Uncategorized

NiceHash defeats Nvidia’s GPU crypto-mining limits, does not appear to be a scam

But the workaround still can't defeat the newest version of Nvidia's LHR tech.

Nvidia began releasing LHR (or “Lite Hash Rate”) graphics cards last year to slow down their cryptocurrency mining performance and make them less appealing to non-gamers. Late last week, crypto-mining platform NiceHash announced that it had finally found a way around those limitations and released an update for its QuickMiner software that promises full Ethereum mining performance on nearly all of the LHR-enabled GeForce RTX 3000-series GPUs.

Unlike past attempts to disable the LHR protections, NiceHash's workaround appears to be the real deal—Tom's Hardware was able to confirm the performance boosts using QuickMiner and a GeForce RTX 3080 Ti.

For now, NiceHash says that the LHR workaround will only work in Windows, with "no Linux support yet." The more flexible NiceHash Miner software doesn't include the workarounds yet, though it will soon. NiceHash also says that the software won't accelerate mining performance on newer GeForce cards that use version 3 of the LHR algorithm, a list that (for now) includes the RTX 3050 and the 12GB version of the RTX 3080 but which will presumably grow as Nvidia releases new GPUs and updated revisions for older GPUs.

Miners have been trying to find ways to circumvent the LHR limitations since they were introduced. The first card to use LHR, the GeForce RTX 3060, was defeated by a botched driver release from Nvidia. Other workarounds have included flashing alternate BIOSes and mining multiple cryptocurrencies on the same card.

But LHR workarounds can also be too good to be true. Another group promised LHR-defeating drivers in February, but they didn't do what they said they'd do and ended up being full of malware.

Whether this LHR unlock has an impact on the pricing or availability of GPUs remains to be seen. Bitcoin and Ethereum prices have been falling recently as interest rate increases and stock market turmoil have pushed investors toward safer bets. Ethereum's "merge," which will switch the currency from a mining-driven "proof-of-work" system to an ownership-driven "proof-of-stake" system, is also allegedly a few months away, though that has been the case for several years now. At this point, buying up a bunch of new GPUs for cryptocurrency mining, even with the performance improvements and price increases, could still be a risky investment.

Wi-Fi 7 home mesh routers poised to hit 33Gbps

Wi-Fi Alliance has promised "at least" 30Gbps.

It's looking increasingly likely that Wi-Fi 7 will be an option next year. This week, Qualcomm joined the list of chipmakers detailing Wi-Fi 7 products they expect to be available to homes and businesses soon.

The Wi-Fi Alliance, which makes Wi-Fi standards and includes Qualcomm as a member, has said that Wi-Fi 7 will offer a max throughput of "at least 30Gbps," and on Wednesday, Qualcomm said its Network Pro Series Gen 3 platform will support "up to 33Gbps." These are theoretical speeds that you likely won't reach in your home, and you'll need a premium broadband connection and Wi-Fi 7 devices, which don't exist yet. Still, the speeds represent an impressive jump from Wi-Fi 6 and 6E's 9.6Gbps.

The next-gen tech is aimed at network-intensive applications, like virtual and augmented reality, video streaming at 4K and higher, and cloud computing and gaming. By making changes to the physical (PHY) layer and medium access control (MAC), Wi-Fi 7 should allow you to enjoy these applications with less latency and jitter.

Like Wi-Fi 6E, Wi-Fi 7 should be able to leverage the 6 GHz band while expanding channel width support from 160 MHz to up to 320 MHz.

Other expected benefits of Wi-Fi 7 include multi-link operation, enabling the simultaneous use of multiple frequency bands. Qualcomm also pointed to 10Gbps enterprise access points and up to 500 users per channel.

Qualcomm announced two tri-band and two quad-band offerings for home and business use, and the company expects Wi-Fi 7 to hit homes as soon as next year, Digital Trends reported. Qualcomm's Network Pro 1620 quad-band option, which can be used in home mesh routers, claims speeds of up to 33.1Gbps.

Even though the Wi-Fi Alliance hasn't finished the standard yet, Wi-Fi chipset makers MediaTek and Broadcom have also flagged 2023 as Wi-Fi 7's expected debut date and are showing off Wi-Fi 7 technologies. Vijay Nagarajan, Broadcom's VP of marketing for the Wireless Communications and Connectivity Division, said that home availability of 33Gbps should come in the second half of 2023.

"Ultimately, Broadcom's Wi-Fi 7 will keep performance a step ahead of the broadband speeds that DOCSIS 4.0 and multi-gigabit PON technologies bring to the home," Nagarajan told Ars Technica.

Razer’s 240 Hz laptop makes OLED more appealing to gamers

Highly expensive PC's display balances speed, resolution, and contrast.

The portability of laptop displays means they often come with compromises in speed, color, or sharpness. For those who want a screen with the deepest contrast and darkest blacks without sacrificing speed, Razer is hoping to have the answer.

The upcoming configuration of the Razer Blade 15 announced Tuesday will be the first laptop to offer a 240 Hz refresh rate in an OLED panel. That makes OLED a more viable option for people, like gamers, who prioritize fast motion processing. Neither OLED monitors, with their high prices, nor OLED TVs have yet to reach such speeds. However, OLED has hit 240 Hz in other designs, such as in this camera's viewfinder. 

The 15.6-inch screen on the upcoming Blade 15 claims a gray-to-gray (GtG) response time of 1 ms, which is as good as it gets for a gaming laptop these days, as well as 100 percent coverage of the wider DCI-P3 color space. As expected with an OLED panel, it's not super bright. At a claimed 400 nits, there are better laptop displays to use outside on a sunny day, for example.

To achieve its high refresh rate, the display is at 2560×1440 resolution, while you'll find many ultralight, non-gaming laptops offering OLED displays with 4K resolution. Razer makes a version of the Blade 15 with 4K OLED that's limited to a 60 Hz refresh rate for $200 cheaper as of this writing. However, 1440p has become a sweet spot for gamers seeking a sharper image without requiring as much graphics horsepower as 4K.

Beyond the panel, the Blade 15 has an RTX 3070 Ti laptop GPU, Intel Core i9-12900H, and up to 32GB of DDR5-4800 memory and a 1TB SSD, plus an empty M.2 slot. There's also good port selection with two USB-C ports, including one Thunderbolt 4, a USB-A port, an HDMI port, and an SD card reader.

Razer's announcement didn't make any battery-life claims about the OLED-equipped clamshell, but with OLED, a high refresh rate, and the machine's gaming heritage, don't expect longevity. The laptop does use Nvidia's Advanced Optimus feature that, among other things, switches between the discrete GPU and the processor's integrated graphics as necessary to conserve battery life.

When all is said and done, you're looking at a wildly expensive $3,500 machine when it comes out in Q4. "First" doesn't come cheap.

Razer's laptop should finally allow users to enjoy OLED's rich contrast, detailed shadows, and nuanced highlights without visual artifacts, like stuttering, deterring from the experience when your GPU is pushing high frame rates. Razer's Blade 15 laptops also come with G-Sync, Nvidia's Adaptive-Sync flavor, to fight screen tears. But if you want the smoothest video processing, laptops (including alternative configurations of the Razer Blade 15) that can hit 360 Hz and carry more powerful GPUs than Razer's upcoming speedy OLED system do exist.

In its announcement, Razer highlighted AAA games, which tend to boast impressive visuals, as a top use for the OLED panel. But unless it's a less graphically demanding esports game, you won't be pushing 240 frames per second to make the most of the screen's refresh rate, especially at 1440p resolution. If you're a gamer who plays a variety of genres, you'll more easily take advantage of the screen's multiple attributes.

Razer also pointed to photo and video editing and movie watching as good uses for the fast OLED display. But if OLED makes you envision turning the Blade 15 into an HDR vehicle, think again. The Blade 15 doesn't note any HDR10 support, which is necessary for HDR on Windows.

But if you want your SDR display to have some of the best contrast around while looking sharp during fast-paced action and supporting pretty high frame rates, Razer is finally making that an option.

Microsoft is testing a free 1GB-per-month VPN service in its Edge browser

Service is free but requires a Microsoft account sign-up.

A couple of years ago, Microsoft reformulated its Edge web browser with a backend based on Google's Chromium codebase. Since then, the company has tried to make Edge stand out primarily by adding on extra features, mostly related to privacy, security, and online shopping.

One interesting new experimental feature that could be coming to Edge soon is a Cloudflare-powered VPN feature, according to a support document published last week. A VPN (or virtual private network) provides an encrypted tunnel for all of your network traffic, shielding it from the view of other devices on the same network.

Using the VPN service, dubbed the "Microsoft Edge Secure Network," requires you to be signed in with a Microsoft account, just like cross-device syncing of bookmarks and extensions and plenty of other features. It provides up to 1GB of data per month, with no option to get more data if you want or need it—Edge will track your data usage and let you know when you're getting close to your limit.

That low data cap means Edge's VPN won't replace a subscription to a VPN service for people who run all of their traffic through VPNs or those who use them to circumvent geographical restrictions on video streaming. But it's enough for occasional coffee shop Wi-Fi use, and that seems to be the type of usage it's geared toward.

Using a VPN to safeguard traffic on a public Wi-Fi network arguably isn't as important as it used to be, given the widespread use of HTTPS and browsers that increasingly prefer HTTPS connections to HTTP ones when they're available. But if you manually force HTTPS connections and browse the Internet normally, you'll still find plenty that's unencrypted, and a VPN can protect things like DNS requests that HTTPS doesn't always obscure.

Any VPN is only as private or as secure as the VPN provider you're using—it can still see all of the unencrypted data you send it since it needs to send you the data your browser is requesting. Microsoft's support post says that Cloudflare "permanently deletes the diagnostic and support data collected every 25 hours" and that the data Cloudflare processes "is subject to the Microsoft Privacy Policy."

Apple releases beta update with fixes for Studio Display’s mediocre webcam

Firmware 15.5 includes fixes for "noise reduction, contrast, and framing."

Apple's Studio Display got dinged for plenty of things in reviews, including its price and its IPS panel technology. But the most consistent criticism was aimed at its built-in webcam, which fell far short of most standalone desktop webcams—and even the front-facing cameras included with iPhones and iPads.

Apple said in March that it would release a firmware update to address at least some of the problems, and that update is currently available in beta form starting today.

"This beta update has refinements to the Studio Display camera tuning, including improved noise reduction, contrast, and framing," an Apple spokesperson told Ars.

To download the Studio Display firmware update, you'll first need to download and install the macOS Monterey 12.4 beta that was also released today. That update is currently only available to developers, but a version for the general public should be available in the next day or so. Once that's installed, the Studio Display 15.5 firmware beta will download and install separately. The Studio Display includes an Apple 13 chip along with 64GB of internal storage—it's essentially an iPad in a monitor's body—so its iOS-based firmware updates are downloaded and installed separately rather than coming down as part of a macOS update.

There's no word on when the final version of macOS 12.4 or the Studio Display firmware update will be released. But we expect them both to come out sometime between now and the time Apple unveils its next major OS updates at its Worldwide Developers Conference in early June.

End of the road: Apple is killing macOS Server, the place where Mac OS X began

Apple says that Server's most popular features have been integrated into macOS.

Apple announced today that it is formally discontinuing macOS Server after 23 years. The app, which offers device management services and a few other features to people using multiple Macs, iPhones, and iPads on the same network, can still be bought, downloaded, and used with macOS Monterey. It is also still currently available at its normal $20 retail price but will no longer be updated with new features or security fixes.

Server was never as widely used as the consumer versions of macOS, but macOS Server has a long history going all the way back to Apple's late-'90s acquisition of NeXT and its NeXTSTEP software. NeXTSTEP was adapted into a project called "Rhapsody," which added support for some longtime Apple software and a more Mac-like user interface and was initially released as Mac OS X Server 1.0 in March of 1999. This initial version of Mac OS X Server shared a lot of underpinnings with what would become Mac OS X but predated important user interface elements like the Dock and the Aqua theme, which would launch two years later in the first consumer version of Mac OS X.

Mac OS X Server remained its own totally separate version of the operating system, from the launch of that initial version through to Snow Leopard Server (version 10.6) in 2009. Starting in Mac OS X Lion, Apple began selling the Server software as a downloadable add-on app for any Mac, coinciding with the death of Apple's last rack-mounted Xserve hardware. This transition also slashed the software's price; a single Snow Leopard Server license cost $499, while the Server app cost just $50.

Apple continued to develop the Server app in the following years, releasing major new versions roughly in lockstep with yearly Mac software updates. But the software gradually began shedding features, starting with services like DNS and mail that weren't specific to Macs. Apple did still offer unique features to Mac and iDevice users in Server: mobile device management for IT admins; a Time Machine backup service that could enforce per-device storage quotas to keep one Mac from filling up a server's entire hard drive; and a Caching service that could save bandwidth by storing and offering app and OS updates to other devices on your server's network rather than downloading things from Apple's servers multiple times.

Apple notes that the Time Machine, Caching, and File Sharing services are all included in all macOS installs now, and they have been since High Sierra was released back in 2017. For mobile device management, Apple points to a pair of pages about choosing third-party MDM software and migrating from one MDM service to another in lieu of making more specific recommendations or offering specific tools that could speed migration from Apple's MDM service to a different one.

Microsoft enters “final phase” of disabling SMB1 file-sharing in Windows 11

Re-adding SMB1 support is still possible but will get harder as time goes on.

Most Windows 11 preview builds focus on adding features, but sometimes Microsoft uses them to remove things. Users installing the latest Windows 11 Home Insider builds will find that support for version 1.0 of the venerable SMB file-sharing protocol is now disabled by default, something that may break file-sharing for older networked storage equipment. A post by Microsoft program manager Ned Pyle details the reasoning behind the change and how it will affect users.

Microsoft had already disabled SMB1 by default in other editions of Windows. The SMB1 server service was removed from all Windows versions starting in 2017, and the client service was disabled in Windows 10 Pro editions starting in 2018. Pyle writes that the client in the Home editions of Windows came last since it will "cause consumer pain among folks who are still running very old equipment, a group that's the least likely to understand why their new Windows 11 laptop can't connect to their old networked hard drive."

SMB1 has long since been replaced with newer and more secure versions of the protocol; SMB2 was introduced in 2007, and version 3.1.1 was added to Windows 10 in 2016. But the original is still occasionally used by old servers and equipment—and if a machine is old enough to rely on SMB1, it's probably old enough that no one is interested in maintaining or upgrading it.

For now, the SMB1 feature can still be installed manually by users and system administrators who need it, and if you're using SMB1 on a PC that you're upgrading to Windows 11, the upgrade won't disable the feature. The next phase of the transition will go one step further, completely removing DLL files and drivers needed for SMB1 support from the OS. The company "will provide an out-of-band unsupported install package for organizations or users that still need SMB1 to connect to old factory machinery, medical gear, consumer NAS, etc.," writes Pyle.

HP is expected to release a 17-inch foldable OLED laptop soon

HP would be the third PC maker to announce a bendable-screen OLED laptop.

While smartphones are having fun with the trend, PCs with foldable screens have yet to become mainstream, partially because there's only one option readily available. But with HP expected to enter the scene, it's possible 'foldable OLED' could become more common laptop lingo.

Lenovo made the bold first step into foldable laptops with its 13.3-inch ThinkPad X1 Fold. According to South Korean electronics website TheElec, HP's take on foldable OLED will be bigger, with a 17-inch panel from LG Display that measures 11 inches when folded up. HP hasn't publicly announced or commented on the rumored PC, but a couple of details make the machine seem at least somewhat plausible. For one, LG Display confirmed work on a 17-inch foldable OLED laptop design in January.

Most recently, TheElec on Monday reported that South Korean company SK IE Technology will make transparent polyimide films to cover the bendy 4K OLED panels. The publication also claimed that LG Display currently has plans to make up to "around 10,000" foldable OLED panels for HP, starting in Q3.

Should the HP foldable bend into reality, it'll have a couple immediate competitors. The ThinkPad X1 Fold is getting older, with the 2020 machine currently available with a 2048×1536 foldable OLED screen, up to an Intel Core i5-L16G7 with one performance core, four efficiency cores, and a clock speed ranging from 1.4-3 GHz; 8GB of LPDDR4X-4267 memory, and 256GB of SSD storage.

In January, Asus also announced its Zenbook 17 Fold OLED for mid-2022. The machine looks like a stronger competitor for the purported HP device, with a 17.3-inch bendable OLED touchscreen, 2560×1920 resolution, and an Intel i7-1250U with a 1.1-4.7GHz clock speed, two performance cores, and eight efficiency cores. Asus also announced the laptop with 16GB of LPDDR5 RAM and a 1TB PCIe 4.0 SSD.

Asus hasn't confirmed, but TheElec claimed today that its foldable machines will also use polyimide films, as well as OLED panels from BOE.

Beyond image quality, productivity performance, price, and durability, features that boost usability of the new form factor (like Asus' detachable physical keyboard) will be key differentiators for enthusiasts.

Although Lenovo's ThinkPad X1 Fold came out two years ago, we've yet to see a direct rival release. But if another one of the world's top-three PC makers does indeed enter this space, don't be surprised to see more laptop-sized OLED screens that bend in half and are flanked by a chassis with ports in the future.

 

Developer logs reveal details about the M2 in several new Macs

Logs appear to confirm core counts for the M2 and M2 Max.

Apple is in the late stages of readying several new Mac models built on the forthcoming M2 chip, according to a report from Bloomberg citing both developer logs and people familiar with the matter.

Over the past couple of years, Apple has been transitioning nearly its entire Mac product line from Intel's chips to Apple's system on a chip, which includes a CPU, GPU, NPU, ISP, and more. The first generation of Macs running on Apple's silicon used the M1 chip and its more powerful variants, the M1 Pro, M1 Max, and M1 Ultra.

Now Apple plans to introduce the M2, an evolutionary step from the M1.

According to the developer logs, Apple is working on the following machines:

  • MacBook Air with the M2, codenamed J413
  • Mac mini codenamed J473 with the same M2, and another codenamed J474 with the M2 Pro
  • Entry-level MacBook Pro with the same M2, codenamed J493
  • 14-inch MacBook Pro with the M2 Pro and M2 Max, codenamed J414
  • 16-inch MacBook Pro with the same M2 Pro and M2 Max, codenamed J416
  • Mac Pro codenamed J180, with a successor to the Mac Studio's M1 Ultra

The M2 seen in the logs has eight CPU cores and 10 GPU cores—that's two more GPU cores than the most robust M1 chip. Meanwhile, the M2 Max has 12 CPU cores and 38 GPU cores with 64GB of memory—two more CPU cores and six more GPU cores than the M1 Max.

Unfortunately, the report doesn’t detail the breakdown of efficiency cores versus performance cores in the M2 or M2 Max., or Bloomberg didn't report it.

This breakdown suggests that the M2 will be iteratively faster and more robust than the M1, but unsurprisingly, it appears it may not be a radical leap or major departure.

This story doesn't include details about ship dates, but it does note that the testing process is "far along in some cases." However, not every model listed above is guaranteed to ship to consumers.

In a previous report, Bloomberg claimed that Apple plans to introduce two new Macs this summer, likely at the company's annual developer conference in June. The most likely candidate appears to be a heavily redesigned MacBook Air with the M2, but an updated 13-inch MacBook Pro and Mac mini are also possibilities, as is a new Mac Pro.

Meta announces plans to monetize the Metaverse, and creators are not happy

Creators will fork over 25% after the 30% already taken by hardware platforms.

Meta, the company formerly known as Facebook, announced some initial plans on Wednesday to allow content creators to monetize its would-be Metaverse platform, Horizon Worlds. Meta's planned revenue share for contributors' creations could add up to nearly 50 percent.

Horizon Worlds is a network of shared 3D spaces that is currently exclusively available on Oculus Quest headsets. (Meta has plans to bring it to mobile, game consoles, and desktop VR in the coming months and years.)

There are already people creating spaces for Horizon Worlds, including a virtual yoga studio and a Second Life-like fast-food brand integration in the form of the "Wendyverse." But to date, Horizon Worlds has not offered the tools for creators to make a living creating that content as they can on similar services like Roblox.

That's not to say creators haven't been making money; some get paid for contract work outside of Meta's system for creating content that brands or other users can use. But Meta's announcements paint a picture of what to expect.

First off, Horizon Worlds will support in-world purchases. A handful of creators will be able to sell virtual items from within user-generated areas. Meta also plans to introduce a creator bonus program that awards money to creators based on how much other users engage with their content.

All this sounds like what you'd expect, but Meta caused a stir in the AR/VR, NFT, and the relevant online communities when it revealed specifics about how much revenue Meta will take from each transaction.

When users purchase an item in Horizon Worlds, Meta confirmed to CNBC that it would take a 25 percent cut—but that's after any amount a hardware platform might take. Right now, that just means Meta's Oculus Store, which takes a 30 percent cut. So content creators will have to hand over 30 percent to the Oculus Store (or the applicable percentage for whatever platform stores Horizon Worlds ends up on later, like Google Play), then they'll have to cede 25 percent of what's left to Horizon Worlds.

That leaves creators with just over half of their content's revenue before any applicable taxes.

The announcement has drawn ire from creators in the loosely related NFT community, who are accustomed to single-digit-percentage platform takes. There are also accusations of hypocrisy from game developers and others who have seen Meta publicly criticize companies like Apple for charging 30 percent on similar transactions around in-game content.So far, Meta has brushed off the controversy. Vivek Sharma, Meta's VP of Horizon Worlds, told The Verge, "We think it's a pretty competitive rate in the market... we believe in the other platforms being able to have their share."

Meta is an enormous company with several disparate objectives and teams behind each of those objectives. It's not that surprising that a Horizon Worlds executive is telling a different story (other platforms deserve their cut) than the parts of Meta that are agitating Apple and others (claiming those platforms' cuts are unreasonable).

But conflicting messaging aside, content creators we've spoken to say the chunk of the pie they'll get on this platform is far less than expected.

Horizon Worlds is just one part of Meta's plan to bring AR, VR, and virtual worlds to the masses. The company also plans to introduce mass-market augmented reality glasses sometime in 2024, placing it in direct competition with Apple and other companies with similar plans.