Thread: Anandtech News

Today we're looking at the flagship PSU series from be quiet!, the Dark Power Pro 10. We received two of the models, the 650W and the 850W versions, but despite being part of the same series and sharing the same features, the six Dark Power Pro 10 PSUs are have significant differences. We're going to look at these two models to explore those differences and gain a better idea of what to expect from the Dark Power Pro series.

More...

The original Raspberry Pi, launched three year ago this month, has been very successful. Raspberry Pi founder Eben Upton noted recently that over 4.5 million Pis have been sold since launch, vastly exceeding the initial goal of 1k units. This popularity can be attributed to many things, but one of them was a consistent hardware platform. Similar to gaming consoles, each Raspberry Pi has almost the exact same hardware, enabling developers to code apps that were guaranteed compatibility. Thus, laymen can find value in Raspberry Pi even without any intention of developing on the platform. In fact, Kodi (XBMC) has announced that the Raspberry Pi is the most popular hardware for their software.
Reaching the $35 price point was difficult however, and one peculiarity of the Raspberry Pi is its downright ancient ARM11 CPU inside the Broadcom BCM2835 SoC. While the ARM11 CPU was a watershed moment for mobile computing when it launched (it was at the heart of the original and 3G iPhones launched in 2007 and 2008), to continue using it in 2015 is a large concession. Not only is it slow, but the ARM11 core does not use the most popular ARM ISA, ARMv7, and instead relies on the older and less popular ARMv6 ISA. Thus, most software does not run natively on the Pi and must be recompiled.
Many online software repositories maintain ARMv7 sections for everything-but-the-Pi as well as a special Raspberry Pi ARMv6 section. Considering the popularity of the Raspberry Pi, this actually works okay, but the lack of a modern CPU has left the door open for competitors. Projects such as Vero, ODRIOD, and Imagination’s CI20 have all differentiated themselves with much more powerful and modern CPUs, however at significantly higher prices.
This morning, the Raspberry Pi foundation changes that by introducing Raspberry Pi 2 at the same $35 price point. Working with Broadcom, the Raspberry Pi foundation has attempted to maintain as much compatibility as possible by changing only two features of the design. The new Broadcom 2836 SoC replaces the older 2835 SoC, but is nearly identical other than the CPU. The 700 MHz ARM11 CPU is replaced with a quad-core set of 900 MHz ARM Cortex-A7s, which use the more standard and modern ARMv7 ISA. Additionally, the memory has been doubled and increases from 512MB to 1GB. Otherwise, the Pi 2 is exactly the same as the Raspberry Pi 1 B+.

	Raspberry Pi 2 Model B	Raspberry Pi Model B+
SoC	Broadcom BCM2836	Broadcom BCM2835
CPU	4x 900 MHz ARM Cortex-A7 (ARMv7 instruction set)	700 MHz ARM1176JZF-S (ARM11 family, ARMv6 instruction set)
GPU	Broadcom VideoCore IV @ 250 MHz	Broadcom VideoCore IV @ 250 MHz
Memory	1GB DDR2	512MB DDR2
USB	4x USB 2.0 Host	4 USB 2.0 Host
Camera Interface	15-pin MIPI (CSI) connector	15-pin MIPI (CSI) connector
Video Output	HDMI 1.3a Composite video 3.5mm jack	HDMI 1.3a Composite video 3.5mm jack
Audio In	via I²S interface	via I²S interface
Audio Output	HDMI 3.5mm jack I²S interface	HDMI 3.5mm jack I²S interface
Storage	via MicroSD	via MicroSD
Network	10/100 Ethernet	10/100 Ethernet
Max Power Consumption	< 5 Watts	3 Watts
Power Connector	MicroUSB	MicroUSB
Size	85.60 mm &times; 56.5 mm (3.37 in &times; 2.22 in)	85.60 mm &times; 56.5 mm (3.37 in &times; 2.22 in)
Price	$35 without MPEG2 and VC1 codecs $40.60 with MPEG2 and VC1 codecs	$35 without MPEG2 and VC1 codecs $40.60 with MPEG2 and VC1 codecs

While the ARM Cortex-A7 is one of the slowest A series CPUs, it is miles ahead of ARM11. We did an overview piece on the Cortex A7 when it launched, and stated that it should offer similar performance to the A8 but at significantly lower power requirements. Comparing the A8 to the ARM11 in the iPhone 3GS review, we found the A8 to be about twice as fast. If you consider the Raspberry Pi 2 contains four Cortex-A7s compared to a single ARM11 in the Raspberry Pi 1, the difference is gargantuan. The Raspberry Pi foundation is claiming a speedup of 6x depending on the workload.
CPUs built on the ARMv7 ISA can execute code written against the ARMv6 ISA. Therefore, unless using hand coded assembly specific to the old Raspberry Pi 1 CPU, most existing software should be compatible with the Raspberry Pi 2. Thus, at launch the Raspbian image supplied by the Raspberry Pi foundation actually still uses ARMv6 libraries in userland. Over the next few months they will be evaluating replacing these libraries with standard ARMv7 Debian libraries.
Raspberry Pi 2’s ARMv7 ISA compatibility also opens the doors to operating systems that have never properly worked on the Raspberry Pi before, notably Ubuntu, Android, and Windows. Ubuntu and Android are a somewhat obvious availability, but Windows is a particularly interesting announcement. The Raspberry Pi foundation states it has been working with Microsoft for the past six months to bring Windows 10 to Raspberry Pi 2, and it will be available free of charge “for makers”, which I assume means non-commercial use.
This is a huge announcement by Microsoft, as providing compatibility for arguably the most popular maker device cements their messaging that Windows 10 can run on anything and they won’t be missing out on IoT. The Windows 10 experience on Raspberry Pi 2 will likely be very similar to Windows RT - i.e. no desktop apps. However, Microsoft is doing the work to make sure that universal applications that run on the Surface Pro 3 also run on the Raspberry Pi 2. This is a key part of the Windows for IoT initiative.
While the ability to run Windows 10 is big, it is also important to keep performance expectations in check. In our original Surface review, we noted that the device felt slow and performance benchmarks were low - and that was with much faster 1.3 GHz ARM Cortex-A9 CPUs. Windows software has improved tremendously since then, but there is no getting around the fact that 4x Cortex-A7s at 900 MHz will be a new low for Windows. However, Windows Phone has used such processors before, so perhaps Microsoft's Windows 10 will carry forward enough optimizations to surprise us. Additionally, I imagine a lot of Raspberry Pi 2 Windows 10 devices will be running purpose-built maker applications, hiding task switching or general OS use performance.
Windows 10 on Raspberry Pi 2 also highlights that Microsoft is artificially limiting Windows 10 from being deployed to the original Surface and Surface 2. If Windows 10 can run on the Raspberry Pi 2, there isn’t anything stopping Microsoft from running it on Surface and Surface 2. It is a real shame to see Microsoft leaving Surface adopters in the dust, even after the negative feedback from Windows Phone 7 adopters left in the same position.
This is an official page up discussing the Windows 10 support, but it does not contain any details yet. Microsoft is directing interested users to join their Windows for IoT development program to receive additional details in the future. Personally, I would be interested to try to make a large touch screen surface with just a cheap Raspberry Pi 2. While it may not run the bevy of x86 Windows applications, it is certainly the most touch optimized OS the Pi 2 can run.
The Raspberry Pi 2 is available today from distributors element14 and RS Components. Software distributions are available at the Raspberry Pi Foundation page.

More...

After arriving in certain European countries not long ago, the Android 5.0 Lollipop update for the Galaxy S5 has finally made its way to the United States. Oddly enough, it's the Verizon variant of the Galaxy S5 that is receiving the update before any of the other carriers. The changes listed on Verizon's page for the update are below, and they include both changes made in Lollipop and Samsung specific changes that relate to Touchwiz.

• Material Design
  
  • Android Lollipop brings the all-new material design offering fluid animations, vivid colored themes, and 3-D views for a more pleasurable experience
  
  
• Basic Interaction
  
  • Bolder color and more fluid animations
  • The on/off toggle button style updated based on material design
  
  
• Lock Screen
  
  • The shortcuts for both Phone and Camera are provided for convenience
  • Material design-inspired notifications
  • Ability to set priority and privacy for notifications on lock screen
  • Provide music control as a card to utilize the real estate of the lock screen UI
  
  
• Touchwiz
  
  • Brighter look and feel, bold colors, fluid animation inspired by material design.
  
  
• Floating Action Button
  
  • Floating action buttons provide easy access to the app’s most important features, such as adding a new contact or favorite, or composing a text message.
  
  
• Interruptions
  
  • Interruptions give you more advanced options to control your sound settings
  • Priority notifications such as events and reminders, phone calls, and messages can be set up to sound or vibrate while all other notifications are muted.
  • Set downtime for specific days and times to control notification sounds automatically
  
  
• Recents
  
  • Easy access to your Chrome tabs right from the Recent apps page • Multi-window icons for ease-of-use during multi-tasking
  
  

The Verizon document also has pictures that accompany the changelog so users can see what visual changes are heading their way, and any interested users can view that PDF in the source below. The rollout is staged, and users shouldn't be concerned if their device currently doesn't show any updates available.

It's good to see more devices receiving the update to Lollipop. Enough devices have been updated for it to show up on Google's platform version dashboard. In January less than 0.1% of Android devices were running Lollipop which meant that it was not displayed in the chart or on the graph. Still, at 1.6% of all Android devices, one can only hope that the adoption rate increases so more users can experience the latest version of Google's operating system.
Source: Verizon via Android Central

More...

Back in September, Microsoft launched what would be the last two Nokia branded Lumias. The Lumia 830 was announced as an affordable flagship, and we reviewed the 830 back in November. The Lumia 730 and 735 were launched with a different feature set, a lower price, and a higher resolution front facing camera. Microsoft announced them as “built for Skype calls and selfies” with the inclusion of the five megapixel front facing camera and put the package in a smaller, lighter body. It was launched with four colors, with the 730 being a dual SIM version and the 735 is a single SIM with LTE capabilities.

More...

Today Razer took the wraps off of their updated version of the Blade. We took a look at the 2014 model last year, and found that it was a very impressive device, with one of the best displays we have ever tested, and a powerful CPU/GPU combo that was capable of driving most games at reasonable framerates. It is packed into a CNC aluminum shell that gives a feeling of quality. However, by the time we were able to review the device (our fault, not Razer's), NVIDIA had already released a new set of mobile GPUs that offered even more performance and better power efficiency with the Maxwell based GTX 980M and GTX 970M.
Razer kept all that was great about last year’s model including the amazing IGZO 3200x1800 display, great aluminum body, and powerful cooling capabilities, and updated the internals to make the system even more powerful. The aforementioned GPU has been updated from the GTX 870M to the GM204 based GTX 970M (with 3GB of VRAM – GTX 970 RAM issues do not affect the mobile part; interestingly, the mobile parts are available with 3GB or 6GB VRAM, but Razer opted for the slightly less expensive option). This should provide quite a bit more punch when gaming, especially with the high resolution display.

They have also bumped the CPU up from the Core i7-4702HQ to the Core i7-4720HQ, which bumps the frequency from 2.2-3.2 GHz up to 2.6-3.6 GHz. The 4720HQ model is new to the Intel Haswell stable this quarter. The move from a 37 watt CPU to a 47 watt CPU is likely possible due to the power efficiency of the GM204 GPU and should give a pretty good boost to performance and allow the CPU to stay at its maximum boost for longer. Finally, to round out the performance updates, Razer is now outfitting the Blade with 16GB of DDR3L 1600 MHz memory, which removes one of my complaints about the 2014 model, since a $2500 laptop should not have 8GB of memory.
Razer is also adding a new option to the lineup. A less expensive model is now available, which features a 1920x1080 resolution IPS display, and 8 GB of memory. It keeps the same CPU and GPU combo though, but only a 256 GB SSD is available. This model will also lose the touch capability, and instead feature an anti-glare matte finish, which some people prefer.

Razer Blade 14-Inch Specifications
	2014 (Last Model)		2015 (New Model)
Processor	Intel Core i7-4702HQ (4x2.2GHz + HTT, Turbo to 3.2GHz, 22nm, 6MB L3, 37W)		Intel Core i7-4720HQ (4x2.6GHz + HTT, Turbo to 3.6GHz, 22nm, 6MB L3, 47W)
Chipset	Intel HM87		Intel HM87
Memory	8GB DDR3L-1600		16GB DDR3L-1600
Graphics	NVIDIA GeForce GTX 870M 3GB GDDR5 1344 CUDA cores, 941 MHz core 5 GHz memory clocks 192-bit memory bus Intel HD 4600 Graphics (20 EUs, up to 1.15GHz)		NVIDIA GeForce GTX 970M 3GB GDDR5 1280 CUDA cores, 924 MHz + Boost 5 GHz memory clocks 192-bit memory bus Intel HD 4600 Graphics (20 EUs, up to 1.20GHz)
Display	14" Glossy 16:9 3200x1800 Sharp LQ140Z1JW01 IGZO Multitouch with LED Backlight Optional Anti-Glare Matte 16:9 1920x1080 LED Backlight non-touch
Hard Drive(s)	128/256/512GB SATA M.2 256GB Only on 1080p model
Optical Drive	N/A
Networking	Intel Wireless-AC 7260HMW Dual Band 2x2:2 802.11a/b/g/n/ac Bluetooth 4.0
Audio	Realtek ALC269 HD audio Stereo speakers Combination mic/headphone jack
Battery	150 watt power adapter 70Wh Lithium-Ion Polymer
Front Side	-
Right Side	USB 3.0 HDMI 1.4a Kensington Lock
Left Side	AC adapter 2x USB 3.0 Combination mic/headphone jack
Back Side	-
Operating System	Windows 8.1 64-bit
Dimensions	13.6" x 9.3 " x 0.70" 345mm x 235mm x 17.8mm
Weight	QHD+ Model: 4.47 lbs 2.03 kg 1080p Model: 4.19 lbs 1.9 kg
Extras	2.0 MP Webcam Razer Synapse 2.0 Software 10-point Multitouch Display on QHD+ Model Backlit anti-ghosting keyboard
Warranty	1 year limited
Pricing	$2199 for 128GB QHD+ Model $1999 for 256 GB 1080p Model

This is a pretty solid update to one of the best Windows laptops you can buy. The device already had excellent performance, but the high DPI display would often bring the GTX 870M to its limits. The move to GM204 should have a significant performance improvement. The addition of 16GB of RAM will also allow this laptop up to be more than a gaming laptop, with the ability to run multiple virtual machines or demanding software. It keeps it in a package that is the same dimensions and weight as the outgoing model, which at 4.47 lbs / 2.03 kg was already slim for the amount of power onboard.

Razer is shipping us a new model of the Blade for review, which we will have up shortly. Check back to see how the new system compares soon. For those that cannot wait for the review, the new Razer Blade is on sale now, starting at $2199 for the 128GB/QHD+ model, and $1999 for the FHD model at www.razerzone.com/blade.

More...

Late last week Intel updated its enterprise SATA SSD lineup with the 'Data Center' DC S3610 and DC S3710. At the time of the announcement the details of what exactly had changed were rather scarce, but I just got off of the phone with Intel and have a better understanding of the announcement.
The S3610 and S3710 are based on Intel's second generation SATA 6Gbps controller, which was first introduced in the high capacity (1.2TB & 1.6TB) and M.2 S3500 models a couple of months ago. Intel is keeping the specifics of the controller close to its chest and the only tidbit of info I was able to get out of Intel is that the new controller runs at a higher frequency, although Intel wouldn't disclose the exact speed (for the record, the first generation controller runs at 400MHz in the S3500/S3700). My guess would be that the second generation controller is merely a die shrink with the architecture being very similar, which would allow a higher frequency without impacting the power consumption. I'm also thinking that the internal caches and the DRAM controller have been upgraded to enable the higher capacities.

Intel SSD DC S3710 Specifications
Capacity	200GB	400GB	800GB	1.2TB
Controller	Intel 2nd Generation SATA 6Gbps Controller
NAND	Intel 128Gbit 20nm High Endurance Technology (HET) MLC
Sequential Read	550MB/s	550MB/s	550MB/s	550MB/s
Sequential Write	300MB/s	470MB/s	460MB/s	520MB/s
4KB Random Read	85K IOPS	85K IOPS	85K IOPS	85K IOPS
4KB Random Write	43K IOPS	43K IOPS	39K IOPS	45K IOPS
Read/Write Power Consumption	2.6W / 4.0W	2.7W / 5.4W	3.0W / 5.9W	3.1W / 6.9W
Endurance	3.6PB	8.3PB	16.9PB	24.3PB
MSRP	$309	$619	$1,249	$1,909

The S3710 replaces the popular S3700 and in addition to the controller the NAND has been switched from 64Gbit 25nm HET MLC to 128Gbit 20nm HET MLC (HET stands for High Endurance Technology, which is essentially Intel's marketing term for eMLC). Despite the larger die (and hence slightly slower) NAND, the performance sees an upgrade from 76K random read IOPS to 85K and from 36K random write IOPS to up to 45K. The improvement is mostly due to the new controller, although Intel told me that there is also some NAND-level optimizations in the mix. The endurance remains at the same 10 drive writes per day for five years and as usual to enterprise drives there's full power loss protection along with AES-256 encryption. Pricing comes in at ~$1.50 per gigabyte and the drive is already available.

Intel SSD DC S3610 Specifications
Capacity	200GB	400GB	480GB	800GB	1.2TB	1.6TB
Controller	Intel 2nd Generation SATA 6Gbps Controller
NAND	Intel 128Gbit 20nm High Endurance Technology (HET) MLC
Sequential Read	550MB/s	550MB/s	550MB/s	540MB/s	500MB/s	540MB/s
Sequential Write	230MB/s	400MB/s	440MB/s	520MB/s	500MB/s	500MB/s
4KB Random Read	84K IOPS	84K IOPS	84K IOPS	84K IOPS	84K IOPS	84K IOPS
4KB Random Write	12K IOPS	25K IOPS	28K IOPS	28K IOPS	28K IOPS	27K IOPS
Read/Write Power Consumption	2.6W / 3.3W	2.7W / 4.7W	2.7W / 5.3W	2.9W / 6.3W	3.0W / 6.4W	3.3W / 6.8W
Endurance	1.1PB	3.0PB	3.7PB	5.3PB	8.6PB	10.7PB
MSRP	$209	$419	$509	$839	$1,289	$1,719

The S3610 is a totally new series in between the S35xx and S37xx lineups and its differentiator is the three drive writes per day endurance (whereas the S35xx is rated at 0.35 DWPD and the S37xx at 10 DWPD). Similar to the S3710, the S3610 also uses 128Gbit 20nm HET MLC, but there's a difference in the binning quality of the chips. Basically, the S3710 gets the best of the best dies with the highest endurance, while the S3610 gets the dies that don't make the cut for the S3710 but are still very high quality. There's a difference in over-provisioning levels too as the S3710 features a higher 30-40% over-provisioning with the S3610 having only 10-20% (in both models the exact over-provisioning depends on the capacity). Random write performance is quite a bit lower as well, but the S3610's focus is in mixed rather than very write-intensive workloads where the S3710 is the best fit. Full power loss protection and AES-256 support are also present, similar to the rest of Intel's enterprise SSD lineup.

Comparison of Intel's Enterprise SATA SSDs
	S3500	S3610	S3710
Form Factors	2.5", 1.8" & M.2 2280	2.5" & 1.8"	2.5"
Capacity	Up to 1.6TB	Up to 1.6TB	Up to 1.2TB
NAND	20nm MLC	20nm HET MLC	20nm HET MLC
Endurance	0.35 DWPD	3 DWPD	10 DWPD
Random Read Performance	Up to 75K IOPS	Up to 84K IOPS	Up to 85K IOPS
Random Write Performance	Up to 18.5K IOPS	Up to 28K IOPS	Up to 45K IOPS
Approximate Pricing	~$0.90/GB	~$1.05/GB	~$1.55/GB

I made a quick table comparing Intel's enterprise SATA SSDs after last week's announcement. The S3500 will remain on the market as an entry-level product and it's also the only one that comes in M.2 form factor. As usual, the performance and endurance scale with price, so the S3500 is mainly aimed for read-intensive workloads (e.g. media streaming), whereas the S3610 and S3710 provide the necessary performance and endurance for mixed and write-intensive workloads respectively.
Intel hasn't sampled the S3610 and S3710 yet, but we hope to get samples soon and we'll be able to put the new drives through our own tests.

More...

Today ARM is announcing three brand-new premium IP designs targeted at high-end mobile SoCs. We're still only starting to get widespread commercial availability of ARM's latest generation of SoCs, which includes the Cortex-A57 in big.LITTLE configuration coupled with the A53 as little cores, and the newest T760 Mali GPUs. But, while those designs are still ramping up through offerings from Samsung, Qualcomm, HiSilicon and co. this year, ARM isn't staying still and already looking forward to 2016 and beyond.
Cortex-A72 - a new high end core

At the center of today's announcements a new high-end performance core which succeeds the A57 in flagship devices. ARM was very vague about the architectural characteristics of the new design, disclosing for now only estimates of the chip's performance and power targets. ARM promises a 3.5x sustained performance increase over the A15 generation of SoCs while remaining in the same power budget. One has to note that we're talking about performance targets on TSMC's 16nm FinFET+ node versus previous nodes such as 28 and 20nm, which in itself should bring large clock and power advantages.

The A72 targets roughly 1.9X the sustained performance of current 20nm A57 SoCs, meaning the Exynos 5433 and the Snapdragon 810 can be taken as the base for comparisons. ARM doesn't yet mention peak performance so we may be talking about overall power efficiency gains that enable future SoCs to throttle much less. ARM will be divulging more information on the architecture of the A72 in the coming months, and we're hoping to have a better picture on the actual IPC and efficiency gains of the new flagship core by then.

The Cortex-A72, being a "big" core, can be partnered up with the already existing A53 LITTLE core architectures. ARM has said in the past that the A53 took in-order designs to new heights, and while work on a successor is underway, it seems that for now we'll be sticking with the A53 architecture for a while longer.
HiSilicon, MediaTek and Rockchip are listed among more than then launch partners which have already licensed the Cortex-A72 processor, so expect to see a variety of vendors offering the new ARM IP in 2016.
CoreLink CCI-500 SoC interconnect

It's been over 3 years since ARM initially announced their CCI-400 (Cache Coherent Interconnect), which saw widespread usage as the corner-stone technology enabling big.LITTLE heterogeneous multiprocessing in all consumer SoCs from the Exynos 5410 to the latest Snapdragon 810. While ARM also offered high-end alternatives such as the CCN-5XX (Cache Coherent Network) range of interconnects, these were targeted more at server-applications and not meant for mobile SoCs in smartphones or tablets.

The CCI-500 is a large upgrade over the CCI-400 as it introduces a variety of new functionality over its predecessor. The largest change in functionality is the addition of a snoop filter on the interconnect itself. Until now snoop control was only possible between CPUs within a single cluster. The addition of a snoop filter on the interconnect allows for power efficiency benefits as the amount of transactions when doing cache lookups is decreased, enabling both reduced overhead on the interconnect and also higher idle residency times on the CPU cores. This reduced overhead also frees up memory bandwidth on the interconnect, and ARM claims this enables for 30% better memory performance on the CPU ports.

The new interconnect also doubles up on its system bandwidth: We now have twice the number of ACE (AXI Coherency Extension) ports, enabling usage of a maximum of four CPU clusters (instead of the two that are possible with the CCI-400). We'll be continuing to see the usage of only two clusters in mobile designs, but the new IP gives licensees the flexibility to deviate according to their needs.
The increased bandwidth and numbers of ports on the interconnect also opens up the possibility of quad-channel memory controllers, resulting in 128-bit memory buses. The Snapdragon 805 was the first mobile product to feature such capability, although Qualcomm used a non-cache-coherent interconnect in their design.
Mali T880 GPU

Lastly, ARM also announced a new member of the T800 series of Mali GPUs. In addition to the T820, T830 and T860 comes the T880. ARM was again light on details of what this new configuration brings, only promising a 1.8x increase performance over 2014 Mali T760 GPUs and a 40% reduction in energy consumption for the same workloads.

With today's announcements, ARM appears to be addressing its weaknesses in mobile SoCs by focusing on sustained performance and efficiency of its big core architecture. We also have the much needed upgrade in the memory/interconnect subsystem and an expansion in its GPU IP offering.

More...

Imagination has sent along word that the Creator CI20 we covered in some detail late last year is now shipping to North America and Europe. For those unfamiliar with this development board, I have reproduced the specification table below and included the Raspberry Pi 2 for comparison. Along with the shipping announcement, Imagination has bumped the on board storage from 4GB of NAND up to 8GB. I updated our original coverage with this welcome unexpected improvement. They have also revised the PCB to be purple, matching Imagination's color scheme.

	Imagination Creator CI20	Raspberry Pi 2 Model B	Raspberry Pi Model B+
SoC	Ingenic JZ4780	Broadcom BCM2836	Broadcom BCM2835
CPU	Dual 1.2GHz XBurst (MIPS32 instruction set)	4x 900 MHz ARM Cortex-A7 (ARMv7 instruction set)	700 MHz ARM1176JZF-S (ARM11 family, ARMv6 instruction set)
GPU	PowerVR SGX 540 @ Unknown MHz	Broadcom VideoCore IV @ 250 MHz	Broadcom VideoCore IV @ 250 MHz
Memory	1GB DDR3	1GB DDR2	512MB DDR2
USB	1x USB OTG (A and Mini connectors) 1x USB Host	4x USB 2.0 Host	4 USB 2.0 Host
Camera Interface	24-pin connector (ITU-R BT.645)	15-pin MIPI (CSI) connector	15-pin MIPI (CSI) connector
Video Output	HDMI 1.4a	HDMI 1.3a Composite video 3.5mm jack	HDMI 1.3a Composite video 3.5mm jack
Audio In	3.5mm jack shared with output	via I²S interface	via I²S interface
Audio Output	HDMI 3.5mm jack shared with input	HDMI 3.5mm jack I²S interface	HDMI 3.5mm jack I²S interface
Storage	8GB NAND 1x full SD slot 1x SD slot via expansion header	via MicroSD	via MicroSD
Network	10/100 Ethernet 2.4GHz WiFi b/g/n + BT 4.0 (BCM4330)	10/100 Ethernet	10/100 Ethernet
IR	Input (IRM-2638A)	-	-
Max Power Consumption	4 Watts	< 5 Watts	3 Watts
Power Connector	5V barrel connector	MicroUSB	MicroUSB
Size	Approx. 90mm x 95mm (3.54 in x 3.74 in)	85.60 mm &times; 56.5 mm (3.37 in &times; 2.22 in)	85.60 mm &times; 56.5 mm (3.37 in &times; 2.22 in)
Price	$65 or &pound;50	$35 without MPEG2 and VC1 codecs $40.60 with MEPG2 and VC1 codecs	$35 without MPEG2 and VC1 codecs $40.60 with MEPG2 and VC1 codecs

The evaluation board provided to me by Imagination is green with 4GB of NAND and I should consider it a collectors' item now. I am working to get some power measurements directly on the VDDCORE line to characterize CPU performance/watt but that requires some delicate PCB modification. Stay tuned.
Also, for the media center users, great news came three days ago with MIPS32 support merged into the master branch of Kodi (XBMC). Therefore, users willing to compile binaries themselves can now run Kodi on the Creator CI20 or any other MIPS32 device. For those willing to wait, the next major release of Kodi should include this support in their provided binaries.

More...

For the past few years, we’ve seen Google place significant emphasis on price as a way of competing with other tablets on the market. The original Nexus 7 managed to deliver a good tablet experience without the conventional 500 USD price for a tablet. The successor to the Nexus 7 was even more incredible, as it pushed hardware that was equal to or better than most tablets on the market at a lower price. However, as with most of these low cost Nexus devices not everything was perfect as corners still had to be cut in order to hit these low price points.
The Nexus 9 is supposed to be the polar opposite. Instead of driving price as the primary differentiator, Google has refocused on the high end tablet market for the Nexus 9. With a new focus on industrial and material design, along with some of the latest and greatest hardware in every dimension. HTC has been brought on as a partner for the first time since the Nexus One to enable this vision. In addition, NVIDIA’s Tegra K1 with Denver CPUs can be found inside as the launch platform for Android Lollipop on 64-bit ARM v8. The Nexus 9 also has a 4:3 aspect ratio on its display, a notable departure from the 16:10 ratio that was shared with phones. There’s also the addition of BoomSound speakers on the front and a metal frame running around the edge of the device for improved in-hand feel. To find out how this device stacks up against the competition, read on for the full review.

More...

As expected (and reported from various news outlets), today FCC Chairman Tom Wheeler has proposed using Title II legislation to help settle the question of net neutrality. The debate over net neutrality – the idea that all packets traveling over the Internet routers, servers, etc. should receive equal treatment, which is how things have generally worked up to now – has been going on for more than a decade, with two primary views. On the one side are those who want this continue, primarily consumer interests along with companies including Google, Amazon, Facebook, and Netflix; on the other side are major Internet Service Providers like Comcast and AT&T.
Proponents argue that the Internet has flourished thanks to net neutrality and an attitude of being free and open. They also point out that these are the rules that have more or less been followed so far, and they have worked well, so the regulation is more of a safeguard to keep the current norm of equal treatment of data in place. Opponents meanwhile have expressed concerns that regulation would hinder their ability and motive to roll out faster broadband service, and others are worried that regulation may open the way to increased taxation and other costs. When President Obama weighed in last year and recommended that the FCC implement “the strongest possible rules” to protect the Internet, today’s eventual recommendation by the FCC was more or less a foregone conclusion.
Wheeler states in his Wired piece, “I am proposing that the FCC use its Title II authority to implement and enforce open internet protections…. I am submitting to my colleagues the strongest open internet protections ever proposed by the FCC. These enforceable, bright-line rules will ban paid prioritization, and the blocking and throttling of lawful content and services. I propose to fully apply—for the first time ever—those bright-line rules to mobile broadband. My proposal assures the rights of internet users to go where they want, when they want, and the rights of innovators to introduce new products without asking anyone’s permission.” The application of the same rules to mobile broadband is particularly important, as according to the FCC around 55% of online traffic now occurs on smartphones and tablets. While Title II regulation has been applied to the cellphone industry since 1993, mobile data services have up until now been exempt.
As for how Title II will be applied, Wheeler proposes modernizing Title II and tailoring it for the 21st century, at the same time continuing to encourage investment into broadband networks via incentives. The goal is to create a system where the necessary returns continue to exist in order for large communications corporations to construct competitive networks. “For example, there will be no rate regulation, no tariffs, no last-mile unbundling.” Elsewhere Wheeler mentions using a “light touch approach” similar to what has been done in the cellphone industry.
What does this all mean to end users? It’s important to note that net neutrality and Title II regulation alone don’t mean faster speeds for lower prices. I for one would love to see more affordable pricing on broadband, particularly for mobile broadband, but that may or may not happen; Title II for now is about making things “fair” for all parties. What this really means is that a company like Comcast, as an example, cannot prioritize their own content over that of competitors – so on demand streaming of the latest movie from Comcast has the same access to bandwidth as streaming of content from Netflix, Hulu, Amazon, etc. It’s also important that net neutrality remain in place not just on the connections to our homes and businesses, but also on the Internet backbone.
There have been examples of ISPs limiting traffic in the past, so it will be interesting to see how this plays out. BitTorrent trafic was often limited to end users, and this may not change if the traffic is not deemed as "legal content" but it would be hard for an ISP to declare that all traffic of a single protocol is not legal, even if some of it is not. Famous examples of legal BitTorent traffic would be things like Linux distribution networks and World of Warcraft, just to name some examples. Netflix traffic has also been degraded in the past due to the lack of sufficent interconnect bandwidth, with Netflix agreeing to pay last mile ISPs to improve service.
As this is all at the proposal stage right now, we’ll need to wait to see what some of this means in actual practice – and voting will take place on Feb. 26.

More...