Unlocking a smartphone securely with one’s face is a very convenient and highly sought-after feature. However, it is not as secure as fingerprints or long passwords because of the technologies involved in various face ID implementations.

Looking at one’s facial features via a 2D camera lends itself to being fooled by something as simple as a printed photo, and we’ve seen that in the past.

More sophisticated methods using a 3D capture of one’s face defeat the printed photo but can still be tricked by a more sophisticated 3D-printed mask. It’s been valid from day 1. The most sophisticated masks can unlock today’s most expensive phones.

Ultra-realistic 3D masks for testing. They can defeat modern phones’ face unlock systems

trinamiX GmbH introduces a new paradigm that makes it much harder to bypass. Its system checks that a real, live person is looking at the sensors, and it’s all hidden under a phone’s OLED screen, with no notches involved. Now, it only works with OLED displays, so more affordable screens using IPS LCD won’t be compatible.

The system is so performant that it is already seen by the FIDO Alliance, the Android Biometric Security, and the IIFA (International Internet Finance Authentication Alliance) as the new mobile biometric security “gold standard.”

According to trinamiX, the system comprises an IR camera, a dot projector, and a flood illuminator. Essentially, this trio works two types of facial recognition in parallel: the first detects the face like other systems do today. However, the second sees the presence of (live) human skin.

And that skin detection is what effectively defeats today’s best masks. Even though these 3d-printed masks can be highly accurate from a geometry and color perspective, they have a weak point: their surface does not reflect light like human skin.

This is a blast from the past because one of my former colleagues worked on real-time skin rendering (see the Advanced Techniques for Realistic Real-Time Skin Rendering chapter of the GPU Gems 3 book). That’s when I was first exposed to the fact that human skin is made of multiple layers, some of which are slightly transparent.

Incoming light enters the skin, bounces inside, and some of it is reflected out, giving the typical skin soft appearance we all know. trinamiX’s hardware design detects this exact human skin subsurface reflectance signature.

It does so thanks to its dot projector and by analyzing how the light diffuses around each dot. Every other material will reflect in a completely different way, and that’s 3d-printed masks can’t fool trinamiX’s face authentication. If you wonder, their system is flexible enough to work with face masks and can be tuned for various situations via software (or AI).

Essentially, trinamiX made it much harder for potential hackers to build a 3D mask that could bypass its liveness detection. Someone might create a “mission impossible” type mask that perfectly mimics the human skin reflectance, but that remains to be proven.

Additionally, trinamiX says its design can be built at the same cost envelope as existing ultrasonic fingerprint readers, using different sets of affordable sensors that OEMs can choose from, depending on their vendor preferences.

trinamiX Face Authentication at the Qualcomm Snapdragon Summit 2022

Finally, this face authentication system has already been demonstrated with Qualcomm’s latest Snapdragon 8 Gen 2 platform, using the Snapdragon Trusted Execution Environment block, which ensures the complete privacy and security of biometric data.

We’ll hopefully see ultra-secure face unlocking in 2023. I sincerely hope we’ll get both face and fingerprint readers hidden under OLED displays because both are incredibly convenient under different circumstances. It looks like Android is going to have the upper hand when it comes to face-unlock security.

The new Lenovo Slim 7i Pro X is an excellent addition to Lenovo’s thin & light consumer laptop line-up because it packs a big punch, thanks to its Intel 12th generation H-Series CPU paired with a discrete NVIDIA GPU. They makes the Slim 7i Pro X (i stands for Intel) a very attractive laptop for Creative users who seek an ultra-compact laptop.Any consumer who likes to have great multimedia and […]

Yesterday at the Snapdragon Summit, Hugo Swart, Vice President, XR Product Management, Qualcomm, unveiled the Snapdragon AR2 Gen 1, a hardware platform made of three chips that will power the next-gen augmented reality glasses.

The new Snapdragon AR2 Gen 1 is the latest addition to Qualcomm’s extended reality (XR) product line.

The system is made of an AR processor, located inside the right arm of the glasses, an AR co-processor located on the front in the middle, right behind the camera module, and a Qualcomm FastConnect 7800 connectivity chip located inside the left temple.

For once, Qualcomm did not integrate all the chips in a single large SoC but has split the hardware into three small packages to help OEMs deliver thin and light designs for their next-gen AR glasses.

To enable the slim form factor, the Snapdragon AR2 distributes the processing workload between the glasses and a Snapdragon-powered smartphone, PC, or other compatible host devices.

According to the chip maker, the main processor occupies a 40% smaller PCB area on glass compared to the Wireless AR Smart Viewer Reference Design powered by Snapdragon XR2.

The AR2 platform achieves a 50% boost in AI performance while consuming 50% less power than the previous generation. Qualcomm designed the AR2 to power glasses that will compute using less than 1W.

The AR processor, optimized for low motion-to-photon latency (sub-9ms), supports up to 9 cameras for scanning the environment and for the user.

The AR processor features a hardware acceleration engine for better user motion tracking and localization, and an AI accelerator to reduce latency for input interactions such as hand tracking or 6DoF. Finally, the chip manages the display output.

The AR co-processor aggregates camera and sensor data and features eye-tracking for foveated rendering, to optimize graphics rendering performance. Additionally, the computing unit delivers iris authentication.

The connectivity chip is based on Qualcomm FastConnect 7800, a connectivity system that supports Wi-Fi 7. Thanks to this tiny connectivity chip, OEMs will be able to deliver sub-2ms latency between the AR glasses and the smartphone host device. Qualcomm provides the FastConnect XR Software Suite 2.0, to help developers better control the XR data, reduce latency and jitter, and prevent interference.

Qualcomm offers the Snapdragon Spaces XR Developer Platform, an OpenXR SDK tailored to Qualcomm XR products that provides commonly needed software functions that most XR apps require.

Since Snapdragon Spaces adheres to the open-source specifications, the AR2 system of three chips could also work with a non-Snapdragon phone or mobile device. We can safely assume that Snapdragon Spaces works optimally with the most recent Snapdragon mobile devices.

History of Qualcomm XR platforms:

Snapdragon XR2+ Gen 1, Qualcomm’s latest chipset for VR headsets, combines Mixed Reality (MR) and Virtual reality (VR), launched on October 11, 2022, and powers the Meta Quest Pro announced at Meta Connect 2022. The XR2 Plus delivers 50% higher sustained power and 30% improved thermal performance over its predecessor.

Snapdragon XR2 5G (XR2 Gen 1) launched on December 4, 2019, an XR platform that powers, among other devices, the Lynx R-1 Mixed Reality (MR) headset, and the Wireless AR Smart Viewer reference design.

Snapdragon XR1 launched in May 2018 that powers the Lenovo ThinkReality A3 smart glasses launched at CES 2021.

At the Snapdragon Summit, Chris Patrick, SVP and GM, Mobile Handset, Qualcomm, unveiled the Snapdragon 8 Gen 2, Qualcomm’s flagship SoC for premium smartphones, and the successor of the Snapdragon 8 Gen 1 and the Snapdragon 8+ Gen 1.

System Performance: Kryo CPU Cluster

Let’s start with system performance, the standard hot topic for any new generation of processors. Qualcomm has improved all performance aspects over last year’s Snapdragon 8 Gen 1.

The new Kryo CPU cluster is now based on one ARM Cortex X3 (Prime, 3.19 GHz), four performance cores (2.8 GHz), and three efficiency cores (2.0 GHz).

Last year’s generation featured one ARM Cortex X2 (Prime, 3 GHz), three Cortex A710 (2.5 GHz), and four Cortex A510 (1.8 GHz), compatible with the ARMv9 Gen 1 CPU architecture.

The extra performance core and the new ARMv9 Gen 2 architecture drive most of the performance boost, while the minimal increase in CPU frequency also provides an extra nudge.

Qualcomm estimates this new CPU cluster to be 35% faster and up to 40% more power-efficient than the Snapdragon 8 Gen 1, the previous market leader for flagship Android smartphones.

The 40% power-efficiency number is surprisingly high, and Qualcomm claims it is due to the Kryo CPU’s new microarchitecture. Without a clearer explanation, I will assume it is based on the best case study, and upcoming benchmarks might shed more light too.

The new Kryo CPU features an 8MB L3 cache, a 33% increase over the 6MB L3 cache in the previous generation. Augmenting the cache size helps to avoid expensive DDR memory access whenever possible.

The new chip offers support for LP-DDR5x memory up to 4200 MHz and with up to 16 GB of memory density.

Artificial Intelligence (AI) Engine – Snapdragon Smart

One of the top areas “that will be critical to Snapdragon 8 Gen 2 is artificial intelligence. This platform is purpose-built for AI. […] At the heart of our AI experience is the Qualcomm AI Engine [which] is made of several technologies including Adreno GPU, Kryo, Sensing Hub, Memory, and Hexagon” said Francisco Cheng, Senior Managing Director, Product Marketing, Qualcomm, during our briefing.

Across all technologies involved in processing AI models, the Qualcomm AI Engine 8th generation improved the performance by up to 4.35X and the performance-per-watt by 60% over the 7th generation.

Hexagon, AI co-processor

Qualcomm defines Hexagon as a “fused AI accelerator architecture,” meaning that the Hexagon logic block is made of several types of computing units. Qualcomm made multiple improvements throughout Hexagon, and the most notables are the 2X size increase of the Tensor accelerator and the addition of a dedicated power delivery system.

The major changes in Hexagon were made to address the ability to run more complex neural networks, particularly around transformer networks, to deliver, among other things, multi-language translation in real-time.

Qualcomm’s Micro Tile Inferencing is a new technique that works by splicing neural networks into ‘tiles’ that will be dispatched to available compute units in Hexagon. Qualcomm did not go into much detail, but in our experience, tiling data is typically utilized to increase parallelism (process the tiles in parallel to maximize hardware utilization) and potentially increase cache coherency and memory bandwidth. Both can make a significant impact on overall performance.

Last year, Qualcomm added the support of mixed precision INT8+INT16, (automatic switch between INT8 and INT16 precisions) to its non-automatic mixed precision technology, named “all precisions support” (INT8, INT16, FP16).

In the new AI Engine, Qualcomm added the support of 4-bit Integer precision (INT4) in the “all precisions” mode, so select AI algorithms and neural network models can process faster across the Hexagon compute units. On top of that, INT4 allows Hexagon to run more models simultaneously (parallel computing).

Each AI model could use a different level of precision, as some AI-driven tasks will deliver good enough results using an INT4 low level of precision for faster processing, while others may require higher levels of precision such as INT8, INT16, FP16, and FP32 (Floating Point stored on 16 or 32 bits).

Qualcomm said that the new AI Engine uses INT4 in “cinematic mode” for real-time blurring of the background (video bokeh), with no visible difference in image quality. Previously similar AI models used INT8.

4th generation Sensing Hub with a dual-AI processor

The Sensing Hub is the independent low-power multi-core logic unit that functions when the rest of the SoC is asleep. It processes in real-time the contextual data streams such as voice, audio, sensors, and connectivity.

Last year, Qualcomm introduced a new feature for the 3rd generation Sensing Hub, the Always-on Camera, which has now become the Always Sensing Camera. The Always-on Camera allowed the phone to automatically lock when the user does not look at it or wake up each time the user glance at the sleeping phone.

With the new feature, the camera can recognize and process a QR code placed on the optical module when the smartphone is asleep (+display off), with the previous processor, the phone had to be turned on to manage a QR code.

Addition of a second AI processor with 50% more memory

Qualcomm has doubled the Sensing Hub’s AI computing capacity by adding a second AI processor. On top of that, 50% more internal memory is now available. The extra memory allows the system to run new AI models which were previously unavailable but also better versions of the ones that existed before.

Qualcomm mentioned 17 AI-based activities tracked by the new Sensing Hub, and the ones which caught our attention are [human] Activity Recognition, Voice Activation, and Eye-Tracking.

Hexagon Direct Link and Cognitive ISP

Qualcomm introduced Hexagon Direct Link for the first time, a feature that provides dedicated physical communication channels (bi-directional buses, in hardware lingo) to other logic blocks such as the GPU, the memory, and the Spectra ISP.

Alongside Hexagon Direct Link, the chip manufacturer announced its first Cognitive ISP, a compute unit located inside Hexagon that works in connection with the Spectra ISP for faster processing of the AI-driven imaging features.

Snapdragon Sight:  upgraded camera performance with Cognitive-ISP

Cognitive ISP for better camera performance

The Cognitive ISP, located in the Hexagon AI co-processor, communicates in real-time with the Spectra ISP that processes the pixels captured by the camera. The Cognitive ISP executes the Multiple Semantic Segmentation AI model on the images sent by the Spectra ISP and tells the Spectra to apply different image-enhancing filters on the different identified zones of the image (up to eight segments).

The Cognitive ISP can locate boundaries and recognize objects in an image such as faces, bodies, skin, clothes, sky, plants, flowers, cats, beaches, buildings, and more.

This new feature, named Realtime Segmentation Filter, enables the system to precisely adjust colors, contrasts, and any aspects of photography to each segment of the image, in real-time. Developers can select different configurations and filters.

In other words, the Cognitive ISP is the brain, and the Spectra ISP is the muscle.

The new architecture allows the camera system to receive the segmentation filter a lot faster than before, and processes AI filters in real-time, even before the user presses the capture button, according to Qualcomm.

Graphics and Gaming – Snapdragon Elite Gaming

At the Snapdragon Summit, Chris Patrick, SVP and GM, Mobile Handset, Qualcomm, said that the new Adreno graphics processing unit (GPU) brings a 25% increase in graphics rendering performance than the previous generation while delivering 45% better power efficiency.

This means the baseline performance will increase for nearly all apps that depend on the GPU, like games.

Adreno GPU Display

The Adreno subvisual system also includes the display, and Qualcomm announced the support of HDR Vivid, a new HDR standard launched in March 2021 by the China UHD Video Industry Alliance (CUVA). HDR Vivid and Adaptive HDR are the latest additions to the other formats already supported by Adreno GPU Display, including HDR10, HDR10+, and Dolby Vision HDR.

Adreno GPU for games

Adreno GPU will be the first mobile platform to support Unreal Engine 5’s Metahuman, a framework that allows game developers to create ultra-realistic humans programmatically. Unreal Engine 5’s Metahuman characters were previously only usable in desktop games. That’s a testament to Qualcomm’s graphics software’s maturity.

The new-gen Adreno supports Vulkan 1.3 and delivers a 30% Vulkan performance boost compared to the previous generation.

Real-time hardware-accelerated ray tracing

Hardware ray tracing was the star feature of the show, as it pushes realism in games to the next level. Bringing real-time ray tracing to mobile gaming is a big deal, as it is compute-intensive, and only powerful gaming rigs can render the more advanced ray tracing techniques at high framerate, like global illumination for the whole 3D scene.

On Snapdragon 8 Gen 2, game developers can integrate basic ray tracing techniques such as soft shadows and accurate reflections. The traditional rendering methods such as shadow maps yielded hard shadows because they were not nearly as accurate compared to ray-traced shadows. Previous reflections rendering methods such as screen-space reflections were limited and could not reflect an object that was off-screen. Ray tracing can.

In short, ray tracing mimics the light’s behavior in the real world by following the path of light in a 3D world from the camera through the objects in the scene and to an eventual light source.

Dave Durnil, Global Head of Gaming & Snapdragon Studios, Qualcomm, showed us a cool demo from the popular game Justice Online Mobile with the ray tracing on and off. We could clearly see the difference, with ray tracing, the reflections in the water and on the curved objects were realistic and we could observe the soft shadows.

He told us that War Thunder Edge and Blue Reflection games for Android will feature Adreno-accelerated ray tracing.

Snapdragon Sound: dynamic spatial audio with head tracking

Sarah McMurray, Staff Manager, Product Marketing, Qualcomm, introduced Spatial Audio with dynamic head-tracking, a new feature of the Snapdragon 8 Gen 2 that delivers more realistic 3D sound with low motion-to-sound latency. Spatial Audio is a technology dedicated to immersive gaming and multimedia experiences.

Connectivity: 5G AI and Wi-Fi 7 support

5G Modem-RF System

The 5G connectivity’s improvements revolve around three new features. The first is the integration of a “5G AI Processor” compute unit inside the 5G system. This essentially helps with “tuning” the 5G modem parameters to obtain optimal performance in any situation. It is the closest you can get to having a (human) modem engineer tweaking things constantly. AI could also help with signal error correction to avoid unnecessary data re-transmission.

Next, the 4X carrier aggregation further enlarges the data pipeline available for 5G communications, resulting in higher peak downloads (an impressive 10 Gbps).

Finally, it is possible to use two 5G SIM at the same time, thanks to the 5G Dual SIM Dual Active feature. It opens the door to even more data parallelism, but you might end up paying two subscriptions of course.

FastConnect Wi-Fi 7 and Bluetooth connectivity system

Snapdragon 8 Gen 2 comes with Wi-Fi 7, the upcoming standard for home connectivity. Qualcomm uses extremely advanced techniques such as high-band multi-link to use as much of the available spectrum as possible.

The company says its Wi-Fi can reach 5.8 Gbps and, more importantly, sustain a latency below two milliseconds in a sustained manner. Low latency could persuade wired ethernet users to finally switch to Wi-Fi.

Along with the Dimensity 9200 SoC, MediaTek has released the MediaTek T800 5G Modem that accompanies the Dimensity 9200. It is slated to be utilized beyond the smartphone market and may find itself in future industrial, IoT, Laptops, and machine-to-machine applications

The T800 5G modem is designed to reach peak speeds of 7.9Gbps and 4.2Gbps (download/uploads). The upload speed is the best we’ve seen, and content creators can be assured that companies like MediaTek care about uploads as much as downloads.

It would be hard to reach these theoretical speeds in the real world, even in ideal conditions. Still, it also has to do with the quantity and quality of the 5G deployment and how radio-friendly the environment is. That said, for places for line-of-sight from the phone to the cell tower for everyone, like a stadium, speeds, and overall bandwidth could be surprisingly high on a per-user basis.

The improved 5G mmWave support drives the extreme speeds, but the T800 5G modem can handle sub-6Ghz bands as well for the broadest possible network coverage.

5G is also becoming a very attractive solution for private and local networks because of its higher security and enterprise-grade configuration options. Chances are that newer high-speed, low-latency industrial networks will utilize 5G for replacing their wired equivalent, which was difficult to deploy, let alone reconfigure if needed.

MediaTek touted the power efficiency of its T800 modem derived from its 5G UltraSave technology. The company says the energy usage can be reduced so drastically that it would make the T800 suitable for IoT (Internet of Things) devices that typically have small batteries. We need more visibility to vouch for it, but it is an exciting prospect, and we should see such products appear if things happen as MediaTek expects.

Some observers claim that WiFi will eventually merge into 5G, but that’s false. The technologies overlap more, but fundamental aspects of how bands are allocated and managed differ. Security is also much more drastic on mobile networks.

On the other hand, WiFi benefits from being a local deployment, hopefully at the end of a fiber network. These differences make WiFi a more cost-effective solution in many instances. What’s important is not that we have “one ultimate network” but that our devices can switch to the best network available in real-time.

There’s also the MediaTek T830 5G modem aimed at the fast-growing Home Internet market. T-Mobile and others are seeing rapid growth in this segment because some cities can deploy 5G more efficiently than cable or Fiber. Even when Cable Internet is available, wireless carriers can offer better deals, thanks to a more data-efficient infrastructure. This advantage will increase over time.

Home Internet boxes built around the T830 will feature some of the finest 5G and WiFi technologies (check MediaTek’s Filogic WiFi 7 products). Their performance will easily outpace the network capabilities for the foreseeable future, making them a sound investment for the next 3-5 years.

Each box can handle a theoretical 7 Gbps peak download speed, and I’d be surprised if wireless carriers could deliver that at scale. We’re delighted with a 1Gbps connection at the office; even 200Mbps symmetrical already feels very nice.

On Thursday, Nov. 10,  at the MediaTek Executive Summit, Adam King, Vice President and General Manager, Client Computing Business Unit, Mediatek, announced the Kompanio 520 and Kompanio 528 chipsets for entry-level Chromebooks. The new chips are the successors of the Kompanio 500, with upgraded CPU, GPU, video codec, better power efficiency, and with support for higher resolution display.

For both products, the heterogeneous 8-core CPU unit combines 2 performance cores Arm Cortex-A76 and 6 efficient core Arm Cortex-A55, a significant upgrade from the Kompanion 500 that features 2 Cortex-A73 and 6 Cortex-A53. The CPU relies on LPDDR4x RAM clocked at 3733MHz.

Adam King, Vice President and General Manager, Client Computing Business Unit, presented the single-core performance improvement of the new Kompanio chips.

The only difference between the 520 and the 528 is the performance cores’ frequency, which is clocked at 2GHz for the 520 and 2.2GHz for the 528.

During the presentation, Adam King said that the single-core performance got an 80% boost while the multi-core performance is 15% better, compared to the previous generation.

Mediatek says it has a sizeable advantage against the Intel Celeron N4500, a common non-heterogeneous dual-core Chromebook processor, especially when it comes to performance-per-watt. Consuming less power to perform a given task could also be seen as emitting less carbon (electricity production emits carbon), and higher power efficiency should translate into better ESG metrics (Environmental, Social, and Governance) in our opinion.

“The sustainability benefits of the MediaTek Kompanio products are significant too, especially in heavy workloads where MediaTek’s low power processors generate 40% – 60% less energy-based greenhouse gas emissions compared to competitor platforms,” said Victor Tyan, Senior Director, Client Computing GTM.

Power efficiency and performance comparison with Intel’s Celeron N4500

On the GPU side, the Arm Mali-G72 MP3 is replaced by the Arm Mali G52 MC2 2EE in both new chips.  “Both GPUs [Arm Mali-G72 MP3 & Arm Mali G52 MC2] are very close in terms of performance and for most people, the performance difference won’t be noticeable in real-world scenarios. But if you still want the better GPU, then Mali-G52 MC2 is the way to go”, according to Tech Centurion (Centurion Mark). We’ll have to test to have more accurate results.

MediaTek claims a 20% GPU speedup

The new dual-core AI processing unit (APU) delivers AI features such as AI-driven real-time low-light video noise reduction and helps with image segmentation and tagging used for virtual green screen and image quality enhancement (i.e. color and contrast tuning optimized for different parts of the image).

The ISP has been improved with up to 32 MP support for the camera. This feature is leveraged from the smartphone SoCs. OEMs might not use the highest resolution in the near future, but they can scale it down easily to 5 MP for example, which is a huge improvement over existing laptop webcams (mostly with 0.7MP webcams).

The upgraded multimedia engine includes hardware VP9 decoding and H.265 encoding, which is an upgrade from the H.264 encoding and H.264, H.265 / HEVC decoding of the last generation (software VP9). The system supports 60fps Full HD video recording now, an upgrade from the FHD 30 fps support in the Kompanio 500. MediaTek claims a 2X performance boost for the multimedia unit.

The Kompanio 520 and 528 each support a Full HD+ display and a Full HD external display, which doubles the resolution support compared to the previous generation chipsets.

ChromeOS is known for its good security track record and Victor Tyan reminded us that MediaTek works closely with Google to integrate the latest security features such as Verified Boot, Built-in Virus Protection, and Multi-User Support with their latest Kompanio 520 and 528 processors.

Conclusion

The Kompanio 520 and 528 processors look very promising in terms of performance and high power efficiency for the uber-affordable Chromebook segment.

MediaTek insisted on the all-day battery life, a feature that helps to alleviate battery life anxiety and allows students to better focus on their work. To prove the point, the chip maker showed a photo of its booth at an education tradeshow, where a range of Kompanio-powered Chromebooks was exhibited without a single power cable lying around.

Schools might like that feature a lot since it is certainly safer for kids to circulate freely in a classroom free of wires and it may be easier for the IT personnel to manage all the computers.

Photo credit: Eliane Fiolet

At its recent Executive Summit, MediaTek talked about its Filogic 880 and Filogic 330 WiFi 7 solutions (announced in May 2023) and, more importantly, showed live demonstrations of both the router and client sides. One of the demos features a speed of 13Gbps between two WiFi 7 devices, which is phenomenal.

If you’re unfamiliar with these two platforms, Filogic 880 is destined to power WiFi 7 in the router, while Filogic 330 is integrated into clients such as computers, Smart TVs, etc.

Even though these demos are set up to show the best-case scenarios, the technological prowess is nonetheless impressive and previews where consumer-level wireless networking is heading in the near term.

In this case, two WiFi-7 router devices were close to each other, and a demonstration laptop had external antennas. It’s not a real-world scenario, but the laptop demo shows the final WiFi 7 speed might be limited by the antenna design rather than the chipset. This again shows that many wireless networking challenges also reside in the analog domain.

The Filogic 880 is meant to communicate with multiple WiFi 7 devices simultaneously, and its maximum theoretical bandwidth is 36Gbps. That’s an imposing number, even for users equipped with fast equipment. I still use wired Ethernet for its reliability and low latency, but these demos made me want to try wireless again on my primary PC.

Filogic 880 has meaningful features such as a VPN accelerator (tunneling offload engine) or its hardware cryptography engine. Thanks to the rise of Work-from-Home employees, more people need these features than ever before.

For the most demanding customers, Filogic 880 supports two 10Gbps wired Ethernet ports, which could be very handy for transferring files between two workstations on the local network. That’s assuming such computers do have 10 Gbps Ethernet support, and that’s unfortunately uncommon without buying an adapter ($100-$300). It’s not cheap, but SMBs and power users would appreciate the option.

All this speed is made possible by some of the nerdiest and most advanced radio and signal processing technologies, such as operating multiple frequencies (MLO, Multiple Link Operation) or supporting all possible WiFi bands, advanced beamforming, and highly parallel processing.

The Filogic 380 is a client-side WiFi 7 chip that typically talks to one router device, or mesh node, at a speed of up to 6.5 Gbps. It achieves such rates by using simultaneous bands and channels. I can’t imagine a real-world scenario where WiFi 7 performance is limited by the chip rather than by the antenna design or external factors WiFi environment or router. This seems like a good long-term investment.

The past few years have been incredible for WiFi with the arrival of WiFi Mesh networks, then WiFi-6, followed by WiFi 6E. The work on WiFi 7 seems well underway, with solid demos from MediaTek. This gives OEMs a great connectivity platform to build upon, and it’s a great incentive to use MediaTek as the overall platform (including SoC) rather than doing some mix and matching. Expect more demos at CES 2023, perhaps with product announcements.

Yesterday, before its Executive Summit, MediaTek announced its flagship SoC, the Dimensity 9200, the successor of the MediaTek Dimensity 9000. Alongside better performance and enhanced power efficiency, the new chipset brings many “firsts” to the mobile semiconductor industry.

The Dimensity 9200 is the first chip made with the TSMC 2nd Gen 4nm Process and to feature an  Arm Cortex-X3 primary core clocked at 3.05 GHz based on the Gen 2 ARMv9 architecture.

8 Cores CPU for performance and power efficiency

The heart of this system-on-chip (SoC) is a heterogeneous CPU cluster combining 4 performance cores and 4 efficiency cores. The Cortex-X3 (3.05 GHz) and 3 Cortex-A715 (2.85 GHz) are used for the workload that requires performance, while the 4 Cortex-A510 (1.8GHz) are selected by the cluster scheduler and OS thread scheduler to process less compute-intensive tasks.

As measured by MediaTek using GeekBench 5.0, the single-core performance has increased by 12%, and the multi-core performance is 10% better than the Dimensity 9000. The improved power efficiency is the most impressive gain over the previous generation. The new chip consumes 25% less power despite the increased performance.

The CPU cores access speedy LPDDR5X RAM with up to 8,533Mbps of bandwidth, an industry first offered by MediaTek in the Dimensity 9200.

For storing data permanently, the UFS 4.0 with Multi Circular Queue (MCQ) storage communication protocol offers the fastest data transmission currently available for direct access to storage.

Immortalis-G715, 11 Cores GPU with hardware ray tracing

Similarly to the Dimensity 9000’s graphics processor Mali G710, the Immortalis-G715 GPU features hardware ray tracing (RT).

We have not seen specific ray tracing performance indicators, but MediaTek’s messaging suggests that RT could be used in 3D games for basic use cases such as soft shadows, reflections, and ambient occlusion.

MediaTek’s GFXBench Manhattan benchmark numbers show that the GPU is 32% faster than the previous generation while using 41% less power.

Gaming and Viewing Experience Enhancements

On the gaming side, the new SoC features MediaTek’s HyperEngine 6.0 Gaming Technology that automatically optimizes system settings to ensure maximum gaming performance.

According to MediaTek, the chipset’s MiraVision 890 display technology intelligently adjusts many display and video stream parameters and provides hardware and software optimizations that improve viewing quality.

Additionally, for a better gaming experience, the new hardware supports Full HD+ up to 240Hz, WHQD up to 144Hz, and 5K (2.5Kx2) up to 60Hz, with an adaptive refresh rate.

APU 690, 6th Gen AI Processing Unit

The AI compute unit named APU 690 is the 6th generation AI Processing Unit developed by MediaTek.

On paper, the APU 690 should bring a 35% performance improvement over the previous generation in the ETHZ 5.0 benchmark and 45% lower power consumption for AI-super resolution tasks (AI-SR).

For example, when you play a 1080p video stored on your smartphone on a 4K monitor or TV, the APU will help the graphics system upsample the video with higher quality. This task is compute-intensive, so it is great that MediaTek made it significantly more power efficient.

The AI architecture has been improved from the previous APU with mixed precision. The 5th Gen used the traditional integer 16 computing method, which means that each whole number (integer) was always encoded with 16 bits. With mixed-precision computing, the APU 690 can now process both floating point (fractional number) and integer (whole number) with various number binary formats, 4 bits, 8 bits, or 16 bits.

For instance, 4 bits encode numbers from 0 to 7 while16 bits encoding reaches 65,535. For numbers from 0 to 7 or from 0 to 255, there is no need to use the 16 bits format. Allowing the arithmetic units in the hardware to operate on more data blocks simultaneously (parallelism). MediaTek claims that the introduction of mixed precision for the APU 690 boosts the computing speed by 98%.

Camera features and Imagiq 890 image signal processor (ISP)

The APU usually plays a vital role for AI-driven image segmentation. The GPU and the ISP can also participate in that task.

The new hardware enables real-time detection of photo and video content to identify the different areas of the image, such as people, sky, ocean, buildings, greenery, foreground, and background. Once image segmentation is executed, the system can apply different filters or algorithms to the various parts of the scene in real time. It is used for saturation, color, and contrast tuning.

AI-driven image segmentation enhances image quality for both the camera and the display.

MediaTek uses AI in the Dimensity 9200 to enable the Dual Stream AI Shutter feature, eliminating the blur in action and sports shots. A feature that MediaTek nicknamed “Motion Unblur.”

The Imagiq 890 is the first RGBW ISP to support the new RGBW sensors. In addition to the traditional RGB pixels, the new sensors have white pixels to capture brightness regardless of color. This helps to shoot better night pictures as MediaTek claims a 30% boost in brightness for night shots.

Connectivity

The Dimensity 9200 is the first SoC to support WiFi 7 with up to 6.5 Gbps data speed.

The chipset supports mmWave 5G and sub-6GHz connectivity and features mmWave SMART Beamforming which is 25% faster than regular beamforming, according to MediaTek. The new mmWave technology delivers an impressive 7.9 Gbps peak speed for data transfer!

The Lenovo ThinkStation P360 Ultra is a highly compact workstation that combines the portability of large workstation laptops with Desktop-level performance. The target audience is professionals working in the field, capturing or creating data on-site, or working in a mobile studio where it is possible to drive potentially large monitors but where engineers are still expected to travel with the computer.Specifications OverviewOur specific review unit came with the following configuration: […]

Logitech recently launched the Logitech MX Mechanical Mini keyboard for Mac, and we were curious to test the PC version, which is nearly identical but with different colors. The only difference is the lack of a USB receiver and USB-C to USB-C cable for the Mac version.

This MX Mechanical Mini is a 75% keyboard, which means it has 75% of the keys of a classic full-size keyboard, like the full-size MX Mechanical and other typical PC keyboards. Visually, you can see the lack of a numeric pad and other keys that would typically be found to the right.

I used to do stuff that required heavy usage of the numeric pad, such as 3D graphics coding, but that’s no longer the case, so the lack of a numeric pad was surprisingly painless. I also expected the arrow placement to require adaptation, but it went fine. So far, I haven’t adapted to using the new Home, end End key placements, but Page Up and Page Down are “OK.”

I do a lot of research and screenshots because I use Slack with my co-workers, so the lack of a “Printscreen” button was felt initially, but I set up Greenshot to use CTRL+` instead, and it’s OK. Coding stuff like /* comments */ without a Numpad is slower, but I could use a shortcut or a macro.

The thing I miss the most from my previous G815 full-size keyboard is the volume knob I use all the time. If I were to use this keyboard permanently, I’d consider getting a USB knob to control the volume. The MX Mechanical Mini has dedicated volume buttons (F12 and two others), but a knob is better (for me).

That’s it. The key sizes are the same as larger keyboards, and this compact design is great for reducing the space your right hand has to travel from the keyboard to the mouse and is more suitable for transportation. However, Logitech does not consider it a “travel keyboard.” More on that later.

Logitech has three tactile options for the keys: Tactile Quiet, Linear, and Clicky. I’ve tested all three on various keyboards, and I consider the “Clicky” option the only one I would consider giving you a “Mechanical” feel, so be warned. Logitech’s Clicky keys are closer to the Cherry Brown and not the super-clicky Blue keys.

Tactile Quiet and Linear feel more like membrane-based keys, and that’s not something I would use. Your preferences may be different. But even the Clickly option is much quieter than Cherry Blue keys.

This keyboard can connect to three devices via Bluetooth, and one of the three channels is used by default to connect to the included Logibolt wireless receiver in the box.

There’s a USB-C port, but it’s for charging only, and you cannot operate the keyboard as a wired device. It’s a pity that Logitech did not include this connectivity option. Although the receiver would typically work just as well, a wired option would have been 100% bulletproof and possibly lower latency.

USB-C connectivity would have been a logical inclusion, especially for a pricey device like this, where we want every possible option. It could also have been a way to connect a 4th device (desktop, laptop, phone, tablet) or reduce latency.

For instance, I keep my USB-C cable connected in the evening, preventing the LED Backlighting from going to sleep. Sometimes, I need to read something or think about something before typing, and when I do, the keyboard goes dark to save battery. That can be a bit testy.

Logitech has pretty good software to manage its devices. You don’t have to install Logi Options+, but it does let you disable keys like CAPS LOCK, etc. It also saves your preferences to a Logitech cloud.

Strangely enough, there’s no place to store the Logibolt connector inside the keyboard. That’s because Logitech considers this product as a desktop accessory, but I beg to differ. People will use it as a travel keyboard, and not having storage for the wireless receiver is a bit annoying. I’m pretty sure there’s ample internal volume room for it.

The Logitech MX Mechanical Mini’s overall design and build quality are what I like most. The Logitech Clicky keys deliver excellent tactile feedback with a click feel and shorter key travel (but not too short) than classic Mechanical keys.

The industrial design is very compact and agreeable to use and looks neat on any desk. Logitech uses quality materials, and the product has a Premium feel that can be lacking in many alternatives. If anything, I’d love to see a TKL version of this keyboard with a volume knob. If you’re looking for a compact keyboard, I highly recommend considering this option.

Logitech MX Anywhere 3

I’ve also tested the Logitech MX Anywhere 3 mouse (official page), which is a familiar design since I started using that line of products when the first MX Anywhere was launched. Back then, the big feature was the ability for the optical mouse to track glass surfaces, which was a game changer for me because I happen to be in hotels and conference rooms often without a mouse pad.

This feature is still present and remains an undeniable advantage over any competitor who would not support a glass surface.

The new MX Anywhere 3 is pretty light, and I would use it as a travel mouse because of its compactness. But beware that Logitech does not consider it a “travel” mouse, and unlike its predecessors, you cannot store the Logibolt receiver inside the mouse.

Again, this should be rectified, and I can’t think of one good reason to remove the internal receiver storage – other than reducing BOM costs (??). I expect every single convenience from an $80 mouse, so I won’t be forgiving with this.

That aside, the mouse can also connect to three devices, which is very convenient because many people now have work + personal computers or desktop + laptop combos.

Logitech also has the option to switch both mouse and keyboard from computer to computer if you install Logi Options and use its Flow feature (yes, that’s different from Logi Options Plus…). Warning, Flow might not work when you run a VPN.

The MX Anywhere MagSpeed scrolling wheel can also switch from smooth to indented scrolling via software. You can set the tactile feedback and get a perfect personalized setting.

There’s a USB-C port for charging only. It’s a good upgrade over the micro-USB port Logitech still has on some of its gaming keyboards, and by now, everyone has USB-C cables all over the place. It would also be nice to have cable connectivity for the same reasons I mentioned for the keyboard.

Overall, the MX Anywhere 3 performance is stellar and very accurate and reliable. In my case, its size is slightly small (hand size: I wear M gloves), but it’s an excellent travel mouse. My primary mouse is a little bit larger. Go try one at Best Buy and see how it feels to you.

Last, the MX series of products is optimized for an office environment, so click noises are reduced or virtually non-existent. These devices can operate across OSes and platforms.

Depending on your work, you can finely tune the page scroll or create macros to expedite repetitive tasks. For example, I configured the mouse’s thumb buttons for copy/paste. Logitech’s wireless technology has also been optimized for locations where everybody uses wireless devices, so the company has paid particular attention to that.

Young-Lee, Minister of SMEs and Startups (photo courtesy of Aving.net)

The 23rd Innovative Technology Show (ITS) was held last week in the COEX center in Gangnam, Seoul. Like last year, Ubergizmo was an ITS Awards jury member (via video conferencing). We carefully reviewed and rated the value proposition and technical excellence of the finalists of the “K-Innovation Company” award.

The ITS opened with a welcoming speech by Minister Young Lee of the Ministry of SMEs and Startups who called the show “Korea’s biggest SME-centered technology innovation event”. This year’s theme was ‘Tech-Innovation, Talk to the Future!’.

10 companies out of 245 made it to the final, so the competition was intense, and we were impressed by the general level of sophistication of the products and technologies used across diverse industries.

Ubergizmo nominated three companies as Ubergizmo’s “Best of ITS 2022” based on their innovation, capacity to export worldwide, and value proposition. They are Everchemtech, SHENB, and NES&TEC Co., Ltd, in this order.

Everchemtech

Everchemtech is a company that develops films made of various materials typically used for industrial purposes, to insulate electronics, for example. However, they came up with a new product that could reach every consumer because it is related to food: NEXRIER.

NEXRIER stands for NEXt BarRIER and is a bio-degradable replacement for plastics used in the food industry. Its goal is to prevent contact between food and oxygen, and it can be a drop-in replacement for traditional plastic wrapping.

It is extremely typically difficult to create a bio-degradable plastic that won’t degrade very quickly when in contact with moisture or food. Everchemtech says it has managed to do so, and although its product will bio-degrade eventually, they are stable during the food’s shelf life.

Additionally, the company says it can produce this NEXRIER 30% to 40% cheaper than the traditional plastic equivalent in the Ethylene-vinyl alcohol (EVOH) family. As far as we know, this is a huge breakthrough that could lead to a drastic and economically sustainable reduction in petroleum usage in this industry.

Surprisingly, the base component of NEXRIER is whey protein, which is normally a by-product of Cheese-making. Everchemtech’s NEXRIER received one of two “Excellence Awards” from the combined jury.

SHENB

SHENB is a medical device maker demonstrating its VIRTUE RF Microneedling device.

Microneedling is a medical procedure typically utilized in the cosmetic industry to rejuvenate or heal people’s skin. For example, it has been used to treat Acne, Hair loss, Scars, Sun damage, Wrinkles, and more.

Microneedling works on the principle that pricking the skin with tiny needles (controlled skin damage) causes your body to heal (rebuild) that area by producing more collagen and elastin. The healed skin might have a better appearance, although it’s not without risk, and experts should do procedures according to local regulations.

SHENB uses RF (radiofrequency) technology in addition to the needle, hence the VIRTUE RF product name. RF turns into heat and can reach deeper into the skin layers, giving even more control and treatment options. Most people don’t realize it, but human skin has three distinct layers.

Other manufacturers also use RF, but SHENB’s novelty comes from a few features. First, it has ACS (advanced cooling system), which cools down the skin as soon as the RF energy has done its job. This makes it possible to use more energy (reach deeper) while avoiding the risk of thermal injury.

Next, IPRS (Integrated Pulsed Radiofrequency System) offers better energy pulse control which turns into a “wider area of skin rejuvenation,” says SHENB. Finally, the device works across multiple frequencies (0.5MHz to 2MHz) and can safely deliver energy for up to 7 seconds (vs. 0.8s previously), the company says.

Apparently, the Virtue RF has already been approved by the U.S FDA and its European equivalent, and SHENB is already working on the 2024 upgrade. SHENB is the other recipient of an ITS 2022 “Excellence Award.”

NES & TEC

NES & TEC Co LTD was pitching its SWID drone (multi-copter UAV system in industry lingo) that has multiple potential purposes. It has all it takes to be utilized as an industrial drone to keep an eye on infrastructures, or by fire departments during forest fires, for example.

However, it also has a strong military potential because it is built to resist jamming and can navigate even when traditional positioning systems (like GPS) aren’t available.

To thwart hackers, NES & TEC uses a proprietary flight control system. While we can argue about the safety level of proprietary vs. open-source code, the company didn’t really have a choice since the South Korean military prohibits existing open-source communications systems. The one used in this drone adheres to the Korean Military Specification-level encryption (KCMVP) standard.

NES & TEC also got security certifications in the USA and Europe but didn’t go into details about which ones. The company mentioned that the drone can resist jamming and can still navigate despite being cut off from GPS. It does so by using a camera-based positioning system as an alternative.

Given recent military developments, this capability is likely to attract attention from security forces worldwide as drones have proven to be extremely valuable reconnaissance tools.

Other finalists

The other great finalists are listed below. Most are excellent as well and worthy of your attention:

• JPS: ITS 2022 Grand Prize winner. (electric band for packaging steel products)
• Kanavi Mobility: scanning LiDAR sensor products
• IntelliVIX: deep learning-based CCTV video analyzer
• Jijun System: intelligent general parking management system
• Secuworks: sound field security sensor
• Ison: drone charging station
• P&S Technology: food and beverage (PET bottle) defect inspection equipment

MediaTek introduced its Dimensity 1080 system on a chip (or SoC), which will power affordable smartphones (mid-range). These phones have much higher sales volumes than your high-end, expensive handset and can affect the balance of power of various underlying OS or app platforms.

Dimensity 1080 takes the slot previously occupied by the 6nm Dimensity 920 launched last year. Overall, not much has changed from the 920: there’s a small CPU frequency bump (~4%), but we’re still looking at a CPU cluster made of two ARM Cortex-A78 2.6GHz (2.5Ghz before) and six ARM Cortex-A55 2.0GHz The Mali-G68 MC4 GPU remains cadenced at 950MHz.

However, MediaTek announces “significantly upgraded camera features compared to its predecessor.” We think the foundation for this statement comes from the announced 200 Megapixel camera support. MediaTek has either upgraded or optimized its ISP (Image Signal Processor) design.

If we go back to the official Dimensity 920 specifications, it says, “Max Camera ISP 20MP + 20MP, 108MP”. If you aggregate the throughput, this means 148 Megapixels at maximum speed. To be precise, this is the number of raw pixels that can be practically processed from the sensor in photography. That’s before processing (stacked photography, filters etc.) that happens after the initial ISP work.

Committing to supporting a 108MP camera to a 200MP one is quite a jump. We’ll assume that no other cameras run simultaneously – that’s expected.

This announcement tells us that the megapixel war is alive and that 200 MP cameras will soon enter the mid-range market segment. Users benefit because these extra pixels can bring additional sharpness to the photos in broad daylight.

Unfortunately, most OEMs hide the full-resolution mode deep in the camera app settings (what a mistake!), and nearly no one uses it. We’ve seen that Samsung has a full-resolution option in “auto” mode. Xiaomi used to have an excellent 27MP “auto” mode but has removed it.

Either way, our Ubergizmo CAMERA HW benchmark isn’t swayed by the megapixel marketing alone and manages to correctly classify camera power based on all relevant technical details, so the megapixel war doesn’t affect it much. We’re planning an upgrade based on measurements we’ve taken with the latest sensors to make it even better. Stay tuned!

Edge Computing World (ECW) is back in person at the Santa Clara Convention Center from October 11th to 12th, 2022. Yesterday evening, the conference hosted the Awards Ceremony where Topio Network, a next-generation business information provider and accelerator of high-tech markets, announced the winners of the Fourth Annual Edge Award 2022. 

The awards recognize excellence in edge computing hardware, software, platform, and applications, and individual excellence and contributions to the edge ecosystems.

The Awards went to:

CruxOCM for Edge Startup of the Year – Recognizes Startups with between US$ 3 million to $15M funding and more than 20 employees. The list of the finalists is here.
CruxOCM uses Edge Computing yo automate Oil & Gas control room operations and improve sustainability

 dRISK for Edge Innovator of the Year Recognizing young companies with up to $3M funding with less than 20 employees – The list of the finalists is here.
dRISK Leverages Edge Computing for testing and training Autonomous Vehicles to make AVs safe in the modern world.

Precision AI for Edge Rising Star of the Year  – Recognizing more mature companies with more than $15M of funding raised. The list of the finalists is here.
Precision AI leverages Edge computing for Precision Spray Drone System to increase yield, reduce inefficiency & overspray using artificial intelligence

Odience for Edge Native Application of the Year
Odience is a remote participation platform, used to create live immersive, and interactive experience for markets such as retail, live music, and sport

Kaniz Mahdi for Edge Woman of the Year
SVP Technology Architecture & Innovation, Deutsche Telekom, recognized for leadership in, and outstanding contribution to, the field of edge computing.

Kaniz Mahdi, Edge Woman of the Year, 2022

Edge Computing Industry Recognizes Women Shaping the Future of Edge with Fourth Annual Awards Program

“Year after year, this award highlights the incredible contributions and accomplishments of women in this swiftly evolving industry,” said Gavin Whitechurch of Topio Networks and Edge Computing World.

The 2022 Edge Women of the Year Award finalists include:

• Amanda Marrs, Senior Director of Product at AMP Robotics
• Azita Martin, VP & GM Artificial Intelligence for Retail and CPG at NVIDIA
• Elaine Liew, Head of Hybrid and Edge Computing at Amazon Web Services
• Giulia Merlo, Edge Computing Senior Portfolio Manager at Vodafone
• Inna Ushakova, CEO and President at AI EdgeLabs
• Kaniz Mahdi, SVP Technology Architecture & Innovation at Deutsche Telekom
• Kuan Collins, Director of Innovation Factory Hub Huntsville. SAIC
• Meryem Simsek, Chief Scientist, Advanced Technologies at VMware
• Nitu Kaushal, Managing Director, Intelligent Edge, Europe Region at Accenture
• Tilly Gilbert, Principal Consultant and Edge Practice Lead at STL Partners

Oliver Kaplan, a young man from Manchester (England), went to the pub with his friend after a long day’s work. After having a good time, he ordered an Uber to get home safely, as he had done many times before. But this time, he woke up with a hangover and a £35,000 (~$39,000) bill for his Uber ride.

Perhaps “fortunately” for him, he had insufficient funds in the bank, and the transaction failed. He contacted Uber’s customer service because he remembered that the quoted price was $11 or $12, which was confirmed by the ride service.

His destination was set to arrive in “Australia,” even though the driver dropped him at the expected place. Therefore, it’s unclear what happened, but it seems like an issue in the back end rather than a user error.

You have to wonder how this could even happen in the first place. If you try plotting a course from the USA to Australia, the app won’t give you a quote or plan a route because… you can’t get there by car.

There have been instances where people entered the wrong address and dropped hundreds of miles from their destination, but it looks like the user was not at fault here. Imagine having to fight a $39,000 charge with your bank. That could be “interesting.”

Fortunately, Uber quickly recognized the problem and charged Olivier the proper amount. Now he has a great story for next Friday at the pub.

Officially launched during my coverage of IFA 2022 in Berlin, Germany, the Honor 70 series, especially the Honor 70 (official page) is designed to bring a high-end design look and feel to the competitive $600+ market segment. We have had time to look deeper, and here is our experience and analysis of this beautiful smartphone.

This phone does not have an official retail channel in the USA, so the $650 street price is based on more expensive import channels. Outside the USA, the official UK price is 499.99 UKP for the base model (8GB RAM, 128GB storage). In Spain and France, the official price is 549 Euros. If you want 256GB of storage, it’s 599 Euros.

Design

The Honor 70 is built around a thin and light chassis with glass on both sides and a “super dual curved screen,” which means that the curve is present on the front and the back, making the edge look and feel extra thin.

This type of design used to be super-expensive to produce, and it’s nice to see it coming into mainstream segments like this $650 phone (street price). Not everyone likes curved screens, but I think they look really cool, and the chassis feels agreeable in hand.

+3

Some people might remember the issues with the grip inducing unwanted touch events, but I haven’t experienced this in recent phones with a curved screen, so that should not be a problem anymore. I’m curious about what other people think.

Our unit is the “Emerald Green,” and it looks great with its anti-fingerprint matte textured surface in the back, However, Honor has other cool back covers, notably the Crystal Silver, Icelandic Frost, and Midnight Black.

The chassis is light (178g) for a 6.67-inch phone and very thin, too: 7.91mm. Many people like that very much, especially when you carry your smartphone in (tight) pants or shirt/vest pockets.

Unfortunately, there’s no water-resistance certification (IP Rating), so please be careful around water. At this price range, it’s not surprising as the IP certification does cost extra. However, it would have been very nice to have some unofficial water protection. It seems that’s not the case.

Sound

A single speaker delivers clear and distortion-free audio on the bottom right. It is very decent but is not loud and good as phones with dual speakers, it’s just physics. At the same time, it’s understandable that resources have to be moved around when building affordable premium phones.

Display

The 120Hz FHD+ (2400 x 1080) 6.67″ OLED display looks great and has all kinds of tuning options in the settings. I like to set the Color and Temperature mode to “natural” rather than “vivid” (default), but the other default options work well for me.

The dynamic refresh rate should provide a good ratio between scrolling smoothness and battery life. The refresh variability is not as granular as more expensive phones and does not go all the way down to 1Hz or 10Hz. However, the overall system power efficiency makes up for that.

As usual with OLED, colors do look great, and Honor has included a default background image that matches the unit’s color (green in our case), a nice touch.

We measured the maximum brightness at 784 NITs with manual settings, which is a great number. There isn’t an extra boost in brightness if you’re in direct sunlight, so that’s pretty much the ceiling. The most expensive displays can spike temporarily to more than 1500 NITs, but it’s temporary because your eyes will adapt and negate that extra brightness if left ON for too long.

Finally, there’s an under-screen fingerprint reader, which shines a strong light to read the fingerprint. In our experience, it worked well and was efficient. The most visible downside compared to an ultrasound reader is the extreme brightness of the fingerprint reader’s light when your eyes are accustomed to the darkness.

Honor 70 camera

The new Honor 70 camera sparked quite a bit of excitement on the web because it is the first to use Sony’s new IMX800 camera sensor in its f/1.9 27mm primary camera.

The IMX800 is a 50 MP sensor with 1.0-micron sensing pixels, which therefore makes it a sensor that’s approximately 50mm² (54 mm² actually) large. That’s similar in size to the Samsung GN5 found in the Galaxy S22’s primary camera, for example.

The Honor 70 ultrawide camera is intriguing and features a 50MP sensor as well, although we think the pixel size is probably very small, perhaps in the 0.6 to 0.7-micron range (that’s our guess, looking at the photos we shot). Ultimately, the f/2.2 13mm lens might not let enough light go through to fully take advantage of this sensor.

The third camera is a depth sensor to assist with portrait mode background blurring.

Our Ubergizmo CAMERA HW score for the Honor 70 is 147, which places it near the new Samsung Galaxy Z Flip4, also equipped with a comparable primary and ultrawide set of cameras.

The primary camera captures good photos that are detailed with proper HDR, even in difficult high-contrast situations, like nature shots that have a mix of extremely bright (sun, sky, cloud) and shadowed (low vegetation) areas.

There’s no telephoto, but the 2X zoom probably uses a 50MP sensor cropping technique that is, in our opinion, similar to an optical 1.7X zoom. It’s more useful for portrait pictures (~50mm) than for actual long-zoom (5X-10X).

The ultrawide camera can capture shots that are sharper and clearer than the iPhone 13 Pro’s ultrawide camera. However, the experience is not always consistent. Honor’s ultrawide camera captures tend to be a bit on the soft (blurry) side for both day and night pictures unless you’re really careful holding the camera still (more than normal). The Ultrawide camera doubles as a macro camera since it can focus on a subject located only 2.5cm away!

Honor has added a cool software feature called “Solo Cut Mode” to build an additional vertical video ready for TikTok, Instagram, and the like. The idea is brilliant: while you film a movie in landscape mode, you can select one person to be tracked. The phone will generate a separate vertical video for that subject, which you can post immediately as a story, no editing required!

Performance

The Honor 70 is powered by a Qualcomm Snapdragon 778G (5G), a computing platform initially designed for more expensive smartphones. In its price category, it does offer good CPU performance, and most users should be completely satisfied with day-to-day utilization and basic gaming.

Not surprisingly, this phone has exactly the same performance as others that run on the same Snapdragon 778G platform, including the Honor 60 and Honor 50 smartphones.

Since this chip has been around since May 2021, it’s not surprising that some phones like the Oppo Reno8 feature a faster processor like the MediaTek Dimensity 1300. That said, the Oppo Reno8 has completely different tradeoffs concerning design and luxurious materials.

Heavy users and gamers might find other phones dedicated to performance. However, they’ll probably have to forgo some of the design or camera advantages (including the selfie camera) of the Honor 70. Again, it’s a balancing act in these price ranges.

For users who truly want extreme high performance, a phone like the OnePlus 10T is a formidable competitor for about the same budget, although its primary camera is not as powerful as the Honor 70’s.

Software

It goes without saying that Honor has complete access to Google Services like other android smartphones. Some people ask because of the past Honor status as a Huawei subsidiary, but Honor is now an independent entity.

This phone runs on Android 12, with Honor’s Magic UI 6.1 as a top layer and custom interface. For those familiar with Honor, things have not changed much, so you’ll feel right at home. The same is true for Huawei HarmonyOS since both software have a lot of the same logic in common.

For those coming from another Android skin, don’t worry, as Magic UI is similar and intuitive enough for any Android user.

Battery

The 4800 mAh is a bit larger than the previous Honor 60, but the overall battery life remains very similar because of the higher display refresh. Objective battery benchmarks like PCMark Work 3.0 (200 NITs) show a 10h20mn battery life under continuous load and on-screen time.

In the real world, this translates to a stress-free smartphone experience, and you should go by your day and come home with battery capacity to spare. By the way, the amount of capacity crammed into such a thin chassis is remarkable (mAh per cubic inch).

When you need to charge, the included 66W AC adapter (a ~$45 value at big brands) will get the battery from 0 to 80% in about 30mn, which is much faster than the iPhone 13 Pro, although not the absolute fastest on the market.

Conclusion

The Honor 70 is a smartphone that caters to users who demand beautiful craftsmanship, premium looks, and a great primary camera experience. And for them, Honor does deliver. Keep in mind there are two other phones in this product line: the Honor 70 Pro and Honor 70 Pro +, so this is the baseline.

Honor-oriented resource utilization to specific materials and components to make this happen. As such, this is not a gaming phone, nor it pretends to be the best camera phone. Honor has other options for that, like the Honor Magic4 series.

With a street price of about $650, we find this phone to deliver interesting value for users who care about its strengths. There’s no question that’s one of the most competitive market segments, but Honor is more likely to sway buyers based on design, brand trust and user experience than sheer technical specifications and benchmark numbers.

Logitech recently introduced its new “Combo for Business Gen 2”, including the new MX Master 3S mouse and MX Keys keyboard. I have owned quite a large number of keyboards, and I like to try new ones that have some potential. Since this is a popular combo in the enterprise space, I wanted to look closer.I had not tried the first generation MX Keys Keyboard, but this new one is […]

The new Lenovo Slim 9i 14 was created with two objectives in mind: first, it is the first carbon-neutral laptop to be commercially available. Secondly, despite being made of many recyclable materials, it should bring premium entertainment and mobile luxury. In this review, we will look at the choices and design strategies Lenovo has adopted to see if the reality matches the pitch.Slim 9i Specifications HighlightsWith the Lenovo Slim 9i […]

Like clockwork, we’re entering one of the busiest parts of the year with an avalanche of cool new devices. The Google Pixel series is a landmark now, and we’re excited about today’s launch of the new Google Pixel 7 and Pixel 7 Pro.

Last year’s Google Pixel 6 and Pixel 6 Pro represented a turn for the better with many improvements over the Pixel 5, with notably a significant escalation of the camera hardware in the Pixel 6 Pro as spotted by our CAMERA HW score. Google came back into the race, competing at the highest levels.

The new Google Pixel 7 series is an upgrade, refinement, and optimization of the Pixel 6 series as Google capitalizes on last year’s herculean effort.

The industrial design is seeing changes but sticks to the overall design language that was well received last year. The changes make the new phones look more modern, and the manufacturing quality has increased.

The new polished aluminum camera bar of the Pixel 7 Pro gives it an extra “premium” look, while the Pixel 7 (non-pro) gets a matte version. It’s excellent positioning using design as an instrument.

Google is known for its “pure Android” experience and long commitment to software updates, and nothing has changed on that front. Design aside, there are aspects of these new Pixel 7 phones we want to attract your attention to, namely the Camera and Computing.

The camera is the most exciting aspect for the majority, so let’s get to it. The Pixel 7 Pro’s camera is configured as follows:

• 50 MP primary (OIS, 1.2μm pixels, ƒ/1.85, 1/1.31” sensor)
• 12 MP ultra-wide (1.25μm pixels, ƒ/2.2)
• 48 MP telephoto (OIS, 0.7μm pixels, ƒ/3.5, 5X optical)
• 8MP selfie (1.22μm pixels, ƒ/2.2, fixed-focus)

This configuration is well-balanced (for the price) and should cover all everyday use cases very well. We’ll compute the CAMERA HW score in our full review, but this looks quite good.

Google is using its telephoto in two ways. It shoots natively at 5X optical zoom in 48MP. Alternatively, it can zoom even further using a sensor-cropping technique, like what Samsung has in some of its Galaxy phones.

Sensor-cropping uses only a fraction of the sensing surface without merging pixels to preserve details. That is like having a smaller sensor with a lens that zooms further. Sensor cropping is typically not as good as having an optical equivalent, but it is much better than software upscaling.

The Pixel 7’s camera is very similar but forgoes the telephoto – a choice is often found between different tiers of phones. It makes a lot of sense because ultrawide is objectively a more common use case than zoom – that’s why OEMs make this tradeoff for more affordable phones.

• 50 MP primary (OIS, 1.2μm pixels, ƒ/1.85, 1/1.31” sensor)
• 12 MP ultra-wide (1.25μm pixels, ƒ/2.2 – different lens)
• 8MP selfie (1.22μm pixels, ƒ/2.2, fixed-focus)

You get almost all of the photographic benefits of the Pixel 7 Pro, except when the subject is far away. In the $599 range, this is very exciting.

Regarding computing, Google uses its Tensor G2 processor to provide excellent support for all imagery activities that rely heavily on stream computing and AI.

Google has demonstrated some long-exposure shots, which were nearly 4X faster, going from 5.25s down to 1.25s. That should also affect near-instant photographs or video recording, which we’re much more interested in.

That said, the competition (Qualcomm, MediaTek) is also capable of doing such things, and there’s no cross-platform benchmark to compare them apples-to-apples.

The Pixel 7 Pro has 12GB of RAM, so that’s the one heavy app users should consider, as the 8GB of RAM in the Pixel 7 should suffice for everyone with more basic needs.

Storage options are 128/256GB for both, but only the Pixel 7 Pro gets a 512GB version if you need that much.

Both phones are available for pre-order today for $599 and $899 on the official Google store, but also from carriers such as AT&T (pre-order here), which also says that in-store units will arrive on Oct 13. As usual, if you sign-up on a new AT&T plan, you could get the Pixel 7 for free, and the carrier says it offers “up to” $800-off on the Pixel 7 Pro

Infinix just launched two Android smartphones aimed at young customers. The OEM has tuned the value proposition with seemingly high specifications for the dollar and has added an NFT marketing strategy (more Infinix NFT info).

The Infinix Zero Ultra ($520, official page) is the more powerful of the two and features a MediaTek Dimensity 920, which we previously covered. It’s an excellent computing platform designed for this use case of bringing features, including 5G and advanced camera features, to a mainstream market.

The Infinix Zero Ultra’s back cover comes in two colors with an original glass texture. The White version features a pearl-like iridescent surface that varies depending on incoming light.

The Black one has a mix of black matte glass and carbon-fiber-like material, giving it a “sporty” look. It’s a great way to address two radically different user tastes.

Among the notable features of the Infinix Zero Ultra, a few caught our attention and are worth highlighting.

• It has a great 8” AMOLED waterfall display (120Hz), but its 165.5x 74.5x 8.76 dimensions make it highly comfortable to hold in your hand, even with the included case.
• The 200 MP “Ultra Vision” camera (with OIS) is the primary image capture tool, but two more camera sensors supplement it
• Its 180W Thunder Charge battery charging speed leaves big brands’ phones in the dust when you need to charge in a hurry. It can charge to 100% in 12mns, according to Infinix. It’s a believable number.

The Infinix Zero 20 ($270) seems extraordinarily competitive and could shake the Android smartphone market in several regions.

Infinix is not yet known in the USA, but it has made enormous strides in Latin America and other markets, thanks to a finely-tuned marketing and communications strategy and high specs in several market segments.

Powered by the proven Mediatek Helio G99, this $270 phone has a 108MP camera sensor, supplemented by two additional sensors, including ultrawide. Although the number of megapixels isn’t the only quality metric, but it’s a sign the Zero 20 uses recently introduced electronics (and software).

That’s a far cry from potential competitors that still use camera sensors that have been around for years.

The 6.7” FHD AMOLED display should prove great for daily usage, including casual gaming, movies, etc. And when you need a charge, the 45W charger can get you from 0 to 75% in 30mn. That’s a charging power equivalent to the most expensive big brand phones, and unlike big names, the high-power charger is included (another $50 value).

The most notable feature of the Infinix Zero 20 might be its 60MP selfie camera. Vloggers / Tiktokers should notice, and OEMs are working hard to sway their phone purchase decision. We see a new selfie-camera arms race, and that’s exciting.

It will be fascinating to compute these phones’ performance/value numbers because, on paper, it looks like they are crushing it in their respective segments.

TJ Walton, Product marketing Manager, Xiaomi International, presenting on stage in Munich

Yesterday in Munich, alongside the 12-gen T-series smartphones and other IoT products, Xiaomi launched its new Smart Band 7 Pro equipped with a beautiful 1.64” AMOLED display. According to Xiaomi, the always-on 280 x 486 pixels, 326 PPI display is 84% bigger than the Smart Band 6.

The device was released this summer in China and yesterday was the global launch. Besides the form factor and the larger screen size, the addition of a built-in GPS (GNSS) is one of the main differences from its sibling, the Smart Band 7.

The Xiaomi Smart Band 7 Pro’s stylish high gloss frame chassis encases the beautiful rectangular display covered with tempered glass and anti-fingerprint coating.

The screen delivers a peak brightness of 500 NITs and has an auto-brightness feature. From the settings, you can choose to turn on the always-on option, however, it would probably affect the battery life. When wearing the tracker, the display wakes up by turning your wrist or gently tapping once on the screen. Fashion enthusiasts can customize it with the 150+ watch faces that Xiaomi has designed to enable clothe matching.

The Smart Band 7 Pro offers a suite of health and fitness features, including more than 110 sport modes, an on-wrist running coach with 10 pre-planned running courses at different levels, activity tracking, heart rate monitoring, blood oxygen level (SpO2) measurement, and sleep tracking.

Like any other smartwatch or fitness band, you get all the standard notifications, from messaging apps, phone calls, calendar reminders, and more. The Smart Band 7 Pro turns into a remote control for your camera (not all smartphone models are compatible) or for the new Xiaomi Vacuum Robot X10+, using Amazon Alexa Voice AI for hands-free interface (see photo below).

Thanks to the 5ATM rating, the device is water resistant for swimming or taking a shower. According to Xiaomi, you will get up to 12 days of battery life with regular usage.

The high-quality large display makes the device look more like a smartwatch than a fitness tracker. In fact, the 99 euros price tag makes the Smart Band 7 Pro a quality smartwatch packed with standard fitness features at a sweet price.

While the body comes in two flavors, Black or Ivory, Xiaomi offers 6 color options for the strap: Ivory, Olive, Orange, Blue, Black, and Pink. There are also two special edition versions: Pine Green and Moon Gray.