JustForex
Loading recent posts...

Mar 26, 2012

Linux Kernel 3.2.13 Is Available for Download




Greg Kroah-Hartman announced a few days ago, on March 23rd, the immediate availability for download of the thirteen maintenance release of the stable Linux kernel 3.2 series.  
  
Linux kernel 3.2.13 includes a few important fixes for the AFS, Perf tools and PowerPC, as well as networking IPv6/IPv4, iwl3945 and nilfs2 fixes. 
  
"I'm announcing the release of the 3.2.13 kernel. All users of the 3.2 kernel series must upgrade." 
  
“The updated 3.2.y git tree can be found at: git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git linux-3.2.y and can be browsed at the normal kernel.org git web browser: http://git.kernel.org/?p=linux/kernel/git/stable/linux-stable.git;a=summary" - said Greg KH in the email announcement
  
Just like previous releases, Linux kernel 3.2.13 is a must-grab update for all users of the Linux 3.2 kernel stable series. 
  


Samsung Galaxy Tab 11.6 Pictured. Maybe




There have been murmurs about a new tablet for the Samsung Galaxy Tab 11.6 series, murmurs that picked up in volume not too long ago. 

Granted, it still isn't known if the alleged Galaxy Tab 11.6 will ever actually become reality. 

Still, people are watchful of what they consider important, and this applies here as well. 

Long story short, a photo of an unreleased tablet has made its way to the web somehow. 

In looks, it resembles the Galaxy Tab 2, the one showcased at the Mobile World Congress. 

The photo actually first showed up on the product page of the Exynos 5250, which is the chip expected to power the galaxy Tab 11.6, hence the assumption that the one in the photo is it. We'll keep an eye out for anything new.


17W TDP Sandy Bridge Based Celeron Coming in Summer




Intel plans to release two new mobile CPUs in Q3 2012, the first one being named Celeron 887 Mobile, with a 1400MHz working frequency, and the second one being a 100MHz "faster" Celeron 807.

It comprises two cores with no HyperThreading involved along with 2 MB or cache memory. As expected, it has Intel HD Graphics built-in and it supports DDR 1333 MHz memory. It can clock its iGPU from 350MHz to 1Ghz  according to the task it is working on.

The CPU is part of Intel’s Ultra-Low voltage family and it will end up inside entry level mobile solutions.

The other mobile CPU in Intel’s future Ultra-Low voltage line is the Celeron 807. Regardless of the numbering scheme that might lead you to believe this CPU is just a lower-clocked Celeron 877 or even cache-deprived, this one is seriously handicapped as it has only one core working at 1500MHz.

It’s certainly 100MHz faster and even supports HyperThreading, but lacking the second core, this one will seriously be underperforming when compared with its Celeron 877 brother. The cache has also been cut to 1.5 MB. The CPU comes with no turbo support although it can alter its iGPU frequency between 350MHz and 950MHz.

Intel claims that the 877 model is 47 percent faster than its 807 little brother and we definitely believe them as half the cores and minus 25% of the cache in 807 cannot be compensated by the 7% increase in working frequency.


Intel's Valleyview Atom SoC to Have Multiple Versions




We learned about the Valleyview system-on-chip devices from Intel a few days ago, but the information was sparse, to say the least. 

VR-Zone managed to provide some actual details on the product series, starting with the correct spelling of the brand name. 

At least, we guess it is the correct spelling. Initially, the SoCs were called Valley View but now the report writes the name in a single word: Valleyview. 

Then again, this is not all that important in the end, since the chips will still be called Atom. 

Valleyview (really Valleyview 2) is actually the name of the Atom core, not the SoC itself. The latter is named Bay Trail in fact. 

There will be two versions of Bay Trail, neither of which is intended for consumers. 

There might eventually be a third, user-friendly one somewhere down the line (Atom N or Atom D), but that is pure speculation now. 

One of the planned SoCs is optimized for in-car infotainment systems, while the other is for standard desktop-style applications (and we don't mean standard PCs here). 

Two or four cores will be available, all of them created with the 22nm manufacturing process technology. 

The performance will be about four to seven times superior to the Atom E600, which wasn't really much to be honest. The PowerVR SGX 535 core only reached 400 MHz after all. 

Still, it's a leap and it also bears noting that many chipset features have been integrated into the Valleyview as well. 

Intel will even develop some versions of the SKU with the MIPI-DSI display interface and the MIPI-CSI camera interface. 

This all boils down to a healthy spec list: four PCI Express 1.0 lanes, two SATA 3.0 Gbps ports, SDIO support, four USB 2.0 ports, two USB 3.0 connectors, Gigabit Ethernet, DisplayPort, eDP and HDMI. 

Finally, though most SKUs will handle DDR3L memory (up to 8 GB) and even ECC (error-correcting code, only in single-channel), the ones for battery-based electronics will only support LPDDR2. Once Valleyview start shipping, the Santa Clara company may decide to phase out its Celeron series.


The Battle for 14nm Manufacturing Begins




The first publicity blow of showing a series 14nm silicon wafer belonged to IBM this year. SemiAccurate’s Charlie is reporting that, in the second week of March, the company has showed one of their first 14nm silicon wafers containing IC dies. 

The piece is certainly a test wafer and the chips on it have not been identified, but IBM has something to show for while Intel’s FinFETs are still a while away from manufacturing.

The VP of IBM’s Semiconductor Research & Development Center, Gary Patton said, at the Common Platform Technology Forum two weeks ago, that the company would use SOI technology for all of its 14nm products, including its own server products as well as for their Fab clients.

Samsung also showed a 14nm wafer last week. Unlike IBM’s new tech, Samsung’s is High-k FinFET based just like Intel’s will be.

The Korean giant has also shown a 20nm wafer, a whole year after IBM first came up with such a technology. It’s not FinFET but it is here now.

Intel promises it will start making 14nm chips in 2013 but, as expected, they won’t have anything to do with any x86 design as more fine tuning is needed working on simpler designs before moving a huge 3 billion transistor design on a new manufacturing process.

Intel’s first tape out might be a 4G LTE design that should ensure a low power envelope for a future smartphone chip that might go along with an x86 design to help make x86 a worthy competitor with its ARM counterparts.

Next to the table will probably be Toshiba, with a new process miniaturization for new flash cells, but there’s no word on that yet.


Nokia Belle FP2 Disappoints in HTML5 Benchmark, Tizen OS Outshines All Other Platforms




Nokia announced it would continue to support its Symbian operating system, especially after the Finnish company launched several smartphones that are meant to strengthen its proprietary OS market share.

Although the company decided to outsource the development of Symbian software and support activities to Accenture, Nokia included the transfer of 2,800 employees in the deal. Nokia employees transferred are to focus solely on developing and supporting Symbian software.
 
In the meantime, the Finnish handset maker decided to change the new platform’s name to Nokia Belle. The next-generation operating system was launched on the market along with several new smartphones, such as Nokia 603, 700 and 701.

Nokia also confirmed that these smartphones would receive a Belle FP1 update sometime in the summer. In this regard, it is also worth mentioning that Nokia 808 PureView will be shipped with Belle FP1 out of the box.

However, it appears that Nokia is already testing the Belle FP2 update, though it’s not clear what new features it will bring. We already know that Nokia Belle FP1 will improve Nokia 603, 700 and 701 CPUs speed.

A recently leaked benchmark for HTML5 shows disappointing results for Nokia Belle FP2 platform tested on a Nokia 701. The list of mobile platforms tested, Tizen OS is in the lead with 387 points followed by Blackberry 10 with 361 points. 

Unfortunately, Nokia Belle FP2 is last with only 242 points. Chrome Beta is third with 343 points, Firefox Mobile on fourth with 318, and on the fifth, outperforming Belle FP2, is Windows Phone 8 with 298 points.

Keep in mind that this HTML5 benchmark only includes results from mobile platforms that are currently in development or beta. This means that they are likely to receive further enhancements, which will improve their benchmark scores. Stay tuned for more updates on the matter.



Apple, Bring Back "Save As" in OS X Mountain Lion




Apple has a mission with OS X Mountain Lion, whether it likes it or not. It has to bring back some of the features that made people fall in love with the Mac. And that also includes one fundamental function in computing called “Save As.”

We published a feature piece on OS X Lion last year highlighting a surprising change in the operating system – the removal of "Save As" from a number of first-party applications, including Preview, and TextEdit.

As it was expected, Apple fans quickly took to the comments to explain how this omission thwarted their workflow.

Everyone had an example of how they used Preview to quickly save a photo with a different name, or how they employed TextEdit to quickly edit documents, then save to a specific location, with a specific name.

In OS X Lion, then the just-released Mac OS, they could no longer do this. Worst of all, Apple never gave anyone any warning that Save As would be removed. The upgrade to the new OS was (and still is) $29.99.

One of the last comments we received should really strike a chord in Cupertino. And not in a pleasant manner either.

reader Namban writes…:

Finally moved to Lion last night.

A Mac Lombard was my first computer and I have owned nothing but Macs since and have advised many, many people to switch to them. Today, I apologized publicly for that.

Mac was a wonderful thing for many years, but that was when "think different" meant something. Now, Mac stands for, "don't think- we'll do it for you". I work with images and Lion has completely destroyed my workflow. No more "save as", really?

So far the replacement, "save a version" has destroyed an original photo and caused what could be done with a click to need several files and many clicks to accomplish.

Finder no longer tells me how many files are selected, or how much space is available on the drive. That is important to me. So far, I've run into a bunch of stuff like that, little irritants with the potential to do real damage, like losing files. Sure, I can learn new workflows, but the point is Apple has veered off in a direction I don't like and won't get better.

If I have to learn new stuff, might as well do it going in a positive direction. Blending Mac OS with iOS will just result in a sugary little toy that looks cute in a coffee shop, but is incapable of doing any actual work. It's been a fun ride, Mac, but the bus is filling up with Bozos and it's time to hop off. I upgraded to Lion because I was about to buy an iPad, but those bucks will go to a new laptop running a Linux distro instead.

While Namban's approach may seem rushed, there are numerous Mac users out there who feel the same. Many of them can be found in the same thread in which Namban wrote his own thoughts on the matter.

With the upcoming arrival of OS X 10.8 Mountain Lion this summer, Apple has a chance to correct this mistake by re-including Save As in the File options.

Mountain Lion is shaping up to become a truly amazing operating system by stitching core features from iOS with the best of desktop computing.

It would be great to see “Save As” re-enabled in the next developer preview of OS X 10.8. This way Apple would also ensure that customers like Namban don’t jump ship before Mountain Lion is released to the public.

A screenshot taken in OS X 10.8 Mountain Lion Developer Preview 1 ("Save As" missing)

NVIDIA Dares Intel to Start Making Chips for ARM Designers




NVIDIA’s sparring with Intel has been well publicized in the last few years. NVIDIA claims their GPUs do a much better job than Intel’s FPUs and Intel is fighting back with better iGPUs and software integration. 

Ever since Intel decided not to allow NVIDIA to make chipsets for their new-generation CPUs some years ago and then basically shut NVIDIA’s  solutions out of their market as Intel’s own CPUs gained GPU capabilities, the two companies have been involved in a serious boxing match.

In the dawn of the first battle, NVIDIA even got a 3 billion dollar payment from Intel and then both signed an agreement. So there is serious friction between the companies.
  
Now, NVIDIA’s Jen-Hsun Huang has been quoted by HotHardware as saying, “Why not be a foundry for all the mobile companies? There’s no shame in that!” Of course, Intel has started manufacturing chips for third party chip designers as they have huge manufacturing capacity and it shouldn’t go idle.

But these chip designers, that are using Intel’s fabs, are mostly FPGA builders and they are in no direct competition with any of Intel’s products.

Companies like Achronix and Tabula build FPGAs (Field Programmable Gate Arrays) are not in the same market with Intel’s Atom line of products while NVIDIA’s Tegra line is a direct competitor and a very successful one at that.

The Tegra designs usually surpass most of the capabilities displayed by their Intel counterparts. Either their graphics processing is better or their power consumption is lower and, most importantly, they’re not held back by the x86 instruction set.

Most industry insiders believe that Intel is not doing it for the money, but rather to keep close any technology or IP that might be interesting for them to license or to acquire.

 With Oualcom’s new Krait design and NVIDIA’s new Tegra 3 / 4 + 1 , Intel could never let all those billions invested in the Atom project go to waste and start building better performing parts designed by its competitors.


Colorful GTX 680 iGame Kudan Graphics Card Pictured




Even though it has already been several days since the official launch of NVIDIA's GeForce GTX 680, there seems to be no stopping the downpour of custom models. 

Then again, this is not so much a report concerning a new video board, but a rumor about a model that has yet to come forth. 

Much like Galaxy, Colorful is developing a special iteration of the Kepler-based graphics card. 

It will be called GTX 680 iGame Kudan and will use the best air cooler that the company has in its collection. 

Called iGame Kudan (hence the card name), the cooling solution is actually modular, meaning that owners can modify it themselves, to some extent. 

The heatsinks can be handled independently (the VRM and memory can be covered with their own) and even the color can be changed (a pastel set is included). 

As people will have surmised by now, Colorful won't waste its best cooling technology on a reference-specced board. 

Unfortunately, though we can say for certain that factory overclocking will be involved, we don't know how much faster than the stock speeds the GPU, shaders and memory will work. 

We presume it won't be in the 2GHz range, but that's just fine, since buyers will get all they need to tweak the performance on their own. 

Speaking of which, the Colorful GTX 680 iGame Kudan will get an 8-phase digital VRM that can be augmented with a 4-phase circuit, through a daughterboard attached to a modified So-DIMM slot. 

The daughterboard lies at the back of the product, parallel to the plane of the PCB. 

People familiar with the company will know that Colorful only sells video controllers in Asia and Europe, so US citizens will have to look for alternatives. 

For the reference specifications that the Kudan seeks to leave in the dust, go here, where we covered the formal release of NVIDIA's invention.




IBM Power 7+ Pictured




IBM is still the packaging king as they’ve been for the last decade or so. They were first to the market with a really big MCM module of “bubble memory” back in the ’70, a huge MCM server design using their Power5 architecture in 2003, and now they are back with a new packaging technology.

On Friday, Charlie Demerjian’s camera from SemiAccurate caught eye of IBM’s new baby, fully exposed. 

The chips pictured here are part of the new Power7+ family and they are, from left to right: four P7+ CPU dies mounted on an interposer with no lid; four P7+ CPU dies with organic packaging, also without lid; and, on the right, one unnamed stacked die with and without the heat spreader. 

Nobody else can show such advanced technology on such a grand scale right now, so IBM is once again on the forefront of silicon IC design. 

It is said that the interposer is an active part that can integrate substantial amounts of eRAM  communicating on a very wide bandwidth.


Twitter Delicious Facebook Digg Stumbleupon Favorites More

 
Design by Free WordPress Themes | Bloggerized by Lasantha - Premium Blogger Themes | coupon codes
`