Intel Core i7-3820 Overclocked to 5.666 GHz

GUN'G'STAR of TeamRussia successfully achieved an overclocking record of 5.666 GHz for the quad-core Intel Core i7-3820 processor; a feat that validates the LGA2011 i7-3820 of being a generally good chip for overclocking, despite being "Limited Unlocked" (BClk multiplier being unlocked only to a few notches above its stock setting). GUN'G'STAR achieved a clock speed of 5,665.99 MHz using a BClk speed of 131.74 MHz, a multiplier value of 43x, and core voltage of 1.6V. The chip was assisted with GeIL-made DDR3-1600 memory sitting on all four channels, and ASUS Rampage IV Extreme motherboard. Only 2 out of 4 cores of the i7-3820 were enabled for the feat, HyperThreading was disabled. Extreme cooling was used. The CPU-Z validation can be accessed here.

Source: Expreview

Mass Shipments of Ivy Bridge Delayed to June

In what could come as a disappointment to some, Intel has reportedly put off mass-shipments of its third-generation Core processors (codenamed "Ivy Bridge") to June. The company, however, will go ahead with its April launch schedule, and will issue small shipments of these chips. This information comes from sources among notebook manufacturers.

This delay, however, doesn't appear to be linked to Intel's ability to ship the new 22 nm chips. Sources say that most notebooks vendors are having a very hard time digesting inventories of current-generation "Sandy Bridge" Core processors, and so the delay is merely to help them ship the last of their notebooks running current-generation chips, and to minimize the impact of the onset of new chips. Notebook vendors are currently seeing lukewarm demand.

Source: DigiTimes, Image Courtesy: LegitReviews

Ivy Bridge Desktop Core i3 Processor Lineup Detailed, Lack PCIe Gen. 3.0

Details of desktop Core i7 and Core i5 "Ivy Bridge" processors in the LGA1155 package have been detailed at lengths, in the past. Core i3 parts based on the same 22 nm Ivy Bridge silicon, however, were relatively known. Tables listing out updated information about the lineup points out that Intel has as many as five Core i3 "Ivy Bridge" desktop processors in the works, all dual-core, and among which two are low-power parts.

The table also suggests that these Core i3 chips will have reduced features, importantly, the lack of PCI-Express 3.0 bus. When connected to these chips, PCI-E 3.0 add-on cards (such as graphics cards) will function in PCI-Express 2.0 mode. Further, these chips will lack support for AES-NI (accelerates encryption), VT-d (enhanced virtualization), and TXT (security). Certain models in the lineup have faster integrated graphics, denoted by a "5" in the end of the model number. These chips also lack Turbo Boost for the x86 cores, but feature HyperThreading.

Leading the pack is Core i3-3240, clocked at 3.40 GHz, with integrated graphics clocked at 650 MHz / 1020 MHz Dynamic. This is followed by the Core i3-3225, which is clocked at 3.30 GHz, packing faster Intel HD 4000 graphics. The Core i3-3220 is likely the cheapest of the lot, also clocked at 3.30 GHz, but with slower HD 2500 graphics. These three chips have rated TDP of 55W. Moving on to the low-power models, which have 35W TDP, we have the Core i3-3240T, with 3.00 GHz clock speed, and Core i3-3220T, with 2.80 GHz clock speed.

Sources: Zol.com.cn, VR-Zone

Intel Haswell Packs DirectX 11.1 Graphics

Intel will launch its new processor architecture, codenamed "Haswell", which will go on to succeed "Ivy Bridge". More than an year away from its market entry, Haswell has already been exhaustively documented, but not many got into the details about its embedded graphics processor. That is, until now. A new internal slide sourced by DonanimHaber details the integrated GPU (iGPU), it appears like Intel has solid plans for home users.

To begin with, Haswell's iGPU will be DirectX 11.1 compliant, which means it will take advantage of API optimizations that improve performance, for typical desktop usage scenarios. Apart from support for a new DirectCompute architecture, it will also support OpenCL 1.2, which speeds up certain GPGPU-optimized applications. More importantly, the iGPU will be designed around a new stereoscopic 3D standard called Auto-Stereoscopic 3D (AS3D), which will take the likes of Blu-ray 3D acceleration, stereo 3D photos, etc., to the masses. Currently, it takes at least an entry-level GeForce or Radeon GPU to for acceptable performance with stereo 3D.

Another significant change is what Intel refers to as "digital display repartition". Until now, Intel processors with embedded graphics relied on the PCH to perform all display I/O functions, the iGPU communicates to the PCH over the Flexible Display Interface (FDI), a special data link dedicated to this function. With Haswell, digital display outputs (such as DVI, HDMI, DisplayPort), will be wired directly to the CPU socket, while analog display (such as D-Sub), will be handled by a RAMDAC located in the PCH, to which a digital signal is sent over FDI. This design ensures there's no performance bottleneck with higher resolution digital displays, while at the same time, ensuring analog display quality isn't affected by the RAMDAC being located in the GPU die, where it's bound for interference by some pretty high-current circuits.

Source: DonanimHaber

AMD "Trinity" APU Models Further Detailed

In the second quarter of this year, AMD will launch its next-generation accelerated processing unit, codenamed "Trinity", starting with quad-core A10 and A8 parts in the quarter, followed by dual-core A6 and A4 parts in the third-quarter of 2012 (more here). DonanimHaber scored a specifications-sheet that details all the models AMD has in store for this year, also giving us an insight to what constitute each of these model numbers.

An interesting revelation here is that AMD does not have triple-core APUs anywhere in its lineup. The A10 lineup consists of quad-core APUs that come with Radeon HD 7660D graphics, while the A8 lineup also consists of quad-core APUs, but with slower Radeon HD 7560D graphics. The A6 lineup consists of dual-core APUs with Radeon HD 7540D graphics, while the A4 lineup has dual-core APUs with Radeon HD 7480D graphics.

The HD 7660D has 384 Graphics CoreNext stream processors, HD 7560D has 256 of these, the HD 7540D has 192, and the HD 7480D has 128. The HD 7480D also lacks dual-graphics support that lets you pair the APU-embedded graphics processor with a compatible discrete GPU to make the two work in tandem. While these GPUs have comparatively lower stream processors than previous-generation "Llano", performance increase is made up for by the more advanced GPU architecture, and higher GPU core speeds.

Leading the chart is the A10-5800K, designed for gamers and overclockers on a tight budget. This chip packs 4 "Piledriver" architecture cores (based in two modules), with 3.80 GHz (stock) and 4.20 GHz (TurboCore) frequency, 4 MB total cache, and Radeon HD 7660D graphics. This chip bears the Black Edition branding, and hence features an unlocked base clock multiplier that helps with overclocking. This is followed by the non-Black Edition A10-5700 clocked at 3.40 GHz with 4.00 GHz TurboCore.

The A10-5800K is not the only Black Edition part, there's also the A8-5600K, which bears clock speeds of 3.60 GHz (stock), 3.90 (TurboCore) and A6-5400K, both of which feature the unlocked base clock multiplier; Just below the A8-5600K is the A8-5500, with 3.20 GHz (stock), 3.70 GHz (TurboCore). The cheapest of the lot will be the A4-5300. The clock speeds of the A6 and A4 parts were not revealed.

Source: DonanimHaber

Core i7-3820 Sandy Bridge-E Starts Selling in Japan

Intel's most affordable processor in the LGA2011 package, the Core i7-3820, surfaced on Japanese stores in the Akihabara electronics shopping district of Tokyo. It is priced around 25,000 JPY (US $322). The retail box of this chip, carrying the S-Spec code "SR0LD", appears to be as big as those of the Core i7-3930K and Core i7-3960X Extreme Edition, its shape indicates that it lacks a bundled cooling solution, and so users should still rely on compatible third-party coolers, or use Intel's RTS2011LC, purchased separately.

The Core i7-3820 is a quad-core part carved out of the Sandy Bridge-E silicon. It has four cores, and eight logical CPUs enabled with HyperThreading Technology. The chip is clocked at 3.60 GHz. Caches include 256 KB L2 per core, and 10 MB shared L3. The chip retains the quad-channel DDR3 IMC present on the more expensive six-core parts. It is also said to be "limited unlocked", meaning it allows multiplier-assisted overclocking to a certain limit. The Core i7-3820 is not officially launched, though it should be unofficially supported by most socket LGA2011 motherboards based on the X79 chipset.

Source: PCWatch Akiba

Intel Haswell In Bound for March-April 2013

As Intel's tick-tock CPU development Juggernaut rolls on, things seem very much on track, looking into the near future. Intel will launch its new "Ivy Bridge" 3rd Generation Core processor family in early-April 2012, which is a miniaturization of what is essentially the "Sandy Bridge" to the new 22 nm process, with IPC and instruction-set improvements, along with a faster graphics controller. The new process will also up clock speeds and overclocking headroom for chips that support it. What's more interesting, though, is that the architecture that succeeds Ivy Bridge, codenamed "Haswell", will be less than an year away in April...well almost.

A roadmap slide sourced by DonanimHaber pins the launch of Haswell to March-April, 2013. Haswell is a brand new CPU architecture that will succeed Ivy Bridge. According to the conventional idea of Intel's tick-tock CPU development strategy, it will be built on the 22 nm fab process, which will have gained some maturity by then. Intel follows a "tick-tock" product development model. Every year, Intel's product lineup sees either of the two. A "tock" brings in a new x86 architecture, a "tick" miniaturizes it to a newer silicon fabrication process. Earlier reports indicated that Haswell Core processors will be based on a newer socket, the LGA1150, and hence it will not be compatible with LGA1155 platforms.

Source: DonanimHaber

Fake Core i7-990X Surfaces on Forums

A scamster has managed to pull off a fake "Intel Core i7-990X" processor. An unsuspecting buyer paid hundreds of dollars for a Core i7-990X six-core LGA1366 processor, only to end up with a badly-done fake. The fake chip is a worthless LGA775 Pentium, on which the IHS markings of a Core i7-990X have been etched. Intel and AMD put tiny windows on their retail processor boxes so buyers could see these markings before breaking open the company seal. At least in the case of Intel's product boxes, the windows aren't big enough to let you see the entire CPU package. In case of this chip, the buyer couldn't have spotted that the chip was an LGA775 (and not LGA1366), and couldn't spot the 3D hologram and serial number sticker that's found on the obverse side of the package, under the IHS. The thread where this fake was reported can be read here.

Source: CPU World

10-core Ivy Bridge-EP Sample Tested

The Ivy Bridge LGA1155 processors inbound for April are mom and pop PC chips in front of the monstrosities Intel has planned for the enterprise (and possibly high-end desktop/HEDT) markets, based on the architecture. An 10-core Ivy Bridge-EP engineering sample, made it to the right hands in Taiwan (wrong hands for Intel), that wasted no time in putting them through some tests.

The 10-core Ivy Bridge-EP/EX chip features 10 next-generation cores clocked at 2.80 GHz, with 256 KB L2 cache per core, 30 MB shared L3 cache, and HyperThreading technology that enables 20 logical CPUs. This chip crunched WPrime 1024M in 158.5 seconds, and scores 41.78X relative speed in Fritz chess when just 8 of its 20 threads are put to use. You can also find some pretty screen shots of CPU-Z with its long processor selection list and Windows 8 task manager.

Sources: Coolaler, ComputerBase.de

AMD A8-3870K Cracks 6 GHz with All Cores Enabled

Overclocker "Christian Ney" of the HWBOT community claims to have cracked the 6 GHz clock speed mark with an AMD A8-3870K APU, with all four of its cores enabled. The feat consisted of a clock speed of 6067.7 MHz achieved using a base clock of 258.2 MHz and multiplier of 47.0x. The memory was clocked at 3505 MHz, just 95 MHz short of a world-record. The chip was put through a scorching core voltage of 2V. Supporting components included GIGABYTE A75-UD4H motherboard, and GeIL EVO Corsa DDR3 memory. The setup was cooled sub-zero. CPU-Z Validator rejected this submission, probably because like us, it finds 2V vCore and 258.2 MHz base clock unreal for Llano, since it uses a common 100 MHz clock domain for BClk and PCIe.

Source: MyDrivers.com

Intel Releases Seven More 32 nm Processors

US chip giant Intel has now added seven new processors to its portfolio, the Core i5-2550K, i5-2450P and i5-2380P targeting desktops, and the Celeron B815, B720, 867 and 797 for mobile PCs.

The Core i5-2550K costs $225 ($9 more than the i5-2500K) and features four cores @ 3.4 GHz (the 2500K has a base clock of 3.3 GHz), four threads, 6 MB of L3 cache and a TDP of 95W. The i5-2450P ($195) and i5-2380P ($177) have similar specs, expect their base clock is 3.2 GHz and 3.1 GHz, respectively.


The Celeron B815 has two cores @ 1.6 GHz, two threads, 2 MB of L3 cache and bears a price tag of $86, while the B720, which costs $70, has one core clocked at 1.7 GHz, one thread and 1 MB of L3 cache.

The ULV (ultra-low voltage) Celeron 867 and 797 cost $134 and $107, respectively. The former packs two cores @ 1.3 GHz, two threads, 2 MB L3 cache, whereas the latter has just one core @ 1.4 GHz, one thread, and 1 MB of L3 cache.

Source

AMD FX-8150 Tested with Latest Windows Hotfixes, Still No Improvement

German tech website TweakPC did a before-after comparison of applying Microsoft's recently-released KB2645594 + KB2646060 Windows updates, which intend to improve performance of systems running AMD FX processors, by improving the way in which the OS deals with Bulldozer cores, using a top-of-the-line FX-8150 processor. The reviewer put FX-8150 through synthetic tests such as AIDA64 (CPU benchmarks, FPU benchmarks), Cinebench 11.5, MaxxPi (multi-threaded PI calculations), WPrime, Twofish AES, 3DMark (Vantage and 11), ComputeMark; and some real-world tests such as WinRAR, Resident Evil 5, and Battleforge. Barring Resident Evil 5, where the patched FX-8150 produced 4% higher performance and WinRAR, where it produced 3% higher performance, there were no significant performance gains noticed. The review can be accessed at the source.

Source: TweakPC.de

Dishonest Intel OEM Engineers To Get Five Years Behind Bars

The Taiwan Criminal Investigation Bureau has been doing a bit of investigating and has arrested four engineers working for Intel's OEM partners (names not disclosed) for flogging Engineering Sample (ES) processors on eBay. ES processors are intended strictly for qualification testing purposes for development of new products by OEM's and are only loaned to them under strict non-disclosure agreements, hence putting these on eBay is illegal. The Bureau searched the suspects houses last month and found 178 ES CPU's, worth around $800,000. Note that this value seems to be somewhat high, as it would make each CPU worth around $4,500. We will update this article if new values come to light. Also, this is not a new operation that has been busted, since the suspects had admitted to selling around 500 ES CPU's since 2009. For their efforts, the fab four now face five years in prison.

It should be noted for anyone contemplating the purchase of such dodgy CPU's on eBay or similar places, that they may be overstressed and contain faults, due to the intensive and sometimes destructive testing they go through. Just don't do it, kids.

Source: TechEYE

Intel to Roll Out 40 New Server Processors in H1 2012

With key enterprise architectures such as Sandy Bridge-EP awaiting launch, Intel has as many as 40 Xeon processor models to release in the first half of 2012, which it will likely launch in two rounds of 20 models, for each of the first two quarters of this year. In Q1, Intel will release six-core Xeon E5-1660 and E5-1650, and a yet to be named quad-core model; the three will be priced at US US$1,080, US$583 and US$294, respectively. In the same quarter, Intel will also launch 7 eight-core Xeon processor models, including the Xeon E5-2690four six-core processors including E5-2640, three quad-core processors including E5-2609 and dual-core processor E5-2637. Intel will also launch some low-power (energy-efficient) Xeon models, including Xeon E5-2450L and six-core E5-2430L priced at US$1,106 and US$662.

In the second quarter of 2012, Intel will launch the other 20 models. These include Xeon processors based on the Ivy Bridge silicon, and in the socket H2 (LGA1155) package. The Xeon E3-1290v2 is an example of such a chip. These high-grade Ivy Bridge chips will be priced in the range of US$189-884. Then on, 7 more E5 series models will be launched, including Xeon E5-2470, priced between US$192-1,440. Towards the end of the quarter, Intel will launch low-power eight-core and six-core Xeon processor models, including Xeon E5-2650L and E5-2630L, priced at priced at US$1,106 and US$662, respectively.

Source: DigiTimes

Intel Plans to Launch Medfield Platform in Q2, Clover Trail-W platform in Q4, 2012

According to the latest information received by industry observer DigiTimes, Intel plans to launch the first processor platform for ultra-thin Android tablets based on its x86 architecture, codenamed "Medfield", in Q2 2012. Tablets based on this will be able to run Android 4.0 "Ice Cream Sandwich". Then in Q4 2012, Intel plans to launch the more powerful Clover Trail-W platform. Currently, Intel's Oak Trail platform consisting of Atom Z670 processor and SM35 chipset drive Windows 7 or Android Honeycomb tablets.

The fundamental difference between Medfield and Clover Trail-W with its predecessor Oak Tail, is that Medfield will be designed for ultra-thin tablets with long batter life, currently only ARM processors provide the kind of performance-per-Watt to achieve this form-factor. Oak Trail and its succeeding Clover Trail-W, are designed for slightly more capable tablets. Oak Trail is opted today, to design tablets that run Windows 7 PC operating system. Microsoft will design performance and UI-optimized Windows 8 variants when its next-gen operating system sees the light of the day next year.

Source: DigiTimes

Intel DZ77RE First and Only to Feature Thunderbolt from Intel's 7-series Board Stable

This week, we got two important bits of news, firstly, that Intel's third-generation Core processor family (codenamed "Ivy Bridge"), is scheduled for launch on April 8, 2012; and secondly, that Intel will use that occasion to launch its new 7-series chipset, and with it, bring its 10 Gbps Thunderbolt interconnect to the PC platform. A road-map slide tabling Intel's own Desktop Board division products based on 7-series chipset, shows that only one motherboard from Intel's stable will feature Thunderbolt, the DZ77RE.

The DZ77RE is likely listed within the Extreme Series, which will be a top-of-the-line product. This is a bit of a contradiction to the idea behind Thunderbolt, and the way Intel's Desktop Board lineup is designed. It typically consists of Extreme Series, geared for gamers and overclockers; Media Series, for content-creation professionals; Classic Series, for office PCs with room for upgrades; and Essential Series, which well, covers the essentials. None of the supposed Media Series boards from Intel's 7-series chipset based Desktop Board lineup is listing Thunderbolt support, when it's the content-creation industry that stands to be the primary user of the interface. Thunderbolt facilitates ultra-high bandwidth for multiple lossless high-resolution video streams which will help in tomorrow's video-editing systems.

Sources: Expreview, DonanimHaber

AMD Announces New A-Series Accelerated Processing Units (APUs)

AMD today updated its A-Series line-up of desktop and notebook Accelerated Processing Units (APUs), further improving its top-performing family of dual- and quad-core APUs. Along with speed and performance improvements, AMD Steady Video update make this unique feature more compelling than ever. For desktop users, AMD extends its overclocking pedigree to the APU; for the first time users can tune both x86 and graphics settings in a single processor for boosted performance.

The updated AMD A-Series APUs combine up to four x86 CPU cores with up to 400 Radeon cores, delivering powerful DirectX 11-capable, discrete-level graphics and dedicated HD video processing on a single chip. These new APUs increase performance and deliver a richer feature set than existing AMD A-series APUs. Plus, only AMD APUs offer AMD Dual Graphics for an up to 144 percent visual performance boost when a select APU is paired with a select AMD Radeon HD 6500 Series graphics card.

The AMD A-Series family of APUs also features AMD Steady Video, designed to stabilize videos during playback – making unsteady, jumpy content look steady and smooth as you watch. On select systems using AMD A-Series APUs, Internet Explorer 9 will include an AMD Steady Video plugin, unlocking one-click control to simplify access to the premium AMD Steady Video feature for video stabilization.

All AMD A-Series processors are powered by AMD VISION Engine Software, a suite of software that provides end-users with regular updates designed to improve system performance and stability, and can add new software enhancements.

New A-Series Desktop APUs launched today:

• A8-3870K: Four CPU cores, 3.0 GHz CPU base (unlocked), 100W TDP, 400 Radeon cores, 600 MHz GPU base (unlocked), 4 MB L2 cache
• A8-3820: Four CPU cores, 2.5 GHz CPU base (2.8 GHz Turbo Core), 65W TDP, 400 Radeon cores, 4 MB L2 cache
• A6-3670K: Four CPU cores, 2.7 GHz CPU base (unlocked), 100W TDP, 320 Radeon cores, 600 MHz GPU base (unlocked), 4 MB L2 cache
• A6-3620: Four CPU cores, 2.2 GHz CPU base (2.5 GHz Turbo Core), 65W TDP, 320 Radeon cores, 4 MB L2 cache
• A4-3420: Two CPU cores, 2.8 GHz CPU base, 65W TDP, 160 Radeon cores, 1 MB L2 cache

New Notebook APUs launched today:

• A8-3550MX: Four CPU cores, 2.0 GHz CPU base (2.7 GHz Turbo Core), 45W TDP, 400 Radeon Cores, 4 MB L2 cache
• A8-3520M: Four CPU cores, 1.6 GHz CPU base (2.5 GHz Turbo Core), 35W TDP, 400 Radeon Cores, 4 MB L2 cache
• A6-3430MX: Four CPU cores, 1.7 GHz CPU base (2.4 GHz Turbo Core), 45W TDP, 320 Radeon Cores, 4 MB L2 cache
• A6-3420M: Four CPU cores, 1.5 GHz CPU base (2.4 GHz Turbo Core), 35W TDP, 320 Radeon Cores, 4 MB L2 cache
• A4-3330MX: Two CPU cores, 2.2 GHz CPU base (2.6 GHz Turbo Core), 45W TDP, 240 Radeon Cores, 2 MB L2 cache
• A4-3320M:Two CPU cores, 2.0 GHz CPU base (2.6 GHz Turbo Core), 35W TDP, 240 Radeon Cores, 2 MB L2 cache
• A4-3305M:Two CPU cores, 1.9 GHz CPU base (2.5 GHz Turbo Core), 35W TDP, 160 Radeon Cores, 1 MB L2 cache
• E2-3000M: Two CPU cores, 1.8 GHz CPU base (2.4 GHz Turbo Core), 35W TDP, 160 Radeon Cores, 1 MB L2 Cache

The new AMD A-Series family of APUs improves the first generation of highly successful and revolutionary desktop and notebook processors, providing an outstanding experience for consumers seeking more responsive multitasking, long battery-life, vivid graphics, lifelike games, lag-free videos and the ultimate multimedia performance.

Desktop APUs in the component channel as well as systems based on the new AMD A-Series APUs will hit the retail market over the next several weeks.

Source

AMD's Own HD 7970 Performance Expectations?

Ahead of every major GPU launch, both NVIDIA and AMD give out a document to reviewers known as Reviewer's Guide, in which both provide guidelines (suggestions, not instructions), to reviewers to ensure new GPUs are given a fair testing. In such documents, the two often also give out their own performance expectations from the GPUs they're launching, in which they compare the new GPUs to either previous-generation GPUs from their own brand, or from the competitors'. Apparently such a performance comparison between the upcoming Radeon HD 7970 and NVIDIA's GeForce GTX 580, probably part of such a document, got leaked to the internet, which 3DCenter.org re-posted. The first picture below, is a blurry screenshot of a graph in which the two GPUs are compared along a variety of tests, at a resolution of 2560 x 1600. A Tweakers.net community member recreated that graph in Excel, using that data (second picture below).

A couple of things here are worth noting. Reviewer guide performance numbers are almost always exaggerated, so if reviewers get performance results lower than 'normal', they find it abnormal, and re-test. It's an established practice both GPUs vendors follow. Next, AMD Radeon GPUs are traditionally good at 2560 x 1600. For that matter, the performance gap between even the Radeon HD 6970 and GeForce GTX 580 narrows a bit at that resolution.

Source: 3DCenter.org

Intel 'Ivy Bridge' Core Desktop Processor Prices Compiled

In 2012, Intel will update its processor lineup up to performance 2 tiers with new models based on the spanking new 22 nm "Ivy Bridge" silicon, which will increase performance over the current generation, and bring some new features to the table. In late November, a list of desktop Core i5/i7 models could be compiled, which were later confirmed on roadmap slides. CPU World compiled retail channel pricing of several of those Core i5/i7 "Ivy Bridge" chips. The prices look to be more or less consistent with current "Sandy Bridge" Core processors, which those chips are bound to replace. The Core i7-3770K, for example, which will replace the Core i7-2600K, will be similarly priced to it.

Source: CPU-World

26 December Launch Date for AMD A8-3870K and A6-3670K Black Edition Unlocked APUs

December thru January looks to be a busy time for AMD. Along with a few new graphics products, AMD will launch new CPUs and APUs. The company has chosen December 26 to launch its A8-3870K and A6-3670K Black Edition "Llano" accelerated processing units (APUs) in the FM1 package, which feature unlocked base clock multipliers for the processor component, making overclocking a breeze. In two of its presentation slides to distributors, it unveiled the swanky new box art of these unlocked APUs.

These chips pack four x86-64 cores based on the "Stars" K10.5 architecture with 1 MB dedicated L2 cache per core, dual-channel DDR3-1866 MHz integrated memory controller, PCI-Express 2.0 root complex, and a "discrete-class" graphics processor that packs 400 VLIW5 stream processors, DirectX 11 support, and the ability to pair with similarly specc'd discrete GPUs. The A8-3870K Black Edition features x86 core clock speed of 3.00 GHz, and Radeon HD 6550D graphics that features all 400 of those stream processors, with 600 MHz GPU clock speed. The A6-3670K Black Edition, on the other hand, has its x86 component clocked at 2.70 GHz, it features Radeon HD 6530D graphics that has 320 out of the 400 stream processors enabled, and a GPU clock of 433 MHz.

AMD presented results of its own tests, where it pitted the two chips against a similarly priced Intel Core i3-2120 dual-core processor. In CPU-centric tests, the dual-core (2 cores, 4 threads) i3-2120 performed more or less on par with the A6-3670K, with only the A8-3870K having a notable lead. It's with GPU-intensive tests where the Radeon-powered APUs stormed past Intel HD Graphics-powered Core i3. Go figure. The A8-3870K Black Edition is expected to be priced at US $135, while the A6-3670K Black Edition is expected to go for $115.

Source: DonanimHaber