Quicky IT news

Some news are worth mention, yet not worth entire thread, much less article covering… so that is the purpose of this thread.

And let's start with todays NDA end on new AMD APU's AMD Beema (GPU 800MHz, CPU 2.4GHz) and Mullins (GPU 500Mhz, CPU 2.2GHz) we can tell some informations. Previous APU (that is AMD call a CPU with GPU in one package) Kabini (GPU 600MHz, CPU 2.0GHz) and Temash (GPU 400MHz, CPU 1.4GHz) are quite outperformed by these new offerings.

The GPU part is basically same as Kabini - GCN architecture with 128 stream processors, 8 texture units and 4 ROPs. However with better design (38% less leakage = less heat and even less power required) and higher clocks is the performance higher and not only by the mere clocks bump. The memory interface (on witch GPU depend heavily) was reworked for the DDR3L standard. It go up to DDR3L-1866 clocks, witch should give some boost at well. (AMD says 5% on 3DMark CloudGate over DDR3-1600)

DDR3L is less power hungry, because it use 1.35V, as oposed to 1.5V to 1.6V used for most DDR3 rams (and hence the ram controller itself).

All in all, 15W APU Beema is in graphic about ~10% faster that 25W APU Kabini and also ~50% faster that 15W Haswell Pentium 3556U 😉 Intel is not going to be happy about this development.

As the CPU part go, the clocks also recieved a healty bump.(mainly the slower "tablet" CPU get raise from 1.4GHz to 2.2GHz, witch is quite a lot). The core architecture is now codenamed Puma+.  19% less leakage. The new "turbo" mode is more aggressive, so now the STAPM technology allow the CPU run at turbo (when need) before the 40°C limit of outside case temperature is reached, and only then the throttling come in action.
In short it means more power when need and less waste of power for powering apps, that did not benefit from turbo.

And the probably most interesting change is that these new APUs containing a whole ARM Cortex-A5 core (Trusted Platform Module, TPM) with own Cryptographic coprocessor.  AMD call this together as PSP (Platform Security Processor) and NSA must be really "happy" about this development.
(time to push for backdoors, you $#%!)

Now only if developers support this, since that is the main question, if this move will prove as valid in long run.

There is also slightly more logical (for customer) labeling of the GPU cores, atc. But that it out of the quick news scope 🙂

Since AMD gaming bundle Never Settle Forever was enhanced to 20+ games since 4.21.2014 for all the R7 240, 250, 260, 260X and R9 270, 270X, 280, 280X, 290, 290X and 295 X5 cards:

…it was "about damn time" nVidia enhance their offers as well. And it did. So if you choose one from these graphic cards - GTX 660, 660ti, 760, 770, 780, 780ti and Titan (surprisingly no word about Titan Z, just like the Asus announced yet not delivered yet non-reference Titan Z, witch was probably delayed, because it did not outperformed the R9 295 X2: http://videocardz.com/50387/asus-announces-geforce-gtx-titan-z ) and you bought it from some of the nVidia partner shops, then you get the game Watch Dogs 🙂

Well, it is not the Never Settle Forever pack by long shot, but the techdemo looks nice at least:
…and sure it have the appeal and added value 🙂

Samsung new enterprise class SSDs with 3bit MLC NAND chips

New Samsung SSD line are marked as PM853T and it offer capacity of 240, 480 and 960GB and it is basically a enterprise class extension over the last year introduced consumer class SSD with 3bit MLC NAND chips. Presumably Samsung made their working specifications good enought for enterprise segment.
We, of course, cannot expect from these SSDs the speeds of much more expensive SLC types, however Samsung expect that these 3bit SSDs will graduately replace the currently used 2bit types.
Samsung in kinda secretive about the used controller, speaking only about that it is an "advanced type." These MLC NAND chips are made using "10nm-class" technology (gotta love these marketing BS) and in reality these SSDs can do 530MB/s reading and 420MB/s writting. IOPS are good 90 000 for random read, while random writting should hold steady at 14 000 IOPS.
So since the first 3bit chips that Samsung come with the 840 EVO line back in 2012, these 3bit chips actually undergo some generations and we see in few years, how the longetivity will look. Simply spoken - adding another bit into the memory cell shrink the size of the manufacturing process and that usually is not good for longetivity. Still it looks to me, that old good SLC chips (preferably made with some old (read - bigger) manufacturing process) will be best in terms of longetivity.

$1500 Google Glass manufacturing cost $80

Teardown server take on the almost exactly year ago released "beta" testing of Google Glass and after the diassembly it searched for cost of used components and discoveded, that they are mere $80. That leave about $1400 sale margin for sellers, witch looks as pretty good for any company to make profit on.
The fact, that the price stays the same over whole year, lead Teardown guys to take on this "biggest intrusion in privacy yet unleashed on mankind thing" to discover, how costy these are to build. Sure some price need to be packaging and manual, but into the nearly $80 dollars price are already calculated even the assembly and testing ($2 per piece).
The distribution model Google choosen is clearly very profitable. It also minimize the risks involved in marketing new products. To put long story short, the exclusivity price already covered the developing costs and therefore when the final time to release the product happen, then Google will be only making more profit, as this "betatesting" is already pretty nicely profitable:

IMHO this product is terrible idea and i think it's true cost is really only one swift punch in the right direction. I value my privacy more that the price of this thing is, but I could be alone with my negative view in times, when people willingly post many revealing photos of themselves (nicely tagged for the NSA: http://www.youtube.com/watch?v=JbQeABIoO6A ) on Facebook already.
Yet still I would fiercely oppose the idea that sticking WiFi transmitter permanently so close to human living brain ( http://www.youtube.com/watch?v=cczGVhd63pM ). IMHO the people that stick WiFi transmiter on the skin of their heads will sooner or later end up with brain tumors.

Of course you can choose to believe that WiFi is safe. The presstitutes tell you anything. For example even that mercury is good for your children: http://www.youtube.com/watch?v=20IhxCzeuDU (never mind the crazy conspiration theoretics that babble something about that mercury is poisonous, they are insane)

nVidia GeForce GTX Titan Z was put on hold for not specified amount of time

At 4.29.2014 it was planed to be released, but it was put on hold. At GDC 2014 we can see the dual GPU Titan card in the hands of nVidia CEO Jen-Hsun Huang, but that was it for now. It was planed to be there slots card, Asus was ready to produce no reference version of Titan Z, but all the hopes are vanished.
And nVidia did not set any new date yet. So looks like that either the card cannot outperform the R9 295X AMD offering, or it is at least not a clear, distinctive winner, so nVidia decided to wait and do some tweaking? Or perhaps technical issues? Dunno what is the cause, but we have to wait for it.

That somewhat suxx.

If we take the 4K resolution, then dual GTX 780ti cards that run at 837/876MHz, is 6% slower that R9 295 X2:
Titan Z got 16% slower clocks compared to single GPU cards and TDP 25% lower. In such situation is expectable, that the difference in speed between R9 295X2 and Titan Z will be about 10 to 20% slower for the Titan Z, witch is almost 2x more pricey. That is probably not something nVidia want to be clear, because that means clear and vera painfull defeat.

Update: Now they claim that Titan Z will be released before July. So quite soon. People that bad mouth nVidia claim that this announcement has a flaw. They did not say before July of witch year. they plan the release.. 🙂

Secong batch of the Maxwell chips (first generation are these GTX 750 ones - GM107 chip) will feature hardware h.265 decoding

According to the Fudzilla, the second wave of the Maxwell chips (maybe even 20nm process will be used?) will feature full hardware support for h.265 decoding, so even 4K videos will be played smoothly w/o big requirments to the CPU. While it is not clear, that the hardware capabilities end with h.265 in 4K, or it will support 8K as well (4x more pixels compared to 4K!), it does not matter much with current display connectivity options.
Because for DisplayPort 1.2 is 4K in 60Hz top resolution, while HDMI 1.4a can do 3840x2160 at only 30Hz.
For support of 4K resolution in 60Hz is planed HDMI 2.0, but more (think 8K) will require support of DisplayPort 2.0, witch use even compression to get the data thru… but support of neither was announced yet for the new wave of Maxwells. And it have to be done...
4K w/o HDMI 2.0 will be suxxking badly (30Hz? Come on!). Also there are rumors about the VP9 compression support, but that is not confirmed and basically spoken, the h.265 and VP9 did not differ that much.

TSMC is extending lead time for 28nm chips notably

The Taiwan Semiconductor Manufacturing Company (TSMC) is well known producer of many chips and now their lead time increased from 10 to 16 weeks for the 28nm process. It seems uninteresting, untill you realize that the customers of TSMC are for example Apple, Samsung Electronics, Qualcomm, Broadcom, MediaTek, STMicroelectronics, Infineon Technologies, Renesas Electronics, Texas Instruments and many many others.

In short that means, that the May orders will not be shipped untill September, maybe even in October. And that could be a huge problem for the marketing and sales.

There are other chip makers, that can do 28nm, such as UMC, GlobalFoundries and SMIC, but they are mostly as overloaded, as TSMC (for example Samsung is making the 28nm chips full time for Apple), or have not reached usable yields. Not to mention that changing the chip maker require finetuning the chip design for the specifical company - and that is no small task.

In short - this is likely to affect launch schedules or evenprices of devices.

To put that information in perspective - as we appraching the limits of our current chip making technology, the problems are starting to get bigger and bigger. It is not hard to see, that at some point around 10nm we might hit a wall of what is possible - unless there will be some major technological break-thru.

Microsoft actively surpressing using Windows Server 2003 x64 patches to work on Windows XP x64 edition!

One thing is "not supporting" some systems anymore. Another is to activelly work to make the patches not working on nearly absolutely the same system. The first is questionable (because the bugses are not what users payed for and shall be fixed), the second M$ way is downright wrong.

First things first - as the Windows XP x64 are just a client variant of the Windows Server 2003 x64, then the same are Windows Vista - just a client variant of Windows Server 2008.
(Windows XP 64-bit Edition is just for Intel Itanium CPUs, even with 32bit version, but that is not relevant as both are ended with the end of Intel Itanium workstations sales)

13.5.2014 there come the patch KB2962488 (Security Bulletin MS14-027) and since it is for Windows Server 2003 x64, it should be working on Windows XP x64. However it does not work, as in the inf file, there is this:


…and the Install Block containing this:


Funny thing is, that in the very same inf file, there are sections that explicitly do support the Windows XP x64:

SP_TITLE = "Security Update for Windows x64 (KB2926765)"
SP_TITLE_XP = "Security Update for Windows XP (KB2926765)"
SP_TITLE_WS03 = "Security Update for Windows Server 2003 (KB2926765)"

Therefore it is fixable and the patch can be used on Windows XP x64. As anyone can see there:
…but the moral part of the story is somewhat disturbing. This is not "non support", but "active denial of support" and that is quite different thing. Anyone should really consider from now on, if their bucks are worth to be sendt to company, that is actively working on blocking the usage of patches for their bugs in the first place.

50 gametitles are going to end - GameSpy is ending by 31 May 2014

Electronic Arts game titles depend mostly on the GameSpy for their online matchmaking service. Several gametitles are still quite popular among fans, but EA offer no solution. They admit, that people still play a lot for example Battlefield 2, Battlefield 1942, Medal of Honor: Allied Assault or the seriers Command & Conquer, but that is going to end soon.
Some possible solutions are emerging, like the GameRanger client: http://www.gameranger.com/download/ but some games are still quite problematic. Among with the most wanted and still missing are Battlefield 2 with all the mods and the there Crysis titles…

The list of games that will end their network playability on the end of May:

Battlefield 1942
Battlefield 2 (for PC and PS2)
Battlefield 2142 (for PC and Mac)
Battlefield Vietnam
Bejeweled 2 (for Wii)
Bulletstorm (for PS3)
Command & Conquer 3 (for PC and Mac)
Command & Conquer (for PC and Mac)
Command & Conquer: Red Alert 3 (for PC and Mac)
Crysis 2  (for PC)
Crysis  (for PC)
Crysis Wars  (for PC)
Dracula: Undead Awakening (for Wii)
Dragon Sakura (for Nintendo DS)
EA Sports 06  (for PC)
F1 2002 (for PC)
FIFA Soccer 08 (for Wii)
FIFA Soccer 08 (for Nintendo DS)
FIFA Soccer 09 (for Nintendo DS)
FIFA Soccer 10 (for Nintendo DS)
FIFA Street 3 (for Nintendo DS)
Full Spectrum Warrior: Ten Hammers (for PS2)
Global Operations (for PC)
Green Day: Rock Band (for Wii)
James Bond: Nightfire (for PC)
Madden NFL 08 (for Nintendo DS)
Madden NFL 09 (for Nintendo DS)
Master of Orion III (for PC)
Medal of Honor: Allied Assault (for PC and Mac)
MySims Party (for Wii)
MySims Racing (for Nintendo DS)
MySims SkyHeroes (for Wii and DS)
NASCAR Sim Racing (for PC)
NASCAR Thunder 2003 (for PC)
NASCAR Thunder 2004 (for PC)
Need for Speed: Hot Pursuit 2 (for PC)
Need for Speed: ProStreet (for Nintendo DS)
Need for Speed: Undercover (for Nintendo DS)
Neverwinter Nights 2 (for PC and Mac)
Neverwinter Nights (for PC, Mac and Linux)
SimCity Creator (for Wii)
Skate It (for Nintendo DS)
Sneezies (for Wii)
Spore Creatures (for Nintendo DS)
Spore Hero Arena (for Nintendo DS)
Star Wars: Battlefront (for PC and PS2)
Star Wars: Battlefront 2 (for PC and PS2)
The Beatles: Rock Band (for the Wii)
Tiger Woods PGA Tour 08 (for Nintendo DS)

Of course GameSpy is not only affecting EA, many other games will be affected as well. Even games like Arma 3 or Operation Arrowhead will need to be changed to work again… Not really a good news.

Sad day for computer world - Mozilla implement DRM, so users can watch Netflix with it

To put long story short, the DRM crap is back. And the online streaming media provides insist on enforcing this crap on us, customers. I would say boycot them, but Mozilla had choice. Stay open source and lose customer base, because to implement the HTML5 video fully, they pushed the DRM in it. And the other browsers already implemented it, so that is why.
The reason for the damn DRM to get back is, that online streaming services, such as BBC, Netflix, Amazon Video, Hulu and others, need to protect the content, as they are bound by the content supliers, that the movies will not end to be for everyone, but just for the customers of these services.
Therefore Mozilla had no choice (except to say - we are out of this) and decided at least to implement the custom code in the safest way possible. Google and Microsoft teamed and produces W3C EME and Mozilla go for the Adobe version of the closed CDM code. And - unlike Explorer and Chrome - it run it in own sandbox and the fingerprint of the device used will generate for it, not allowing the closed code to grab many, possibly even private, data. That way Mozilla at least stays in control about how much data are given to the DRM check. These informations can and will be verified by the provides, and the closed CDM code will come externally for Mozilla.

The Mozilla hopes, that once these DRM implementations as W3C EME will vanish in the history, but that depends on how the customers will buy into all this privacy invasion.

IMHO all protections will be sooner or later broken and become only useless obstacles for paying users. That day come and that will mean, that the customers will by paying for all the hassles w/o any reason. At it depends on how people accept this. If there will be significant opposition, then the DRM will go away. If not, then it stays with us. Like I stated above - not a good day for computers and humanity. DRM are back.

Google Glass will get competition - Samsung is preparing own version of this product

Rumors are, that they will be called Gear Glass (no relevance to the illustrative only picture), but facts says, that they will be available in September 2014. The difference between Google Glass will be minimal. Only they will use in-ear headphones, of course Samsung own operating system Tizen and they will be actually released.
Not like Google "betatesting" their product on premium paying customers, this will can buy anyone. Also the price is expected much lower that the whooping $1500 for the exclusivity of Google Glass.

All in all, I did not see that as any good development, because I would like to not be videorecorded everywhere. The fact, that this spying glases can be manufactured under $100 I see as warning sign. Not even George Orwell can invison society, where governments will be able to turn most humans on the planes into their own cameras. And do believe, that NSA sure force Google to have backdoor into these glases. If past experience with them (mainly I talking about NSA being able to use cameras and microphones w/o user consent and even being alarmed by the activity diode getting even on) is any indication, then they will already have a way to acces the raw feeds… not to mention sticking WiFi transmiter near one brain is IMHO planed suicide.
See you WiFi guys in 20 years, then we talk.

Do you want graphic card with 6GBy of videoram orther that nVidia Titan? eVGA GTX 780 is the answer!

Currently, the top dogs in nVidia stable end up with 3GBy of videoram (GTX 780/780ti). There are some AMD cards with more videoram, tough:
Radeon R9 290 / 290X - 4GBy
GeForce GTX 770 - 4GBy (however this is slower card too)
Radeon R9 280X - 6GBy (also notably slower card that GTX 780/780ti)
GeForce GTX Titan / Titan Black - 6GBy (quite pricey for +3GBy of more videoram)

But then eVGA come and feature 6GBy on classic 780 chip, superclocked to 967/1020MHz (reference 780 clocks are 863/900MHz) with ram running (sadly) at reference 6GHz clock.

As at is now, NewEgg  offer 3GB GTX780 SuperClocked ACX for 519$ and for the 6GBy model it ask 599$, so that is still way less expensive that 999$ Titan for those, that are in reall need of 6GBy of videoram.

How much faster is GPU integrated on 73W i5 Clarkdale over 4W Bay Trail Atom?

IGP Comparison - Bay Trail-T (Z3740) vs. Clarkdale (i5-660) – 03:59
— Brue Computing

Quite surprisingly not much. The GPU's are some generations apart - so let's compare Atom Z3740 Bay Trail-T to Core i5-660 from the Clarkdale generation. The specs ans follows:

Intel Core i5-660 - Clarkdale 32/45nm (GPU) - 2 cores, 4 threads - 3.33/3.6GHz - 12 DX10 GPU blocks - 733MHz - 73W TDP
Atom Z3740 - Bay Trail 22nm - 4 cores, 4 threads - 1.33/1.86GHz - 4 DX11 GPU blocks - 311/667MHz - 4W TDP

And surprisinly, both Atom and Core i5 get around the same 7-11fps in most scanes of Unigine Heaven. Yes, there are some cases where Core i5 is 1fps faster, but also few cases where the Atom is 1fps faster! Giving that it have only 1/3 of the GPU processing blocks and lower clock and very restricted TDP, then it is absolutelly stunning result.
Of course the CPU power of the Atom is quite much lower (sometimes the Atom is just 1/3 of the Core i5), but the GPU's are comparable. So comparing a 45nm and 22nm technology and some generationg of development, the difference is like night and day. So soon we see, how the 28nm APU AMD Carrizo go against the 14nm Intel Broadwell.

14nm Intel Broadwell you shoudl be able to buy before Xmass 2014

Or at least that is what Intel chief Brian Krzanich claims. He claim, that the CPU's will be there in time, so that the companies can make their products ready for the Xmass selling frenzy. That imply the Broadwell must be available by around half of October to December at least.
September launch is not impossible, by Brian, but hard to reach. So probably nope.
Still, Broadwell should be there long time ago, yet since Intel have no reall threat to push him around in the high-end CPUs, then Intel got a bit lazy. Recent TOCK to Haswell (Refresh) was not at all impressive.

So let's hope the Broadwell, as the first 14nm x86 CPU, will bring not only new ADXa instruction and better power saving. Still, the "fanless" on the NTB segment means with SSD also dead-silent and that is amazing!

YouTube quietly moved from Adobe Flash to HTML5 video play

If anyone check their browser capability there: https://www.youtube.com/html5
It is becomming clear, that Chrome do support it all, while other browsers seriously lag behind:

Chrome 32.0.1700.107 m

Firefox v10.0.10 ESR

Opera 11.64

To put long story short, the Chrome in Win, as well, as in Linux does now support HTML5 video, H.264, VP9 and other codecs need for HTML5 video playback. So the crappy lame flash time (with countless of security holes) time are near it's end. And Firefox is not much ready for that.
If you start questioning why Firefox v10 was used in "news", then I have bad news for you - in latest Firefox are the HTML5 video support almost exactly the same as in v10 (v10 just lack H.264 in Win, as latest Firefox in Linux):

In Linux, the latest Firefox even lack the H.264 support…!

So web is moving towards HTML5. If you did not like that, there is a Chrome plugin to disable the YouTube HTML5 player: https://chrome.google.com/webstore/detail/disable-youtube-html5-pla/enmofgaijnbjpblfljopnpdogpldapoc