ETH Zurich has created a wearable muscle or “exomuscle” that can give additional strength to its wearer. It was primarily designed for patients with muscle impairments, but it can potentially help healthy people by providing extra motion endurance.

Myoshirt is something people wear, and its artificial muscle fibers (a cable) run parallel to the wearer’s muscles. Sensors and intelligent algorithms contract the synthetic fibers when the user moves for additional strength. The company describes the cable as an “artificial tendon.”

The system is designed to leave the wearer in control, says ETH Zurich, and it’s possible to tune the strength to the user’s preference. This particular setup helps with arm and shoulder motion, and you can see the kind of motion supported in the video below.

I like that the system is intuitive and doesn’t require large equipment, as often seen in professional medical therapy devices. According to ETH Zurich, all testers found the equipment straightforward to use, and all (impaired or healthy) have shown some benefits.

This is still a lab prototype, but the next step is to try it outside the lab. It’s fair to assume that it will take a while (and lots of paperwork) before this could become a medical device for everyday patients. However, it is a very promising lead that could induce more future innovations.

For patients who no longer have the strength to lift things such as a gallon of milk or other everyday items, such a device could be life-transformative. More details in Nature (subscription required)

Researchers at ETH Zurich have developed a #wearable textile #exomuscle that serves as an extra layer of muscles. They aim to use it to increase the upper body #strength and #endurance of people with restricted #mobility. #myoshirt https://t.co/0kl8X4NkkD pic.twitter.com/roaiktydGY

— ETH Zurich (@ETH_en) June 23, 2022

Samsung continues to announce ultra-high resolution camera sensors aimed at affordable markets. The company just launched the Samsung ISOCELL HP3, a 200 Megapixel camera sensor featuring exceptionally small sensing pixels. The thing is: smaller pixels aren’t always better, as their size makes it increasingly more difficult to gather photos.

Each pixel is only 0.56 μm (square), which is tiny, so there’s a 4:1 pixel-binning mode that would equate to a 50 Megapixel sensor with 1.12μm pixels, or even a 16:1 pixel binning that approximates a 12.5 MP sensor with 2.24μm pixels.

It is also very impressive that the sensor features quarter-resolution pixel-level phase detection autofocus. Each group of 2×2 pixels requires an insanely small lens to achieve that.

Samsung calls this its pixel-binning feature “Tetra pixel,” but the principle is the same as we’ve seen before. Each sensor maker has a different implementation and would create a brand, perhaps for intellectual property or marketing reasons.

It’s important to understand that binning sixteen 0.56 μm pixels into one virtual 2.24μm pixel isn’t the same as one physical 2.24μm pixel. That’s because each 0.56 μm pixel requires some supporting electronics, and some sensing area of the chip might be lost to that. The more pixels you have, the more wires and other things might eat up some sensing area.

The additional electronics might also increase the chance of introducing (electronic) noise that might affect the sensor’s performance. Smaller pixels might also affect the effective dynamic range of the sensor, but fortunately, that might be compensated by capturing more HDR frames using stacked photography.

In the past, we’ve seen degrading performance with some 0.7 μm, so I wonder if this sensor would bump into similar issues.

These ever-smaller pixels allow smartphone OEMs to market affordable camera phones with high Megapixel specifications. That said, it remains to be seen if we’re going to see proportional image quality gains, especially in low-light.

Still, depending on where it is used, this sensor could be an improvement over past sensors at the same price point. For example, it could be slightly larger or capable of faster readouts, allowing 8K recording (or 4K in slow-motion) smartphones at a given price range were not capable of before.

When you encounter a single-cell bacteria visible to the naked eye, you know there’s something very uncommon going on. That’s what biologist Jean Marrie Volland experienced when he stumbled upon this bacteria specie that can be 20mm big! (the image above is an artistic rendering, Mangrove photo by Pierre Yves Pascal; Illustration by Susan Brand/Berkeley Lab)

It turns out that these bacteria from Guadeloupe were spotted before by another researcher. Guadeloupe is a French territory located not far from Puerto Rico. Unsurprisingly, such mega bacteria come from a tropical area as it’s where I would imagine they thrive.

It seems unimaginable to find single-cell creatures that big, but there’s an explanation. This bacteria feeds on sulfur and is oversized because the mangroves where it was found generate copious amounts of that substance. Their family is called Thiomargarita Magnifica. Additionally, there’s no natural predator for bacteria this size.

Scientists believe that there might be even bigger bacteria out there. The good news is that this particular one is a scientific curiosity but doesn’t pose any dangers to humans. It is just a curious exception of nature, a bacteria you can catch with tweezers. If you swim around mangroves, keep your eyes open, you might spot one.

Amazon has announced that the Alexa digital assistant can now technically imitate anyone’s voice based on a few clips, about 1 minute of recording.

If it works as advertised, it would be a technical milestone as computer-generated voices don’t really fool humans and are still a poor choice for video voice-over, for example. You can watch the live demo in the video below (timecode 1:02:38) :

In reality, I found it to transform the voice’s print/tone from the typical machine-generated agent to match the target human voice print. It isn’t easy to quantify the demonstration’s success without knowing the original voice, but it seemed reasonably successful, although still sounding a bit robotic.

The selected sentence was undoubtedly well-chosen for the demo as it lends itself to a slow-paced, almost robotic, reading. The technology is similar to AIs used to transform your pictures in Picasso paintings but applied to an audio stream.

It might sound fun to have Alexa speak with the voice of your favorite celebrity, friend, or family member. However, the Internet has quickly turned its attention to using voice clips from past-away family members. That’s the use case that an Amazon executive was putting forward in the above video.

On the one hand, it may sound like a healing experience to hear the voice of someone close who’s no longer with us. However, it is also a potentially slippery slope with unintended consequences. Many people started to question whether the technology could be misused to impersonate living people and whether we have the right to use voices without consent.

The answer is probably “it depends,” based on the situation and the users. However, one thing is sure: these technologies exist and will get better and better. It’s only a matter of time before synthesized voices are indistinguishable from human ones.

MediaTek has introduced the Dimensity 9000+, a new performance-kicker chip, to supplement the existing Dimensity 9000. As the name suggests, this new processor based on the Dimensity 9000 could be characterized as “the same, but a bit better.”

We are primarily talking about a 5% speed gain for the CPU and 10% for the GPU. The extra CPU performance comes from a higher clock speed for the Prime ARM Cortex X2 core, although the clock speed for the smaller Cortex 510 ones was not disclosed.

Since it’s based on the same design logic and manufacturing process as the original Dimensity 9000, the rest of the specifications remain the same. OEMs already using the 9000 should be able to drop this new one into their mid-year update mobile phone.

This kind of performance delta might sway some customers in the high-end segment of the market, especially gamers or high-end photo/video enthusiasts.

The Dimensity 9000+ is also a technological benchmark for MediaTek as the company wants to show high-performance prowess, especially since Qualcomm has recently launched their Snapdragon 8 Gen1+ competitor.

Desktop workstation computers are usually bulky and heavy because they are optimized solely for maximum performance and expandability. The ThinkStation P360 Ultra takes the opposite direction and is primarily optimized for space. Lenovo tries to cram as much computing power as possible within that sub-4-Liters space.

This computer is a Small Form Factor (SFF) workstation that claims to be 15% smaller than the competition and has key energy-efficiency certifications such as ENERGY STAR or EPEAT Gold. The 8.7×7.9×3.4 inches dimensions make it certainly smaller than Intel’s NUC 12 Extreme which is 14.1×7.4×4.7 inches big.

Inside, you will find an Intel gen12 desktop CPU (16-core max) with a maximum RTX A5000 GPU configuration (16GB VRAM). That’s very much like a mobile RTX 3080. The mobile nature of the GPU better matches the cooling and power envelopes feasible on such a small computer.

Note that the GPU board connects to the motherboard via a proprietary connector, which means that only Lenovo can supply a GPU upgrade. That said, I would not be surprised if customers would in fact never upgrade the GPU, like laptop users.

Four DDR5 RAM slots for the primary system should be enough for most people interested in an SFF workstation. It should be easy to configure 64GB of RAM that way.

Like other SFF computers, the 300W power supply is external, and I would be curious to see if the highest CPU and GPU options would be limited by power. But extra speed and power also mean more heat, so it’s probably fair to think of this workstation as having better cooling than a laptop workstation like the ThinkPad P15 Gen2 we reviewed.

The ThinkStation P360 Ultra has lots of ports. It is also considerably more affordable (and speedy) than a similarly priced ThinkPad workstation. At the same time, it is relatively (trans)portable if you include a ThinkVision M15 mobile display, for example. I wish there were a couple more USB-C or Thunderbolt 4 ports, though.

You can easily extend the connectivity at a fixed location with a Thunderbolt 4 dock to access extra storage and accessories linked to that workspace. Having a one-cable setup is extremely agreeable if you connect/disconnect often.

In conclusion, it’s easy to like the SFF nature of this workstation and its relative affordability. It would look amazing on any desk, making it easy to create a neat environment and provide some level of mobility. That said, it’s for the space-conscious crowd rather than your typical workstation customer, so we’re looking forward to hearing about real-world experiences.

We’re alerted by news organizations now and then about an impending “massive solar flare” that might fry the Internet or even the electric grid. In theory, it’s possible, but there’s also a click-bait factor to this gloom and doom coverage.

But first, what’s a solar flare? Solar flares are immense bursts, or eruptions, of electromagnetic radiation from the Sun. How immense? NASA says it’s “as much energy as a Billion one-megaton nuclear weapons.”

Fortunately, the Sun is far, and space is vast. All that energy is not concentrated in a finite space human can relate to. Instead, the flare will dissipate, and a solar storm might head toward earth in minutes.

Can it do some severe damage? Yes, absolutely. In 1859 telegraph lines caught a fraction of this energy, leading to telegraph paper igniting in some places in the U.S and Europe. In 1972, AT&T experienced some communications shutdowns and redesigned its infrastructure to cope with solar storms.

In 1989, six million Canadians were left without power for nine hours after a solar flare melted transformers in the grid. These events are severe but nowhere near the apocalyptic tone of some reports.

Fortunately, the earth’s surface is protected by its electromagnetic field and atmosphere. Only a tiny portion of the initial energy reaches through to cause the damage mentioned above. The most significant solar flare ever recorded happened on April 2, 2001; no one remembers that event.

Now, damages tend to happen in orbit, where satellites are above the atmosphere and more exposed. There were some instances of satellites being damaged by this type of event, but overall, solar flares concern space agencies rather than the ordinary person on earth.

In San Francisco, where Ubergizmo is located, we’ve had more electricity shutdowns due to wildfire or speculation than solar flares. Our Internet went down for more than 9 hours because a car ran into the street pole carrying the cables.

Next time you read about the big upcoming solar flare catastrophe, it will most likely be all right.

Some U.S. lawmakers have sent a letter to the U.S. Secretary of Commerce to “develop a comprehensive plan that will protect both consumers and the environment by addressing the lack of a common U.S. charging standard.”

The letter continues, “We cannot allow the consumer electronics industry to prioritize proprietary and inevitably obsolete charging technology over consumer protection and environmental health.” (PDF link)

Lawmakers refer directly to the European Union USB-charger 2024 mandate that includes “all smartphones, tablets, cameras, headphones, portable speakers, and handheld videogame consoles.” The goal is to reduce the amount of e-Waste and avoid producing unnecessary products that contain plastic.

In the U.S., lawmakers don’t directly mention the USB-C connector, but it is evident that USB-C is the most widely used standard and that Apple would be the OEM that needs to make some changes. That said, we’ve been hearing about this for many years now, and this has yet to materialize.

Even if Apple switched to USB-C, that would likely not impact its licensing business. Apple already uses a microchip to ensure that all Lighting cables are licensed and working well. The company could simply do the same with USB cables, although it remains to be seen if lawmakers would enforce complete interoperability with other brands.

Many Android OEMs such as Samsung have already stopped shipping phones with chargers and cables. The primary downside is that you don’t get the fastest possible charging unless you buy a high-power charger accessory.

But enforcing a standard such as USB-C would have tremendous advantages too. USB Power Delivery supports extremely high wattage (100W+), and owners of a high-power charger can use it with any USB-C device, from laptops down to toothbrushes.

Phones aside, there is still a large swatch of smaller electronic devices such as razors, clippers, and more that could use USB-C but don’t. Regarding e-waste, some huge gains would come from there, as the Android ecosystem is already going in that direction. However, this has yet to be quantified.

Finally, the “stifling innovation” talks don’t seem very credible regarding charging activity. Although “some” innovation exists, none has come from the connector itself. It will depend on how any legislation will be worded and how much leeway OEMs will have going forward. As long as backward compatibility is present, like with USB-C, I don’t see any innovative limitation – for charging.

Perhaps, it could be reasonable to enforce such regulations on low-wattage (65W and below) or low-speed (based on Mbps) devices, where charging and data transit innovation are less likely to occur.

During its recent WWDC developer conference, Apple mentioned that the next-generation Apple M2 processor CPU cluster would be “18%” faster than the current M1. Numbers found in the Geekbench database are the first 3^rd party data points that corroborate Apple’s claim.

It’s not surprising because chipmakers usually provide reliable and accurate performance estimates for their upcoming chips. But for prudent would-be buyers, it’s always a relief to see multiple confirmations that their next hardware would perform as expected. CPU aside, Apple also says the M2 GPU gets a 35% boost.

CPU Benchmarks have leaked for Apple's M2 chip!

3.49GHz CPU clock vs M1's 3.2GHz
Single-core performance gain vs M1: 11.56%
Multi-core performance gain vs M1: 19.45%

A little bit better than my estimates. I'm impressed!https://t.co/TGHOHw77Ds
Thanks to @amoss_137 for sharing. pic.twitter.com/NS9xODnOdX

— Vadim Yuryev (@VadimYuryev) June 15, 2022

The industry was agreeably surprised by the introduction of the M1 platform, which turned out to be formidable in absolute performance and power efficiency. It looks like Apple’s architecture can scale and has more legs for the foreseeable future.

The M2 chip is at the core of the newly announced M2 MacBook Pro and M2 MacBook Air. The latter has a fan-less design, which is a testament to the low heat dissipation of Apple’s processor. However, the MacBook Pro has a fan and can sustain high-performance better.

Japan has turned to the oceans to produce continuous, reliable, green energy. After years of development and real-world testing, the “Kairyu” (Seadragon?) underwater power generator will enter its commercial phase in the next decade.

Kairyu was a demonstrator designed to harness energy from undersea currents, which is different from tidal power generators. Undersea currents have a slower flow but might occur in a much larger area. It means more generators could be deployed, making the idea entirely scalable.

At a depth of 50 meters, the generator is also located in a safer location than tidal generators. Japan is hit by a significant number of typhoons every year, and they can generate massive waves that could put generators at risk.

Depth and orientation are maintained by a controlled buoyancy and turbine blades that make it easy and low energy to keep steady or come to the surface for repairs and maintenance.

Any undersea project is fraught with challenges as the oceans remain a harsh place for hardware. However, the tests showed this project and strategy could be one of the most cost-efficient, reliable, and scalable ways to harness clean energy.

It is much more efficient than wind energy and would be immensely less intermittent than solar. Japan isn’t an ideal place for solar energy anyway, and tidal generators are difficult to deploy because there are so many naval activities around Japan. Ultimately, these challenges have led Japanese researchers to create a considerably better opportunity.

Geek out with all the details from this IHI PDF document.

The Netherlands-based startup Lightyear has unveiled an electric car with five square meters of built-in solar panels and therefore does not need to be plugged in to charge. In an optimum situation, a range of 43 miles is achievable daily.

The integration of solar panels directly into the car’s chassis is something EV enthusiasts have been asking for many years. However, it is a tough problem as solar cells are typically not designed to be laid out on a curved surface. These cells should also be protected by a layer of hardened glass that tends to be flat.

In any case, Lightyear has solved some of the technical challenges at a price: the Lightyear 0 costs $260,000 and will be produced in limited quantities (less than 1000). However, the company promises that more cost-effective models will come in the future.

Forty-three miles doesn’t seem like much, but the car is capable of more since its battery has enough capacity for a 388 miles range. If your commute is short enough, you might not need to plug the car for months at a time, according to Lightyear.

That said, the maximum range should be considered in light of the relatively modest 170HP produced and the top 62MPH speed. This car is an efficient commuter that might not be the most fun to drive.

Overall, I like where Lightyear is going, and I want them to succeed and push this technology further. After all, it looks like only 1/3 of the car is covered in solar panels, so there’s more room for improvement.

On the other hand, I have to admit it’s much easier and more economical to install solar at home and charge from there or use public solar infrastructure. Finally, the cost of building the car (including carbon footprint) must be considered.

Every organization must reduce its carbon footprint, and the US Military is no exception. As one of the largest buyers and consumers of fossil fuels, there are a lot of opportunities to improve. However, it’s easier said than done, and you want to have plenty of energy to get the job done in potential life and death situations.

Nuclear energy is reliable and can output immense quantities of power, 1 to 5 megawatt. But to make it available to the theater of operations, it also needs to be… mobile! The US military has selected BWXT and other contractors to build such a generation IV transportable nuclear reactor to prove the concept.

The design calls for a nuclear reactor that is small enough to fit in a 6-meters long shipping container for easy transportation. It should also be easy to put into operations, and ideally, it would be up and running within a week from the shipping time.

The reactor could also be deployed in disaster zones that have lost power, and hopefully, it won’t create a disaster of its own. The idea that it could get targeted by enemy forces is not reassuring, but risk management has gone a long way since this idea was considered in the 60s and 70s.

The fuel is not as enriched (less radioactive), and the containment vessels and technologies are much better these days. The DoD says that even if the containment vessel were to fail, it would not be “catastrophic.”

The electrification of the military will take a long time, but having vast amounts of electric power could prove to be very useful not for logistical and net-zero carbon reasons. It could also power direct energy weapons such as lasers or microwave beams. These could prove to be the ultimate projectile interceptors.

As smart television gets ever more powerful, their hardware makes them a prime target for game streaming, effectively turning them into gaming devices, as long as you have the proper subscription.

2022 Samsung Smart TVs will be compatible with the Xbox Game Pass Ultimate, according to Microsoft. All it takes is to install the Xbox client app and log in. Samsung and Microsoft had already worked on making the same 100+ games accessible to Galaxy phones, and this is an extension of that partnership.

Microsoft is actively exploring similar deals with other TV makers, but there are no specifics now. Without a doubt, Samsung has more engineering resources to make this happen early, while others might require help from Microsoft to make it happen.

The Xbox app is compatible with Xbox game controllers connected to the TV. Other controller brands might work, including PlayStation ones, and the whole process is pretty straightforward.

Streaming games can be very rewarding and perhaps even more cost-effective than buying hardware if you have an excellent Internet connection. For example, I rarely play because I want to check a few new games here and there. I can upgrade my account when I want to, and overall I find game streaming to work very well.

I hope more TV manufacturers will join Samsung in making Xbox games playable via streaming, and hopefully, other clients such as GeForce Now can be supported as well.

NAVER, South Korea’s leading search engine, is turning its new 1784 building into a giant place where “humans work in harmony with robots.” The plan is to have 100 robots autonomously roaming around the building to provide various services to humans. 1784 is a reference to the year the industrial revolution started.

The ROOKIE robot will be the primary platform for this experiment. Robots will even have their robots-only elevator to go floor to floor quickly. Robots will be “brainless,” meaning that they will be controlled by ARC, a centralized artificial intelligence via a dedicated 5G network. NAVER Cloud powers ARC.

ARC stands for Ai Robot Cloud and encompasses the AI software, the low-latency network, and the cloud computational resources. For instance, the AI includes ARC eye and ARC brain, two critical neural networks that enable the operating of the robots. The robots use ARC eye to recognize where they are and find the best way to their destination. No GPS is needed.

This architecture is contrarian to the “edge AI” utilized by others. With Edge AI, the robot hosts the computing platform and may not need constant network access. On the other hand, NAVER’s centralization might make software updates and incremental improvements easier, which is critical in the testing phase. Additionally, it can take advantage of the denser computational power of a data center, and make the robots more efficient, both in cost and battery life.

NAVER even built a “digital twin” to get the AI development going. Digital Twins are essentially virtual 3D worlds that mimic the physical facility in which robots will operate. Within these worlds, robots can be taught and tested safely and quickly. Many scenarios can be devised, and it’s even possible to run the simulations much faster at many times the speed of the real world. “Digital twins” is leading-edge technology.

Robots aside, the CLOVA FaceSign device also identifies people walking in the facility, even if they are wearing masks. NAVER even envisions having AI-based physical exams and contactless smart assessments based on camera footage. From a privacy perspective, it would be very interesting to see how people react to that.

The 1784 building has been designed with the latest technology and trends, including the recent pandemic experience. For example, each floor has its HVAC system, so pathogens don’t spread to other floors. For convenience, it also has a fully-featured clinic.

The double-skin façade and radiant cooling (cooling the ceiling) should make it energy-efficient, especially in the scorching hot Korean summer.

This project seems like a grand experiment, and I’m looking forward to testing it in the real world. I wonder what services these robots will provide and how they will make life easier for people in 1784. That might be a sneak peek of where things are going.

With the new Honor 70 series launch in China, we’re getting a peek at what’s coming later to the global market from the company: three new phones priced between $400 and $650, the Honor 70, Honor 70 Pro, and Honor 70 Pro+.

The differentiation between these is relatively straightforward, with mainly the screen size, processor, and camera hardware getting more potent as the price goes up. Even the base Honor 70 looks great. It is powered by a Snapdragon 778G+ with (up to) 12GB of RAM and (up to) 512GB of storage.

There are two high-performance camera sensors in the back supplemented by a dedicated 2MP bokeh sensor for portrait shots. I like that Ultrawide cameras are now getting powerful sensors, almost like primary cameras. That makes a lot of sense since Ultrawide is the second-most-popular use case in mobile photography.

At the top of the 6.67-inch display (120Hz), a 32MP selfie camera ensures good quality video calls, well beyond what personal computers can typically do.

The 4800 mAh battery is close to the maximum 5000 mAh found on high-end large phones, plus it can be replenished at 66W of power, which is significantly faster than other established brands.

The Honor 70 Pro is the next upgrade with a larger 6.78-inch display and, more importantly, a dedicated telephoto 3x zoom lens with stabilization (OIS). It is also powered by the MediaTek Dimensity 8000, which is generally more performant than the Snapdragon 778G+.

The 4500 mAh battery is slightly smaller than on the non-Pro Honor 70, perhaps because of the telephoto camera module’s thickness? However, the battery can charge at 100W power, making it extremely fast to replenish.

Finally, the Honor 70 Pro+ comes with the high-end MediaTek Dimensity 9000 SoC platform. It is significantly faster than the Dimensity 8000 and offers the best image processing hardware that MediaTek offers. It would be interesting to see if perceptible image quality (IQ) is induced by the increased computing power of the Dimensity 9000’s image signal processor (ISP).

Half of the year has already gone, and it’s time to check which five phones are the best at slightly different price points, going from $899 to $1199. We use many criteria, including the industrial design, speed, camera prowess, and user experience, among other things.

We evaluated the camera potential using our proprietary Ubergizmo CAMERA HW score.

You can learn more by going to the benchmark page.

Editor’s note: prices in Euros are from the official web stores in France, as of May 27, 2022.

5 – Xiaomi 12 Pro ($999, €1099.99)

The Xiaomi 12 Pro is a very agreeable and good-looking smartphone with a fingerprint and scratch-resistant back cover that keeps it looking clean. The phone is powered by the latest Snapdragon 8 Gen 1 platform that ensures it can compete at the highest levels among Android phones. This phone did really well in the synthetic CPU and GPU benchmarks, and this might be your handset of choice for sustained performance.

Its camera has three 50MP sensors (Primary, Ultrawide, Zoom), giving it a strong camera hardware configuration. However, some sensors have tiny pixels (0.7 microns or less), which could limit the amount of captured light under challenging conditions. We computed a CAMERA HW score of 177 for the Xiaomi 12 Pro.

The zoom lens is optimized for portrait photography (48mm), not for long zoom (200mm+) performance. There’s nothing wrong with that, but we’ve noticed that most people find long-zoom shots very useful.

Unfortunately, this handset is not certified as waterproof, and tests show that its battery life is slightly more limited than other high-end smartphones. Xiaomi makes up for the battery life with an uber-fast 120W charging, one of the fastest.

4 – OnePlus 10 Pro ($899, €919.00)

The OnePlus 10 Pro is an exciting Android phone because it has excellent performance and top-notch battery life, thanks to its 5000 mAh battery and Snapdragon 8 Gen 1 processor. Additionally, if you ever find yourself in a battery emergency, it can quickly charge at 65W, which is very fast (but not the fastest) in the industry.

Our data shows this phone packs the highest battery density (mAh per cubic inch) on the list, partially explaining its endurance.

The impressive-looking camera module hides a triple-camera system in the rear, with a 48MP primary, a 50MP ultrawide, and an 8MP zoom camera. OnePlus works with Hasselblad in the development and tuning of the camera. This cooperation leads to particular color profiles that could sway Hasselblad fans.

On a more objective note, our CAMERA HW score of 164 indicates that the

OnePlus 10 Pro’s primary camera could benefit from having better hardware. Its zoom and ultrawide camera modules are also not as impressive as its competitors in this list.

3 – Apple iPhone 13 Pro ($999, €1159)

The iPhone 13 Pro represents the best Apple can muster for now, and it boasts impressive performance and battery life. Most benchmarks show an unquestionable lead in CPU performance and longevity, so kudos to Apple. The scores were a bit low in some graphics benchmarks, but actual games should run fast enough.

On the camera front, Apple has done a great job updating its camera system with a potent primary camera that seriously competes against Samsung and Google. If you care about iOS and photo quality, we recommend getting the 13 Pro instead of the non-Pro models. It’s worth it, trust us.

This camera snatches a CAMERA HW score of 191, clearly in the top percentile, even at this price range. Unfortunately, it’s not the absolute best overall camera on the market, and you would need to switch over to Android for that.

On another topic, the lack of fingerprint unlock might be frustrating for some users. If you need a larger display and battery, you can pick the iPhone 13 Pro Max. It has the same performance and camera experience as the smaller iPhone 13 Pro.

2 – Google Pixel 6 Pro ($899, €899.99)

The Google Pixel 6 Pro is the most potent Pixel phone currently available. People love it for its “pure” Android OS experience, but we think that the photo editing and camera style are also among the most popular reasons to own a Pixel phone.

We find the industrial design neat and easy to distinguish from the competition, thanks to the iconic camera bar in the back, which is also a convenient way to integrate a large camera module into the phone.

Powered by the Google Tensor processor, the Pixel 6 Pro comes with beefier AI capabilities for its generation, contributing to more advanced image processing effects. Performance is comparable to Snapdragon 888, and while it’s not the fastest today, it is very respectable and sufficient, even for advanced users.

For classic 3D graphics as measured by the GFXBench Manhattan 3.1 test, the results are not very different from the S22 Ultra. However, in more modern graphics tests, Snapdragon 8 Gen 1 can render up to 50% faster, according to our tests. If you play the most modern 3D games, a Snapdragon 8 Gen 1 might be preferable.

We like Google’s investment in the camera, as it scores a superb 197 in our CAMERA HW benchmark, which is much higher than the Pixel 5’s 121 score. Experts would agree that the photo quality is proportionally better, so kudos to Google for recognizing that it’s not “all about camera software” and that physics matters (a lot) in photography.

The Pixel 6 Pro’s battery life could be better, and it slightly loses out to some high-end Android phones in various battery tests, but most people should get through the Day.

1 – Samsung Galaxy S22 Ultra ($1,199  €1,259.00)

The Samsung Galaxy S22 Ultra is currently the best Android phone for power users and the most expensive in this selection.

Like others, it is powered by the Snapdragon 8 Gen 1 processor, but on top of that, it delivers world-class industrial design and display technologies. It also features the best e-pen support in the smartphone market, and the pen user experience is unchallenged. The Galaxy S22 Ultra is an absolute multimedia and productivity beast.

That’s not all. Its camera hardware is also one of the best in the world and slightly outpaces its S21 Ultra predecessor. Reviewers are raving about the camera’s prowess, and we’re not surprised.

The Galaxy S22 Ultra has the highest CAMERA HW score on this list, at 203. The overall photo quality is high, especially the mighty zoom lens that packs a serious punch. Unlike many others, this camera is very balanced and shows excellent performance in wide, ultrawide, and zoom shots.

The battery life is slightly shorter than last year’s S-series, but it has the largest

display compared to other phones on this list. Fortunately, Samsung has increased the charging power to 45W, so you can replenish quickly and be on your way. Unfortunately, you’ll have to purchase that fast-charger as none is provided in the box.

Conclusion

All these handsets are high performers, but synthetic CPU and graphics tests clearly show some differences (see the charts below), mostly in favor of the iPhone 13 Pro and at the expense of the Google Pixel 6 Pro. Only gamers will really feel the speed difference between these phones, and they are all plenty fast for the rest of us.

Apart from their respective design and sizes, the camera quality might be one of the most critical differentiators, with the Galaxy S22 Ultra and iPhone 13 Pro leading the charge. The Google Pixel 6 Pro’s camera is slightly behind but significantly less pricey.

The Xiaomi 12 Pro and the OnePlus 10 Pro are worthy camera contenders, but they are objectively slightly behind. On the other hand, both have exciting battery characteristics: the OnePlus 10 Pro has an excellent battery life while the Xiaomi 12 Pro has a crazy-fast battery charging, leaving other phones in the dust.

Each phone has its pros and cons, but you can’t go wrong with any of them as they are all high-quality products. There has never been a better time to upgrade your phone.

With the new Dimensity 1050 SoC platform, MediaTek is launching its first 5G processor with an integrated modem that supports mmWave and Sub-6 5G. As a result, users will experience higher average 5G performance with potentially extraordinary speed spikes if a mmWave base station is in line of sight. In the USA, many are installed in stadiums and other high-traffic venues. In time, their density will increase.

The Sub-6 GHz spectrum enjoys much broader network coverage, although at lower speeds than mmWave, but it is still faster than 4G. The Dimensity 1050 can switch seamlessly between different spectrums depending on network conditions.

Indoors, this processor supports WiFi 6E and should provide fast and low-latency connectivity. If anything, many users’ home internet won’t be fast enough to saturate the Dimensity 1050 capabilities. Gamers might be the primary beneficiary of WiFi 6E, but the average experience is well above the previous generations.

From a computing standpoint, this is a chip destined for the mid-range market, with an 8-core CPU (2 performance, 6 efficient cores) and a Mali G610 MC3 graphics unit that can control a 1080P 144Hz display.

It benefits from some advanced imagery hardware introduced with the Dimensity 9000 but with a smaller performance envelope due to its size, process node (6nm), and transistor count. From the specifications, it looks like MediaTek has tuned it to be competitive and should appear in mass starting from Q3 2022.

Qualcomm recently announced its new Wireless AR Smart Viewer reference design for developers. The idea is to demonstrate what the next-generation AR devices could (or should) look like and what the Snapdragon XR2 Platform can do.

It is beautiful and highly functional for something that’s supposed to provide a technological baseline. At the very least, it can provide AR functionalities similar to the very cool Lenovo ThinkReality A3, which can replace five external monitors.

But this Wireless AR Smart Viewer does it without the cord and feels much lighter on you. They have a 40% thinner profile, and the 1080p displays (one per eye) provide an excellent AR experience on par with bulkier devices.

When I tried it, it was a world away from your typical AR experience in which the headset is distracting because of its bulk and weight.

It was like wearing a pair of sports sunglasses, and the lack of wire (even going from the glasses to the phone in my pocket) introduced an unprecedented sensation of freedom. (Thanks to Judie @ Geardiary for the photo!)

Qualcomm has correctly identified primary AR pain points and is working aggressively to address them within its ecosystem by using a clever mix of computing and telecommunication.

Electronics embedded in the glasses package the sensor data and communicate it via Wi-FI 6E (Qualcomm FastConnect 6900) and Bluetooth 5.3. On the other end, a smartphone receives the stream and does the heavy computing (3D, AI, video encode), all with a 20ms motion to photon latency, according to Qualcomm. The glasses decode the video stream sent by the smartphone, and the process repeats for every frame.

The communication happens on a dedicated WiFi band, so your WiFi Internet access is not occupied by the communication between glasses and phone. By the way, the 650 mAh battery can keep the experience going for ~30mn. It’s also possible to use USB-C for power.

Enabling such a design is an essential step toward the mass-market adoption of AR glasses. The most critical aspect of any wearable technology is to get people to…. Wear it. Previous generations were simply too bulky.

Researchers at Northwestern University are working on (PDF link) sub-millimeter robots that are capable of “walking” without complex hydraulics and other systems that would be impossible to build at such a small scale for now.

This 8-legged crab (+2 arms) can walk sideways, thanks to materials specially developed. The technology works by building legs that remember their original engineered shape. When heated up with a laser, the leg changes shape, allowing a motion to occur. Repeat the process, and you’ve got a “walk.”

The laser effectively controls how the robot moves, and it’s a way to offload some of the power sources and mechanics away from the robot. Depending on the material and laser strength, some designs could even “jump,” but this one moves relatively slowly.

This kind of technology is one of the ways tiny robots can move and is popular in the nano-robots world. It is believed that eventually, these devices will be able to operate in ordinarily inaccessible places, including the human body.

It has been long theorized that nanorobots would someday perform surgery tasks in a completely non-invasive way, such as destroying tumors or repairing tissue damage.

In science fiction stories, you could “back up” your body’s state and have such machines continuously restore it, making you effectively immortal. Fun, but let’s not get ahead of ourselves.

Qualcomm just held a launch event in China called “Snapdragon Night,” where the new Snapdragon 8+ Gen 1 platform was announced.

The “plus” update is a performance-kicker version of the Snapdragon 8 Gen 1 hardware platform and brings noticeable performance and power efficiency improvements.

These improvements come from basically a two-pronged approach to the chip’s development. First, Qualcomm uses a new 4nm semiconductor process at TSMC, which is even better than the previous Samsung’s 4nm one. Both platforms will be produced in parallel as part of Qualcomm’s strategy of manufacturing diversification.

Secondly, the Qualcomm physical design team has had more time to optimize the chip’s layout and internal datapaths that might have maxed up at higher frequencies. As a result, Snapdragon 8+ Gen 1 reaches between 6% and 12% higher frequency clocks across various units.

For context, the “physical design” of the chip refers to how the transistors and wires are physically laid out. At the same time, the overall logic of the design remains identical to the original version. In short, it is better built and manufactured.

These improvements lead to increased performance, with Qualcomm claiming a 10% CPU performance boost, a 10% GPU clock increase, and a much-improved power efficiency (performance per Watt): up to 30%, says Qualcomm.

This product’s power efficiency is awe-inspiring as we usually don’t see these numbers jump that much within the same generation. Using less power for the same task also reduces the heat and improves sustained performance. It means real-world gaming/XR isn’t as throttled due to heat build-up. The handset cooling system significantly affects this as well.

I’ve had some hands-on time with a couple of Qualcomm’s development handsets, and we saw benchmark numbers in line with the claimed speed improvements.

This new platform will allow Qualcomm’s customers to build ultra-premium products with the absolute best Android performance and gross margins. You can expect the usual OEMs such as Motorola, Honor, Xiaomi, or OnePlus to build products based on this new SoC (complete specs in the official PDF).

Simultaneously, Qualcomm launched the new Snapdragon 7 Gen 1 platform, which serves the Premium smartphone market (vs. High-End).

The goal is to bring many high-end features to a lower price point, including Gaming or Camera features that are highly coveted and are differentiators from an OEM marketing perspective.

For example, Qualcomm says that Snapdragon 7 Gen 1 can render 3D 20% faster than its predecessor, the Snapdragon 778G. Also, the new chip can support Camera capture of up to 200 MP vs. 66MP before). These are all very appreciable gains within a similar price envelope.

Finally, Qualcomm mentioned during its briefing that its Frame Motion Engine had been integrated into Snapdragon 7 Gen 1. It’s a hardware frame-interpolation feature that can increase the perceived games FPS without increasing the graphics computations proportionally. It’s great to preserve battery life as well.