intel on Track with Processors Featuring Graphics Core in 2009.

Intel to Roll Out Processors with Graphics Cores This Year

Despite some over-conservative expectations and reports, Intel Corp. is on track to release its central processing units (CPUs) with integrated graphics cores this year, the company said during its most recent conference call with financial analysts.
Answering the question from Hans Mosesmann – an analyst from Raymond James, a financial services holding company – Stacy Smith, chief financial officer of Intel, said that the code-named Havendale processor with integrated graphics core for desktop and mobile computers will be available “in the second half” of this year.
Intel’s Havendale processor is a multi-chip module (MCM) in LGA1160 form-factor containing Nehalem micro-architecture-based dual-core CPU in addition to graphics and memory controller hub (GMCH) that features dual-channel DDR3 memory controller, PCI Express 2.0 x16 interface to connect add-on graphics cards as well as integrated graphics core. It is projected that both chips on the MCM are made using 45nm process technology.
Since Havendale CPU line has memory controller as well as PCI Express interconnection inside, there will be no need for GMCH (or North Bridge) on the mainboard. Instead, the new processors will connect directly to code-named Ibexpeak platform controller hub (PCH) that will carry hard drive controller, wired and wireless network controllers, monitor physical interfaces, PCI controller and other input/output as well as platform-related capabilities.
Earlier some web-sites reported that Intel’s processors featuring integrated graphics cores will only be out in 2010. Still, even despite the fact that Mr. Smith said that the chips will ship for revenue in 2009, it remains to be seen, whether products on their base will emerge this year.
“You can expect volume production of Nehalem into mainstream price points for desktops and notebooks in the second half of this year,” said Paul Otellini, chief executive of Intel.

Atom N280 netbooks coming

Intel is preparing to start shipping Atom N280 + GN40 chipsets to netbook manufacturers.
Asustek, Acer and Gigabyte are expected to introduce netbooks based on the new chips sometime between the second and third quarter of this year. While performance gains are not expected to be spectacular, there should be some improvement due to the faster FSB (667MHz vs. 533MHz) and presumably the GN40 chipset will be less power hungry than the 945GSE Intel has been trying to get rid of. There is no change in CPU frequency (1.66GHz).
The price for the cpu + chipset combination is expected to be $14-$19 higher, with the new set costing $60-$65 (compared to the old 945GSE+Atom270 at $46)

Multi-Core Processors Outpacing Key Business Software

The relentless doubling of cores per microprocessor chip will drive total processor counts in upcoming generations of servers well beyond levels for which key business software has been engineered, a market research firm said Wednesday.
Technologies that will be affected by this evolution include operating systems, middleware, virtualization tools, and other applications, Gartner said. As a result, companies and other organizations will be faced with "difficult decisions, hurried migrations to new versions and performance challenges."
Looking at the specifications for these software products, it is clear that many will be challenged to support the hardware configurations possible today and those that will be accelerating in the future," Gartner analyst Carl Claunch said in a statement. "The impact is akin to putting a Ferrari engine in a go-cart; the power may be there, but design mismatches severely limit the ability to exploit it."
On average, organizations get double the number of processors in each chip generation about every two years, Gartner said. The increase is accomplished through some combination of more cores and more threads per core.
For example, this year's 32-socket, high-end server with eight-core chips in each socket would deliver 256 processors. In two years, with 16 processors per socket expected in the market, the machine swells to 512 processors total. Four years from now, the server would host 1,024 processors.
Gartner said organizations need to pay attention to this evolution because there are limits on the ability of software to make use of all this horsepower.
"Most virtualization software today cannot use all 64 processors, much less the 1,024 of the high-end box, and database software, middleware, and applications all have their own limits on scalability," Claunch said. "There is a real risk that organizations will not be able to use all the processors that are thrust on them in only a few years time."
Gartner points out that software programs have both hard and soft limits on the number of processors they can effectively handle. The hard limit is within the product documentation. The soft limit, however, can only be uncovered through word of mouth or real-world cases. Due to characteristics of the software design, application performance could decline as more processors are added.
"Most virtualization software today cannot use all 64 processors, much less the 1,024 of the high-end box, and database software, middleware, and applications all have their own limits on scalability," Claunch said. "There is a real risk that organizations will not be able to use all the processors that are thrust on them in only a few years time."

Intel Denies Trying To Derail AMD Spin Off

Intel (NSDQ: INTC) on Friday denied claims from rival Advanced Micro Devices (NYSE: AMD) that the larger rival is trying to derail AMD's plans to spin off its manufacturing operations.
The two companies started trading barbs Thursday when AMD sent an e-mail pointing news organizations to a filing with the Securities and Exchange Commission that disclosed AMD had received a formal letter from Intel requesting a meeting on the current cross-licensing agreement between the rivals.
Intel contends that the separate manufacturing company AMD plans to form with the Advanced Technology Investment Co., formed by the Abu Dhabi government, would not be covered under the agreement, and the two companies need to enter negotiations to come up with a separate deal. AMD, on the other hand, believes the spin-off qualifies as a subsidiary, which would be covered under the agreement.
"We believe that the letter is another attempt by our competitor to cause uncertainty as we approach our Asset Smart deal closing next month," AMD spokesman Drew Prairie said in the e-mail. Asset Smart is the marketing term AMD uses for its spin-off strategy.
Intel on Friday denied it was trying to stop AMD's plans with ATI. The letter was meant as a formal request for a meeting, since AMD has ignored Intel's other requests, an Intel spokesman told InformationWeek.
"Intel has not and does not intend to block or any other way hinder the formation of the Foundry Co.," the Intel spokesman said. "What we're concerned about is the licensing rights."
He insisted that Intel is looking to settle the disagreement. "We certainly would be open to discussion and finding a way to resolve this," he said.
The Foundry Co. is the temporary name of the business AMD plans to open with ATI next month. The manufacturing company would handle AMD's processor production, as well as that of other chip-design firms. AMD would have a 34.2% stake in the company with ATI controlling the rest. AMD would receive $700 million from ATI in the deal.
AMD sent the e-mail the same day it reported a $1.4 billion loss in the fourth quarter and a 33% drop in revenue, which the chipmaker blamed on the slump in the PC market brought on by the global economic downturn. While AMD's loss was large, it was narrower than the same quarter a year ago.
Intel last week reported that fourth-quarter profits plunged 90% and revenue fell 23% as a result of the economic slowdown.

AMD Rounds Out 45-Nm Opteron Line

Advanced Micro Devices (NYSE: AMD) on Monday launched low-power and high-end 45-nanometer Opteron server processors, rounding out the product line that previously offered only mainstream chips.
The latest chips, all quad-core, cover two-, four-, and eight-socket servers and can plug into the same motherboards running AMD's older 65-nm Opterons, which means customers can get a speed boost without replacing hardware, Brent Kerby, senior product marketing manager for AMD, told InformationWeek. A BIOS upgrade is needed for the new chips.
AMD's new products are in the 2000 and 8000 series, which are for two-socket servers and four- and eight-socket servers, respectively. They also carry the designations SE and HE, which indicate high performance and low power, respectively.
The models and prices per chip in batches of 1,000 include the 2386 SE, $1,165; the 8386 SE, $2,649; the 2376 HE, $575; the 2374 HE, $450; the 2372 HE, $316; the 8376 HE, $1,514; and the 8374 HE, $1,165. The high-performing chips consume a maximum of 105 watts and the low-power models consume 55 watts. AMD's mainstream Opterons run at a maximum of 75 watts.
The HE processors have a maximum clock speed of 2.3 GHz, and the high-end products a maximum of 2.8 GHz. Faster models of the latter chips are scheduled to ship in the second quarter. AMD also plans to ship 45-nm Opteron processors tailored for cloud computing environments this year, but has yet to release details. In late 2009, AMD is scheduled to ship a six-core server chip.
The latest products deliver 44% better performance than previous generation Opterons of the same type, according to AMD. The chipmaker is hoping that customers with tight budgets in the economic slowdown will choose to swap the new chips with older models in the same hardware to take advantage of the higher performance while keeping the capital expenditure to a minimum.
"We believe IT managers are going to be a lot more conservative this year," Kerby said.
The latest models introduce new power-saving capabilities that will be part of all future 45-nm Opterons. The features include so-called CoolCore technology that shuts down portions of the processors' L3 cache that aren't being used. In addition, there's a PowerCap manager that allows IT pros to go into a server and select a lower power level for the chips, trading lower energy consumption for less performance. The new feature is accessible through the BIOS setup, Kerby said.
AMD launches its new products in a tough economic environment that has hammered all segments of the PC market. The slump contributed to a 33% drop in revenue for AMD in the fourth quarter of last year. AMD rival Intel also has been hit hard, reporting that revenue fell 23%. Global PC sales fell in the fourth quarter for the first time in six years, according to IDC. Shipments dropped 0.4% from a year ago and 2.5% from the third quarter. The decline followed a half-dozen years of rising shipments, with the last five averaging increases of 15%.

Intel Set to Detail Octa-Core Xeon Processor at Forthcoming Conference.

Intel Corp. will disclose details regarding its Intel Xeon processor with eight cores at the IEEE International Solid-State Circuits Conference (ISSCC) that will be held in San Francisco, California, from the 8th to the 12th of February, 2009. This will be the first time that Intel will discuss its eight-core chips in detail and it is also likely that the company will unveil its approximate time of availability.
According to the official description of the session where Intel plans to announce details regarding the new processor, the 8-core 16-thread enterprise Intel Xeon processor based on code-named Nehalem micro-architecture has 2.3 billion transistors and is made using 45nm process technology. The chip that is known under Beckton code-name will have four point-to-point quick path interconnect links to connect to other processors as well as system I/O operating at up to 6.4GT/s. Like all Nehalem-class chips, this processor for expandable systems will feature built-in memory controller, which will have four channels. The processor will require a new platform with LGA-1567 sockets.
It is interesting to note that currently available code-named Bloomfield (Intel Core i7) processor features 731 million transistors, yet it has four cores, 1MB of L2 cache (256KB per core) as well as unified 8MB L3 cache. Considering the fact that the enterprise-class eight-core Intel Xeon chips have three times more transistors, it is highly likely that their cache sizes will be dramatically larger - some sources point to 24MB - and certain additional performance enhancements may be implemented.
When Intel showcased its code-named Nehalem microprocessor at Intel Developer Forum Fall 2007, the company indicated plans to ship its eight-core Xeon microprocessors in 2008.
“At the largest configuration that we'll ship in 2008, they'll be an eight-core product. Eight core on one die, and each core will have two threads. So, each eight-core die will be supporting 16 threads,” said Paul Otellini, chief executive officer and president of Intel Corp., at that time.
Recent confidential documents by Intel seen by X-bit labs did not contain any mentions of the octa-core chips designed for expandable enterprise-class systems due in 2009. Moreover, Intel also did not have plans to update the existing Intel Xeon 7400-series family of processors with four or six processing engines this year.

Intel Prepares Assault with 35W Quad-Core, Dual-Core Processors for Desktops

Intel Readies Broad Lineup of Ultra Low Power Processors for Desktop Computers
In addition to low-power quad-core processors for desktop computers, Intel Corp. also intends to release a broad family of chips with 35W thermal envelope designed for small form-factor, yet high-performance, desktop systems. The processors will be the same as the company’s mobile offerings in terms of specs, but will be compatible with desktop infrastructure.
As reported earlier, the first breed of 35W chips for high-performance desktops made using 45nm process technology will be released already this month, whereas additional chips will be added into family on the 28th of December, according to certain documents. The family of ultra low-power chips for desktop computers will include both dual-core and quad-core central processing units, which are projected to be compatible with rather rarely available FCPGA6 infrastructure.
Last year Advanced Micro Devices released a broad lineup of desktop processors with 35W or 45W thermal envelopes, but the company has not updated that family for some time now and many chips have become outdated.
Even though the line of Intel’s 35W chips consists of microprocessors developed for mobile computers and it was not hard for Intel to start offering the very same chips in FCPGA6 form-factor for desktops, the new products may face demand from numerous system builders and end-users, who are looking forward truly high-performance PCs in very small form-factors. Since AMD is unable to supply up-to-date offerings with 35W power consumption, Intel’s line will have no rivals on the market and the CPUs will not be very affordable.
Diversification of the product mix will also help Intel to boost its revenues and profits amid global economic downturn.