Our phones have become do-it-all gadgets. They enable us to keep in touch with one another, track how many steps we’re taking each day, snap photos of everything, and access the collective knowledge of humanity. All that universal functionality extends across both our work and personal lives.

The gadget that pings you on Slack when your colleague has a question is the same one that alerts you when your friend has just outscored you at Wordle. That blurring of boundaries can be a problem.

There is actually a solution to this, at least if you’re using an Android phone. Android supports the use of multiple profiles (based on Google accounts), so you can switch between two different ones as needed. The most obvious use case for this is to split up your work and personal usage.

It’s essentially like having two phones on the same physical device: The profile that’s active dictates the apps you can use, the notifications you see, the browsing history you can dig into, and the files you’ve got access to, all dictated by the Google account.

Note that while this feature has been around in Android for a long time and is available on many handsets (including Pixel phones), not every Android manufacturer has implemented it. It’s not available on Galaxy phones, for example, though Samsung does offer something similar with the Secure Folder.

Get Started

Adding a new user.

Photograph: Courtesy of David Nield

To begin with, you’re going to need two Google accounts, but many of us already have a Google account managed by work anyway. If not, you can set up a new Google account, free of charge, to use as the basis of the work profile on your phone: Head to the Google account portal on the web, then click Create an account.

The details you’ll need to provide include a first name, your birthday, and a Gmail address you’d like to register to use with the account—this is what you’ll use to sign in to this account on your phone and on the web. You get some suggestions for an available address based on your name, but you can also come up with something custom as long as no one else has taken it first.

Once you’ve got a second account you can use, open Settings, then choose System > Users. Tap Add user, then Next, and you’re able to give a name to the second user account. With that done, select it from the Users screen, choose the Switch to… option, and you’ll be taken through the process of signing in to your secondary Google account.

The DoJ said the app has made significant strides in privacy protections since the 2024 suit.

Most people who remember personal computers from the late 1980s think first of the Macintosh and its clean black-and-white interface. Apple had already shipped something that looked and behaved a lot like it, only this version ran in color on a machine that still carried the Apple II name. That system was GS/OS, and it arrived in 1988 as the heart of Apple IIGS System Software 4.0.

GS/OS was built from the start in 100% 16-bit code for the 65816 CPU inside the IIGS, which was a significant shift from previous versions. Most other software on this computer was still inadequately optimized for 8-bit routines carried over from the previous Apple II generation. The improved code resulted in significantly faster disk access, screen rendering, and program loading. Customers who had upgraded from ProDOS 16 were particularly affected.

Apple 2026 MacBook Air 13-inch Laptop with M5 chip: Built for AI, 13.6-inch Liquid Retina Display, 16GB…

• BUILT FOR COLLEGE. AND BEYOND — MacBook Air with the M5 chip packs blazing speed and powerful AI capabilities into an incredibly portable design…
• TEAR THROUGH TOUGH ASSIGNMENTS — With its faster CPU and unified memory, the M5 chip delivers even more performance and fluidity across apps, making…
• MAKE QUICK WORK OF YOUR TO-DO LIST — Apple Intelligence helps you write, express yourself, and get things done effortlessly — whether it’s for…

GS/OS’s Finder was crucial to the system. Do you remember those desktop icons? Windows that can be moved around, resized, and stacked? Pull-down menus at the top of the screen and a trash can waiting patiently for files you no longer wanted, as it all looked pretty familiar to anyone who’d used an early Mac, but here’s the thing: all of this was happening in glorious color on hardware capable of displaying thousands of different hues on screen at any given time. The control panels allowed you to change all of the minor settings without leaving the desktop, such as mouse speed and clock.

The system’s technical foundation was especially impressive, as the modular architecture drew on many brilliant ideas from prior Apple systems. You could add or uninstall device drivers without rewriting the entire operating system. File System Translators allowed you to access disks formatted for other machines, such as Mac HFS volumes and various MS-DOS formats. Even resource forks appeared, allowing applications to store extra data alongside the main file, and let’s not forget those progress bars that filled up like a thermometer and proportional scroll bars, which are now considered standard features but were a real novelty at the time.

As you might expect, storage support increased with each subsequent iteration. Later versions of the system could handle hard disks containing hundreds of megabytes, eventually supporting volumes of up to four gigabytes. Then there was AppleShare, which added networking to the IIGS, allowing it to connect to the same office network as Macintosh machines and share files. Installation and disk utilities were eventually made far less onerous than the previous command-line approaches.

Apple continued to update the software into the early 1990s. System 5 improved the networking and speed even further. System 6, which was released in 1992 and completed with version 6.01 in 1993, introduced some polish to the Finder, improved font handling, and a slew of other tweaks that made day-to-day use a little easier, but that was the extent of the official revisions. By the end of 1992, the IIGS was no longer on Apple’s pricing lists, and the focus had turned entirely to the Macintosh series.

The hardware behind GS/OS was rather remarkable. The IIGS included a mouse, Apple Desktop Bus connectors, and a sound chip capable of playing sounds that was far more fascinating than anything previously available on the Apple II. Its graphics modes offered far more colors than the original Macintosh, but the system itself couldn’t escape the shadow of its more famous sibling. Software developers were focusing their efforts on the Mac, and the IIGS was really a niche product, even if its operating system silently packed in capabilities that felt current at the time.

Nobody appears to have informed the Compact Disc that its funeral was held years ago.

U.S. CD sales increased 16% to 16.3 million units during the first half of 2026, according to Luminate, and that growth still reached 6.7% when K-pop sales were removed from the equation. At the same time, the hardware industry has suddenly developed a serious case of silver-disc fever, with new CD players and transports arriving from Marantz, NAD, FiiO, Quad, Shanling, Cayin, Mission, TEAC and others at prices ranging from under $200 to several thousand dollars.

The latest manufacturer to join that increasingly crowded party is hardly some Kickstarter operation with a transparent plastic lid and Bluetooth.

Accuphase has introduced the DP-470, a dedicated high-end CD player that replaces the DP-450 and borrows technology from the considerably more ambitious DP-570S SACD/CD player. It goes on sale in Japan in late August for ¥528,000 including tax. Accuphase has not announced U.S. pricing or availability, and we will update this article when those details become available.

There is something wonderfully 2026 about one of Japan’s most traditional high-end manufacturers introducing a new machine whose primary purpose is extracting 16-bit/44.1kHz audio from a plastic disc introduced in 1982.

Apparently progress occasionally involves finding your jewel cases again.

This Is Not Just a DP-450 With a New Number

The DP-470 remains a CD-only player. It does not play SACDs, which Accuphase reserves for models such as the DP-570S higher in the range. But calling it a conventional Red Book player understates what Accuphase has done with the digital section.

The mechanical side starts with a high-rigidity, low-center-of-gravity CD drive designed to isolate the optical pickup from vibration. Accuphase uses an elastic damping system between the transport and loading mechanisms, an eight-pole neodymium magnetic disc clamp and a three-layer bridge structure intended to suppress chassis resonance and noise generated by the spinning disc. The tray itself is machined from an aluminum block.

None of that makes the disc contain more data. The objective is to give the optical system the quietest and most mechanically stable environment possible while it reads what is already there.

This is Accuphase, after all. Nobody was expecting the transport to be held together with two screws and optimism.

The More Important Changes Are Inside the DAC

Accuphase retains the ESS ES9026PRO used by the DP-450, but employs it through the company’s four-circuit parallel MDS+ conversion architecture. Accuphase says operating four DAC circuits in parallel improves characteristics including noise, distortion and linearity compared with using a single circuit.

The more significant change is the addition of ANCC — Accuphase Noise and distortion Cancelling Circuit — in the I/V conversion stage. The circuit uses a secondary amplifier to cancel noise and distortion generated by the main amplifier. The company has also redesigned the DAC board and optimized the power supply architecture. ANCC is technology already used in the more expensive DP-570S, so this represents some meaningful trickle-down rather than changing the display font and calling it a new generation.

Accuphase also uses its Direct Balanced Filter architecture with independent signal paths for the RCA and balanced XLR analog outputs, reducing interaction between them. Output level can be adjusted from 0 dB to -60 dB in 1 dB increments, and the balanced outputs include a phase selector.

Accuphase DP-470 Key Specifications

• Dedicated Red Book CD player
• ESS ES9026PRO DAC
• Four-circuit parallel Accuphase MDS+ D/A conversion
• ANCC noise and distortion cancellation in the I/V conversion stage
• High-rigidity, low-center-of-gravity CD transport
• Machined aluminum disc tray
• Independent Direct Balanced Filter circuitry for RCA and XLR outputs
• RCA and balanced XLR analog outputs
• Coaxial and optical digital outputs
• USB Type-B, coaxial and optical digital inputs
• USB PCM support up to 384kHz/32-bit
• USB DSD support up to 22.5MHz using ASIO Native
• DSD up to 11.2MHz using DoP
• Coaxial PCM input up to 192kHz/24-bit
• Optical PCM input up to 96kHz/24-bit
• Adjustable analog output level from 0 to -60 dB
• 465 x 151 x 393mm
• 13.7kg / approximately 30.2 pounds
• Five-year warranty in Japan

Your Streamer Is Invited Too

This may be the most important thing separating the DP-470 from purist machines such as the Hegel Viking.

Accuphase has turned the DP-470 into a high-resolution standalone DAC as well as a CD player. A computer can feed the USB input, while a network streamer or other digital source can use the coaxial or optical connections. Owners can therefore use the same Accuphase DAC and analog output stage for both physical discs and other digital sources.

That makes considerably more sense in a modern high-end system than requiring someone to buy an expensive CD player and then duplicate its DAC circuitry in another box for streaming.

The irony is difficult to miss. Streaming was supposed to eliminate the CD player. The new generation of CD players is responding by becoming the DAC for the streamer.

Who Is the DP-470 Competing With?

Until Accuphase announces U.S. pricing, positioning the DP-470 precisely against American-market competitors requires some restraint. Converting ¥528,000 into dollars and pretending that figure represents a U.S. MSRP would be misleading because distributor margins, tariffs, freight and regional pricing can substantially alter the final number.

But the competitive neighborhood is already obvious.

The Hegel Viking, currently selling for about $5,500 in the U.S., takes the opposite approach. It is unapologetically devoted to 16-bit/44.1kHz Red Book CD playback and does not provide digital inputs for using its DAC with outside sources. Hegel’s argument is purity; Accuphase’s is versatility without abandoning serious mechanical engineering.

The Luxman D-03X is probably the closest conceptual rival. It is another Japanese dedicated CD player with balanced outputs and a high-resolution DAC section that accepts USB, coaxial and optical sources. Its USB implementation supports PCM up to 384kHz/32-bit and DSD to 11.2MHz.

Further down the price ladder, products such as the Rotel DT-6000 combine dedicated CD playback with a high-quality DAC, while the new Marantz CD 70 at $750 demonstrates just how quickly the category is filling up at more attainable prices. The point is not that a $750 Marantz buyer is cross-shopping an Accuphase. It is that manufacturers at almost every level have reached the same conclusion: people are buying discs again, and quite a few of them would apparently like something better than the DVD player hiding in the basement.

Why the DP-470 Is Interesting

The DP-470 is unusual because Accuphase is not treating the CD revival as retro fashion.

There is no Bluetooth transmitter, transparent lid, rechargeable battery or attempt to make the player resemble something that escaped from a 1997 Sony catalog. Accuphase has instead applied some of its current high-end digital engineering to a format with a fixed 16-bit/44.1kHz specification and then made the DAC useful enough to justify its existence in a system that probably streams as well.

That tells us something about where the CD revival is going.

The first wave was inexpensive used discs and younger buyers rediscovering physical ownership. The next brought portable players from FiiO, Shanling and Cayin. Marantz, NAD, TEAC and Mission followed with new home components. Now Accuphase is willing to invest in an entirely new high-end player.

Nobody builds a 30-pound CD player with a machined aluminum tray because they expect the format to disappear next Tuesday.

The Bottom Line

The Accuphase DP-470 is not interesting because CD playback suddenly required DSD 22.5MHz support. It is interesting because the company is investing serious engineering resources in a dedicated CD player at the exact moment the supposedly obsolete format is showing unexpected sales growth.

U.S. CD sales rising 16% in the first half of 2026 does not mean streaming is in trouble, and it certainly does not mean vinyl has been dethroned. It does mean that millions of listeners still want physical music they actually own, while manufacturers have recognized that those listeners range from teenagers buying $200 portable players to audiophiles willing to put an Accuphase on the equipment rack.

The DP-470 gives the latter group a substantial transport, balanced analog stage and high-resolution DAC without trying to turn itself into a music server or streaming platform.

Thirty-four years after people started predicting the death of the Compact Disc, Accuphase has built another one.

The CD remains extremely poor at following instructions.

Price & Availability

The Accuphase DP-470 will be released in Japan in late August 2026 for ¥528,000 including tax. Accuphase’s Japanese website officially lists the player, while the company’s international site currently identifies the DP-470 as a forthcoming product.

U.S. pricing and availability have not been announced. We have not converted the Japanese price into a speculative U.S. MSRP because that would not accurately reflect Accuphase’s regional distribution and pricing structure.

We will update this article as soon as official U.S. pricing and availability become available.

For more information: Accuphase CD Players

Related Reading:

The memory crisis has already pushed up prices across phones, laptops, consoles, and cameras, and now it is starting to reach smaller everyday gadgets. Amazon has raised prices on several Echo, Kindle, Fire TV, and Eero products, and Fortune reports that the company has blamed the increases on rising memory and storage component costs.

Some of the jumps are pretty steep. The Echo Dot, which previously cost $49.99, is now $79.99. That is a 60% increase on a device that was supposed to be one of the cheapest ways into Amazon’s smart home ecosystem.

Even the cheap stuff is getting caught up in it

Amazon’s streaming hardware is feeling the pressure too. The Fire TV Stick 4K Max has climbed from $59.99 to $84.99, while the cheaper Fire TV Stick HD is now $39.99 instead of $34.99. The 16GB Kindle has gone from $109.99 to $149.99, and the Kindle Paperwhite is up another $40 to $199.99. Home networking gear has not escaped either, with the Eero Pro 7 rising by $100 to $799.99.

Amazon confirmed the increases, saying the consumer electronics industry is facing “significant increases in memory and storage component costs.” The company says it held off on passing those costs to customers for as long as it could.

The RAM crisis is spreading well beyond PCs and phones

Most of the damage from rising memory prices has so far been easiest to spot in products where RAM and storage make up a noticeable part of the hardware cost. But smart speakers, streaming sticks, e-readers, and Wi-Fi systems still need memory and storage to function.

For months, analysts and industry executives have warned that skyrocketing memory costs would eventually spread beyond smartphones, computers, and consoles to the wider consumer electronics market. We are now seeing more of those warnings play out, with Amazon’s latest price hikes showing how far the memory crunch has started to spread.

Nvidia published some interesting new research on Friday suggesting it’s the harness, more than the underlying model, that is far more important when asking an AI to do long-horizon tasks. A harness is the software wrapper around an AI model — the tools, memory management, and rules that turn a raw model into something that can act on its own.

The TL;DR: Simply by using a custom harness tweaked to handle memory well and including a “supervisor” boss-like component, researchers got Claude Opus 5 to achieve a 100% score on the interactive reasoning benchmark ARC-AGI-3 — a set of 2D games with no instructions, where the model has to figure out how to play and win, similar to how a human would. (That’s a benchmark that has particularly irked rival frontier lab OpenAI.) Without the harness, Opus 5 scored 30%, which was the top result among all the models tested.

Nvidia’s research is another indicator that, while model choice does matter, the model itself — the part that acts as the agent’s “brain” — is a smaller part of an agentic system than many AI users realize, especially for long-horizon tasks. The harness is what makes a model an agent: It handles memory, context, and feedback.

“Generally speaking, the world interprets an agent almost as an API of the model,” Adel El Hallak, vice president of product in Nvidia’s AI unit (pictured above), tells TechCrunch. But an agent is actually more than that. “It is the model. It is the scaffolding around the model, which we call the harness, i.e. the set of tools that it utilizes. It is the runtime and the associated skills and libraries that we give it access to.”

Long-horizon tasks are those that require stringing many decisions together, sometimes over days, to produce completed work. This is in contrast to an AI just spitting out a response to a prompt. Figuring out how to get an AI to do long-horizon tasks without getting distracted and going off in la-la land is one of the holy grails in agentic research.

For example: Microsoft published research in April that tested 19 LLMs on long-horizon tasks involving document editing and discovered that all the models, including frontier ones, filled the documents with errors. (If humans produced work like that, they would be promptly fired.)

Models stringing decisions together on their own have also been caught deleting their users’ files or even whole databases or turning to criminal behavior, from collusion to hacking, to achieve their objectives.

The choice by Nvidia researchers to use this interactive reasoning benchmark for their tests is particularly meaningful, almost funny. A 100% score means that the model can beat the games as well as humans.

OpenAI was so flustered by its models’ abysmal scores (less than 10%) on ARC-AGI-3 that it conducted its own research last month. Like Nvidia, OpenAI discovered that simply by tweaking two settings on the harness, its models tripled their scores.

But none of the models came close to hitting a 100% score, like Nvidia’s researchers achieved. They showed that the harness needs a “supervisor” component that prods the agent in the right direction if it gets stuck.

“The more interesting part was introducing a supervising agent in addition to your main agent that’s doing the work,” El Hallak said. It “almost acts like a CEO to nudge the agent when it goes off direction or starts exploring a path that it might lead to a dead end, or re-explore a path that it had previously trod.”

While the concept of the supervising agent isn’t exactly new, today most agent users are relying on only one layer for their harness, like Claude Code, Codex, or Hermes. Nvidia researchers created their own souped-up harness called the Agentic Variation Operators (AVO).

Note that this isn’t a new Nvidia product. Nvidia instead produces lots of open bits and pieces of tech for building harnesses under the Nemo brand. Some of that tech is commercial, while much is openly available.

Still, Nvidia’s results add to the growing evidence that model choice is far from the only factor in agentic performance. In July, for instance, Databricks published some stunning research that shows that the harness, more than the model, dramatically impacts AI costs.

“You can pick the same model but different harnesses, and you get significantly more cost if you use the wrong harness,” Databricks CEO Ali Ghodsi told TechCrunch. “So you think, oh, this is an expensive model. This is a cheap model. But wait, which harness are you using? That itself can 2x your cost.”

Nvidia’s larger point is to show that open harnesses, like open models, put users in control far more than they realize.

“We believe, and we’re demonstrating with the ecosystem, how open harnesses allow you to turn a lot more knobs to drive up that accuracy,” El Hallak said. “It relates to OpenAI slowing down the training of their models,” as a result of models creating security breaches.

“We believe in having an open agent stack — where you have control across the harness, across the infrastructure, across the runtime — is what’s required for us to usher the ecosystem forward and securely,” he added.

Watch Hull vs Man Utd live streams as the Tigers face a huge test in their first game back in the Premier League against a United side looking to build on a hugely impressive run under now-permanent manager Michael Carrick.

Back in the top flight for the first time since suffering relegation in 2017, Hull surprised many last season by defying a transfer embargo and overcoming Middlesbrough in the play-off final. They’ll start the new campaign as relegation candidates but have significantly improved the squad over the summer, with talented Aussie centre-back Luca Herrington the pick of the signings. However, goals could be an issue, having found the net just once in their final three friendlies.