Many everyday objects are now ‘smart’ and connected to the internet. If you feel the need to have everything automated or you want to check your email using a toaster, then you might be in luck. Google are working on a new operating system, ‘Brillo’, for low powered devices that could run on as little as 32 megabytes of RAM.
To bolster Google’s attempt to have a finger in every pie, the global giant are looking to make a new OS that would run on low amounts of RAM and power. Their aim is to develop software that can power any electronic device that can connect to the internet, even if it doesn’t have a digital screen. To put things in perspective, the aim is to have the software run on 32 megabytes of random-access memory, but at the moment its fully fledged older brother, Lollipop, is aiming at phones with a minimum of 512MB RAM. So there is quite a step down from its sibling, but this could be good news for embedded devices like the Arduino and other embedded tech.
The team developing the OS are linked to the company’s Android unit, so it isn’t a giant leap to think that it will be released with the Android brand. However, for now they are calling the software ‘Brillo’.
Google aren’t the first to get into the Internet of things market, Microsoft is releasing Windows 10 IoT core, Huawei announced recently an OS for IoT gadgets that is only 10kb in size and Samsung have announced a series of chips designed for IoT devices. So with all these large companies jumping on the bandwagon, it’s safe to assume that the Internet of Things is just around the corner and our truly connected lives are about to begin.
TDK is a company that I haven’t thought of in years, actually many years as I don’t think I’ve seen the name since the time of audio cassettes (MC), but they’re still around and still producing a lot of products in various regions of the electronic sectors. One of their newest products is a SATA III SSD built around their own in-house SSD controller, the SDS1B series of SSDs.
TDK developed their own controller for this drive, working from the base of their old controllers. The new GBDriver GS1 NAND flash controller features the recovery and refresh functions of the old controllers as well as a power interruption tolerance algorithm and integrated power supply protection circuit; it will work with both MLC and SLC NAND flash memory.
The new SSDs presented here come without a DRAM or other type of cache, meaning they will run slower than what we’re used to from consumer class drives. The new product can achieve read access speeds up to 440 MB/s and write access speeds up to 240 MB/s with MLC and up to 340 with SLC flash.
These drives are intended for embedded and industrial applications and reliability is more of an issue than the extra speed. Each feature and chip added to something is another part that can fail and stop the drive from functioning and by not adding a caching function, TDK eliminates one more of these.
TDK SDS1B features include global static wear leveling algorithm for the entire drive and improved power interruption tolerance with an internal power supply protection circuit to prevent collateral data errors. The auto-refresh function reads all data including little-used data and automatically performs error correction (ECC) when needed and TDK also added a function for specifying a data range and completely deleting all data in it. The drives feature AES 128/265-bit hardware encryption and also TDK’s proprietary security function that requires mutual authentication from both host and SSD. SMART and TRIM are also present as it should be on any drive.
TDK will demonstrate the drive at the Embedded Systems Expo (ESEC) to be held in Tokyo this week and the new products will be available from August 2015.
Don’t feel bad if you haven’t heard of Habey before, most people haven’t. They have however been in the business for quite some time and probably longer than some of our readers have been alive, so they know a thing or two about creating solutions for embedded and small form factor systems. Their newest product, the mini-ITX motherboard dubbed the MITX-6670 is somewhat of a swiss army knife of the small form factor with a lot of onboard functions and even more upgrade functions.
As a base, you’ll get a mini-ITX motherboard with a passive cooled Intel Baytrail quad-core J1900 processor running at 2.4GHz and two SO-DIMM slots for up to 8GB total DDR3 memory. There are two onboard Intel LAN port and there is an onboard header for an optional POE module. You can connect two SATA drives or two mSATA drives, but not both at once. The selection is made by jumper and one of the mSATA ports support full size where the other only allows half-size mini-PCIe.
But it doesn’t stop here as you can customize even more. There is a default PCIe x1 slot, but the really cool feature is the onboard header that allows you to expand with what you, may it be an extra HDMI port or a DisplayPort connector instead.
The built-in iGPU supports HD resolutions and the Realtek ALC892 chip powers the audio part. There’s also support for a Trusted Platform Module (TPM 1.2) and you can even insert a sim card for 3G, 4G, and LTE connectivity.
Dual Intel I210/I211 Gigabit EthereEthernet
Built-in Ethernet port pin-header ready for POE (Power over Ethernet) setup
HDMI + VGA and LVDS/HDMI/DP/eDP add-on interface card options
Dual mSATA SSD support with easy jumper selection
Build-in Inﬁneon Trusted Platform Module and 8-bit GPIO
Realtek ALC892 high-performance 7.1-channel HD Audio
Low power, fanless passively cooling
Habey’s MITX-6770 has a manufacturer suggested retail price starting at $159, which actually isn’t much considering the features you get and compared to other available boards with fewer features and functions.
Super Talent Technology announces its newest and third mSATA SSD, the new JS2 solid state drive designed for industrial and embedded applications. The mSATA drives might not be the fastest on the block, but they have the big advantage of a tiny form factor and wide compatibility. There aren’t many compact or embedded systems that don’t come with at least one mSATA slot.
The mSATA SJ2 is an updated version of the SJ1 solid state drive and it provides fast speeds of up to 480 MB/sec reads and 160 MB/sec writes with access times down to 0.2ms. The drives are available with 16GB, 32GB, 64GB and 128GB capacity so you can pick just the right one for your system, why get more space than the system needs. The drives are based on the Mini PCIe and architecture and utilize the SATA-III interface. From the provided image it looks to use a SandForce SF-2141 SSD Controller and Micron NAND packages.
Thanks to SuperTalent for providing us with this information
Hong Kong based DRAM manufacturer, I’M Intelligent Memory, has announced its 8 GB DDR3 components with a single chip, which doubles the amount of memory per chip compared to other DRAM devices on the market. Based on the latter chip, the company is said to have introduced the 16 GB DDR3 UDIMM and SO-DIMM memory modules, having EEC error-correction technology as an optional upgrade.
It is said that the JEDEC specification JESD9-3 has always allowed a 8GB capacity for DDR3 devices. However, it seems that most manufacturers are waiting for the 2x nm process in order to fit smaller chips and bring high memory capacity. I’M Intelligent Memory has apparently made the leap by developing their own way of manufacturing 8 GB DDR3 components with a single chip using existing 30 nm technology.
The company states that their memory modules are compatible with the JEDEC standard pinout, timing and row/column/bank addressing. In addition to the latter, the company has made available devices including x8 (1Gx8) configuration in FBGA 78 ball package, a x16 (512Mx16) type in FBGA 96 ball package, a x32 (256Mx32) configuration in FBGA 136 ball package, as well as providing DDR3L low-voltage 1.35V versions, all of which are currently available on the market.
Given the 8 GB device, the company has released its first 16 GB DDR3 240 Pin unbuffered DIMMs and 204 Pin SO-DIMMs on the market, while also having them available in 72 Bit width for EEC error correction. The latter modules are said to be compatible with processors and micro-controllers from AMD, Cavium, Freescale, Tilera and others.
While not all processors used in desktop PCs are compatible with the high-capacity memory, Intel has offered support for the Atom C2000 series and Atom E3800 series with a new BIOS version available to download now. Also, ASUS has confirmed support for the latter memory on its X79-DELUXE, RAMPAGE IV BLACK EDITION and other ASUS X79 motherboards. Other manufacturers, such as ASRock, Supermicro, AIC and Portwell have already verified and approved the IM 16 GB DDR3 memory modules for a variety of their motherboards based on ADM, Tilera, Intel’s C2000 series and other processors.
I’M Intelligent Memory apparently has noticed the potential embedded markets, networking and telecommunication applications, as well as PCs and laptops, allowing all to reach a memory capacity previously untouched by any manufacturer out there.
AMD has announced their first discrete graphics card based on Graphics Core Next architecture, the AMD Embedded Radeon E8860 codenamed “Adelaar”, which is said to advance the visual growth and parallel processing capabilities of embedded applications.
Compared to its predecessor, the AMD E8860 is said to deliver 3D and 4K graphics to embedded devices, such as digital signage, commercial aerospace and conventional military, medical imaging, gaming machines, and other embedded applications, having a 33 percent higher single precision floating point over the previous generation.
Also, having an output of 768 GFLOPS (single precision), it enables the AMD E8860 to tackle the most complex parallel applications like terrain and weather mapping, facial and gesture recognition and biometric and DNA analysis.
“The demand for rich, vibrant graphics and enhanced parallel compute capabilities in embedded systems is greater than ever before, and is expected to continue to grow as we enter the Surround Computing era,” said Scott Aylor, corporate vice president and general manager, AMD Embedded Solutions. “Legacy graphics no longer meet the needs of embedded solutions for today and tomorrow. With unprecedented performance-per-watt, the AMD E8860 GPU addresses the need to drive multi-displays, 3D and 4K content and GPGPU compute across small form factors, harsh environments and mission critical applications.”
AMD’s E8860 is said to feature 2GB of GDDR5 frame buffer having a 128-bit bus interface, advanced GPGPU capabilities for parallel processing, AMD APP technology and improved performance with 92 percent higher 3D graphics performance-per-watt than the previous generation, sporting a 2689 score in 3DMark 11. Other features span from the ability to have support for multiple displays with the help of AMD’s Eyefinity technology, to support for DirectX 11.1, OpenGL 4.2 and OpenCL 1.2.
In terms of operating system support, the E8860 is stated to run on Microsoft Windows 7, Windows Embedded 7 Standard, Windows 8/8.1, Windows Embedded 8 Standard, Linux and real-time and safety-critical operating systems supported by CoreAVI’s suite of embedded software drivers.
The AMD E8860 GPU is stated to have a seven-year longevity supply guarantee and is available as a mobile PCI Express module as well as a PCI Express add-in board. No price has been announced at the moment, though the GPU is said to start shipping in Q1, which most probably will be March.
Thank you Guru3D for providing us with this information
AMD has related its financial situation today at the Q4 2013 Earnings Conference, where they have detailed their milestones, earnings and overview of their company’s financial situation. They have stated that they were successful in executing their “3 step strategic turnaround”, which ultimately lead to “transform” AMD.
According to AMD’s officials, they have accelerated their business growth during the 2013 year, leading to profit growth in the second half of 2013 by diversifying their product range from semi-custom SoCs, ARM and all the way to the PC component market.
In Q4, AMD reported a $1.59 billion revenue, a 9% sequential increase and 38% annual increase since 2012. They have also exceeded the goal for the semi-customed and embedded market with a 20% revenue growth in this sphere for the Q4 2013, ultimately leading to their expectations of reaching a 50% increase in revenue by 2015.
The embedded and semi-customed market propelled the business in Q4, where the demand for Sony and Microsoft had been met and both companies combined sold and estimate of 7 million next-gen consoles in less than 2 months (double than the previous number of consoles sold in their first quarter).
In terms of graphics solutions, AMD pointed to an excellent year in growth revenue, pointing to the dual FirePro graphics solution inside Apple’s Mac Pro. Though, the R7 and R9 chips have also been brought up, where they related a sequential increase in revenue from the chip demands, which they are expected to see in 2014 as well.
APUs and FX CPUs have also seen a steady increase during the third and all the way to the fourth quarter of 2013, having the launch of the Kaveri chipset market as a turning point and an exponential increase in revenue. AMD is aiming to provide a performance oriented solution at a reasonable price, having mentioned the A8 and A10 series of Kaveri CPUs as the best alternative to other competitors, which we mainly think is Intel’s Haswell.
The hardest part comes last, where AMD is aware of the 2013 decrease in the PC market, which they felt as well, and are continuing to cautiously approach it during 2014. They have predicted a slight decrease in 2014 as well and have stated to tackle the decrease quarter by quarter, while managing their expansion into other areas where AMD is not strongly represented to compensate their estimated loss in the PC sphere this year.
All in all, AMD has been keeping in on the green last year, and are going to do the same in 2014 by getting more involved in the semi-customed SoC, embedded and ARM markets.
Intel’s Haswell technology hasn’t had a chance to permeate all parts of Intel’s CPU portfolio. The embedded sector is still rather lacking of the new Intel architecture but from September we can expect to see them arrive with Intel’s 4th generation Shark Bay platform. The new releases of embedded CPUs will contain five new processors. These will be the following:
(Note that Thread counts are just educated guesses as this information was not disclosed.)
Core i7 4700EQ @ 1.7GHz with 4C/8T and a 37W TDP
Core i5 4400E @ 2.7GHz with 2C/4T and a 37W TDP
Core i5 4402E @ 1.6GHz with 2C/4T and a 37W TDP
Core i3 4100E @2.4GHz with 2C/2T and a 37W TDP
Core i3 4102E @ 1.6GHz with 2C/2T and a 25W TDP
All of these new embedded chips are based on a mobile Haswell design and are embedded into QM87 or HM86 chipset motherboards. Consequently the feature set of the platform will vary between the different motherboards that these CPUs are embedded on. The CPUs come in a BGA 1364 package. You can check out more detailed information here and here courtesy of EKF-System.
Image #1 courtesy of Intel and Image #2 courtesy of EKF-System