of the HDD?

Commentators have long exaggerated the immediacy of the threat to the hard drive market posed by SSDs. Although HDDs are dramatically slower, the cost per capacity remains lower, which means organizations that need to store massive volumes of data still stand to benefit. HDDs are still relatively common in non-premium consumer devices too.

However, as solid state drives become faster, cheaper and more capacious, the number of use cases for the HDD is undoubtedly shrinking.

Earlier this year, for example, Micron became the first company to ship 176-layer QLC NAND flash at volume, a development that has the potential to bring SSDs to even the cheapest laptops.

Microsoft is also preparing to force PC manufacturers to abandon HDD boot drives, presumably in an effort to increase the level and consistency of performance across Windows 11 hardware. Although the ban applies to boot drives exclusively, dual-drive systems are a rarity, which means HDDs will effectively be pushed to the fringes of the PC market.

In an enterprise context, meanwhile, it is expected that the maximum SSD capacity will jump significantly (to perhaps 400TB) off the back of new technologies, closing the gap on hard drives from a cost per capacity perspective.

HDDs are also being squeezed from the opposite direction; increases in the capacity of magnetic tape (the current generation, LTO-9, has a native capacity of 18TB) means the case for using hard drives for archival purposes is increasingly weak.

Although analysts maintain businesses are best served by maintaining a balanced storage stack comprised of tape, HDDs and SSDs, which should cover off every use case in the most economical manner, the hard drive will only find itself in an increasingly perilous position.

Via Storage Newsletter

Apple Safari patched to fix potentially dangerous zero-day flaws

Apple has moved fast to patch its Safari browser against a serious security vulnerability that is affecting a number of its operating systems.

Safari 15.6.1 for macOS Big Sur and Catalina is available to download now, with anyone using those versions advised to upgrade immediately.

The fix for CVE-2022-32893 patches an out-of-bounds write flaw in WebKit, the engine of Safari that is also used by other apps with web access.

Out of bounds write flaw

Apple has confirmed the flaw is allegedly already being exploited in the wild, and when abused, the flaw allows threat actors to execute remote code on a vulnerable device, remotely.

"Processing maliciously crafted web content may lead to arbitrary code execution. Apple is aware of a report that this issue may have been actively exploited," Apple said in a security advisory

An out-of-bounds write flaw happens when a threat actor forces an input program to write data before the beginning, or after the end, of the memory buffer. That crashes the program, corrupts the data, and allows threat actors to remotely execute code. The fix for Big Sur and Catalia is in the same vein as the one for Monterey - through improved bounds checking. 

Given that the flaw is being exploited in the wild, Apple is staying tight-lipped on the issue until most endpoints are patched. 

The company said it had been tipped off to the flaws by an anonymous user, adding that it had now improved its bounds by checking for both bugs.

Apple has had its hands full fixing zero-days this year. In January 2022, it fixed two such flaws, namely CVE-2022-22578, and CVE-2022-22594, which allowed arbitrary code execution with kernel privileges. 

A month later, it fixed another zero-day, affecting iPhones, iPads, and Macs, and allowing threat actors to crash the OS and run remote code execution, and in March, Apple patched CVE-2022-22674, and CVE-2022-22675, two zero-days abused to execute code with Kernel privileges.

Via: BleepingComputer

Malicious PyPi packages turn Discord into password-stealing malware

Python developers are under attack once again, with attackers looking to steal Discord account details along with data stored in various browsers

Cybersecurity researchers from Snyk have recently spotted a dozen malicious packages, uploaded to PyPi, the biggest Python code repository out there, with more than 600,000 active users. 

The packages were uploaded almost a month ago, by a threat actor called “scarycoder”. They claim to provide the users with various functionalities, Roblox tools, thread management, and others. Instead, the researchers have found, all the packages do is steal sensitive information.

Stealing passwords 

Different packages are capable of stealing different things. Some are focused on data stored in browsers such as Google Chrome, Chromium, Microsoft Edge, Firefox, and Opera. The data includes stored passwords, browser history, cookies, and search history. Others are installing backdoors directly into the Discord client, stealing authentication tokens, Nitro status, billing information, and credit card data.

One of the malicious programs attacks Roblox, it was further said, stealing account cookies, user IDs, Robux balance, and Premium status. 

PyPi’s administrators are relatively slow to respond, the publication states, adding that it’s probably not due to negligence, but rather due to the fact that the entire project is run by a handful of volunteers who simply can’t keep up with a tidal wave of malware uploads. 

Still, the slow response means many of Python developers will remain exposed to various viruses, malware, and other forms of attacks.

Experts from Spectralops recently found 10 malicious packages on the PyPi platform. All of these were given names that are almost identical to the names of legitimate packages in order to dupe developers into downloading, and adopting, the tainted ones. The practice is called typosquatting, and it’s quite a common occurrence in the developer community.

Via: BleepingComputer

Google says it has blocked another huge DDoS attack

Google claims to have stopped on of the largest Distributed Denial of Service (DDoS) attacks ever seen.

In a blog post, the company's Senior Product Manager for Cloud Armor, Emil Kiner, and Technical Lead Satya Konduru, said its tool stopped a Layer 7 HTTPS DDoS attack that peaked at 46 million requests per second (rps), making it 76% larger compared to the previous record-holder. 

“To give a sense of the scale of the attack, that is like receiving all the daily requests to Wikipedia (one of the top 10 trafficked websites in the world) in just 10 seconds,” the blog explained.

Tor exit nodes used

The attack reached its peak some ten minutes in but lasted more than an hour (69 minutes). The researchers speculate that the attackers stopped when they saw that their efforts weren’t producing the desired outcome. 

From the technical side of things, it seems the botnet used in the attack was relatively powerful. All in all, 5,256 source IPs were used, originating from 132 countries. 

The attack used encrypted requests (HTTPS), meaning it took extra computing resources to generate - it was quite an expensive endeavor. Almost a quarter (22%) of all source IPs (1,169) corresponded to Tor exit node endpoints, although their request volume represented just 3% of all attack traffic. 

“While we believe Tor participation in the attack was incidental due to the nature of the vulnerable services, even at 3% of the peak (greater than 1.3 million rps) our analysis shows that Tor exit-nodes can send a significant amount of unwelcome traffic to web applications and services,” they added.

The top four countries contributed almost a third (31%) of the total attack traffic. 

Google’s experts could not definitely confirm the threat actor behind the attack, but are under the impression that this was the work of Mēris, given that the geographic distribution and the types of unsecured services leveraged in the attack match its patterns.

That Coinbase job offer could actually be North Korean hackers

Experts have warned that the dangerous Lazarus group is now targeting Web3 developers on Mac devices. 

The North Korean state-sponsored threat actor recently went after blockchain developers with fake lucrative job offers that turned out to be nothing more than infostealers and malware

While these attacks were limited to Windows users at first, cybersecurity researchers from ESET have now discovered they are expanding into Apple territory, too. 

Intel and Apple chips attacked

The campaign is pretty much the same for both platforms. The group would impersonate Coinbase, one of the largest and most popular cryptocurrency exchanges in the world, and reach out to blockchain developers via LinkedIn and other platforms with a job offer. After a little back-and-forth, and a few rounds of “interviews”, the attacker would serve the victim what seems to be a .pdf file with the job position’s details.

The file’s name is Coinbase_online_careers_2022_07, and while it looks like a .pdf (icon and all), it is actually a malicious DLL that allows Lazarus to send commands to the infected endpoint. The file is compiled for Macs with both Intel and Apple processors, the researchers further discovered, suggesting that the group is after both older, and newer device models. 

Detailing the attack via Twitter, the researchers said the malware drops three files: the bundle FinderFontsUpdater.app, the downloader safarifontagent, and a decoy PDF called “Coinbase_online_careers_2022_07.pdf”. 

Lazarus Group is no stranger to fake job offer attacks, and it’s conducted these attacks in the past with much success. In fact, one of the largest cryptocurrency heists in history, the $600+ million-heavy attack on the Ronin bridge, was done in that exact manner. 

After reaching out to a software engineer and luring him into downloading the fake .pdf file, the attackers from Lazarus found their way into the system, obtained the necessary credentials, and siphoned out millions in cryptocurrency tokens.

In this case, however, the malware was signed on July 21, with a certificate issued to a developer going by the name Shankey Nohria. The team identifier was 264HFWQH63. While the certificate had not been revoked on August 12 when it was checked, BleepingComputer reports, the researchers did find that Apple didn’t scan it for malicious components. 

Via: BleepingComputer

❌