Server - Page 2 of 34

Arduino Nano RP2040 Connect Review: Built-in Wi-Fi and BLE

Our Verdict

Great hardware with lots of power in a tiny package. The price may put some off but if you need Wi-Fi and the RP2040 then this is the board for you, but for how long?

For

+ Small form factor
+ Lots of analog inputs
+ Built-in Wi-Fi and BLE
+ Built-in microphone, IMU

Against

– Pin numbers hidden away
– High Cost
– No USB-C
– Software is buggy

When you think of an Arduino, the first board in your head has to be the classic Arduino Uno, with its Atmel microcontroller. You wouldn’t think of an Arduino powered by ‘Pi Silicon’ but the last of the initial wave of RP2040 partner boards is the Arduino Nano RP2040 Connect and this is the board we have been waiting for. You see the Arduino Nano RP2040 Connect is the only partner board to come with built-in Wi-Fi and Bluetooth. Neatly integrated into the familiar Nano layout is a Nina W102 uBlox Wi-Fi module which provides 2.4 GHz WI-Fi, Bluetooth Low Energy, an IMU capable of detecting movement and gestures and a microphone which can be used to measure audio levels and more importantly used in machine learning projects.

Should the Arduino Nano RP2040 Connect be the heart of your next project? Well that depends on your needs and budget. If you are happy to blink LEDs and make things move, then a $4 Raspberry Pi Pico is more for you. But if you want to build Internet of Things (IoT) devices with a web interface then Arduino’s latest board may just be what you are looking for. But we don’t truly understand how a board works, and more importantly how it is supported until we use it, and use it we shall!

Arduino Nano RP2040 Connect Hardware Specifications

System on Chip	RP2040 microcontroller chip designed by Raspberry Pi
	Dual-core Arm Cortex M0+ processor, flexible clock running up to 133 MHz.
	264KB of SRAM, and 4 / 16MB of on-board Flash memory
GPIO	22× multi-function 3.3V GPIO pins
	1× SPI, 1 × I2C, 1 × UART, 8 × 12-bit ADC, 20 × controllable PWM channels
	8× Programmable I/O (PIO) state machines for custom peripheral support.
	1x User LED (GPIO 13)
	RGB LED, accessible only via the Nina W102 module.
	Castellated module allows soldering directly to carrier boards.
Extra Features	Nina W102 uBlox Wi-Fi module
	ST LSM6DSOXTR 6-axis IMU
	ST MP34DT06JTR MEMS Microphone
	ATECC608A-MAHDA-T Cryptographic Coprocessor
Dimensions	1.77 x 0.7 inches (45 x 18mm)

Design and Use of the Arduino Nano RP2040 Connect

(Image credit: Tom’s Hardware)

The design of the Arduino Nano RP2040 Connect is quite clearly based on Arduino’s Nano range of boards such as the Nano 33 IoT and Nano Every. The Nano range of boards share the same pin out, enabling the Arduino Nano RP2040 Connect to be a drop-in replacement for other Nano boards. The bare board has the castellated edges that have become the norm for an RP2040 board and the underside of the board is unpopulated, enabling it to be surface soldered into a project. Of the 22 GPIO pins we have 20 pins which can be used for PWM and of those we have 8 analog inputs, the most of any RP2040 board as the Pico comes with three and most third-party RP2040 boards have four. If you are looking for the I2C pins they are hidden away on pins A4 and A5 and the onboard IMU is accessible on the I2C bus.

The eagle eyed amongst you will have spotted a micro USB connector, something common on many newer Arduino boards. We would’ve liked to have seen a USB-C connector, because it offers a foolproof connection and the mechanical connector is stronger than micro USB. As this is the most expensive RP2040 board, we think that Arduino could’ve included this without impacting the overall cost.

(Image credit: Tom’s Hardware)

The big feature of the Arduino Nano RP2040 Connect is Wi-Fi and the Nina W102 chip provides us with 802.11 b/g/n 2.4 GHz Wi-Fi and Bluetooth 4.2. The antenna is located opposite the micro USB port. The Nina W102 chip is itself a powerful microcontroller with 520KB SRAM and a 240MHz 32bit Dual Core Xtensa LX6 CPU, but in this board it is relegated to handling communications.

As this is an Arduino, our natural inclination is to use the Arduino IDE, and with a recent update bringing official support for the RP2040 to the IDE, we can easily create projects. We tested the Arduino Nano RP2040 Connect with the Arduino 1.8.15 and 2.0 beta 7 IDE and after installing the board via the board manager we were able to flash the example blink sketch to prove that we had connectivity. To access Wi-Fi we needed to install the WiFiNINA library and then create a secrets file to contain our Wi-Fi login details. We ran the Simple Web Server WiFi example and it flashed with no issue, but we later learned that we needed to change the default GPIO pin used in the example from 9 to “LED_BUILTIN” in order to control via a web interface.

The connection details should appear in the serial monitor, but no matter what we did, we could not connect to the serial monitor and all we saw was an error informing us that the port was busy. This is not unique to the Arduino Nano RP2040 Connect; we have seen this behavior with other RP2040 boards and the Arduino IDE. Our only solution was to close the Arduino IDE and use another application to open a serial terminal. The same issue affected our test of the IMU, so we opened the serial terminal, reset the board and saw raw accelerometer data scroll across the screen. We later discovered that the issue lay with our operating system, Ubuntu 18.04, we had to uninstall modemmanager in order to see the serial data on the port. This issue has not impacted any of our other official and clone boards.

We also tested the onboard microphone, the same model as used in the Nano 33 IoT. We followed the official documentation to flash the test sketch to the Arduino Nano RP2040 Connect, then opened the Serial Plotter hoping to see a visualization of sound. Alas all we saw was a constant flatline value of -128. We were unable to locate any guidance on how to fix this issue.

(Image credit: Tom’s Hardware)

We had other issues when using the Arduino’s IoT Cloud, a service that we have used before with MKR WIFI 1010 and Nano 33 IoT boards. our Arduino Nano RP2040 Connect was detected, but could not be configured. We followed the official documentation, which at the time of writing looks to be incomplete, and still drew a blank. This is a real shame, as the IoT Cloud is Arduino’s new service to enable IoT “Things” (projects) to be created and monitored via a web interface. It is most likely teething troubles for now, and something that will be fixed in the future. For now the smart maker will stick with the desktop IDEs.

(Image credit: Tom’s Hardware)

You are probably thinking that this is an RP2040, so I can run other programming languages? Correct! We can use CircuitPython and MicroPython with the Arduino Nano RP2040, the caveat is that you will need to source your own libraries for the onboard microphone, IMU and WiFi. A quirk exists between the pin numbering and the actual GPIO number. The Arduino pin number, used in the Arduino IDE for a pin is say D2, but the RP2040 pin is GPIO25. Another example is the built in LED is D13, but in our MicroPython tests it was GPIO 6.

We tested the Arduino Nano RP2040 Connect with CircuitPython and it was a much more pleasant experience and thanks to the plethora of libraries, especially the ESP32SPI library we quickly had the Arduino Nano RP2040 Connect online and receiving data from a remote API. We also tested with MicroPython and it behaved in much the same manner as a Raspberry Pi Pico, but we had no drivers to use the onboard Wi-Fi.

This is where we noticed a quirk that prevented us from setting the board ready to accept new firmware. Typically we would reset the board and press BOOTSEL. For the Arduino Nano RP2040 Connect we would double press the reset button, but this did not work. After reaching out for help on Twitter we were told that we could connect the REC and GND pins together to force the board into the mode.

Use Cases for the Arduino Nano RP2040 Connect

The most obvious use case is an IoT appliance and this is made quite straightforward via the Arduino IoT Cloud. With the IoT Cloud we can quickly create GUI applications to read data from a remote device. For example we can create a weather station using the Arduino Nano RP2040 Connect and have live data streamed to the cloud which can be accessed from any web connected device.

The same Arduino IoT Cloud user interface can be used to remotely control the Arduino Nano RP2040 Connect, for example a web controlled robot wouldn’t be much more difficult to build than a non web enabled robot.

If you are not an Arduino IoT Cloud user, you can still make the most of this board using the traditional IDE. There are libraries available for HTTP, MQTT and Bluetooth which have existed for some time and can be ported to run on the RP2040. With these libraries we can build web controlled robots, wireless weather stations etc.

Bottom Line

It comes down to the price. If you need Wi-Fi and the RP2040 then the Arduino Nano RP2040 Connect is a solid piece of hardware but it is more expensive than any other RP2040 solution we’ve tested, including buying a Raspberry Pi Pico and a Pimoroni Pico Wireless. The extra cost gives us access to GPIO pins, which are sadly blocked with the Pico Wireless. If you need the smallest package possible for an embedded IoT project, then this could be the board, but right now the documentation is not up to Arduino’s usual standards despite this being the last board in the first wave of RP2040 powered devices.

The additional analog inputs mean that the Arduino Nano RP2040 Connect has the most of any RP2040 device, useful for connecting many different components. The smart money for those wanting to connect their RP2040 project to the world is with this board, but for how long? Seeed’s own RP2040 board comes with Wi-Fi and it is expected to retail for half the price of the Arduino.

AMD’s Discrete GPU Sales in Q1: Up Sequentially, Down Year-over-Year

(Image credit: Nvidia)

Sales of discrete GPUs in the first quarter were up both sequentially and year-over-year due to high demand for PCs in general and gaming PCs in particular, according to Mercury Research. Nvidia continued to dominate the market,

AMD has been gaining CPU sales and market share in the recent years particularly in the gaming segment as many of its processors do not have any direct rivals. But at the same time sales of the company’s discrete GPUs have been a roller coaster and its market share has been declining. In the recent months the situation with AMD’s standalone GPU supplies were tougher because of chip shortages and its market share in Q4 2020 declined to a multi-year low, according to Jon Peddie Research.

Shipments of discrete graphics processors in the first quarter reached 22 million units (which includes graphics chips for desktops and notebooks as well as for crypto mining), a 1% increase sequentially, according to data by Mercury Research cited by Capital Markets (which was cited by Barron’s). Typically, shipments of standalone GPUs decrease in Q1 versus Q4, but this was not a usual quarter.

Nvidia commanded 81% of standalone GPU shipments in Q1 2021, which was 5.2% higher than in the same period a year ago, based on data from Mercury Research. AMD commanded 19% of shipments, almost flat with the previous quarter and a decline compared to Q1 2020. Meanwhile, since unit sales increased sequentially and AMD’s share remained flat, it means that the company sold more GPUs in Q1 2021 than it did in Q4 2020.

GPUs used to mine Ethereum contributed to additional GPU sales in the first quarter, said Dean McCarron, the head of Mercury Research. Yet, it is hard to estimate how many of GPUs sold in Q1 are now used to mine cryptocurrency. McCarron estimates that sales of Nvidia’s custom crypto mining processors (CMP) totalled 1 – 2 million in the first quarter, which is well below his expectation of around 4.5 million.

Both AMD and Nvidia said recently that their shipments in Q1 2021 were constrained by shortages of production capacity at foundries as well as tight supply of components. Since AMD has contract obligations to supply its custom SoCs for the latest game consoles from Microsoft and Sony as well as CPUs to PC and server makers, it is not particularly surprising that the company could not increase shipments of its own GPUs significantly in the first quarter. Meanwhile, AMD hopes to boost sales of its graphics processors later this year.

Google Replaces Millions of Intel’s CPUs With Its Own Homegrown Chips

(Image credit: Intel)

Google has designed its own new processors, the Argos video (trans)coding units (VCU), that have one solitary purpose: processing video. The highly efficient new chips have allowed the technology giant to replace tens of millions of Intel CPUs with its own silicon.

For many years Intel’s video decoding/encoding engines that come built into its CPUs have dominated the market both because they offered leading-edge performance and capabilities and because they were easy to use. But custom-built application-specific integrated circuits (ASICs) tend to outperform general-purpose hardware because they are designed for one workload only. As such, Google turned to developing its own specialized hardware for video processing tasks for YouTube, and to great effect.

However, Intel may have a trick up its sleeve with its latest tech that could win back Google’s specialized video processing business.

Loads of Videos Require New Hardware

Users upload more than 500 hours of video content in various formats every minute to YouTube. Google needs to quickly transcode that content to multiple resolutions (including 144p, 240p, 360p, 480p, 720p, 1080p, 1440p, 2160p, and 4320p) and data-efficient formats (e.g., H.264, VP9 or AV1), which requires formidable encoding horsepower.

Historically, Google had two options for transcoding/encoding content. The first option was Intel’s Visual Computing Accelerator (VCA) that packed three Xeon E3 CPUs with built-in Iris Pro P6300/P580 GT4e integrated graphics cores with leading-edge hardware encoders. The second option was to use software encoding and general-purpose Intel Xeon processors.

Google decided that neither option was power-efficient enough for emerging YouTube workloads – the Visual Computing Accelerator was rather power hungry itself, whereas scaling the number of Xeon CPUs essentially meant increasing the number of servers, which means additional power and datacenter footprint. As a result, Google decided to go with custom in-house hardware.

(Image credit: Google)

Google’s first-generation Argos VCU does not replace Intel’s central processors completely as the servers still need to run the OS and manage storage drives and network connectivity. To a large degree, Google’s Argos VCU resembles a GPU that always needs an accompanying CPU.

Instead of stream processors like we see in GPUs, Google’s VCU integrates ten H.264/VP9 encoder engines, several decoder cores, four LPDDR4-3200 memory channels (featuring 4×32-bit interfaces), a PCIe interface, a DMA engine, and a small general-purpose core for scheduling purposes. Most of the IP, except the in-house designed encoders/transcoders, were licensed from third parties to cut down on development costs. Each VCU is also equipped with 8GB of usable ECC LPDDR4 memory.

The main idea behind Google’s VCU is to put as many high-performance encoders/transcoders into a single piece of silicon as possible (while remaining power efficient) and then scale the number of VCUs separately from the number of servers needed. Google places two VCUs on a board and then installs 10 cards per dual-socket Intel Xeon server, greatly increasing the company’s decoding/transcoding performance per rack.

Increasing Efficiency Leads to Migration from Xeon

Google says that its VCU-based machines have seen up to 7x (H.264) and up to 33x (VP9) improvements in performance/TCO compute efficiency compared to Intel Skylake-powered server systems. This improvement accounts for the cost of the VCUs (vs. Intel’s CPUs) and three years of operational expenses, which makes VCUs an easy choice for video behemoth YouTube.

Offline Two-Pass Single Output (SOT) Throughput in CPU, GPU, and VCU-Equipped Systems

System	Throughput (MPix/s)	Throughput (MPix/s)	Performance/TCO	Performance/TCO
	H.264	VP9	H.264	VP9
2-way Skylake	714	154	1x	1x
4x Nvidia T4	2,484	–	1.5x	–
8x Google Argos VCUs	5,973	6,122	4.4x	20.8x
20x Google Argos VCUs	14,932	15,306	7x	33.3x

From performance numbers shared by Google, it is evident that a single Argos VCU is barely faster than a 2-way Intel Skylake server in H.264. However, since 20 VCUs can be installed into such a server, VCU wins from an efficiency perspective. But when it comes to the more demanding VP9 codec, Google’s VCU appears to be five times faster than Intel’s dual-socket Xeon and therefore offers impressive efficiency advantages.

Since Google has been using its Argos VCUs for several years now, it clearly replaced many of its Xeon-based YouTube servers with machines running its own silicon. It is extremely hard to estimate how many Xeon systems that Google actually replaced, but some analysts believe the technology giant could have swapped from four to 33 million Intel CPUs for its own VC. Even if the second number is an overestimate, we are still talking about millions of units.

(Image credit: Google)

Since Google needs loads of processors for its other services, it is likely that the number of CPUs that the company buys from AMD or Intel is still very high and is not going to decrease any time soon as it will be years before Google’s own datacenter-grade system-on-chips (SoCs) will be ready.

It is also noteworthy that in an attempt to use innovative encoding technologies (e.g., AV1) right now, Google needs to use general-purpose CPUs even for YouTube as the Argos does not support the codec. Furthermore, as more efficient codecs emerge (and these tend to be more demanding in terms of compute horsepower), Google will have to continue to use CPUs for initial deployments. Ironically, the advantage of dedicated hardware will only grow in the future.

Google is already working on its second-gen VCU that supports AV1, H.264, and VP9 codecs as its needs to further increase the efficiency of its encoding technologies. It is unclear when the new VCUs will be deployed, but it is clear that the company wants to use its own SoCs instead of general-purpose processors where possible.

Intel Isn’t Standing Still

Intel isn’t standing still, though. The company’s DG1 Xe-LP-based quad-chip SG1 server card can decode up to 28 4Kp60 streams as well as transcode up to 12 simultaneous streams. Essentially, Intel’s SG1 does exactly what Google’s Argos VCU does: scale video decoding and transcoding performance separately from the server count and thus reduce the number of general-purpose processors required in a data center used for video applications.

(Image credit: Intel)

With its upcoming single-tile Xe-HP GPU, Intel will offer transcoding of 10 high-quality 4Kp60 streams simultaneously. Keeping in mind that some of Xe-HP GPUs will scale to four tiles, and more than one GPU can be installed per system, Intel’s market-leading media decoding and encoding capabilities will only become even more solid.

Summary

Google has managed to build a remarkable H.264 and VP9-supporting video (trans)coding unit (VCU) that can offer significantly higher efficiency in video encoding/transcoding workloads than Intel’s existing CPUs. Furthermore, VCUs enable Google to scale its video encoding/transcoding performance independently from the number of servers.

Yet, Intel already has its Xe-LP GPUs and SG1 cards that offer some serious video decoding and encoding capabilities, too, so Intel will still be successful in datacenters with heavy video streaming workloads. Furthermore, with the emergence of Intel’s Xe-HP GPUs, the company promises to solidify its position in this market.

Google Pixel Buds A-Series review: Sounds familiar

(Pocket-lint) – Google has announced a new version of the Pixel Buds, its true wireless headphones that originally launched in 2017 – the first-gen weren’t all that, though, while the second-gen Buds 2 stepped things up a little in 2019.

The third model belongs to the A-Series, picking up on the A series that we’ve seen in Google’s phones, presenting an affordable choice of true wireless headset.

What’s different to the previous Pixel Buds?

To look that, there isn’t a huge difference between the A-Series and Buds 2: both have the same overall styling and come in a case that’s smooth, much like a pebble.

Both have the same earbud design with a little promontory at the top to help keep them secure, and a round touch-control area on the outside.

Pocket-lint

The Pixel Buds 2 have wireless charging, however, and inside the case and on the inner part of the ‘buds have a matte finish to the plastics, while the A-Series is glossy. That means the older version looks slightly higher quality.

The A-Series also lacks the option to change the volume via gestures – instead you have to use voice for that – and there are a few minor feature differences. Otherwise, the experience is much the same – but the A-Series is much cheaper.

Design & Build

Earbud: 20.7 x 29.3 x 17.5mm; 5.06g
Colours: Dark Olive / Clearly White
Case: 63 x 47 x 25mm; 52.9g
IPX4 water-resistant
Three ear tip sizes

The Buds A-Series’ case, for all intents and purposes, is the same as that of the Buds 2: it’s the same size, has the same feel, and that same satisfying action when you open and close the lid. Both have a USB-C charging port, a manual connection button on the rear, but the A-Series is slightly lighter.

There’s a satisfying magnetic action when you drop the ‘buds into the case to charge and don’t worry about mixing these up if you happen to have the older version too – the A-Series has two charging contacts inside, the Pixel Buds 2 has three.

Pocket-lint

There are two colours to the A-Series – Clearly White or Dark Olive – and opening the lid reveals the colour you’re looking at, as it’s the touchy smooth round end of these Buds, carrying the ‘G’ logo, which makes them really distinctive.

The A-Series ‘buds have the same design as the previous model, with the body of the earbud designed to sit in the concha of the ear, while sealing into the canal with a choice of three different ear tips. These are round – Google seemingly hasn’t been tempted to move to oval as seen on some rivals.

There’s an additional rubber arm that sticks out the top of the buds that is designed to slot into one of the folds at the top of your ear to help keep things secure. We weren’t a fan of it on the previous version and we have the same reservations here: you can’t remove it from the ‘buds and we’re not convinced it’s necessary. As for us, the Buds A-Series sit securely in the ear anyway – even when exercising.

Indeed, if we rotate the earbuds to get that blobby rubber ant to engage with our ears, the sound from the headphones gets worse because they then don’t sit in the best position for our ears. That’s one thing to consider: all ears are different, so this might work for some people and not for others.

Pocket-lint

The great thing about these earbuds’ design is that they don’t hang out of your ear, so you don’t need to worry about pulling a hat over the top or anything else – we think they look a lot better than the ear-dribble style of Apple’s AirPods and all those who copy them. We find the Google design more comfortable for wearing over long periods, too.

Connection, setup and control

Native Pixel support
Pixel Buds app
Touch controls

Google Fast Pair means you just have to lift the lid of the case and your nearby Android phone will detect the Pixel Buds A-Series and allow you to connect with one tap. It’s essentially the same as Apple’s system with the AirPods and iPhone, linking the Buds to the Google account you register them with so they are then available on other devices too.

If you’re using a Pixel phone then you’ll have native support for the Buds; if using another brand Android device you’ll be prompted to download the Pixel Buds app, which will provide access to firmware updates and details on how to use all the features, as well as some options.

Pocket-lint

As far as setup is concerned, that’s all there is to it: you’ll be asked to walk through things like Google Assistant, and you’ll be prompted to allow notifications access, so you can unlock the potential of the Pixel Buds.

The touch-controls are fairly easy to master, too, with both left and right sides offering the same function: single-tap to play/pause; double-tap to skip forward; triple-tap to skip backwards; press-and-hold to get a notifications update.

The last of those is interesting, because you’ll get a report of the time and then you’ll be told about your notifications – with the option to reply, needing a press-and-hold to speak your reply, before it’s confirmed and then sent.

Pocket-lint

Missing from this selection of touch-controls is volume: unlike the Pixel Buds 2, you can’t swipe to change the volume, you have to ask Google Assistant to do it or you have to thumb the volume controller on your device instead.

This, we feel, is the biggest flaw of these headphones: volume control is pretty important when you’re listening to something, so having to ask Google using voice just isn’t appropriate in all situations.

Google Assistant and smart features

Google Assistant integration
Adaptive Sound

With a lack of volume control, Google pushes its Adaptive Sound option as a solution. This is designed to adapt the volume to the ambient sound levels. As the external noise goes up, so does the volume of the headphones. That’s fine in principal and works when you move from and area of consistent background noise to another – from a quiet library to a server room with whirring fans, for example – but it’s hopeless when you have varying noise levels.

Just walk along a busy street with Adaptive Sound on and you’ll find the volume of the headphones yo-yoing, because it’s not constant noise, it depends on what’s driving past at that moment. This could be corrected by a software update with Google reducing the frequency of volume changes. If you manually adjust the volume then it suspends the system for a bit and leaves the control to you, but in reality, it’s just too irritating to use in many situations and you might as well turn your phone volume up instead.

Pocket-lint

As we’ve said, Google Assistant is fully integrated into the headphones, so you can ask Google anything that you might on your phone or Nest Hub at home. For fans of the system, that’s a great addition, because you don’t need to fish your phone out of your pocket first. Sure, there are lots of headphones out there that offer Google Assistant, but naturally, Google puts Google first and the experience is nice and smooth.

It’s also a two-way experience, with Google Assistant notifying you of incoming messages and it’s able to read them out to you too – with the option to speak a reply. You can disable messages from any apps you don’t want in the Pixel Buds app, to maintain privacy (or, indeed, a barrage of non-stop voiced messaging). You can also trigger message sending through voice – and you’ll get to confirm the message that’s being sent.

Thanks to Voice Match, it will only respond to your voice – and that also means you can access things like your calendar and so on. It’s plain sailing all round.

Sound quality and performance

Buds: 5 hours battery life
Case: 19 hours extra
Spatial Vents
Bass Boost

When it comes to the performance, Google is taking a bit of a gamble. Rather then pursuing isolation from the outside world, it wants to provide an experience that lets some of the ambient sound in, so you don’t feel cut off.

Pocket-lint

Google uses what it calls Spatial Vents, while claiming that the headphones provide a gentle seal rather than trying to block everything out. We’re not huge fans of this approach and with the rise in headphones offering active noise cancellation (ANC), it suggests that generally speaking that’s what people are buying.

Needless to say, there’s no ANC here and you’ll be able to hear what’s happening around you a lot of the time. At home that’s perhaps useful – you can hear the doorbell or the dog bark – but out on public transport, you’ll hear every announcement, door crash, clatter of the wheels on the tracks, and that’s not something we want. This is exactly the same experience as the previous Pixel Buds and whether that suits you will depends very much on where you wear your headphones. If that’s a busy place, the A-Series might not be the best for you.

Aside from that, in quiet conditions, the sound quality is actually very good. The Pixel Buds A-Series benefits from the Bass Boost option that Google added as a software update to the previous Buds in late 2020, so they offer better performance for tracks which want a driving bassline. In quiet conditions at home we have no complaints: the Pixel Buds A-Series is a great pair of headphones, especially at the asking price and given the smart options they offer.

Pocket-lint

When it comes to calling there are two beam-forming mics on each ‘bud, but they still let noise through to the caller. This is reduced, but they’ll hear every car that drives past as a hiss. If you’re after a better calling experience, the Samsung Galaxy Buds Pro offers a far better experience, providing a better veil of silence when making calls.

The Pixel Buds A-Series provides battery life of 5 hours, which we’ve found to be accurate – although we found the left ‘bud to drain slightly faster than the right one. The case supplies 24 hours of life, recharging the buds when they are back in it, and charged itself via USB-C. This isn’t the longest battery life on the market, but it matches the Apple AirPods.

Best Bluetooth headphones 2021 rated: Top on-ear or over-ear wireless headphones

By Mike Lowe
·
3 June 2021

Verdict

The Pixel Buds A-Series have a lot to offer considering the price: Google Assistant integration, comfortable design, a lovely case, plus great audio performance when in quieter conditions.

The biggest downsides are the lack of on-bud volume controls and the design decision to not strive for isolation from external noise. The Adaptive Sound – which auto-adjusts volume – is a good idea in principle to compensate for this, but it sees the headphones’ volume yo-yo unnaturally.

Compared to the older Pixel Buds 2, we’d pick the Pixel Buds A-Series every time: they do the important things just as well but the price is much more approachable, meaning you can forgive the omissions given the context of price.

Also consider

Pocket-lint

Samsung Galaxy Buds Pro

Samsung’s Galaxy Buds Pro offer great noise-cancelling – which is especially effective when making calls – while also offering a great set of features.

Read the full review

squirrel_widget_3816695

Jabra Elite Active 75t

These headphones are a little more bulky, but they offer noise-cancellation that will almost entirely eliminate external noise. If you want silence, Jabra delivers it.

Read the full review

squirrel_widget_176733

Writing by Chris Hall.

Seeed reServer: Compact and Powerful Edge Computing

(Image credit: Seeed)

Seeed today announced the first member of its reServer System today. The reServer x86 series is designed for edge computing and IoT applications. The systems come powered by an 11th-Gen Intel processor, and higher-spec models feature Intel Iris Xe graphics.

Image 1 of 4

(Image credit: Seeed)

Image 2 of 4

(Image credit: Seeed)

Image 3 of 4

(Image credit: Seeed)

Image 4 of 4

(Image credit: Seeed)

Seeed’s reServer is a compact piece of kit measuring 4.8 x 5.2 x 9 inches (124 x 132 x 230 mm), giving it enough height to house two 3.5 inch SATA drives and plenty of space for cooling using heat pipes and vapor chambers. The lowest spec reServer is the 11th-Gen Core i3 model with Intel UHD Graphics. If you want Intel Iris Xe graphics, then you should look at the i5 and i7 models.

No matter which version you opt for, each has a TDP of 28W. The mainboard has M.2 expandability for SSD and 4G/5G connectivity, and you can upgrade the RAM to a maximum of 64GB of DDR4-3200. The onboard HDMI 2.0b port supports up to 4k60, and the Displayport connector provides up to 7680×4320 at 60Hz. An additional LCD eDP 40 pin connector provides another means to output video. In total, there are four simultaneous display outputs if we count the USB-C Thunderbolt 4 connection on the i5 and i7 models. An Intel I225-V Ethernet controller powers two 2.5Gb Ethernet ports. If wireless communications is more your thing, reSpeaker offers Wi-Fi 6, LoRaWAN and 4G / 5G via an optional module.

The i3 model has a single USB 2.0 and USB 3.0 port, whereas the i5 and i7 models have two USB 3.2 Gen 2 and two USB 2.0 ports. Also present are RS-232 / 422 and 485 COM ports along with a 28-pin breakout for an ATSAMD21G18 32-Bit Arm Cortex M0+ microcontroller which is programmable via the Arduino IDE.

To power the unit, we rely on an external power brick that supplies the needed 12V DC. For i5 and i7 models, we have the option of USB PD via the USB-C port.

Edge computing is computing performed near the source of the data rather than in the cloud. This means that a project or organization does not rely on the cloud infrastructure provided by an external company.

Seeed is aiming the reServer at many different applications, including the traditional small office/home office server, but more interestingly, they see applications in the industrial automation, smart vision/city and healthcare industries.

Prices start from $669 for the i3 model with 8GB RAM, 256GB SSD and Wi-Fi 6 and a 16GB 512GB SSD model is available from $769.

Samsung Unveils ZNS SSD: QLC V-NAND for Datacenter-Grade Endurance & Performance

(Image credit: Samsung)

Samsung has introduced its first Zones Namespaces (ZNS) solid-state drives that combine high performance, long endurance, relatively low price enabled by the company’s QLC V-NAND, and improved quality of service (QoS) for datacenters. To use Samsung’s new PM1731a ZNS SSDs, datacenters will have to deploy new storage systems and software.

ZNS SSDs work differently than conventional block-based drives and have a number of advantages. ZNS SSDs write data sequentially into large zones and have better control over write amplification, which reduces over-provisioning requirements by an order of magnitude. For example, some enterprise drives rated for 3 DWPD (drive writes per day) reserve about a third of their raw capacity for over-provisioning, but for ZNS SSDs about 10% of that is enough. In addition, since ZNS uses large zones instead of many 4KB blocks, garbage collection is not needed as often as traditional SSDs, it also improves real-world read and write performance.

Samsung’s 2.5-inch PM1731a ZNS SSDs are based on the company’s proprietary dual-port controller as well as 6th Generation QLC V-NAND memory. The drives will be available in 2TB and 4TB capacities.

(Image credit: Samsung)

Samsung says that its new PM1731a ZNS drives will last up to four times longer than conventional NVMe SSDs, which will reduce their total cost of ownership (TCO) and will simplify server infrastructure.

But enhanced endurance and performance come at a cost. ZNS ecosystem requires new software infrastructure that is not yet widely available. In a bid to make ZNS more widespread, Samsung is participating in a number of open-source ZNS projects. The manufacturer also plans to make its ZNS technology available to xNVMe and participate in the Storage Performance Development Kit (SPDK) community to enable NVMe and SPDK users to implement ZNS more easily.

Samsung will start mass production of its PM1731a ZNS SSDs in the second half of the year. The company is the second major maker of SSDs, the first being Western Digital, to unveil a ZNS SSD.

“Samsung’s ZNS SSD reflects our commitment to introducing differentiated storage solutions that can substantially enhance the reliability and lifetime of server SSDs,” said Sangyeun Cho, senior vice president of the Memory Software Development Team at Samsung Electronics. “We plan to leverage quad-level cell (QLC) NAND technology in our next-generation ZNS drives to enable higher thresholds for storage performance and capacity in the enterprise systems of tomorrow.”

ARMv8 has been around for years, it is the first 64-bit version of the ARM architecture. Now it’s time to move on to ARMv9. No, it’s not jumping to 128-bit, but it does bring the end of 32-bit designs. ARM is planning to take Cortex-A cores 64-bit only by 2023 and it is working with its partners to ensure that the software ecosystem is ready to wave goodbye to 32-bit apps (a process that ARM hopes will be completed by the end of this year).

Bits aside, ARM unveiled new CPU and GPU designs that will be featured in future chipsets as well as new support hardware to tie them all together in varying configurations that will be used in everything from laptops, through phones to smart TVs and other multimedia appliances. There are some exciting high-end designs, but it’s the entry-level stuff that may prove to be a game-changer.

Cortex-X2, A710, A510 CPU cores

The Cortex-X2 is the second generation of the new breed of high performance ARM cores, it promises a 16% speed boost over an X1 core built on the same node and running on the same frequency.

However, these new designs are expected to be on the market next year, so more than likely they will be fabbed on upgraded nodes compared to the 2021 designs. ARM has optimized the X2’s peak performance and has doubled its machine learning (ML) performance. According to ARM, X2’s peak single-thread performance is 40% higher than a 2020 laptop CPU – a 15W Intel i5-1135G7.

Also, the new DSU-110 (“DynamIQ Shared Unit”) allows up to 8 Cortex-X2 cores to be used in high-performance designs (this is more for laptop/desktop and even server use, not for phones). The L3 system cache is doubled to 8 MB and ARM has worked hard to reduce the latency of data access.

The big cores of the A-series also have a new design, however, with the high-performance X-core around they have transitioned to a more middle ground position. So, the new Cortex-A710 is only 10% faster than an A78 (on the same node and frequency), but energy efficiency has gone up 30% (and, again, machine learning speed has doubled).

That is selling the A710 a little short. ARM predicts that for gaming workloads a chipset featuring an A710 CPU and Mali-G710 GPU will see a 33% boost in CPU performance, 20% in GPU performance and will be 15% more efficient compared to a current A78/G78 design. That’s a pretty big generational improvement.

For the first time since 2017 ARM has designed a new little core too – the Cortex-A510 may actually be the most exciting part of today’s announcement. It is 35% faster than the old A55, 20% more energy efficient and with 3x the machine learning capabilities. According to ARM, the A510 is close to older big A-cores in terms of performance. This will be a huge improvement for entry-level and even mid-range phones.

How ARM achieved this is also interesting – with the new design, two A510 cores actually share bits of hardware to make them physically smaller (and thus cheaper to manufacture). A similar design idea didn’t work out great for AMD, but ARM has found a way to make it work (at least for the needs of a high-efficiency core).

ARMv9 focuses on security and the ability to crunch large amounts of data. The Memory Tagging Extension will make it easier to suss out and prevent memory leaks and other memory issues that both hurt performance and pose a security threat.

Mali-G710, G610, G510 and G310 GPUs

Did you know that Mali is the #1 GPU in terms of shipments? Over 1 billion Mali GPUs were shipped in 2020. They power about half of smartphones out there and around 80% of smart TVs. And today ARM is bringing out the widest range of GPU designs that will fit every nice of the market.

The new Mali-G710 sits at the top. It is 20% faster, 20% more energy efficient than older designs and gets a 35% boost in ML tasks. The G710 will be featured in future smartphone flagships but also Chromebooks. The ML speed boost will come in handy for new improved image enhancements and enabling new video modes.

Below that is the Mali-G610, which is based on the G710, though it targets a lower price point and can be used in high-end phones.

The Mali-G510 is twice as fast and 22% more energy efficient than older designs (ML performance is doubled too). This will become a mainstay of mid-range phones, smart TVs and set-top boxes.

The Mali-G310 is the second most exciting part of today’s announcement after the A510. These two will really change the experience on the lower end. It promises a 4.5x uplift in Vulkan performance, the texture units are 6x faster, Android UI rendering performance is doubled.

	Zen 4*	Alder Lake*	Zen 3
Codename	Raphael	Alder Lake	Vermeer
Branding	Ryzen 6000	Core 12000-series	Ryzen 5000
Lithography	5nm	10nm	7nm
Socket	LGA1718	LGA1700	AM4
Maximum Cores	24	16	16
Memory Support	DDR5	DDR5 / DDR4	DDR4
PCIe Support	PCIe 4.0 x 28	PCIe 5.0 x 16, PCIe 4.0 x 8	PCIe 4.0 x 24
Maximum TDP	170W	?	105W

keyword: Server

Arduino Nano RP2040 Connect Review: Built-in Wi-Fi and BLE

Our Verdict

For

Against

Arduino Nano RP2040 Connect Hardware Specifications

Design and Use of the Arduino Nano RP2040 Connect

Use Cases for the Arduino Nano RP2040 Connect

Bottom Line

AMD’s Discrete GPU Sales in Q1: Up Sequentially, Down Year-over-Year

Google Replaces Millions of Intel’s CPUs With Its Own Homegrown Chips

Loads of Videos Require New Hardware

Increasing Efficiency Leads to Migration from Xeon

Offline Two-Pass Single Output (SOT) Throughput in CPU, GPU, and VCU-Equipped Systems

Intel Isn’t Standing Still

Summary

Google Pixel Buds A-Series review: Sounds familiar

What’s different to the previous Pixel Buds?

Design & Build

Connection, setup and control

Google Assistant and smart features

Sound quality and performance

Also consider

Samsung Galaxy Buds Pro

Jabra Elite Active 75t

Seeed reServer: Compact and Powerful Edge Computing

Samsung Unveils ZNS SSD: QLC V-NAND for Datacenter-Grade Endurance & Performance

Micron Ships First 1-Alpha DRAM: Extreme Density Without EUV

Google Reveals ‘Half-Double’ Technique for Exploiting Rowhammer in DDR4

Marvell’s New PCIe 5.0 SSD Controllers Have 14 TB/s of Throughput

AMD’s Zen 5 CPUs, APUs Will Likely Tap TSMC’s 3nm Process Node

ARM unveils Cortex-X2, A710, A510, new Mali GPUs as it prepares to go 64-bit only

Elite Dangerous Odyssey launches to ‘mostly negative’ reception from players

Mockup of AMD’s Zen 4 Raphael CPU Appears With Potential Specs

AMD Zen 4 Raphael Specifications

Russian Filing Suggests 100+ New Asus GPUs

Frontier Supercomputer to Get World’s Fastest Storage: 75 TB/s, 15 Billion IOPS, 700 PetaBytes

The Orion Global File Storage System Layer: 700PB Capacity at 10TB/s

The In-Storage Layer: Up to 75TB/s at 15 Billion Read IOPS

Summary