Sign up for our Technology Newsletter 
Will Arm/Atom CPUs Replace Xeon/Opteron?
Comments Off on Will Arm/Atom CPUs Replace Xeon/Opteron?
Analyst are saying that smartphone chips could one day replace the Xeon and Opteron processors used in most of the world’s top supercomputers. In a paper in a paper titled “Are mobile processors ready for HPC?” researchers at the Barcelona Supercomputing Center wrote that less expensive chips bumping out faster but higher-priced processors in high-performance systems.
In 1993, the list of the world’s fastest supercomputers, known as the Top500, was dominated by systems based on vector processors. They were nudged out by less expensive RISC processors. RISC chips were eventually replaced by cheaper commodity processors like Intel’s Xeon and AMD Opteron and now mobile chips are likely to take over.
The transitions had a common thread, the researchers wrote: Microprocessors killed the vector supercomputers because they were “significantly cheaper and greener,” the report said. At the moment low-power chips based on designs ARM fit the bill, but Intel is likely to catch up so it is not likely to mean the death of x86.
The report compared Samsung’s 1.7GHz dual-core Exynos 5250, Nvidia’s 1.3GHz quad-core Tegra 3 and Intel’s 2.4GHz quad-core Core i7-2760QM – which is a desktop chip, rather than a server chip. The researchers said they found that ARM processors were more power-efficient on single-core performance than the Intel processor, and that ARM chips can scale effectively in HPC environments. On a multi-core basis, the ARM chips were as efficient as Intel x86 chips at the same clock frequency, but Intel was more efficient at the highest performance level, the researchers said.
Qualcomm surpasses AMD
May 30, 2013 by admin
Filed under Uncategorized
Comments Off on Qualcomm surpasses AMD
It’s no secret that the mobile boom is taking a toll on makers of PC components and AMD is one of them. According to data from IC Insights, Qualcomm and Samsung have managed to pass AMD in microprocessor sales last year.
Intel still dominates the market, with $36.9 billion sales and a 65.3 percent market share. However, Qualcomm has managed to squeeze into second spot, with $5.3 billion in sales and a 9.4 percent share. Samsung ranked third, with $4.66 in sales and an 8.2 percent market share. Qualcomm and Samsung also recorded plenty of growth, 28 and 78 percent respectively.
However, AMD slumped 21 percent to take 6.4 percent of the market, with $3.6 billion in sales. It was still ahead of Freescale and Nvidia, as well as Texas Instruments and ST Ericsson.
It should be noted that about 83 percent of Samsung’s revenue came from chips churned out for Apple. In other words, had Apple built the chips on its own, it would have tied with AMD for the third spot.
WD And Sandisk Join Forces
Western Digital and Sandisk have teamed up to create Western Digital’s first hybrid storage device that uses Sandisk’s iSSD and Western Digital’s Caviar Black hard drive.
Western Digital, which has dabbled in solid state disks (SSDs) for the enterprise market, has stayed away from hybrid drives that use relatively small SSDs to act as cache for hard drives. Now the firm has teamed with Sandisk to create its WD Black Solid State Hybrid drives with 500GB capacity.
Western Digital is pitching its hybrid drives at laptop makers, offering units with 5mm, 7mm and 9.5mm heights. The firm said Sandisk’s iSSD uses 19nm NAND flash and claimed it is the world’s “smallest and most advanced semiconductor manufacturing process”, a claim that Intel might question.
Kevin Conley, SVP and GM of client storage solutions at Sandisk said, “By combining SanDisk’s unparalleled flash memory expertise and technology with the hard drive know-how of Western Digital, WD Black SSHDs [solid state hard drives] offer outstanding hard drive-like capacity, and the slim form factor and the level of performance that you will only get with flash memory solutions.”
Seagate was first to introduce hybrid drives with its Momentus XT range, which offers an impressive performance boost over mechanical hard drives for certain workloads. The problem for Western Digital and Seagate is that hybrid drives are merely a stop-gap rather than a long term strategy, with SSD prices falling rapidly due to competition in the SSD industry as opposed to the hard drive industry, where Seagate, Western Digital and Toshiba have a comfortable ride.
Qualcomm Sticks With Windows RT
Tim McDonough, Vice President, Marketing at Qualcomm, was Qualcomm´s commitment to Windows RT. Ever since Microsoft announced Windows RT, ARM supporters had high hopes and Windows RT has yet to live up to some.
Tim confirmed Qualcomm´s commitment to Windows RT and future releases, saying “we are here for the long run”. He describes the partnership as the beginning of a long journey and of course Qualcomm is going to continue rolling out chips that will run great with Windows RT.
Qualcomm mentioned that Samsung ATIV and Dell XPS 10, both of which use Qualcomm’s S4 dual-core APQ8060A chips, run really nice. Tim told us that he is a real fan of both devices and that he is currently using one of them.
We also learned that Snapdragon 600, the one used in the HTC One and some versions of Samsung’s Galaxy S4, is 40 per cent faster than the S4 Pro, adding that Adreno 320 graphics core is significantly faster than the Adreno 225 used in the S4 APQ8060A chip. Another number we got is that the Adreno 330 is up to four times faster than the 225, which is a huge leap forward. Let’s not forget that Snapdragon 800, which is up to 75 per cent faster than Snapdragon S4 Pro, is also coming in mid-year, second half of 2013. The 800 will be Qualcomm’s first chip with Adreno 330 graphics.
One can easily conclude that there should be some Snapdragon 600 and 800 Windows RT convertible tablets at some point in the future. To stay on the safe side, Qualcomm just confirmed that new and exciting things are coming in the next months and quarter and they are Windows based.
We have to notice that most people in the tablet world get really excited talking about convertible tablets in all shapes and sizes, as the physical keyboard is definitely an accessory you want to have.
openSUSE Lacks Resources For ARM
Opensuse said that its ARM development is being limited by a lack of resources to build software despite having launched its Open Build Service (OBS).
Last month the Opensuse project announced the release of Opensuse 12.3, which brought ARM support to the same level as x86 and AMD64. While the project is working on bringing ARMv7 and more importantly ARMv8 support to its Linux distribution, Jos Poortvliet, community manager at Opensuse, said that the project’s ARM development has been limited by the lack of build resources.
Opensuse announced a collaboration with Samsung to create the OBS, which it was hoped would speed up the development life-cycle. However Poortvliet said, “ARM development is limited by available build resources required for compiling each iteration of new software and while the OBS helps by bringing a lot of build power in one place, the use of QEMU meant that build resources were shared with native x86_64 builds, which turned out to be a performance limitation.
“With fast and dedicated ARM hardware we can reserve build power for ARM builds and make use of the more efficient KVM virtualization.”
However in better news, Poortvliet said that the project had managed to deploy KVM – the Linux kernel based virtual machine – on ARM hardware. He added that parent firm Suse has assigned more resources to building ARM software on OBS and forecast that all packages would be built in two weeks.
While Canonical and Red Hat have been vocal about their ARM developments, Suse and its Opensuse project have been quietly going about their business, though given Poortvliet’s comments regarding a lack of resources, perhaps they have been going about it too quietly.
Although ARM vendors are not expected to converge on the server market until next year, even ARM thinks that most servers using its chips will run open source software.
Unless Suse manages to get its act together, it might find that Canonical and Red Hat have already carved out a significant chunk of the market.
Windows RT Devices To Drop In Price
Prices of Windows RT devices have started to decline, signaling an attempt by PC makers to quickly clear out stock after poor adoption of tablets and convertibles with the operating system.
Microsoft released Windows RT for ARM-based devices and Windows 8 for Intel-based devices in October last year. The price drop is an acknowledgement that Windows RT has failed, analysts said.
Prices of popular products usually don’t fall, but Windows RT devices were not in demand, and prices fell, analysts said.
The starting price for Dell’s XPS 10 is now US$449 for a 32GB model, scalping $50 off the original launch price. The 64GB model is $499, which is a drop from the original $599 price. By comparison, the price of the Latitude 10 tablet with Intel processors and Microsoft’s Windows 8 OS remained stable at $499.
Asus’ VivoTab RT, which is largely sold through retailers, is being offered by Amazon.com for $382 with 32GB of storage, which is a heavy discount from the $599 launch price. Retailers like Best Buy, Staples and Office Depot have also dropped the price of the tablet by $50, now selling it for $549.
Newegg is listing VivoTab RT as having been discontinued. Asus did not respond to a request to comment on whether the company was still offering the tablet.
Lenovo is offering the IdeaPad Yoga 11 for $599 as part of a seven-day deal, which is a drop from the original $799 price. However, TigerDirect is offering an IdeaPad Yoga 11 model for $599 on its website, while Amazon is selling a model for $499.99.
Samsung did not ship its Windows RT tablet, Ativ Tab, to the U.S. market.
However, the starting price of Microsoft’s Surface RT remained consistent at $499 on its online store. Microsoft also offers Lenovo’s IdeaPad Yoga 11 through its store, but has stopped offering tablets like the VivoTab RT on its website. The company last month said it stocks its store with RT devices based on availability and demand.
TSMC Testing ARM’s Cortex A57
ARM and TSMC have manufactured the first Cortex A57 processor based on ARM’s next-gen 64-bit ARMv8 architecture.
The all new chip was fabricated on TSMC’s equally new FinFET 16nm process. The 57 is ARM’s fastest chip to date and it will go after high end tablets, and eventually it will find its place in some PCs and servers as well.
Furthermore the A57 can be coupled with frugal Cortex A53 cores in a big.LITTLE configuration. This should allow it to deliver relatively low power consumption, which is a must for tablets and smartphones. However, bear in mind that A15 cores are only now showing up in consumer products, so it might be a while before we see any devices based on the A57.
In terms of performance, ARM claims the A57 can deliver a “full laptop experience,” even when used in a smartphone connected to a screen, keyboard and mouse wirelessly. It is said to be more power efficient than the A15 and browser performance should be doubled on the A57.
It is still unclear when we’ll get to see the first A57 devices, but it seems highly unlikely that any of them will show up this year. Our best bet is mid-2014, and we are incorrigible optimists. The next big step in ARM evolution will be 20nm A15 cores with next-generation graphics, and they sound pretty exciting as well.
Is Ubuntu Linux Spyware?
Richard Stallman has asked a South American free software association not to promote Ubuntu Linux at its events because it “spies on its users” by collecting its users’ desktop search activity and selling the data to Amazon.
Canonical released Ubuntu 12.10 last October with Amazon search integrated into its Dash desktop search function.
Although Ubuntu users can opt out and Canonical claims it anonymises users’ search information before sending it to Amazon, the change resulted in Ubuntu users being shown Amazon ads in response to desktop search queries.
The ‘feature’ has attracted a lot of criticism and might have led some users to defect to other Linux distributions.
When Stallman’s request was denied by the FLISOL event organiser with the excuse that it would limit user freedom of choice, Stallman fired off a response to the organisation’s entire mailing list on Sunday. Parts of his email are quoted below, as translated by Groklaw.
“The issue I raise is about what should happen at FLISOL events. Give away copies of Ubuntu or not? Promote Ubuntu or no? I asked the organisers of the event that they, as a policy, not distribute or promote Ubuntu.
“Freedom of users is something else, and there isn’t a conflict between a user’s freedom and my request. If someone decides to install Ubuntu, I would consider it a mistake, but it’s his own choice to do it. What I ask is that you don’t participate, help or suggest that he do it. I didn’t request that you block him from doing so.
“As a matter of principle, I don’t believe anyone has a right, morally, to distribute proprietary software, that is, software that deprives the users of freedom. When the user controls his own software, he can install what he wants and no one can stop him. But today’s issue isn’t about him, what he does, but rather what you do with him.”
As Stallman sent his email only yesterday, it’s not yet known whether FLISOL has reconsidered promoting Ubuntu at its free software events.
These points might seem like splitting hairs, but apparently Richard Stallman – the author of the GNU General Public Licence (GPL), as well as the founder and president of the Free Software Foundation – is serious about them.
TSMC And Imagination Team Up
TSMC and Imagination Technologies announced the next step in their tech collaboration in an effort to develop Imagination’s next generation PowerVR 6-series GPUs.
The new GPUs are still not ready for prime time, but they should be used in future SoC designs, including those stamped out using TSMC’s 16nm FinFET process. The two companies will work to create new reference system designs, utilizing high bandwidth memory standards and TSMC’s 3D IC technology.
As GPU muscle becomes more important for next generation SoCs, designers need more advanced and more complex processes, such as TSMC’s 16FinFET.
“Through advanced projects initiated under this partnership, Imagination and TSMC are working together to showcase how SoCs will transform the future of mobile and embedded products,” said Hossein Yassaie, CEO of Imagination.
TSMC VP Cliff Hou argued that the need for high performance mobile GPUs will drive silicon processes in the future, much in the same way CPU development pushed new processes in the nineties.
Are CUDA Applications Limited?
Acceleware said at Nvidia’s GPU Technology Conference (GTC) today that most algorithms that run on GPGPUs are bound by GPU memory size.
Acceleware is partly funded by Nvidia to provide developer training for CUDA to help sell the language to those that are used to traditional C and C++ programming. The firm said that most CUDA algorithms are now limited by GPU local memory size rather than GPU computational performance.
Both AMD and Nvidia provide general purpose GPU (GPGPU) accelerator parts that provide significantly faster computational processing than traditional CPUs, however they have only between 6GB and 8GB of local memory that constrains the size of the dataset the GPU can process. While developers can push more data from system main memory, the latency cost negates the raw performance benefit of the GPU.
Kelly Goss, training program manager at Acceleware, said that “most algorithms are memory bound rather than GPU bound” and “maximising memory usage is key” to optimising GPGPU performance.
She further said that developers need to understand and take advantage of the memory hierarchy of Nvidia’s Kepler GPU and look at ways of reducing the number of memory accesses for every line of GPU computing.
The point Goss was making is that GPU computing is relatively cheap in terms of clock cycles relative to the time it takes to fetch data from local memory, let alone loading GPU memory from system main memory.
Goss, talking to a room full of developers, proceeded to outline some of the performance characteristics of the memory hierarchy in Nvidia’s Kepler GPU architecture, showing the level of detail that CUDA programmers need to pay attention to if they want to extract the full performance potential from Nvidia’s GPGPU computing architecture.
Given Goss’s observation that algorithms running on Nvidia’s GPGPUs are often constrained by local memory size rather than by the GPU itself, the firm might want to look at simplifying the tiers of memory involved and increasing the amount of GPU local memory so that CUDA software developers can process larger datasets.