Intel’s Chief Executive Paul Otellini said Nokia would find it hard to differentiate using the Windows platform: “It would have been less hard on Android, on MeeGo he could have done it.”

“We will find another partner. The carriers still want a third ecosystem and the carriers want an open ecosystem, and that’s the thing that drives our motivation,” he said. MeeGo was created last year by the merger of two Linux-based platforms; Nokia’s Maemo and and Intel’s Moblin. MeeGo runs on both ARM and Intel processors.

The MeeGo tablet demo looks great, but will there be enough apps, asks C/Net. Intel’s MWC press release, claims that MeeGo has made “great strides” since the OS’ announcement a year ago, and “gained strong industry momentum with software vendors, system integrators and operators, as well as OEMs and products shipping today in multiple form factors, including netbooks, tablet, set-top boxes and in-vehicle infotainment systems in cars.”

Meanwhile Otellini said during a keynote at Mobile World Congress that a number of smartphone models would be launched this year using the firm’s upcoming Atom processor, dubbed Medfield. ARM currently holds close to 90% of the mobile market, so convincing hardware manufacturers to embrace Medfield won’t be easy. Medfield runs MeeGo, of course.

Otellini said in Barcelona that open systems had the edge over closed systems: “Some closed models will certainly survive, because you can optimize the experience, but in general, if you harness the ability of all the engineers in the world and the developers in the world, open wins.

Otellini said that Nokia dropped the MeeGo operating system after Microsoft offered “incredible” amounts of money,” reports Reuters.

Nokia’s boss Stephen Elop said on Sunday that Microsoft essentially won a bidding war against Google to supply software to the world’s largest handset maker worth “billions”.

Later in the week, however, it appears like the “billions” figure is soft costs, R&D savings and marketing around Windows’ phone.

Nokia Plan B was a hoax. The person behind the “Nokia Plan B” shareholder revolt turned out to be just one very bored engineer who likes his iPhone.”

It was grounded in real shareholder concerns about the Microsoft-Nokia partnership, however. Nokia stock has plummeted 25 percent since Elop and Microsoft CEO Steve Ballmer announced the partnership Friday.

Google Chief Executive Eric Schmidt said at the Mobile World Congress in Barcelona on Wednesday that he had held extensive talks to try to woo Nokia.

View full post on dailywireless.org

Intel will gain access to Nvidia’s patents while paying the graphics chip supplier $1.5 billion in licensing fees, reports C/Net. The two companies canceled a Dec. 6 trial date to hash out lawsuits they had pending against each other.

“For the future use of Nvidia’s technology, Intel will pay Nvidia an aggregate of $1.5 billion in licensing fees payable in five annual installments, beginning Jan. 18, 2011,” Nvidia announced today. Nvidia and Intel have also agreed to drop all outstanding legal disputes between them.

Intel and Nvidia had both sued each other in early 2009. Intel had contended nVidia’s cross license does not extend to Intel’s future-generation processors. Nvidia countersued blocking access to its patent portfolio.

In effect, Nvidia was barred from building Intel-compatible chipsets beyond the Core 2 Duo generation of processors. For example, the second generation of Apple’s MacBook Air used an Nvidia chipset along with Intel’s Core 2 Duo processor.

The agreement announced Monday still bars Nvidia from using any of Intel’s x86 technology and, as a result, Nvidia cannot build x86-compatible chipsets, according to Intel. But Nvidia CEO Jen-Hsun Huang said, “We’ve already said many times that we have no intention to build chipsets for Intel processors.”

Microsoft confirmed last week that the next version of Windows will support ARM-based computer chips, used in smartphones and tablets. Windows 8 will work with ARM-based systems from partners Nvidia, Qualcomm, and Texas Instruments.

According to nVidia’s calculations, desktop and laptop computing is history. New super phones using multi-core CPUs are the future. NVIDIA also announced Project Denver, an ARM CPU for the desktop. Nvidia will also manufacture the “high-performance ARM core” for supercomputing and eventually high-end PC use. NVIDIA also licensed ARM’s current Cortex-A15 processor for its future-generation Tegra mobile processors.

View full post on dailywireless.org

Note: This Intel Core i5-2500K review also reviews the Sandy Bridge architecture as a whole, as well as the high-end Intel Core i7-2600K processor.

Ready or not, here they come. Intel is rolling out a thoroughly overhauled range of PC processors based on its new Sandy Bridge microarchitecture. Our first taste of the new chips comes in the form of the Intel Core i5-2500K and Intel Core i7-2600K desktop CPUs.

Thanks to the baffling array of chips, sockets and brands, we’ve barely got to grips with Intel’s existing CPU range. Certainly Intel’s main rival, AMD, has no answer in outright performance terms to the chips Intel already offers, but the relentless march of technology must go on.

So, ignore the familiar Core i5 and Core i7 branding. These are all new processors and they’re ready to roll.

As it happens, Intel could actually do with more powerful and, crucially, more power efficient processors for laptop PCs. Deep down, that’s what Sandy Bridge is really about. However, as we’ll learn, Sandy Bridge has a lot to offer for the desktop, too, including exciting new features such as a hardware video transcoding engine and much-improved integrated graphics.

At launch, it’s not quite a full range of processors. In total, Intel is rolling out 14 new desktop CPUs based on Sandy Bridge under the Core i3, Core i5 and Core i7 brands. Pricing extends from around £75 to £250.

The very top and bottom of Intel’s desktop range will remain unchanged. The six-core Gulftown derivative of the Core i7 and the entry-level Pentium chips live on for now. Everything else is effectively for the chop.

With new chips inevitably come new chipsets, but what you might not expect is the new processor socket. Known as LGA1155, it’s almost identical to the existing LGA1156. It loses just one pin. Problem is, the two sockets are entirely incompatible.

You cannot drop Sandy Bridge processors into any existing motherboard. Nor can you use any previous Intel processors with Sandy Bridge-supporting motherboards.

Although Intel will absolutely deny it, we believe this change in sockets is unlikely to have been a technical necessity.

Instead, at best, we suspect it represents a disregard for existing customers, and at worst a tactic designed to force whole a platform upgrade rather than drop-in CPU changes.

Unfortunately, the new socket isn’t the only dubious new feature that debuts with Sandy Bridge. Intel has also made a fundamental change to the chip’s architecture that has killed traditional processor overclocking using the CPU bus stone dead. We’ll come to the details later, but the bottom line is that Intel has brought overclocking under strict control.

It’s not completely gone, but it’s only available where and when Intel allows it.

However, don’t let those negatives convince you the new Intel Core i5-2500K and Intel Core i7-2600K chips are stinkers. Compared to their closest equivalents from Intel’s existing range – the Intel Core i5-760 and Intel Core i7-870 – the new chips are arguably in a different league.

In fact, even the mighty six-core Core i7 980X isn’t safe from Intel’s remarkable new Sandy Bridge architecture. It’s that good.

In terms of basic layout, the Sandy Bridge architecture that underpins the new Intel Core i5-2500K and Intel Core i7-2600K desktop processors is not a revolutionary redesign. For launch, Intel has cooked up dual- and quad-core derivatives.

All Sandy Bridge processors are based on Intel’s latest 32nm production process. Thus, in Intel’s “Tick-Tock” jargon, we’re talking Tock. In other words, Sandy Bridge is a new architecture on an existing silicon node.

Familiar features such as HyperThreading and Turbo Boost reappear with a bit of spit and polish. Likewise, most of the so-called uncore features look like a carry over. The memory controller remains dual channel with support for DDR3 DIMMs up to 1,333MHz.

As before, the on-die PCI Express discrete graphics interface delivers a total of 16 lanes. Multi-GPU platforms from AMD and Nvidia are supported, but again only in dual-card configuration with eight lanes per card.

Dig deeper, however, and you’ll dredge up some seriously significant changes. For our money, one of the most important is the newly integrated clock generator. In operating frequency terms, this change effectively binds the entire platform to the baseclock.

Bump up the baseclock and all core and uncore elements of the chip, including everything from the PCI Express bus to the memory controller and the cores themselves, gain frequency in proportion.

The upshot is that overclocking via the baseclock is no longer a goer. At best you’ll manage an overclock in the low single digits above the standard 100MHz frequency. Instead, significant frequency adjustments with Sandy Bridge is done through the CPU multiplier.

If you recall, Intel recently introduced a number of K series chips with unlocked multipliers. Until now, the special Ks didn’t make much sense, such was the ease and effectiveness of baseclock tweaks.

However, it now seems clear Intel was preparing the ground for the baseclock-locked Sandy Bridge architecture.

For Sandy Bridge, the overclocking drill goes like this. K series chips are fully unlocked up to a theoretical 5.7GHz while all other Sandy Bridgers are unlocked to the tune of four speed bins. Given the baseclock of 100MHz, that means a maximum overclock of 400MHz for all non-K models.

That’s a major reduction compared to the typical 1GHz overclock possible via the bus or baseclock with most existing Intel processors.

Overclocking aside, the other big change with Sandy Bridge involves video processing. For starters, Intel has given its HD Graphics a major kick up the backside. The result, Intel claims, is a more than doubling in performance compared the previous generation of Intel integrated graphics.

Architecturally, the GPU has been has brought completely on-die.

Remember, previous Intel processors with integrated graphics actually contained two separate chips in the CPU package, one chip with cores and cache, the other with graphics and uncore features such as the memory controller.

At launch, all Sandy Bridge processors have on-die graphics. In terms of 3D performance, the detail spec changes include an upgrade from DX10 to DX10.1. However, much of the performance increase comes from the boost in clockspeed from a maximum of 900MHz to a maximum of 1,350MHz, model dependent.

Things get a little more complicated if you further examine the different models. Among desktop variants, only the two K series chips tested here get the full-fat version of the new graphics core with 12 execution units, known as Intel HD Graphics 3000. All other models make do with the 6-unit Intel HD Graphics 2000.

Odd as it may seem, we believe the other chips in the range do actually contain the full 12 execution units. Most likely for marketing reasons, Intel has decided to disable them on some models.

The other big news involving the graphics core is a new video transcoding engine known as Quick Sync Video. It’s a dedicated hardware core for accelerating 2D video encoding and it’s in every single Sandy Bridge chip. Intel says it delivers at least twice the performance compared with previous architectures encoding video in software, but a little Intel birdie tells us you can often expect much more than that.

Like pretty much every new Intel architecture Sandy Bridge also brings revised vector instructions. This time round, the new instructions are known as AVX and essentially compliment the existing SSE vector instructions.

Oh and, if you care, the new graphics core supports HDMI 1.4 and, in turn, stereoscopic 3D video including Blu-ray 3D.

Finally, regarding the chips themselves, a quick lesson in Intel’s new nomenclature is essential.

All Sandy Bridge processors get some kind of 2000 suffix. Core i3-2100 chips are dual-core processors with HyperThreading and are therefore quad-thread. The Core i5-2300, 2400 and 2500 series are a mix of dual-core with HyperThreading and quad-core without HyperThreading.

Meanwhile, the Core i7-2600 series is quad-core with HyperThreading. If that wasn’t complicated enough, you have the aforementioned K series chips with fully unlocked multipliers. Then there are T and S series models with lower operating voltages and power ratings.

Overall, it’s pretty baffling stuff.

As for motherboards, all Sandy Bridge processors drop into the new LGA1155 socket. It’s not compatible with any previous Intel socket. In total, five new chipsets arrive with Sandy Bridge, but only the P67 and H67 are really relevant to consumers.

There’s not a great deal to choose between them, but the key differences include support for the integrated graphics core and video out with the H67, while the P67 exclusively gets the latest version of Intel’s Performance Tuning application and a few more options in terms of multiplier adjustment.

Another motherboard-related tweak is the introduction of the EFI firmware with all Sandy Bridge boards. Developed by Intel and already used by Apple in all its Mac computers, EFI replaces the traditional BIOS firmware with a much more powerful platform that supports high resolution graphical user interfaces, faster boot times, larger boot disks (over 2TB) and more.

With every new architecture comes a certain amount of intrigue. Intel’s Sandy Bridge brings much higher clocks, but it’s also interesting to note how performance compares on a clock-for-clock basis.

To find out, we compared the new Core i7-2600K running at 2.9GHz with Turbo disabled to the old Core i7-870, again at 2.9GHz with Turbo disabled. Also worth noting is the performance of the new video transcoding core and the integrated 3D graphics. In their own way, both are frankly staggering.

Stock clocked benchmarks

Cinebench R10 multi-thread

Intel Core i5-2500K: 39s
Intel Core i7-2600K: 35s
Intel Core i7-870: 46s
Intel Core i5-750: 59s
Intel Core i7-980X: 32s

Cinebench R10 single-thread

Intel Core i5-2500K: 2m 22s
Intel Core i7-2600K: 2m 19s
Intel Core i7-870: 2m 58s
Intel Core i5-750: 3m 17s
Intel Core i7-980X: 3m 00s

x264 HD video encode

Intel Core i5-2500K: 27fps
Intel Core i7-2600K: 35fps
Intel Core i7-870: 28fps
Intel Core i5-750: 19fps
Intel Core i7-980X: 43fps

World in Conflict

Intel Core i5-2500K: 91fps
Intel Core i7-2600K: 96fps
Intel Core i7-870: 78fps
Intel Core i5-750: 75fps
Intel Core i7-980X: 90fps

Memory Bandwith

Intel Core i5-2500K: 18.6GB/s
Intel Core i7-2600K: 18.7GB/s
Intel Core i7-870: 17.3GB/s
Intel Core i5-750: 16.9GB/s
Intel Core i7-980X: 17.8GB/s

Idle power consumption

Intel Core i5-2500K: 125W
Intel Core i7-2600K: 130W
Intel Core i7-870: 115W
Intel Core i5-750: 110W
Intel Core i7-980X: 145W

Peak power consumption

Intel Core i5-2500K: 190W
Intel Core i7-2600K: 215W
Intel Core i7-870: 210W
Intel Core i5-750: 180W
Intel Core i7-980X: 245W

Overclocking

Maximum overclock

Intel Core i5-2500K: 4.4GHz
Intel Core i7-2600K: 4.6GHz
Intel Core i7-870: 3.9GHz
Intel Core i5-750: 4.1GHz
Intel Core i7-980X: 4.2GHz

Cinebench R10 multi-thread

Intel Core i5-2500K: 34s
Intel Core i7-2600K: 29s
Intel Core i7-870: 30s
Intel Core i5-750: 43s
Intel Core i7-980X: 26s

x264 HD video encode

Intel Core i5-2500K: 32fps
Intel Core i7-2600K: 41fps
Intel Core i7-870: 33fps
Intel Core i5-750: 27fps
Intel Core i7-980X: 51fps

World in Conflict

Intel Core i5-2500K: 98fps
Intel Core i7-2600K: 102fps
Intel Core i7-870: 90fps
Intel Core i5-750: 89fps
Intel Core i7-980X: 94fps

Clock-for-clock comparison (Turbo disabled)

Cinebench R10 multi-thread

Intel Core i7-2600K @ 2.9GHz: 46s
Intel Core i7-870 @ 2.9GHz: 51s

Cinebench R10 single-thread

Intel Core i7-2600K @ 2.9GHz: 3m 5s
Intel Core i7-870 @ 2.9GHz: 3m 28s

x264 HD video encode

Intel Core i7-2600K @ 2.9GHz: 27fps
Intel Core i7-870 @ 2.9GHz: 25fps

World in Conflict

Intel Core i7-2600K @ 2.9GHz: 88fps
Intel Core i7-870 @ 2.9GHz: 77fps

Integrated graphics

Call of Duty 4 @ 1,280 x 800

Intel Core i7-2600K – HD Graphics 3000: 32fps
Intel Core i5-655K – HD Graphics: 14fps

World in Conflict @ 800 x 600

Intel Core i7-2600K – HD Graphics 3000: 22fps
Intel Core i5-655K – HD Graphics: 8fps

Quick Sync Video transcode engine

720p H.264 encode in software

Intel Core i7-2600K: 45s

720p H.264 encodewith Quick Sync Video

Intel Core i7-2600K: 12s

First a quick précis of our review chips. The Intel Core i5-2500K is a quad-core, quad-thread chip with 6MB of L3 cache memory and clocks in at 3.3GHz standard and 3.7GHz maximum Turbo frequency. It’s yours for approximately £160 unboxed and effectively replaces both the old dual-core Core i5-655K and quad-core Core i5-760.

A partially locked Core i5-2500 is also available at the same clockspeeds for about £10 less.

The £225 Core i7 -2600K, meanwhile, sports four cores, eight threads, 8MB of cache as standard and Turbo frequencies of 3.4GHz and 3.8GHz respectively. It slots in approximately where you would previously have found the likes of the Core i7-870 and Core i7-875K processors in the price list. Again, there’s also a marginally cheaper Core i7-2600 alternative that lacks the fully unlocked multiplier.

With all that logged away, it’s time for some performance analysis. In a word, these chips are stunning.

At stock clocks, the new Core i5-2500K comprehensively blows away an old Core i7-870. Put another way, the new mid-range chip hammers the old high-end offering. It’s literally faster in every benchmark including video encoding, gaming and professional rendering and sometimes by a margin of over 10 per cent. As for the comparison to the old Core i5-760, it’s brutal. The new chip is as much as 30 to 40 per cent quicker.

Unsurprisingly, then, the new Core i7-2600K absolutely hoses its progenitor, the Core i7-870. We’re talking 25 to 30 per cent more performance absolutely, positively everywhere.

If that wasn’t astonishing enough, the i7-2600K actually gives Intel’s mighty Core i7-980X six-core flagship processor a real scare. The 980X is, for instance, only about 10 per cent quicker in the Cinebench rendering test and 20 per cent faster in HD video encoding. But it’s slower for gaming.

Remember, the i7-980X is a £750 processor.

If you’re wondering where the performance gains come from, much is down to increased clockspeeds, but the new processors would still be 10 per cent quicker even running at the same frequencies. Somehow, Intel has made the fastest CPU architecture on the planet significantly faster.

As for overclocking, the news only gets better. Our i5-2500K hits the wall at 4.4GHz, while the i7-2600K motors on to 4.6GHz. That’s courtesy of an air cooler. It’s silly numbers, frankly.

As if that wasn’t enough, the Quick Sync Video hardware transcoding engine adds yet another dimension. Using a special build of Cyberlink MediaShow Expresso, it’s possible to compare video encoding with and without the transcoder enabled. Our test video is crunched in 45 seconds in software mode. Flick the switch on the transcoder and it drops to just 12 seconds. Bonkers.

Finally, there’s the performance of the new Intel HD Graphics 3000. Here again, it’s a story of massively improved grunt. Put simply, it’s over twice as fast as the previous Intel HD graphics and massively quicker than any other integrated solution. It’s genuinely up to the task of playing older games such as Call of Duty 4 at decent detail settings.

Newer titles, however, require lower resolutions and defenestrated eye candy.

We’ll make no bones about it. Intel’s latest processors present us with a major headache. On the one hand, there’s a hell of a lot we don’t like about the Intel Core i5-2500K and Intel Core i7-2600K. All of which has to do with what we believe are probably silly marketing decisions, not engineering issues.

We don’t like the fact that Intel has locked down proper overclocking to certain models. We hate that Intel has limited the faster HD Graphics 3000 integrated 3D core to these two K series processors. We’re thoroughly cheesed off that the new chips require a new socket, chipsets and motherboard.

And we’re borderline homicidal with rage regarding Intel’s ludicrous branding scheme.

But here’s the thing. In simple CPU performance terms, the Intel Core i5-2500K and Core i7-2600K are staggering. They overclock like the clappers. The integrated graphics is way better than anything before and the transcode engine threatens to annihilate the very idea of running some software on discrete graphics chips before it’s even taken hold.

They’re so good, in fact, they make nearly all of Intel’s supposedly high-end chips for the LGA1366 socket instantly redundant. Bloomfield-based quad-core Core i7 processors are toast. Even six-core Gulftown Core i7s look marginal.

Put simply, these new chips bring massive performance to a lower price point than ever before. How can you argue with that?

Related Links

• Read more processor reviews
• TechRadar’s Reviews Guarantee

View full post on TechRadar: All PC & Mac feeds

Note: This Intel Core i7-2600K review also reviews the Sandy Bridge architecture as a whole, as well as the mid-range Intel Core i5-2500K processor.

Ready or not, here they come. Intel is rolling out a thoroughly overhauled range of PC processors based on its new Sandy Bridge microarchitecture. Our first taste of the new chips comes in the form of the Intel Core i5-2500K and Intel Core i7-2600K desktop CPUs.

Thanks to the baffling array of chips, sockets and brands, we’ve barely got to grips with Intel’s existing CPU range. Certainly Intel’s main rival, AMD, has no answer in outright performance terms to the chips Intel already offers, but the relentless march of technology must go on.

So, ignore the familiar Core i5 and Core i7 branding. These are all new processors and they’re ready to roll.

As it happens, Intel could actually do with more powerful and, crucially, more power efficient processors for laptop PCs. Deep down, that’s what Sandy Bridge is really about. However, as we’ll learn, Sandy Bridge has a lot to offer for the desktop, too, including exciting new features such as a hardware video transcoding engine and much-improved integrated graphics.

At launch, it’s not quite a full range of processors. In total, Intel is rolling out 14 new desktop CPUs based on Sandy Bridge under the Core i3, Core i5 and Core i7 brands. Pricing extends from around £75 to £250.

The very top and bottom of Intel’s desktop range will remain unchanged. The six-core Gulftown derivative of the Core i7 and the entry-level Pentium chips live on for now. Everything else is effectively for the chop.

With new chips inevitably come new chipsets, but what you might not expect is the new processor socket. Known as LGA1155, it’s almost identical to the existing LGA1156. It loses just one pin. Problem is, the two sockets are entirely incompatible.

You cannot drop Sandy Bridge processors into any existing motherboard. Nor can you use any previous Intel processors with Sandy Bridge-supporting motherboards.

Although Intel will absolutely deny it, we believe this change in sockets is unlikely to have been a technical necessity.

Instead, at best, we suspect it represents a disregard for existing customers, and at worst a tactic designed to force whole a platform upgrade rather than drop-in CPU changes.

Unfortunately, the new socket isn’t the only dubious new feature that debuts with Sandy Bridge. Intel has also made a fundamental change to the chip’s architecture that has killed traditional processor overclocking using the CPU bus stone dead. We’ll come to the details later, but the bottom line is that Intel has brought overclocking under strict control.

It’s not completely gone, but it’s only available where and when Intel allows it.

However, don’t let those negatives convince you the new Intel Core i5-2500K and Intel Core i7-2600K chips are stinkers. Compared to their closest equivalents from Intel’s existing range – the Intel Core i5-760 and Intel Core i7-870 – the new chips are arguably in a different league.

In fact, even the mighty six-core Core i7 980X isn’t safe from Intel’s remarkable new Sandy Bridge architecture. It’s that good.

In terms of basic layout, the Sandy Bridge architecture that underpins the new Intel Core i5-2500K and Intel Core i7-2600K desktop processors is not a revolutionary redesign. For launch, Intel has cooked up dual- and quad-core derivatives.

All Sandy Bridge processors are based on Intel’s latest 32nm production process. Thus, in Intel’s “Tick-Tock” jargon, we’re talking Tock. In other words, Sandy Bridge is a new architecture on an existing silicon node.

Familiar features such as HyperThreading and Turbo Boost reappear with a bit of spit and polish. Likewise, most of the so-called uncore features look like a carry over. The memory controller remains dual channel with support for DDR3 DIMMs up to 1,333MHz.

As before, the on-die PCI Express discrete graphics interface delivers a total of 16 lanes. Multi-GPU platforms from AMD and Nvidia are supported, but again only in dual-card configuration with eight lanes per card.

Dig deeper, however, and you’ll dredge up some seriously significant changes. For our money, one of the most important is the newly integrated clock generator. In operating frequency terms, this change effectively binds the entire platform to the baseclock.

Bump up the baseclock and all core and uncore elements of the chip, including everything from the PCI Express bus to the memory controller and the cores themselves, gain frequency in proportion.

The upshot is that overclocking via the baseclock is no longer a goer. At best you’ll manage an overclock in the low single digits above the standard 100MHz frequency. Instead, significant frequency adjustments with Sandy Bridge is done through the CPU multiplier.

If you recall, Intel recently introduced a number of K series chips with unlocked multipliers. Until now, the special Ks didn’t make much sense, such was the ease and effectiveness of baseclock tweaks.

However, it now seems clear Intel was preparing the ground for the baseclock-locked Sandy Bridge architecture.

For Sandy Bridge, the overclocking drill goes like this. K series chips are fully unlocked up to a theoretical 5.7GHz while all other Sandy Bridgers are unlocked to the tune of four speed bins. Given the baseclock of 100MHz, that means a maximum overclock of 400MHz for all non-K models.

That’s a major reduction compared to the typical 1GHz overclock possible via the bus or baseclock with most existing Intel processors.

Overclocking aside, the other big change with Sandy Bridge involves video processing. For starters, Intel has given its HD Graphics a major kick up the backside. The result, Intel claims, is a more than doubling in performance compared the previous generation of Intel integrated graphics.

Architecturally, the GPU has been has brought completely on-die.

Remember, previous Intel processors with integrated graphics actually contained two separate chips in the CPU package, one chip with cores and cache, the other with graphics and uncore features such as the memory controller.

At launch, all Sandy Bridge processors have on-die graphics. In terms of 3D performance, the detail spec changes include an upgrade from DX10 to DX10.1. However, much of the performance increase comes from the boost in clockspeed from a maximum of 900MHz to a maximum of 1,350MHz, model dependent.

Things get a little more complicated if you further examine the different models. Among desktop variants, only the two K series chips tested here get the full-fat version of the new graphics core with 12 execution units, known as Intel HD Graphics 3000. All other models make do with the 6-unit Intel HD Graphics 2000.

Odd as it may seem, we believe the other chips in the range do actually contain the full 12 execution units. Most likely for marketing reasons, Intel has decided to disable them on some models.

The other big news involving the graphics core is a new video transcoding engine known as Quick Sync Video. It’s a dedicated hardware core for accelerating 2D video encoding and it’s in every single Sandy Bridge chip. Intel says it delivers at least twice the performance compared with previous architectures encoding video in software, but a little Intel birdie tells us you can often expect much more than that.

Like pretty much every new Intel architecture Sandy Bridge also brings revised vector instructions. This time round, the new instructions are known as AVX and essentially compliment the existing SSE vector instructions.

Oh and, if you care, the new graphics core supports HDMI 1.4 and, in turn, stereoscopic 3D video including Blu-ray 3D.

Finally, regarding the chips themselves, a quick lesson in Intel’s new nomenclature is essential.

All Sandy Bridge processors get some kind of 2000 suffix. Core i3-2100 chips are dual-core processors with HyperThreading and are therefore quad-thread. The Core i5-2300, 2400 and 2500 series are a mix of dual-core with HyperThreading and quad-core without HyperThreading.

Meanwhile, the Core i7-2600 series is quad-core with HyperThreading. If that wasn’t complicated enough, you have the aforementioned K series chips with fully unlocked multipliers. Then there are T and S series models with lower operating voltages and power ratings.

Overall, it’s pretty baffling stuff.

As for motherboards, all Sandy Bridge processors drop into the new LGA1155 socket. It’s not compatible with any previous Intel socket. In total, five new chipsets arrive with Sandy Bridge, but only the P67 and H67 are really relevant to consumers.

There’s not a great deal to choose between them, but the key differences include support for the integrated graphics core and video out with the H67, while the P67 exclusively gets the latest version of Intel’s Performance Tuning application and a few more options in terms of multiplier adjustment.

Another motherboard-related tweak is the introduction of the EFI firmware with all Sandy Bridge boards. Developed by Intel and already used by Apple in all its Mac computers, EFI replaces the traditional BIOS firmware with a much more powerful platform that supports high resolution graphical user interfaces, faster boot times, larger boot disks (over 2TB) and more.

With every new architecture comes a certain amount of intrigue. Intel’s Sandy Bridge brings much higher clocks, but it’s also interesting to note how performance compares on a clock-for-clock basis.

To find out, we compared the new Core i7-2600K running at 2.9GHz with Turbo disabled to the old Core i7-870, again at 2.9GHz with Turbo disabled. Also worth noting is the performance of the new video transcoding core and the integrated 3D graphics. In their own way, both are frankly staggering.

Stock clocked benchmarks

Cinebench R10 multi-thread

Intel Core i5-2500K: 39s
Intel Core i7-2600K: 35s
Intel Core i7-870: 46s
Intel Core i5-750: 59s
Intel Core i7-980X: 32s

Cinebench R10 single-thread

Intel Core i5-2500K: 2m 22s
Intel Core i7-2600K: 2m 19s
Intel Core i7-870: 2m 58s
Intel Core i5-750: 3m 17s
Intel Core i7-980X: 3m 00s

x264 HD video encode

Intel Core i5-2500K: 27fps
Intel Core i7-2600K: 35fps
Intel Core i7-870: 28fps
Intel Core i5-750: 19fps
Intel Core i7-980X: 43fps

World in Conflict

Intel Core i5-2500K: 91fps
Intel Core i7-2600K: 96fps
Intel Core i7-870: 78fps
Intel Core i5-750: 75fps
Intel Core i7-980X: 90fps

Memory Bandwith

Intel Core i5-2500K: 18.6GB/s
Intel Core i7-2600K: 18.7GB/s
Intel Core i7-870: 17.3GB/s
Intel Core i5-750: 16.9GB/s
Intel Core i7-980X: 17.8GB/s

Idle power consumption

Intel Core i5-2500K: 125W
Intel Core i7-2600K: 130W
Intel Core i7-870: 115W
Intel Core i5-750: 110W
Intel Core i7-980X: 145W

Peak power consumption

Intel Core i5-2500K: 190W
Intel Core i7-2600K: 215W
Intel Core i7-870: 210W
Intel Core i5-750: 180W
Intel Core i7-980X: 245W

Overclocking

Maximum overclock

Intel Core i5-2500K: 4.4GHz
Intel Core i7-2600K: 4.6GHz
Intel Core i7-870: 3.9GHz
Intel Core i5-750: 4.1GHz
Intel Core i7-980X: 4.2GHz

Cinebench R10 multi-thread

Intel Core i5-2500K: 34s
Intel Core i7-2600K: 29s
Intel Core i7-870: 30s
Intel Core i5-750: 43s
Intel Core i7-980X: 26s

x264 HD video encode

Intel Core i5-2500K: 32fps
Intel Core i7-2600K: 41fps
Intel Core i7-870: 33fps
Intel Core i5-750: 27fps
Intel Core i7-980X: 51fps

World in Conflict

Intel Core i5-2500K: 98fps
Intel Core i7-2600K: 102fps
Intel Core i7-870: 90fps
Intel Core i5-750: 89fps
Intel Core i7-980X: 94fps

Clock-for-clock comparison (Turbo disabled)

Cinebench R10 multi-thread

Intel Core i7-2600K @ 2.9GHz: 46s
Intel Core i7-870 @ 2.9GHz: 51s

Cinebench R10 single-thread

Intel Core i7-2600K @ 2.9GHz: 3m 5s
Intel Core i7-870 @ 2.9GHz: 3m 28s

x264 HD video encode

Intel Core i7-2600K @ 2.9GHz: 27fps
Intel Core i7-870 @ 2.9GHz: 25fps

World in Conflict

Intel Core i7-2600K @ 2.9GHz: 88fps
Intel Core i7-870 @ 2.9GHz: 77fps

Integrated graphics

Call of Duty 4 @ 1,280 x 800

Intel Core i7-2600K – HD Graphics 3000: 32fps
Intel Core i5-655K – HD Graphics: 14fps

World in Conflict @ 800 x 600

Intel Core i7-2600K – HD Graphics 3000: 22fps
Intel Core i5-655K – HD Graphics: 8fps

Quick Sync Video transcode engine

720p H.264 encode in software

Intel Core i7-2600K: 45s

720p H.264 encodewith Quick Sync Video

Intel Core i7-2600K: 12s

First a quick précis of our review chips. The Intel Core i5-2500K is a quad-core, quad-thread chip with 6MB of L3 cache memory and clocks in at 3.3GHz standard and 3.7GHz maximum Turbo frequency. It’s yours for approximately £160 unboxed and effectively replaces both the old dual-core Core i5-655K and quad-core Core i5-760.

A partially locked Core i5-2500 is also available at the same clockspeeds for about £10 less.

The £225 Core i7 -2600K, meanwhile, sports four cores, eight threads, 8MB of cache as standard and Turbo frequencies of 3.4GHz and 3.8GHz respectively. It slots in approximately where you would previously have found the likes of the Core i7-870 and Core i7-875K processors in the price list. Again, there’s also a marginally cheaper Core i7-2600 alternative that lacks the fully unlocked multiplier.

With all that logged away, it’s time for some performance analysis. In a word, these chips are stunning.

At stock clocks, the new Core i5-2500K comprehensively blows away an old Core i7-870. Put another way, the new mid-range chip hammers the old high-end offering. It’s literally faster in every benchmark including video encoding, gaming and professional rendering and sometimes by a margin of over 10 per cent. As for the comparison to the old Core i5-760, it’s brutal. The new chip is as much as 30 to 40 per cent quicker.

Unsurprisingly, then, the new Core i7-2600K absolutely hoses its progenitor, the Core i7-870. We’re talking 25 to 30 per cent more performance absolutely, positively everywhere.

If that wasn’t astonishing enough, the i7-2600K actually gives Intel’s mighty Core i7-980X six-core flagship processor a real scare. The 980X is, for instance, only about 10 per cent quicker in the Cinebench rendering test and 20 per cent faster in HD video encoding. But it’s slower for gaming.

Remember, the i7-980X is a £750 processor.

If you’re wondering where the performance gains come from, much is down to increased clockspeeds, but the new processors would still be 10 per cent quicker even running at the same frequencies. Somehow, Intel has made the fastest CPU architecture on the planet significantly faster.

As for overclocking, the news only gets better. Our i5-2500K hits the wall at 4.4GHz, while the i7-2600K motors on to 4.6GHz. That’s courtesy of an air cooler. It’s silly numbers, frankly.

As if that wasn’t enough, the Quick Sync Video hardware transcoding engine adds yet another dimension. Using a special build of Cyberlink MediaShow Expresso, it’s possible to compare video encoding with and without the transcoder enabled. Our test video is crunched in 45 seconds in software mode. Flick the switch on the transcoder and it drops to just 12 seconds. Bonkers.

Finally, there’s the performance of the new Intel HD Graphics 3000. Here again, it’s a story of massively improved grunt. Put simply, it’s over twice as fast as the previous Intel HD graphics and massively quicker than any other integrated solution. It’s genuinely up to the task of playing older games such as Call of Duty 4 at decent detail settings.

Newer titles, however, require lower resolutions and defenestrated eye candy.

We’ll make no bones about it. Intel’s latest processors present us with a major headache. On the one hand, there’s a hell of a lot we don’t like about the Intel Core i5-2500K and Intel Core i7-2600K. All of which has to do with what we believe are probably silly marketing decisions, not engineering issues.

We don’t like the fact that Intel has locked down proper overclocking to certain models. We hate that Intel has limited the faster HD Graphics 3000 integrated 3D core to these two K series processors. We’re thoroughly cheesed off that the new chips require a new socket, chipsets and motherboard.

And we’re borderline homicidal with rage regarding Intel’s ludicrous branding scheme.

But here’s the thing. In simple CPU performance terms, the Intel Core i5-2500K and Core i7-2600K are staggering. They overclock like the clappers. The integrated graphics is way better than anything before and the transcode engine threatens to annihilate the very idea of running some software on discrete graphics chips before it’s even taken hold.

They’re so good, in fact, they make nearly all of Intel’s supposedly high-end chips for the LGA1366 socket instantly redundant. Bloomfield-based quad-core Core i7 processors are toast. Even six-core Gulftown Core i7s look marginal.

Put simply, these new chips bring massive performance to a lower price point than ever before. How can you argue with that?

Related Links

• Read more processor reviews
• TechRadar’s Reviews Guarantee

View full post on TechRadar: All PC & Mac feeds

A group of Intel researchers has pioneered a messaging system that would allow multiple cores to communicate.

View full post on Macworld

Intel today announced it has invested $32 million in an Israeli start-up that is about to release an SSD product based on consumer NAND flash but that is…

View full post on Macworld

Augmented reality startup Layar has raised roughly $14 million (€10 million) in its second round of funding, says TechCrunch, led by Intel Capital. This brings the total of capital raised by the Dutch company to $17.4 million.

Layar says the new funding will be used primarily to enhance its open mobile AR platform to third-party developers and brands to increase available content. The company will continue developing its mobile application, dubbed the Layar Reality Browser, which is currently available for iPhone and Android.

View full post on dailywireless.org

Taiwan President Ma Ying-jeou said Thursday that Intel will continue its support for Taiwan’s development of WiMAX, after a meeting with Intel President and CEO Paul Otellini in Taipei.

Ma said the government has released six licenses for WiMAX operators and has invested US$220 million in developing related technologies and applications.

An apparent lack of communication between Intel and the Taiwan government over the closing of Intel’s WiMax Program Office in Taiwan erupted into a media shit storm this summer.

Intel insists it was all a misunderstanding. VMAX launched their commercial WiMAX network in Taiwan in March, and now covers more than 3 million people.

VMax expects to spend around $47 million to expand its wireless broadband services in Taipei through 2010. VMax is also offering mobile access through taxi fleet operators in the Taipei area. The number of taxis having a WiMax-enabled device from VMax may increase to 5,000 at the end of 2010 and to 20,000 a year later.

Five of Taiwan’s licensed WiMAX operators – Global Mobile, First International Telecom (Fitel), Vmax Telecom, Tatung Telecom and Far EasTone Telecommunications (FET) are rolling out WiMAX services in Taiwan.

Global Mobile, Vmax and FITEL won licenses for northern Taiwan, while Far EasTone, Tatung and Vastar Cable TV System won licenses for the south. FITEL expected to have 52 Mobile WiMAX base stations operating in Taipei City by the beginning of June. The Fitel deployment is part of the massive M-Taiwan project to unwire the country. Fitel, headed by Charlie Wu, operates a Japanese-style PHS system — personal handyphone system.

VMax expects 40,000 subscribers to its service by end-2010, which may increase to 80,000 in 2011. VMax started its WiMax network on Jan. 26, 2010, covering 85% of Taipei City.

The WiMAX Forum hopes that Clear, in the United States, UQ in Japan, KT in South Korea, VMAX in Taiwan and Packet One of Indonesia, among others, will hang in there and that India will come around.

It may be an uphill battle.

The last, best hope of “4G” spectrum is now going up for auction. Available frequencies in the 2.5-2.7 Ghz band and the “digitial dividend” spectrum, using (now freed-up) broadcast television frequencies are the battle ground.

Traditional cellular carriers now see a TD-LTE standard emerging for the unpaired frequency slots, and appear to have the bucks and the motivation to bid and win TD-LTE solutions.

Separately this week, visiting Intel president and CEO Paul Otellini announced a series of new and expanded partnerships with Taiwan’s government.

Intel announced plans to team up with Taiwan on a cloud computing initiative, setting up a multi-million dollar Internet computing research laboratory there.

Intel said it will work with the island’s National Science Council and a leading Taiwanese university to establish a “cloud computing” research centre with an estimated cost of 23.5 million US dollars over the next three to five years. Intel did not specify the amount it plans to spend.

Intel’s cloud computing initiative uses Internet-based computing resources, sharing software and providing computer power on demand.

Intel’s Cloud 2015 vision has three key elements: a “federated” cloud that allows enterprises to share data across internal and external clouds; an “automated” network that automatically allows the secure movement of applications and resources to significantly improve energy efficiency in data centers; and PC and device-savvy “client-aware” clouds that know what types of applications, commands and processing should take place in the cloud or on your notebook, smartphone or other device – thus taking a user and specific device’s unique features into account to fully optimize an online experience.

Related Dailywireless articles include; WiMAX: Good News, Bad News, Qualcomm India: For Sale?, Qualcomm Gets Indian Partners, Vendors Scramble for Indian Backhaul, India’s Broadband Auction: It’s Done, Yota: Planetary LTE Swap, Yota Dumps WiMAX, WiMAX Forum: Not Dead Yet, WiMAX Forum: In Trouble?, Russian WiMAX, Battle for Britain, European 2.5 GHz Auctions & the Global Market, BT’s European WiMAX Plan, WiMAX Roundup, Australia Unwired, India 2nd Largest Mobile Market, Intel: $500M for M-Taiwan, AT&T: More Transpacific Cable, Satphones: Merger Ahead?, Malaysian WiMAX: Now or Never, WiMAX Auctions: NZ & Hong Kong, and Cellular Penetration: Half the World.

View full post on dailywireless.org

GladiatorStocks.com Daily Investment Stock Report Alerts: Amazon.com Inc., Ford Motor Company, Intel Corp., Nokia Corp.
NEWARK, DELAWARE– (Marketwire – Oct. 22, 2010) – GladiatorStocks.com, a premier micro cap research firm provides subscribers with comprehensive research, in depth due diligence and technical analysis to provide our subscribers with high octane stocks that are typically overlooked by your “Average Joe” Wall St. investment advisors. Ask yourself what is easier to find, a stock that moves from …

Read more on Marketwire

FreePennyAlerts.com Daily Investment Stock Report Alerts: Ford Motor Company, Intel Corp., Nokia Corp., Sprint Nextel …
NEWARK, DELAWARE–(Marketwire – 10/21/10) – FreePennyAlerts.com, a premier micro cap research firm provides subscribers with comprehensive research, in depth due diligence and technical analysis to provide our subscribers with high octane stocks that are typically overlooked by your “Average Joe” Wall St. investment advisors. Ask yourself what is easier to find, a stock that moves from $.01 – $1 …

Read more on Marketwire via Yahoo! Finance

Intel announced today that the company will invest between $6 billion and $8 billion on future generations of manufacturing technology in its American facilities.

The action will fund deployment of Intel’s next-generation 22- nanometer (nm) manufacturing process across several existing U.S. factories, along with the construction of a new development fabrication plant in Oregon. The Oregon factory should be ready in 2013 and will primarily produce chips for research and development as Intel advances its designs.

Washington County’s new development fab will be called D1X and is scheduled for research and development startup in 2013. Upgrades are also planned for the D1C and D1D fabs on the Hillsboro campus. The projects will support 6,000 to 8,000 construction jobs and result in 800 to 1,000 new permanent high-tech jobs.

HIGHLIGHTS

• Intel will spend $6-8 billion in manufacturing to support future technology advancements in Arizona and Oregon.
• The investment supports the creation of 6,000-8,000 construction jobs and 800-1,000 permanent high-tech jobs, and also allows Intel to maintain its current manufacturing employment base at these U.S. sites.
• The investment will fund a new development fab in Oregon, as well as upgrades to four existing fabs to manufacture the next-generation 22-nanometer (nm) process technology.
• Intel’s next-generation, 22nm microprocessors will enable sleeker device designs, higher performance and longer battery life at lower costs.

Intel’s announcement indicated a new development fab will be built in Oregon (above), along with upgrading four other U.S. fabs (Fab 12 and Fab 32 in Arizona and D1C and D1D in Oregon).

Oregon Live reports that Intel will remake its advanced Ronler Acres campus for the new Fab. The new Hillsboro plant — Intel’s first new plant since 2007 — is scheduled to start production in 2013, reports the NY Times.

Oregon Governor Ted Kulongoski said, “This is probably the largest private investment during this last two or three years in this country.”

Intel has approximately 15,000 employees in Oregon, making it the largest private employer in the state (pdf). Intel’s operations in Oregon stretch across six campuses in Washington County west of Portland. It is the company’s largest and most comprehensive site in the world, a global center of semiconductor research and manufacturing and the anchor of Oregon’s economy.

Intel’s upcoming 32-nanometer “Sandy Bridge” Core architecture got much of the attention at the company’s developer show last month. Sandy Bridge chips, built using 32 nm architect, will be out early in 2011. Ivy Bridge is the codename given to the 22 nm die shrink of Sandy Bridge.

During his keynote address at the Intel Developer Forum, Sept. 13, CEO Paul Otellini said that the company’s Ivy Bridge” processor, a 22-nm chip, is on schedule and will be delivered in the second half of 2011. The chips are already in sample production, he said.

“Last year at IDF, I showed you the first working wafer with SRAM devices on it,” Otellini said during his talk. “This year, I’m happy to report that our first microprocessor designed for 22 nanometers is moving through our fabs as we speak here today.”

Intel is also building its first production facility in China, reports Bloomberg. Intel is vying with Samsung Electronics to be the industry’s biggest spender on production, budgeted $5.2 billion for plants and equipment in 2010. The company’s microprocessors run more than 80 percent of the world’s personal computers. Rival Samsung is the biggest maker of memory chips.

Moving to 22-nanometer could also help the company produce chips with lower power consumption to better compete in smartphones—where designs from ARM currently dominate. Intel launched the Atom platform two years ago. Now executives are looking to aggressively expand the reach of the Atom chips, into tablets, handheld devices and phones.

View full post on dailywireless.org

Intel the ARM Wrestler
Intel plans to pull out all the stops to conquer the tablet PC market, CEO Paul Otellini announced this week. Intel will use all of the assets at its disposal to win this segment, he said. Intel will offer tablets running Meego, the operating system that’s a melding of its and Nokia’s technologies.

Read more on TechNewsWorld.com

Intel revs up in fourth quarter
Intel Corp posted stronger-than-expected third-quarter results on Tuesday and said its sales could rise about 2.7 percent sequentially in the final quarter of the year

Read more on National Post

Intel: No MeeGo Handsets until 2011
Nokia has been saying all year that the first MeeGo phones will go on sale but the end of 2010, but now an Intel executive is throwing a wrench in the works: in an interview with Forbes, Intel VP Dough Fisher says MeeGo-based handsets won’t hit the market until the first half of 2011. Fisher also says Meego-based tablets wwill land in 2011—although there is already one MeeGo-based tablet on the …

Read more on Digital Trends via Yahoo! News

Intel will provide a tool to help port iPhone apps to handsets and tablets with Intel chips.

View full post on Macworld

Intel and Korea Telecom are collaborating to accelerate the adoption of WiMax-like services in South Korea, reports the Wall St Journal

Intel Capital also announced an investment of USD 20 million in Wibro Infra (WIC), a joint venture with KT, Samsung and KBIC, which builds and operates Wibro infrastructure in Korea, to accelerate KT’s wireless broadband infrastructure.

The KT Chairman Lee reportedly believes WiBro is a “persuasive answer” to better handle wireless data traffic than LTE.

In April, KT signed a memorandum of understanding with Intel to collaborate on WiMAX technology. The deal called for Intel to produce a chipset just for Korea’s WiMAX flavor, called WiBro, which uses the 2.3 GHz band with 8.75MHz wide channels. It is otherwise similar to WiMAX (802.16e). WiBro could then allow international roaming with WiMAX compatibility.

SK Telecom, South Korea’s largest cellular operator, was also a WiMAX partner. But recently the mobile operator said it will provide commercial LTE services in Seoul in 2011 and plans to expand the service to 6 Metropolitan Cities in 2012 with nationwide LTE coverage by 2013. SK Telecom is also an investor in U.S.-based LightSquared, the 4G-LTE/satellite venture backed by Harbinger Capital Partners.

KT will expand the Wibro service this week into five new cities of Busan, Daegu, Gwangju, Daejeon and Ulsan and the expressways of Gyeongbu, Jungbu, Honam and Yeongdong, in addition to its current coverage in the metropolitan areas of Seoul, Inchon and Suwon,

By March 2011, KT aims to offer WiBro services to 82 cities, covering 85 percent of South Korea’s population.

Starting today, Korean customers will be able to purchase Intel-based laptops and netbooks from Samsung, LG and Acer. Subscribers will be able to use WCDMA (GSM), Wi-Fi and WIBRO/WiMAX starting today. KT is also migrating its Wibro network onto 10MHz Wimax channels that allow more interoperability and roaming with Wimax networks worldwide.

Korea Telecom (KT) has the largest portion of the South Korean local telephone and high-speed Internet business. KT had 7.264 million internet customers in August, up from 7.222 million million in July, but the number of Wibro subscribers has slipped to 344,000 from 345,000, while their Wi-Fi user base decreased to 278,000 from 280,000 a month.

KT’s IPTV subscribers totals 1.714 million, up from 1.640 million in July. KT’s PSTN subscriber base slipped to 17.041 million from 17.138 million, while the company’s VoIP subscriber base rose to 2.377 million in August from 2.294 million in July.

In other WiMAX news, Sprint Nextel has pulled three of its top execs, including CEO Dan Hesse, from Clearwire’s board of directors, the companies said today. They cited antitrust concerns as the motivator.

Unstrung is skeptical and speculates that Clearwire may be paving the way for new investors. It has been in talks with T-Mobile USA about securing an equity investment to use its WiMax spectrum for its own 4G service. T-Mobile calls WiMAX a ‘niche play’.

Meanwhile, Clear’s 4G WiMAX service was officially launched in Pittsburgh (820,000 people) and Minneapolis/St. Paul (2.1 million), this week.

Clear’s 4G service is currently available in 56 markets across the United States, serving 66 million people. By the end of 2010, CLEAR says their WiMAX service will also be available in major metropolitan areas such as New York City, Los Angeles, the San Francisco Bay Area, Denver, Miami, Cincinnati and Cleveland, with a market coverage around 120 million.

Clearwire and UQ Communications of Japan have launched international 4G roaming services. Clearwire had a similar roaming deal with Russian WiMAX provider Yota, but Yota is now moving to LTE.

The WiMAX Forum hopes that Clear, in the United States, UQ in Japan, KT in South Korea, VMAX in Taiwan and Packet One of Indonesia will hang in there and that India will come around.

When the Indian BWA spectrum auctions concluded earlier this June, it was assumed that WiMAX would be the technology of choice for the 2.3GHz TDD spectrum. The two state-owned telcos, BSNL and MTNL, made early commitments to the wireless standard. But the broadband wireless standard for 1.2 billion people is still largely undecided.

Related Dailywireless articles include; India’s Broadband Auction: It’s Done, Yota Dumps WiMAX, Clearwire: New Mobile Hotspots, Clear: No Limits, WiMAX Forum: Not Dead Yet, WiMAX Forum: In Trouble?, Sprint’s WiMAX Phone Launched, SK Telecom Buys 25% of Packet One, Compare “4G” Carriers in the U.S., LTE for Sprint? and MIMO: The Paper War

View full post on dailywireless.org

Intel CTO Justin Rattner described the next generation of computers, which he dubbed context-aware, during a Wednesday keynote address at the Intel Developers Forum.

View full post on Macworld

The Intel Developer Forum in San Francisco this week will provide the first details of Sandy Bridge, its new 32nm microprocessor architecture that will power a broad range of its chips over the next two years.

Intel is also expected to show off Moorestown, the shrink of current Atom processors for tablets and smartphones. Many expect it to show Groveland, an Atom variation aimed at set-top boxes.

At least one and maybe more Atom-based SoCs are expected, including Groveland a chip aimed at set-top boxes, says EE Times.

IDF will showcase USB and PCI Express 3.0 and there will be a smattering of news tidbits about support for parallel programming in Intel multicore processors.

At the pre-show Sunday, Hardware Secretssaw demos for the following ideas:

• A photo-recognition system for smart phones
• The ability to run programs (“apps”) on your TV at the same time you watch live shows
• A projected screen on your kitchen countertop that responds to the touch and is capable to recognize objects on the countertop and interact accordingly
• A face-recognition system implemented inside a car, allowing the car to automatically load the driver’s preferred settings, and also to identify if the driver is sleepy or driving in an unsafe manner
• Using the embedded camera of netbooks for activities with early elementary students

But the big news is Sandy Bridge. Sandy Bridge is Intel’s next microarchitecture, or redesign, of its processors–which the chipmaker does every two years. The current design, Nehalem, was introduced in November 2008 and is used in all Core i3, i5, and i7 processors.

Anand Lal Shimpi, a noted hardware reviewer at Anandtech, said he thinks entry-level GPUs will be unnecessary in the wake of Sandy Bridge’s introduction.

Sandy Bridge is Intel’s first architecture to merge graphics and x86 cores on a single die. It uses 32nm process and is the first Intel chip to support vector graphics and AES security instructions. It is expected to be released in Q1 2011.

In Sandy Bridge, Intel improved graphics speed by integrating graphics functions onto the CPU. Sandy Bridge will include new instructions for handling multimedia tasks, called Advanced Vector Extension (Intel AVX). They will assist in accelerating a host of multimedia tasks, including video and audio processing. Sandy Bridge will also include silicon dedicated to handling the transcoding, or converting, of data from one format to another.

Keynotes slated this week include:

• Day 1: Opening keynote with Paul Otellini, president and CEO. “Performance Computing: Making the Extraordinary Ordinary” with David Perlmutter, executive vice president and general manager, Intel Architecture Group.
• Day 2: “Creating the Experience Continuum on Intel® Architecture” with Renee James, vice president and general manager, Software and Solutions Group. “Smart, Connected, Transformational – Fueling the Continuum of Computing with the Intel® Atom™ Processor” with Doug Davis, general manager, Embedded and Communications Group.
• Day 3: “Context: How it Will Really Change Everything” with Justin Rattner, vice president and director, Intel Labs, and chief technology officer and Intel Senior Fellow

The CE 4100 (Sodaville, above), is a follow on to Intel’s first SoC for TVs, the Canmore chip announced last year at IDF. The new chip replaces an 800 MHz Pentium M core with a 1.2 GHz Atom, increasing L2 cache from 256 to 512 Mbytes.

This week we will likely find out if GoogleTV products from Sony and others will use an Atom-based SoC for set-top boxes and tvs.

Meanwhile, just weeks ago, archrival Advanced Micro Devices gave the first peak of Bobcat and Bulldozer, its first new x86 cores designed from a clean sheet of paper in nearly a decade.

By mid week everyone should have a clearer view of what the technical battle lines will be between Intel and AMD for the next couple years.

View full post on dailywireless.org

Speedier performance for more or less the same money is always welcome. That’s exactly what you get from the new Intel Core i5 760. In almost every regard, it’s identical to the existing Core i5 750 processor.

It’s therefore a quad-core chip based on Intel’s 45nm Nehalem processor architecture. What it does offer, however, is a speed bump from 2.66GHz to 2.8GHz. Not exactly a spectacular explosion in performance, we admit. But then Intel is only asking for an extra fiver.

In any case, even after a year on sale, the Core i5 750 is still an excellent all round performer. The slightly quicker Core i5 760 is only going to be better. The only snag is the arrival of AMD’s new Phenom II X6 1055T beast. Six cores for a similar £150 or so price tag is awfully hard to resist.

Fire up any of Intel’s Core i5 processors and you’ll be greeted by four little green graphs in Windows Task Manager. That indicates the presence of four logical processors and therefore support for crunching a quartet of software threads in parallel.

But not all logical processors are equal. Intel’s Core i5 600 series chips are actually dual-core models that support two threads per core thanks to HyperThreading. Core i5 700 series CPUs lack HyperThreading. Instead, you get four bona fide execution cores and one helluva lot more performance. Effective as HyperThreading is, it’s no substitute for a real processor core.

Despite a relatively modest 2.8GHz stock clockspeed, it’s therefore no surprise to find the Intel Core i5 760 flattens the Core i5 661 and Core i5 655K in literally every performance benchmark. You have to wonder why anyone would want to pay more for one of those Core i5 600 series chips. Intel would no doubt justify the price premium by pointing to the Core i5 600′s integrated graphics.

Anyway, a much tougher nut for the Core i5 760 to crack is AMD’s new six-core processor. The Phenom II X6 1055T is actually slightly cheaper despite giving the Core i5 760 a pretty solid schooling in our multi-threading tests. As ever, however, the 760 hits back with awesome gaming performance. It also shows the Phenom chip who’s boss in our file decompression test.

It’s also disappointing to find the new 760 does not appear to offer any additional overclocking headroom. In our testing, the older Core i5 750 actually hits higher frequencies.

We liked:

If you fancy four cores for the price of two, it’s hard to argue against the Core i5 760. It’s cheaper than many of Intel’s latest dual-core processors and hammers them for all round performance. At this price point, it’s also the finest gaming CPU on the market.

We disliked:

Intel has a nasty habit of artificially hobbling some of its mainstream processors. So it is for the Core i5 760. By switching off HyperThreading, Intel has compromised the 760′s performance in highly threaded applications such as video encoding and handed the advantage to AMD’s similarly priced six-core Phenom II X6 1055T. Shame.

Verdict:

Replaces the Core i5 750 as our favourite gaming chip. No HyperThreading is a bit of a bummer.

Follow TechRadar Reviews on Twitter: http://twitter.com/techradarreview

Related Links

• Read more processor reviews
• TechRadar Reviews Guarantee

View full post on TechRadar: All PC & Mac feeds

Six cores, 12 threads, one billion transistors and a whole lotta processing power. That’s the Intel Core i7 970. It’s a fabulous technological achievement. But can it really be worth nearly £700?

Admittedly, the Core i7 970 is a bit cheaper than Intel’s top six-core chip, the Core i7 980X. But it’s still a tough ask given that AMD will sell you its own six-core Phenom II X6 1055T CPU for around one third the price. What the AMD chip doesn’t give you, however, is the very latest technology. The Core i7 970 sports literally the most advanced PC processor die on the planet.

Part of the 970′s advantage comes down to microprocessor architecture. Known as Nehalem, it’s as good as it currently gets thanks to features such as HyperThreading, a triple-channel memory controller and the super-fast QPI interconnect.

The Core i7 970 is also hewn from Intel’s latest 32nm silicon. That means the chip itself is physically smaller than most current quad-core processors including Intel’s own Core i7 870 and AMD’s Phenom II X4 965 Black Edition. In theory that means it should be cheaper to manufacture. For now, Intel isn’t passing on those cost savings to customers.

There’s no doubting the sheer power of the Intel Core i7 970. Try this for starters. In our HD video encoding benchmark it bangs out nearly 42 frames per second. That absolutely blows away both the quad-core Intel Core i7 875K (27 frames per second) and the six-core AMD Phenom II X6 1090T BE (26 frames per second).

It’s a similar story in the professional 3D rendering test. The 970 is miles ahead. Arguably even more impressive are the game and file decompression results – benchmarks that do not normally favour CPU parallelism. The Core i7 970 is usefully ahead of the Core i7 870 and more than 50 per cent faster than the Phenom II X6 1090T.

In performance terms, therefore, no compromises are required. That’s because the 970 backs up its hefty parallelism (six cores and 12 threads) with a 3.2GHz clockspeed. Whatever the workload, it’s damn quick. In fact, the only CPU that operates in the same ballpark is the Intel Core i7 980X, which is of course a slightly higher clocked version of the same six-core processor die.

The one area where you might expect all those cores and threads to cause a problem is overclocking. But even here the Core i7 970 puts on a decent show, ramping up to 4GHz courtesy of nothing more than an air cooler and some voltage tweaks. It’s all very impressive until you remember the price. It’s possible to buy a very nice desktop PC, complete with processor, graphics and screen, for the cost of this CPU.

We liked:

In a word, performance. Whether it’s games, media encoding or just multi-tasking, the Core i7 970 has delivers rude amounts of it. But what else did you expect from six of the most advanced processor cores money can buy. The damn thing even overclocks well.

We disliked:

In a word, price. Awesome as the Intel Core i7 970 is, it’s also awesomely expensive and simply poor value compared to quad-core processors costing one third the price. In fact, if you’re willing to spend this much, you may as well go all the way with the Core i7 980X.

Verdict:

A six-core masterpiece in 32nm silicon. Shame about the grand-master pricing, though.

Related Links

• TechRadar Reviews Guarantee
• Read more processor reviews

View full post on TechRadar: All PC & Mac feeds

Want the finest CPU money can buy? Look no further than the Intel Core i7 980X. It’s the undisputed world heavyweight champ among chips. Just remember you’ll need a lot of it. The money, that is.

But what a processor you get in return. The Core I7 980X is, of course, a six-core beast. But this is six-cores Intel style, so that’s two threads per core and a dozen of those little green graphs when you fire up task manager. That’s unparalleled, er, parallelism in a PC processor.

In that context, even AMD’s Phenom II X6 1090T Black Edition looks rather ordinary. It may have six cores. But each is single-threaded only. Moreover, AMD’s underlying CPU architecture is pretty ancient while the Intel Core i7 980X is literally the latest thing, right down to is impossibly tiny 32nm underpinnings. What it isn’t however, is unique. The recently released Core i7 970 is largely the same six-core, 12-thread processor at a slightly less offensive price point.

Among the super rich of this world, there’s a level of wealth colloquially known as screw-you money. Put simply, if you’ve got enough dosh, you can say “screw you” without consequences to anyone, any time. For people like that, it’s results that matter, not cost. For people like that, Intel has a very special PC processor. It’s the Core i7 980X.

As our benchmark results show, it’s without exception the fastest PC processor on the planet. So, it’s much quicker in our multi-threaded tests than AMD’s best six-core effort, the Phenom II X6 1090T BE.

It’s also an absolute monster in benchmarks that traditionally favour chips with fewer cores but higher clocks, such as gaming and file decompression. For that you can thank the 980X’s 3.33GHz operating frequency. In performance terms, there really are no downsides with this six-core beast.

Indeed, you might expect a chip with over a billion transistors to run hot and not take kindly to overclocking. If so, you’d be forgetting that the Core i7 980X is based on Intel’s fancy new 32n production process. It’s actually smaller than Intel’s quad-core processors. What’s more, it hits an astonishing 4.3GHz when overclocked.

So, that’s the most cores, the highest thread count, the most advanced microarchitecture and the highest clockspeed. No wonder the Core i7 980X is a prince among processors. It’s just a shame it has a princely price tag to match. We also worry about the 980X’s LGA1,366 CPU socket. Intel plans to drop it next year, putting a definite limit on any upgrade path.

We liked:

Can a CPU be considered aspirational? If so, the Intel Core i7 980X is it. It’s a smorgasbord of cutting edge technology and clearly the fastest and finest CPU you can buy. It’s all of Intel’s know how in a multi-threaded masterpiece. We want one. And we want it badly.

We disliked:

Placing a premium on performance is all very well. But charging nigh on £800 for a PC processor is plain bonkers. It’s sad, because it renders the Core i7 980X effectively irrelevant. Virtually nobody can afford it. Given Intel’s track record in regards ongoing socket support, the LGA1,366 form factor is a worry, too.

Verdict:

The fastest PC processor we’ve ever tested. But so expensive, nobody can afford it.

Follow TechRadar Reviews on Twitter: http://twitter.com/techradarreview

Related Links

• TechRadar Reviews Guarantee
• Read more processor reviews

View full post on TechRadar: All PC & Mac feeds

Giving a dual-core processor performance to rival many quad-core chips is quite an achievement. Intel has managed just that with the Core i5 661. Problem is, it also costs as much as many quad-core processors.

The challenges for the dual-core Core i5 661 begin with Intel’s own Core i5 750 and 760 chips. Both are cheaper than the 661 but pack a pair of extra cores. That’s a fundamental advantage that even the Core i5 661′s clever HyperThreading technology will struggle to overcome.

However, things get really silly when you consider the opposition from AMD. Not only does AMD offer a slew of quad-core chips for less money, including the range topping Phenom II X4 965 Black Edition. Even the six-core Phenom II X6 1055T is cheaper. Crazy.

Of course, what none of the chip above can offer is the Core i5 661′s integrated graphics core. Exactly how much that is worth, however, is another matter.

Technically, the Intel Core i5 661 is a thoroughly intriguing chip. Exceptionally quick for a dual-core processor, it also packs a graphics core integrated into the processor package itself. If industry analysts are to be believed, this kind of CPU-GPU “fusion” processor is the way of things to come.

Back in 2010, the reality is more mundane. The Core i5 661′s 3D rendering abilities are feeble. Admittedly, motherboard-based integrated cores from the likes of AMD or NVIDIA are not much better. But the fact remains this latest integrated graphics core from Intel is effectively worthless for gaming. Fortunately, its 2D video decoding prowess is pretty solid.

Anyway, of more significance is the Core i5 661′s performance as a traditional CPU. For a dual-core processor, it’s awesome. Thanks to HyperThreading technology, a healthy standard clockspeed of 3.33GHz and the Turbo feature which boost clocks by a further few hundred MHz or so in certain conditions, the 661 beats other dual-core chips with an extremely large stick in highly-threaded software such as video encoding. It even gives quad-core chips such as AMD Athlon II X4 620 a scare.

That sounds impressive until you realise the Core i5 661 is more expensive than any quad-core processor from AMD and even some Intel quad-core chips. The Core i5 750 and 760 are both cheaper and faster. Then there’s the AMD Phenom II X6 1055T. Despite being cheaper, it packs six cores and absolutely annihilates the i5 661 in heavily threaded apps.

In other words, the addition of an integrated GPU isn’t nearly enough to justify the Core i5 661′s premium pricing. That’s especially true when you note that supporting motherboards are no cheaper. In fact, if you really wanted to pair a six-core AMD processor with an integrated motherboard, you could. And it would still work out cheaper than the 661 and its fancy fusion graphics.

We like:

Intel’s Nehalem CPU architecture makes for an uncommonly quick dual-core CPU. For most PC users, most of the time, it’s all the CPU they’ll ever need. It’s also power efficient, cool running and offers a glimpse of the future courtesy of its CPU-GPU fusion architecture.

We dislike:

In a word, price. The Core i5 661 is at least £50 over priced. Yes, it has a graphics core integrated into the CPU package. But it’s a decidedly feeble graphics core of little value to gamers. Moreover, AMD offers a six-core processor for the same price. ‘Nuff said.

Verdict:

Fusion may be the future. But for now, this CPU-GPU chip costs too much and delivers too little.

Follow TechRadar Reviews on Twitter: http://twitter.com/techradarreview

Related Links

• TechRadar Reviews Guarantee
• Read more processor reviews

View full post on TechRadar: All PC & Mac feeds

Late last summer, Intel rolled out the Core i5 750 based on Lynnfield, its mainstream Nehalem-class CPU. It became an instant favourite. Sure, it was short of a feature or three compared to Intel’s finest, including HyperThreading. But for the money, there was nothing to touch it.

Fast forward 12 months and the Intel Core i5 750 is hanging in there. That’s despite the arrival of the new Core i5 760 [link to review], a similar but slightly faster quad-core model, and several new dual-core, quad-thread Clarkdale chips such as the Intel Core i5 661.

The Core i5 750 has also soaked up an onslaught from AMD in the form of both quad-core chips such as the AMD Phenom II X4 965 Black Edition and the latest six-core AMD Phenom X6 1055T. Somehow, however, this plucky little quad remains relevant, especially if gaming grunt is high on your list of priorities.

At its stock 2.66GHz clockspeed, the Intel Core i5 750 puts on a decent show. Thanks to four cores, 8MB of cache and a dual-channel memory controller, there’s plenty of multi-threaded muscle for media encoding or image rendering.

In fact, compared to Intel’s new dual-core processors, such as the Core i5 661 and Core i5 665K, the Core i5 750 is an absolute bargain. We’d also take it over any of AMD’s Phenom II X4 chips. That includes the excellent Phenom II X4 965 BE in its most recent C3 stepping.

However, inject the AMD Phenom II X6 1055T into the equation and things get a little more complicated. For just one pound more than the 750, the 1055T gives you six cores and a lot more grunt than the Core i5 750 in highly threaded applications such as video encoding.

The arrival of the new Intel Core i5 760 hardly makes life any easier, either. After all, the 760 is essentially the same chip clocked 133MHz higher. But it only costs an extra fiver. Surely that makes it a no brainer?

Not necessarily. In our overclocking tests, the Core i5 750 actually has the edge over the new 760 model. Of course, mileages vary when it comes to overclocking results. Pick another pair of chips off Intel’s production line and the results could very well be reversed. But the main point to absorb is that the newer, slightly pricier chip isn’t necessarily faster in extremis.

What’s more, both at stock clocks and overclocked, the Core i5 750 beats seven shades out of the six-core AMD Phenom II X6 1055T in software that benefits more from the strength of individual cores rather than the sheer number of them – think games or file decompression.

We liked:

In an age where core counts and fancy features grab all the headlines, the Core i5 750 proves that four cores and a solid underlying CPU architecture still gets the job done. As a gaming chip, it remains one of our favourites. It’s hardly a slouch when it comes to media encoding, either.

We disliked:

If content creation and other highly threaded applications are your bag, there’s no doubting AMD’s Phenom II X6 1055T is the weapon of choice at this price point. Likewise, the arrival of the new Intel Core i5 760 has left the 750 looking a little redundant.

Verdict:

A great gaming chip that’s just beginning to show its age.

Follow TechRadar Reviews on Twitter: http://twitter.com/techradarreview

Related Links

• TechRadar Reviews Guarantee
• Read more processor reviews

View full post on TechRadar: All PC & Mac feeds

Intel today announced a $1.4 billion acquisition of Infineon Technologies AG’s wireless unit. Infineon’s Wireless Solutions (WLS) business provides baseband processors, radio-frequency transceivers, and power management integrated circuits (ICs) for wireless devices.

It comes less than two weeks after Intel announced it would buy McAfee for $7.68 billion. The Infineon buy, however, is a hardware deal. Infineon’s unit makes chips for Apple’s iPhone and Samsung’s Android-based Galaxy S phones.

Infineon is Europe’s second largest maker of semiconductors. The unit, officially known as Infineon’s Wireless Solutions Business, will go to Intel in an all-cash deal, expected to close in the first quarter of 2011.

According to a statement from Intel, it will operate as a stand-alone business and its technology will be used in Intel’s Core processor-based laptops and various Atom processor-based devices, including smartphones and netbooks.

“As more devices compute and connect to the Internet, we are committed to positioning Intel to take advantage of the growth potential in every computing segment, from laptops to handhelds and beyond,” Otellini added.

View full post on dailywireless.org