AMD has managed to rehire Jim Keller from Apple. Keller was the lead architect of the K8 architecture and of the x86-64 specification, he will leave his post as a director in the platform architecture group at Apple to take up a position as AMD Corporate Vice President and Chief Architect for CPU cores.

AMD announced today that Jim Keller, 53, has joined the company as corporate vice president and chief architect of AMD's microprocessor cores, reporting to chief technology officer and senior vice president of technology and engineering Mark Papermaster. In this role, Keller will lead AMD's microprocessor core design efforts aligned with AMD's ambidextrous strategy with a focus on developing both high-performance and low-power processor cores that will be the foundation of AMD's future products.

Intel's Core i7-3970X Extreme Edition processor will arrive in the final quarter of this year. This chip will feature a 3.50GHz clockspeed, a 4GHz Turbo, and 15MB L3 cache.

The six-core chip is based on the all too familiar 32 nm "Sandy Bridge-E" silicon, and built in the LGA2011 package. Its feature-set is consistent with that of the Core i7-3960X, with 15 MB shared L3 cache, HyperThreading, and unlocked base-clock multiplier.

Intel announced a second-quarter profit of $2.8 billion, or 54 cents per share, on revenue of $13.5 billion. For the full year 2012, Intel now anticipates revenue will be between 3 to 5 percent higher than 2011, down from previous expectation for high single-digit growth.
Intel Corporation today reported quarterly revenue of $13.5 billion, operating income of $3.8 billion, net income of $2.8 billion and EPS of $0.54. The company generated approximately $4.7 billion in cash from operations, paid dividends of $1.1 billion and used $1.1 billion to repurchase stock.

Business Outlook
Intel's Business Outlook does not include the potential impact of any business combinations, asset acquisitions, divestitures or other investments that may be completed after July 17.

Q3 2012 (GAAP, unless otherwise stated)

Revenue: $14.3 billion, plus or minus $500 million.

• Gross margin percentage: 63 percent and 64 percent Non-GAAP (excluding amortization of acquisition-related intangibles), both plus or minus a couple of percentage points.
• R&D plus MG&A spending: approximately $4.6 billion.
• Amortization of acquisition-related intangibles: approximately $80 million.
• Impact of equity investments and interest and other: approximately zero.
• Depreciation: approximately $1.6 billion.

Full-Year 2012 (GAAP, unless otherwise stated)

• Revenue up between 3 percent and 5 percent year over year, down from the prior expectation for high single-digit growth.
• Gross margin percentage: 64 percent and 65 percent Non-GAAP (excluding amortization of acquisition-related intangibles), both plus or minus a couple points.
• Spending (R&D plus MG&A): $18.2 billion, plus or minus $200 million, down $100 million from prior expectations.
• Amortization of acquisition-related intangibles: approximately $300 million, unchanged.
• Depreciation: $6.3 billion, plus or minus $100 million, down $100 million from prior expectations.
• Tax Rate: approximately 28 percent, unchanged.
• Full-year capital spending: $12.5 billion, plus or minus $400 million, unchanged.

Intel's Business Outlook is posted on intc.com and may be reiterated in public or private meetings with investors and others. The Business Outlook will be effective through the close of business Sept. 14 unless earlier updated; except that the Business Outlook for amortization of acquisition-related intangibles, impact of equity investments and interest and other, and tax rate, will be effective only through the close of business on July 24. Intel's Quiet Period will start from the close of business on Sept. 14 until publication of the company's third-quarter earnings release, scheduled for Oct. 16. During the Quiet Period, all of the Business Outlook and other forward-looking statements disclosed in the company's news releases and filings with the SEC should be considered as historical, speaking as of prior to the Quiet Period only and not subject to an update by the company.

Intel's latest roadmaps show that the next-gen Valleyview Atom SoC and the corresponding BayTrail platform are delayed until the end of 2013, and won't see widespread availability in products until 2014.

On paper, the whole platform looks promising. We have Bay Trail-T and Valleyview-T for mobile phones (until then, you're stuck with Oak Trail-based Atom Z6xx processors) which will support LPDDR3 memory from Day 1. Valleyview-T will be available in quad-core format as well, which means that by Q4 2013 both Qualcomm and TI need to have their processors candy-dandy, given the lead from HiSilicon, NVIDIA and Samsung.

The only problem is that taking a look at competitor's roadmaps, they seem to be more than ready for continuously-delayed "Intel chips that will change the world". Ultimately, Intel plans to utilize its 22nm fabs to the full extent, with the future migration to 14nm, but the mobile platform is looking to lag one process node behind the traditional desktop/notebook/server processors. The official line is that the company still prepares 22nm mobile node, but that node slipped deep into the 2013 and as we all know, you can buy 22nm processors from Intel

DigiTimes learned that AMD delayed its desktop Trinity APUs from August to October in order to make some adjustments to the processors' designs.

Downstream motherboard makers are also rushing to make adjustments to their new motherboard projects set to release in the second half, according to sources from motherboard makers. In addition to production and design issues with the Trinity processors, AMD's previous-generation Llano processor's high inventory levels which need to be digested are reportedly one of the reasons that AMD decided to delay the launch of Trinity. However, AMD did not confirm the speculation. Many Socket FM2-based processors including A10-5800k will be delayed.

In other news the site also mentiones that three new AM3+ Vishera chips are expected in October -- In October, AMD also plans to launch three new AM3+ Vishera-based processors (Volan) including FX-8350 (125W/95W) and FX6300, with FX-4320 to be launched at a later time.

AMD is informing its partners that some of their 45nm processors have reached the end of the life, that includes five Phenom II X4 and X2 processors. The EOL date is December 2012.

The AMD 3.4 GHz Phenom II X4 965 will be EOL'ed even sooner at September 2012 followed by Phenom II X4 955 with and Phenom II X2 parts in December 2012. AMD is clearly moving away from, the SOI stack and has moved onwards tol its 32nm products.

X-bit Labs explains that that Intel slashed the pricing of some of its ultra low-voltage processors by just over 10 percent to spur adoption of ultrabooks:

All-in-all, Intel has reportedly decided to reduce the prices of its ultra low-voltage Core i3 "Ivy Bridge" central processing units in order to make ultrabooks more affordable without further cutting ODMs' profitability or using cheap cases that may eventually crack and returned to stores, which will damage reputation of ultrabooks and PC suppliers. The reduction on Core i3 ULV chips will be in $25 - $27 range, said Cody Acree, an analyst with Williams Financial Group, citing Chinese-language Commercial Times news-paper.

At present Intel has only one mobile Core i3 chip with 17W thermal design power that is suitable for ultrabooks: the model i3-3217U with two cores, 1.80GHz clock-speed, 3MB cache and Intel HD 4000 graphics core. Right now Intel sells the chip for $225, hence the $25 - $27 slash will be over 11%, which seems to be a rather significant, yet not a dramatic price-cut that changes everything.

VR Zone delivers news about Intel's 2013-2014 Haswell-EP platform. This 22nm server platform may be the first to use DDR4 memory. The first and most important portion of a leaked slide reveals the arrangement of the various key components in the platform. It reveals an arrangement that is not much different from today's Sandy Bridge-EP platforms, in which n-number of CPU sockets are wired to each other using fast QuickPath Interconnect (QPI). Intel will use dual-channel QPI, which doubles bandwidth over what is available with today's enterprise platforms. Assuming each channel is clocked at 6.4 GT/s, the cumulative bandwidth would amount to 51.2 GB/s.

Each socket will have four DDR4 DRAM channels. The DDR4 DRAM specification will introduce a fundamental change in the topology of DRAM components, which will be arranged "point-to-point". Each DRAM "channel" from the memory controller will support just one DRAM module, but there will be greater scope for DRAM makers to scale up densities of the modules with advancements in technology and silicon fabrication process. Much like Sandy Bridge-EP, Haswell-EP will see a bulk of the platform's PCI-Express lane budget being care of the processors' system agents; and will require external chipset for peripheral connectivity. 

AMD is working on three new FX series chips; the 8350, 6300 and 4320. The FX 8350 is a new eight-core model with unlocked multiplier and 16MB cache, clockspeeds are still unknown but they should be higher than the FX 8150. This part is coming in Q3 2012 as fud reports today:

The FX 6300 is a six-core, basically a faster iteration of the 3.8GHz clocked FX 6200 and it

AMD has introduced a cashback programme for APU and FX processors. It's for several  EU countries only. The cashback ranges from 10 to 20 euros, depending on the number of cores in your processor.

High-end AMD Llano APUs and the Guinness World Record breaking FX Processors, with unrestricted power and unbeatable pricing, are now available at even better value. AMD has launched a cashback programme on the A8 and A6 APUs and also a range of FX CPUs including the FX 8120 and FX 8150.

The eligible products available in the

AMD senior vice president and chief technology officer Mark Papermaster revealed the firm will fully switch from the existing SOI manufacturing process to 28nm bulk CMOS:

As for GPU manufacturing, AMD does not plan to make any changes. The current Southern Islands series already adopt Taiwan Semiconductor Manufacturing Company's (TSMC's) 28nm process, while its upcoming Sea Islands series will continue to adopt the same process. The Sea Islands series GPUs have already entered tape-out and should start manufacturing at the end of 2012 and will be announced in the first quarter of 2013.

Commenting about whether the Heterogeneous System Architecture (HSA) Foundation was formed to compete against Intel and Nvidia, Papermaster said that the company's cooperation with ARM is mainly to satisfy client demand for comprehensive functions and to allow quick product development, and is not founded to target any specific competitor.

Intel Corp reiterated Thursday that the entry-level model of its new Ivy Bridge processors will begin shipping as soon as this month, following media reports that the products are scheduled for delivery in August.

"The Intel 3rd-generation Core i3 processor will be available after June 24," Intel's Taiwan branch said in a statement.

The company also reiterated its earlier announcement made a year ago at Computex 2011 that it planned to shift 40 percent of the consumer notebook market to Ultrabooks by the end of 2012, saying that this goal has not changed.

The clarification came a day after local media reports said that the shipping date of the faster Intel Core i5 and i7 processors had been delayed from May to June 3, while the entry-level Core i3 processors are scheduled for delivery in August.

Taiwan's Acer Inc. said it has revised its forecast for the penetration rate of the company's Ultrabooks because of the one-month delay in the Core i5 and i7 processors, while the global economic prospects remain gloomy, according to the reports.

AMD announced that Kaveri, its third-generation APU, is set to deliver a computing performance of 1 teraflops. Kaveri is expected to be released in 2013. At the AMD Fusion Developer Summit, AMD stated that the Kaveri APU will have at least 1 TFLOPS of total compute performance. The Trinity APU that was just released is currently rated at 726 GFLOPS.

This is around a 38% performance improvement, which shows the heterogeneous architecture AMD has adopted is headed in the right direction. Kaveri will be the first APU with true shared, unified address space between the CPU and GPU.

AMD's second generation of APU's called Trinirty where being showcased on Computex show floor but it looks like its availability has been delayed. APU

AMD revealed at its Fusion Developer Summit that Trinity's successor, Kaveri, will have true unified memory between the CPU and the GPU as VR Zone reports:

After the show, a source confirmed that Kaveri will be a true Southern Islands GCN part in the GPU department, without any remnants of the HD6XXX VLIW-4 architecture, and that the Steamroller CPU core parts there would solve a couple of remaining major drawbacks of the Piledriver architecture.

Whether it includes - finally - each core getting its own FPU and instruction scheduler, or something else like common L3 cache for CPU and GPU, non existent till now, was not answered. The point is, Intel's Haswell desktop with supposedly vastly improved graphics (GT3, 40EUs and 64MB cache) - but likely no such high level enhancement like true shared memory - will be out at the same time. So, getting out Kaveri as early as possible should be an imperative for the AMD APU time.

Whether it includes - finally - each core getting its own FPU and instruction scheduler, or something else like common L3 cache for CPU and GPU, non existent till now, was not answered. The point is, Intel's Haswell desktop with supposedly vastly improved graphics (GT3, 40EUs and 64MB cache) - but likely no such high level enhancement like true shared memory - will be out at the same time. So, getting out Kaveri as early as possible should be an imperative for the AMD APU time. And, yes, since the sockets are changed everytime, how about quad memory channel option this time, as DDR4 support is too early then, yet more bandwidth is needed for the GPU at least?

Details about AMD's upcoming Trinity APUs for desktop PCs have surfaced at tech report, they feature Piledriver CPU cores, updated graphics and higher clockspeeds. The site says the first All-in-One PCs with Trinity will arrive this month.
All the CPUs feature quad cores (two modules), 4MB of L2 cache, and support for DDR3 memory speeds up to 1866MHz. The chips slip into AMD's FM2 socket, which means new motherboards will be required. Overclockers will want to pay attention to the K-series models, which feature unlocked multipliers.

While the A10-5800K's 4.2GHz Boost clock is impressive, that probably won't be enough to match the CPU performance of Intel's dual-core Ivy Bridge processors. We've already seen the Core i7-3770K dominate the FX-8150, an eight-core Bulldozer part that also peaks at 4.2GHz. It seems unlikely Trinity's per-clock performance improvements will be substantial enough to close the gap.

Sources at notebook players revealed to DigiTimes that Intel's Sandy Bridge processors will remain in the market until September, before being gradually phased out. The site points out that Sandy Bridge chips are still popular because they're still about $67-$100 cheaper than Ivy Bridge CPUs.

With Intel releasing its new dual-core Ivy Bridge processors, the company's Sandy Bridge processors are expected to gradually be phased out of the market starting in September, according to sources from notebook players. The sources pointed out that the phase-out of Sandy Bridge processors will trigger a new wave of PC replacement in September to meet the back-to-school demand for this year with shipments from the supply chain expected to start heating up in June.

In response, Intel said it is already shipping third generation Core processors to OEM customers and expects to see systems in the market beginning from June 24. The company expects to ship over 50% more volume of third generation Intel Core units to the market in the first two quarters of production in 2012 as compared to the first two quarters of production of Sandy Bridge in 2011.

Word out on the street is that that Intel plans to trade TDP for performance. Over the Ivy Bridge and Haswell generations, the chip giant is anticipated to roll out high-end workstation parts with a TDP of 150W and possibly even as high as 180W, reports vr-zone:

So, the sources pointed out to us that, over the Ivy Bridge and Haswell generations, there will be the high end CPU TDP increase across the board to 150W and then 165W - and here we're talking also about generic high end MP servers with 4 sockets or more, including the 'E7 v2 (Ivy Bridge-EX)' and 'E7 v3 (Haswell-EX)' generations. This is a noticeable jump over 135 watts allowed up to now, and could signal even more TDP for the workstation-specific 'W' parts.

For instance, the general expectations - at the same TDP level - is that Intel would provide roughly 40% performance increase when going from Sandy Bridge EP to Ivy Bridge EP Xeon E5 series. The first 25% of that would obviously come by jumping from 8 to 10 cores with the associated cache increase to 25 MB L3. The next 10% on top of that 25% (i.e. extra 12% from the base line) would come from the frequency increase - for instance, the successor to the 2.9 GHz Xeon E5-2690 would be a 3.2 GHz 10-core part in the same 135W TDP bracket. The remaining few % is, of course, minor CPU enhancements and DDR3-1866 server memory support.

An overclocker from Taiwan by the name of HiCookie has overclocked Intel Core i7 3770K E1 (Ivy Bridge) past 7 GHz. He used Gigabyte GA-Z77X-UD3H motherboard, Corsair memory, Corsair AX 1200 power supply and of course liquid nitrogen to achieve this. 3 cores were disabled while HT was enabled, FSB was 111.63 MHz, 63x multiplier was used and the core voltage was a whopping 1.956V. Before this, the record was held by nkdfactory from Japan who achieved the highest frequency of 6931.1 MHz.

The AMD processors can run high clock frequencies, this one it even beyond that, 8805.64 MHz, to be precise. 

A Thai overclocker under the name "ksin" took dibs and was to able to run the chip using a base clock speed of 303.29 MHz, and multiplier of 29.0x. A scorching 1.86V core voltage was used.

On the motherboards side of things an ASUS Crosshair V Formula was used with 4 GB (2x 2 GB) of ADATA memory (1417 MHz). No details such as cooling has been shared, but you can betcha it was sub-zero.