Feed aggregator

APYs are falling, so the sooner you open one of these top CDs, the more interest you stand to earn.

Mobile carriers argue that locking devices helps them lower costs. Consumer protection groups see it a little differently.

Entri, a software provider, accuses GoDaddy, a major domain registrar, of anti-competitive practices, interference with business contracts, and limiting market choices. The court’s recent denial of GoDaddy's motion to dismiss the case signals that Entri has presented plausible claims, setting the stage for further legal proceedings. After the recent ruling by the court, Abe Storey, founder and CEO of Entri spoke to TechRadar Pro, stating:

“We're glad the court recognizes that GoDaddy's attempt to deprive their customers of the freedom to use their preferred software is precisely the harmful, anticompetitive actions the Sherman Act forbids. GoDaddy's continued efforts to regulate a web domain's use through conduct outside its role as a registrar endangers the open internet as we know it today.”

The case Entri, LLC v. GoDaddy.com, LLC reflects a complex legal battle in the U.S. District Court for the Eastern District of Virginia, revolving around competition in the domain registration and configuration market.

Judge approves Entri’s major points for trial

GoDaddy holds a dominant position in the U.S. domain registration market, controlling approximately 40% of all registered domains. Traditionally, users who lease domain names can configure them manually or via third-party services using Domain Connect, GoDaddy’s integration protocol. In 2021, Entri introduced Entri Connect, a software solution that automates DNS record configuration, offering smoother integration for SaaS providers and their users. Entri Connect quickly gained traction, with many third-party users preferring it over GoDaddy's Domain Connect.

Initially, GoDaddy allowed Entri Connect to operate within its ecosystem. However, tensions arose when GoDaddy introduced new policies in 2023 that prohibited the use of aggregator services like Entri Connect, restricting users to either Domain Connect or manual configuration. Entri claims that GoDaddy’s actions disrupted its business by terminating existing contracts and discouraging potential deals.

Entri argues that GoDaddy's policies violate Section 1 of the Sherman Antitrust Act by instituting a negative tying arrangement. This restriction forces users to abandon competing products like Entri Connect, reducing market competition. Entri also claims GoDaddy deliberately interfered with its business contracts by notifying SaaS companies about its restrictive new policies, leading some companies to cancel or modify existing agreements.

Anthony J. Trenga, Senior US District Judge, said, "GoDaddy’s negative tie cuts off all aggregator services, including Entri Connect, from competing in the DNS records configuration market. Instead, the only option is choosing between GoDaddy’s Domain Connect protocol or requiring users to manually update their DNS records. This deprivation of choice is precisely the kind of anticompetitive harm that the Sherman Act forbids"

GoDaddy’s market dominance allows it to set terms that impact smaller players like Entri, demonstrating the fine line between legitimate business practices and anti-competitive behavior. This case serves as a critical reminder that while companies can set policies to protect their platforms, they must not use those policies to stifle competition unlawfully.

More from TechRadar Pro

• These are the best cloud backup services
• AI is becoming increasingly vital in software development
• Take a look at some of the best internet security suites

Striking machinists at Boeing vote Wednesday on a new contract offer. Even before the strike, the company was grappling with production and quality control problems that led to billions in losses.

(Image credit: Jason Redmond)

Don't earn a paltry APY with a standard savings account when the top HYSAs still earn up to 5.25% APY.

Malcolm Campbell remembers Don Lawson, the high school civics teacher who helped him understand the true value of education.

(Image credit: Malcolm Campbell)

Vice President Harris is banking on the support of young voters. But new polling shows potential roadblocks for Harris in replicating the historically high youth support President Biden received in 2020.

(Image credit: DOMINIC GWINN/Middle East Images/AFP via Getty)

With seven states up for grabs, here are the likely scenarios that would put either candidate in the White House.

(Image credit: Spencer Platt/Getty Images)

A new Biden administration effort to raise Head Start teacher wages could force the federally funded preschool program to serve fewer children.

(Image credit: Cory Turner)

Braddock, Pennsylvania lost most of its population over the past almost 100 years, due largely to U.S. Steel's decline. Here's how some locals feel about the company's potential sale to the Japanese.

(Image credit: Nate Smallwood for NPR)

In the long history of computer crime, the players, goals and tactics have seen a lot of change.

Early computers were fairly isolated systems reserved for niche applications, mainly in academic environments. The first instances of security "attacks" were examples of tinkering that went too far rather than malicious activity.

Today’s world is different. Computers power many aspects of our day-to-day lives. They are faster than ever and highly inter-connected. They are in our pockets, homes and offices, but also in our toothbrushes and refrigerators. They even power our critical infrastructure. This now widespread reliance on computers (and the data they process) attracts new kinds of malfeasants.

Over time, computer-based crime has become organized. What started as low-tech cons and scams, or clever technical feats by small groups has been gradually replaced by more professionalized, more damaging, and more hurtful collectives, such as state-sponsored groups. There is one sort of attack that illustrates this transition better than most: ransomware.

The simple effectiveness of ransomware

Ransomware is an extremely lucrative example of computer crime going "corporate": incentivized by the will of making more money by investing less effort.

Most ransomware attacks follow a simple pattern: 

1. They start by running a malicious tool, an encryptor, on the target system. True to its name, the encryptor will then encrypt the whole disk (or disks) and delete the key. If the perpetrators intend to make the data recoverable, they will keep a copy of the key on their files, away from the affected system. 

2. Then, they make their presence known, from red screens to timers. Ransomware campaigns go great lengths to communicate with their victims because they get their money only if the victims believe that paying is the best chance they have to recover their data. 

3. After payment, an "honorable" ransomware gang will provide the victim a decryptor tool with the secret key.

There are some instances of ransomware that do not encrypt the data. Instead, the attackers threaten the victims by disclosing data publicly, which could cause embarrassment or leak industrial secrets.

Challenging attackers

However, with ransomware attacks, there are two steps that are somewhat challenging for the attackers:

Challenge #1: Getting the encryptor into the target system. Unfortunately, attackers can (still) benefit from a very simple tactic: asking nicely. Phishing attacks are popular ways of distributing ransomware encryptors because many victims eagerly click links on emails without verifying the origin or giving it a thought. Technical entry points traditionally used to deliver malware remain a useful alternative: if there is an open file share, the attacker can deploy the file into the target system, then find another vulnerability to execute it. WannaCry, the attack considered by many as the most damaging ransomware campaign to date, is an example of this.

Challenge #2: Receiving the ransom payment without betraying the attacker's identity. Fifteen years ago, this challenge alone would have hindered the expansion of ransomware gangs. They would need to pay in cash, which is hard to scale and would be geographically restricted to the area of influence of the gang, or they would need to rely on digital payments and withdraw the money fast, creating a trail of evidence leading directly to the gang. However, the rise of cryptocurrency presented a solution to this challenge.

While authorities have succeeded in tracking down malicious businesses who took ransom in cryptocurrencies, the international availability of a means of payment that is not linked to an actual identity has made it much easier for criminals to receive their payments and much harder for law enforcement to follow the tracks.

Preventing disruption using backups

Many of the mechanisms that help to prevent ransomware attacks involve general practices that also help to prevent various types of cyber-attacks. Awareness training supports by warning employees about clicking random suspicious links, hardening at the network and operating system level, deploying updates quickly, malware scanning, etc.

There is also great importance in building a sturdy resilience plan, underpinned by a well-defined and tested backup strategy. Of course, backups are a usual control against accidental data loss and conventional disruptive hacking like, say, website defacement. You detect the incident, roll back your data or your environment to a certain previous point in time, and get back to business with (ideally) minimum data loss.

This backup model relies on a few assumptions. To put it simply, it expects backups to work (to contain enough information to allow for a clean rollback) and to be valid (the rollback would clean up any damage made by the attacker). Reality often challenges both assumptions.

Many companies have backup processes in place. Fewer have data recovery plans describing what to do with the backups to return to a working state. Only a small minority of companies test regularly those backups to ensure that they can, in fact, be relied upon. This makes the recovery process clunky and often unsuccessful.

Ransomware attacks also challenge the second assumption. For example: if the backups are hot (that is, constantly connected to the target system), the encryptor could also encrypt the backup drives, rendering the backup unusable. Or the encryptor could be installed at a certain point, stay idle for a few months, then encrypt the data. A backup taken after the initial compromise could recover the data of the system, but could restore an infected state, allowing a reinfection to occur.

To summarize: a robust backup strategy needs to rely on both hot and cold backup locations, sufficiently isolated from each other to keep an attack on the main system from spreading undeterred to the backups, both of which are regularly and rigorously tested. If the downtime requirements of a given system are particularly stringent, the ability to get back up with minimum data loss must be part of those tests.

Wrapping up

At a technical level, ransomware is not a terribly novel threat. The disruptive aspect of it lies in the economic incentives it introduces, leading to more organized criminal structures with the freedom to act more ruthlessly and at a larger scale, and to attack sensitive industries with the hope of maximizing their payment. It is a threat worth considering because it is increasingly prevalent and, for companies caught unprepared, could wreak havoc on their infrastructure. Just remember: do not pay the ransom.

Check out the best cloud antivirus.

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

There are plenty of internet providers in Irving. But based on our research, these are our top picks in the area.

Canada's deputy prime minister said she believes Justin Trudeau has the support of a majority of Liberals in Parliament as some prepare to confront him Wednesday in the hopes of convincing him to step down.

(Image credit: Adrian Wyld/AP)

This review first appeared in issue 361 of PC Pro.

Armari has been producing some of the best workstations in the UK market for decades and, together with Scan, is often the first to send us the latest technology. The Magnetar M64T7-AW1650G4 isn’t the only system this month to sport one of the recently released non-Pro AMD Ryzen Threadrippers, but it’s the only one to opt for the 64-core iteration.

The Threadripper 7980X has a base 3.2GHz clock across its 64 cores (with 128 threads) and a maximum boost of 5.1GHz. The most obvious difference between the two variants of the Threadripper is that the Pro version supports eight-channel memory, and the non-Pro variant has a quad-channel controller, so will have lower RAM bandwidth. However, the Pro also has 128 PCI Express 5 lanes whereas the non-Pro only offers 48, plus 24 PCI Express 4 lanes. So the Pro will support a greater number of fast peripherals, such as networking and GPUs. Unless you need the memory and peripheral bandwidth, the non-Pro is much cheaper and therefore more cost-effective.

The Armari gave us some of the highest test results we’ve ever seen (Image credit: Future)

Armari has taken advantage of the quad memory channels by providing four 32GB DDR5 DIMMs running at 5,600MHz, for a total of 128GB. This will be enough for now, although several manufacturers this month have provided a lot more. The Asus Pro WS TRX50 Sage WiFi motherboard supports up to 1TB of memory.

Although the vanilla Threadripper is cheaper than the equivalent Pro, it’s still expensive, so Armari has been more conservative about its graphics acceleration choice. Instead of the range-topping AMD Radeon Pro W7900, the GPU is the W7800 that sits one rung below. This has 4,480 Stream processors instead of 6,144, and 32GB of GDDR6 frame buffer instead of 48GB, with 576GB/sec of bandwidth instead of 864GB/sec.

Also, while there are two storage drives in this system, Armari hasn’t supplied a smaller, faster one for the operating system and applications alongside a larger, slower one for media. Instead, there are two 2TB Crucial T700 M.2 NVMe devices operating at PCI Express 5 speeds. These have been combined using the motherboard’s built-in array capability into a single 4TB RAID0 device. This configuration delivers a staggering 16,134MB/sec of sustained reading and 22,655MB/sec writing (as measured by CrystalDiskMark 8).

One of Armari’s unique features is that it designs its own chassis and cooling solutions, although not for all its systems. The Magnetar M64T7-AW1650G4 sports the latest iteration of Armari’s M60G4 case, which now has some unique airflow tunnelling but most significantly has custom liquid cooling warrantied to deliver 700W for AMD’s Precision Boost Overdrive. Armari systems come with the latter enabled to obtain the maximum possible performance while remaining entirely reliable and stable. The chassis also has side-accessible easy-swap caddies for adding up to three 2.5/3.5in drives and another with a U.2 connection that supports SAS or up to three PCI Express 4 NVMe drives. The 1,650W PSU should be more than capable of supplying a fully stocked system without missing a beat, however.

Armari’s M60G4 case provides custom water cooling (Image credit: Future)

With that 64-core AMD Ryzen Threadripper CPU, the Armari was sure to dominate in most of the processor-focused benchmarks. The overall PC Pro media benchmark score of 949 is the highest we’ve seen, and although the image-editing result of 281 isn’t top of the pack, 869 in video encoding and 1,224 in multitasking trounce the competition. Likewise, 6,294 in Cinebench 2024 multi-CPU rendering was more than three times faster than any of the £4,500 systems this month, and 80% quicker than anything else. It dispatched the Blender Gooseberry render in 89 seconds on CPU, way ahead of anything we’ve seen before. The Adobe Media Encoder OpenCL-enhanced render took only 105 seconds.

The GPU, however, is merely great rather than dominant. SPECviewperf 2020 results include 185 in 3dsmax-07 and 817 in maya-06, 139 in catia-06, 221 in creo-03, 601 in snx-04 and 455 in solidworks-07, showing 3D animation and engineering workloads will be dealt with easily. But the W7900 is significantly faster in all areas. GPU rendering and AI inference is similarly good but beaten by a few competitors.

Overall, though, this is another incredible system from Armari, and a worthy Labs Winner.

Check out the best business computers.

This review first appeared in issue 361 of PC Pro.

Last year, Intel stopped developing its popular NUC mini PCs, signing an agreement with Asus that it would “manufacture, sell and support the Next Unit of Compute (NUC) 10th to 13th generation systems product line, and to develop future NUC systems designs”. I assumed that meant Asus would have the monopoly on the NUC name, but not so: MSI has entered the fray with its Cubi NUC line, and assures me that more will follow in the coming months.

You can buy the Cubi NUC 1M as a barebones system, with the Core 3 100U version costing £349 inc VAT. Core 5 120U and Core 7 150U versions are available for £449 and £519 respectively (prices from scan.co.uk). Or you can buy Cubi NUCs as complete systems, supplied with RAM, SSDs and Windows 11 Pro. Here, I test the top-end version, with a Core 7 150U, 16GB of RAM and a 1TB SSD.

The Cubi NUC 1M is a great mini PC provided that all-out power isn’t a priority (Image credit: Future)

Arguably, this is a 14th gen NUC as Intel released the 150U earlier this year. However, Intel seems intent on adopting more confusing names for its chips with every release, and the 150U actually has much more in common with the 13th generation Core i5-1335U than the new Core Ultra line. It still uses the Raptor Lake architecture, includes two P-cores and eight E-cores, and there’s no NPU for local AI tasks.

Like all U series chips, the 150U is designed for maximum efficiency rather than all-out speed. When I pushed it to extremes, the Cubi’s fan kept at a low-pitched but audible hum, but its power consumption peaked at 61W (it’s 15W when idle). That compares to 101W from the Geekom A8 Mini PC (see issue 359, p62) when I tested it with an AMD Ryzen 9 8945HS.

While the Core 7 150U holds its own in single-core tasks, outperforming the Ryzen 9 8945HS in Geekbench 6 with a score of 2,635 versus 2,561, there’s a gulf in benchmarks that test all cores. In Geekbench, the Geekom A8’s 13,237 return is light years away from the MSI’s 8,651, and the gap grew in Cinebench R23 with 16,615 against 6,342.

(Image credit: Future)

The gulf in performance continued into gaming, as illustrated by the Geekom’s return of 3,215 in 3DMark Time Spy, almost twice the MSI’s 1,711, and where the Ryzen’s 780M graphics proved good enough for 35fps in Shadow of the Tomb Raider (at 1080p High settings) the MSI could only rustle up 22fps.

So, if all-out speed is your priority, the Cubi NUC is not your best choice. However, few people need such power, especially in businesses where MSI is certainly targeting this mini PC. For instance, it includes two 2.5GbE ports, one of which can be used for out-of-band management, while those IT directors who need to show they’re making environmental strides will appreciate that 43% of the plastics are PCR. MSI takes renewables seriously, too, with 92% of its factory’s power coming from renewable sources.

I also asked MSI about its customer support. You’ll be buying from a reseller rather than direct from MSI, but for businesses its approach is to have a ready supply of buffer stock so that faulty units can be swapped out and then fixed at MSI’s recently established UK service center. It’s hardly Dell next-business-day on-site support, but there is a plan.

The Cubi’s impressive range of connectivity includes two 2.5GbE ports (Image credit: Future)

The Cubi should be of particular appeal to any business wanting to mount mini PCs on a monitor. If you buy a compatible MSI display then its power switch can also switch on the PC, but otherwise MSI supplies an external power switch that sits at the end of a cable and means users don’t need to reach round the back. You can even power the Cubi via USB-C, with up to 100W of power, massively simplifying cable management. There are two USB-C ports here, both supporting Thunderbolt 4, and when you add Wi-Fi 7, Bluetooth 5.3, four USB-A ports, a microSD card reader and two HDMI 2.1 ports it’s hard to think of something that isn’t included.

Those looking for a home entertainment system will be pleased to discover those HDMI ports support CEC (so universal remotes will work), and there are plenty of ways to extend storage. A second M.2 socket sits empty, albeit M.2 2242 rather than 2280, and you can mount a 2.5in SSD externally via four screw slots on its bottom. Or plug fast external storage into the USB-C ports.

As ever with NUC designs, the big advantage of the Cubi is its ability to squeeze into any situation. Here, you could use the Cubi for digital signage (the Core 3 version will be fine for such tasks), in schools, in businesses and in homes. It isn’t the fastest mini PC around, nor the smallest, but for versatility it’s an absolute winner.

We've rated the best business computers.

NAND is a vital component of the future of electronics. It’s everywhere, driving the storage capacity, performance and power efficiency in everything from data center servers to the smallest mobile devices – such as phones, drones, cameras and other portable devices.

As these systems and electronic devices add more features and perform more complex tasks like AI, data storage needs will continue to grow – making NAND flash memory a critical component of future innovations.

As a result, the race is on to build higher capacity NAND with better performance and lower power. Many people believe that a higher layer count is the only way forward. But the truth is there are many vectors of NAND innovation and higher layer counts aren’t the only way to increase NAND flash bits and storage capacity.

This new era of NAND is driving a period of change, where the layers-focused race is behind us. The emphasis is shifting toward strategically timing the introduction of new, longer lasting nodes optimized for specific use cases and applications. Not all applications need the latest node with the highest capacity or performance. Making each layer denser, rather than simply stacking more layers, enhances power efficiency, performance and capacity while managing cost for specific customer needs.

Traditional Vertical Scaling

The “layers race” is the notion that more layers means more bit density and capacity, leading to a cost advantage – therefore the NAND with the highest number of layers must be best. But with 3D NAND, it’s no longer that simple.

Scaling NAND is similar to adding capacity at a hotel. Simply adding more floors may seem like a good idea, but you have to remember that building up leads to an increase in operational costs and complexity, including costs to buy and move equipment, build floors, etc. At some point, there diminishing returns to adding additional floors. Intuitively, the proportional cost reduction provided by adding ten floors to a hundred-floor building is better than adding the same number of floors to a five-hundred story building. But the capital needed to add the extra might be higher to build those additional 10 floors on top of a five-hundred-floor building.

Making each floor denser, by shrinking rooms and using space more effectively, can provide the same increase in occupancy in a much more efficient and cost-effective way.

The same logic applies to NAND architecture. Simply adding NAND layers on top of each other may not be the only way to build more bits or capacity. Like floors of a hotel, it becomes more expensive and difficult to build usable NAND as layer counts grow. For example, stacking layers leads to increased processing time, additional capital for the advanced tools needed to ensure we can reliably manufacture NAND die with high quality.

Scaling Smarter by Leveraging Multiple Vectors

While layer count will continue to grow, it is no longer the core innovation driver. Instead, innovation spans across multiple vectors and there are other ways to scale NAND architecture in addition to vertical scaling, including lateral, logical and architecture scaling methods.

Lateral scaling works by packing every single memory layer while removing some of the redundant support structures. It’s like squeezing more rooms on the same floor of a hotel room or reducing the number of stairs and elevators in a building. For example, starting with lateral scaling allows you to optimize the available space before adding another layer. This phased approach is much more efficient, saving costs while reducing risks. It also allows customers to get to a certain capacity point at the right time, with consistent supply and quality. And when it’s decide to add more layers, the benefit is multiplied by the increased efficiency of the layers added.

Logical scaling increases the number of logical bits that can be stored on a physical device. In the hotel room scenario, this would be akin to squeezing more guests into the same hotel room without causing disturbances.

Finally, architecture scaling optimizes the way circuits support memory arrays – such as positioning circuits next to the array, tucking them underneath or perhaps implementing them on a separate wafer. In a hotel, this could be where the parking lot is put for needed guests – on the side of the building, underneath, or above the building (with a cost-effective way to airlift cars, of course).

A combination of all four

An approach that uses a combination of all four of these scaling vectors is a much smarter way of adding NAND bit growth without sacrificing performance and power efficiency for the widest range of uses cases and devices. And it has the additional benefit to optimize node-to-node cost reduction and minimize capital needed for transitions.

And while NAND technology is complex, the manufacturing processes that create viable NAND nodes, and eventually products, are even more so. These conditions are exacerbated by supply and demand dynamics in an emerging era where new applications, especially AI, will greatly increase the need for both compute- and storage-intensive flash-based solutions.

For example, this AI Data Cycle framework shows the virtuous cycle where storage feeds AI models, and AI in return demands more storage. This AI Data Cycle will be a significant incremental growth driver for the storage industry.

Performance, power, and capacity

Performance, power, and capacity play a major consideration at every phase, as each stage demands something different. Whereas the initial stages need massive capacity to contain as much data as possible for model training, as data goes through the cycle, speed and performance may be the more important factors. And power is increasingly becoming a critical factor in any AI application.

In this new era of NAND, NAND nodal migration paths should also be based on the needs of the customer, not a one-size-fits all approach of the past.

Different needs for different customers are starting to bifurcate and the role of NAND suppliers in addressing these needs is becoming much more interesting. Ultimately, what a customer builds will dictate how the flash inside it should operate—how big it should be, how much capacity it holds, and how much power it will consume. It’s not about how many layers the product has. Focusing on the features that are most important to customers—performance, capacity and power— is the winning strategy.

We list the best SSDs: top solid-state drives for your PC.

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

Egypt has been fighting malaria for nearly 100 years. WHO declares a country malaria-free when the disease has not been present for at least three consecutive years before the designation.

(Image credit: James Gathany/AP)

Around the world, seeds of regenerative cities have already been planted. As they grow, they will shape the metropolises of tomorrow.

Get 10% off a website or domain, on any plan, with this exclusive Squarespace promo code from WIRED. Save on your next big project this October.

The chip technology company Arm has given Qualcomm 60 days notice, according to Bloomberg.