Quad Core Chip Technology

AMD has decided to take a small step back in the race for quad-core supremacy.  On Monday, AMD executives confirmed this weekend’s rumors and admitted that the company will indeed introduce a triple-core microprocessor sometime during the first quarter of 2008. “We believe that triple core is the right product at the right time to serve a broad swatch of market,” said AMD vice president Bob Brewer, at Monday’s briefing in San Francisco.

AMD stressed the update does not affect its planned launch date for a consumer-oriented quad-core chip, codenamed Phenom, which is slated for Q4 of this year.  According to AMD, the the tri-core processor makes sense from both a marketing and retail point of view. Due to the slow adoption rate of quad-core, as well as the lack of multi-threaded apps that take advantage of such processors, AMD is banking on the fact that a tri-core processor will offer desktop consumers an attractive middle-of-the-road option.

In particular, Brewer and other AMD executives cited a recent study from Mercury Research pointing out that quad-core processors represented less than two percent of desktop shipments during the second quarter, while dual-core took up the remaining 98 percent. Until demand ramps up for quad-core, triple core will serve as something of an intermediary, according to AMD. At the same time, the company hopes it will also stimulate broader multi-core adoption.

“We left the megahertz race for the core race,” said John Taylor, product communications director at AMD. “The sweet spot is now four (cores) and down … and there are a lot of scenarios where three cores outperforms dual core,” he said.

In terms of architecture, Brewer also confirmed on Monday that the yet-to-be named processor will basically be a quad-core processor with one core disabled, and that it will feature the company’s Direct Connect architecture, as well as a shared L3 cache, and other architectural selling points that Intel currently lacks.  “It’s simply a better intermediate step,” Brewer said, adding that triple core is about product management and design flexibility.

Conveniently enough, it’s also something AMD’s larger competitor, Intel, can’t do at the moment — or at least do well. “Do I rub my hands together a little about that fact? Sure,” said Brewer, “but ultimately this is about scaling.”

Another potential benefit Brookwood points to is production efficiency. If AMD finds it is having many of its quad-core Opteron processors coming off the production line with only three of the four cores fully functioning, then it can still market those as tri-core instead of chucking them into the trash can.

Of course, pricing will be crucial to the success or failure of AMD’s tri-core processors. On that topic, company executives wouldn’t comment other than to say that prices would be competitive.   They better, especially if Intel decides to drop the price of its quad-core processors to match or beat AMD tri-core chips.

Regardless of Intel’s reaction, AMD seems to be using its core strengths — namely, the five years of research and development that went into its Opteron line (formerly codenamed Barcelona) — to its advantage.  AMD executives promised last week that core features from the new quad-core architecture would trickle down to the consumer level. It’s now making good on that promise.  Sometimes slow-and-steady wins the race and it might be the case here.

That’s right. It’s time to clear the air: most people will not even slightly benefit from the quad-core monstrosity when it becomes available.

For those of you looking at upgrading, you may have yourself in a knot deciding when and what to purchase. There are now extremely fast and viciously-priced dual-core processors on the market — this is, ordinarily, a damn good thing. However, you seemingly can’t ignore the quad-core CPUs on the horizon coming from both AMD and Intel. Surely these upcoming beasts will make roadkill of the existing line?

My unconcealed answer is no. There’s an elephant disguised as a lampshade in this room and it stinks.

Above is a quote from an article in August of 2006 and it has been proven very, very wrong.  I think many techies are sick of having to upgrade every six months to keep up with the cpu curve but with faster cpu’s comes better and more useful software, so stop your whining and get with the program.

After speaking with many developers including some big names, we learned that they can’t make much sense out of quad or more core CPUs. They confirm that they can put two cores to good use but not much more. The main problem is that the performance is far from scalable.   You have to spend both time and money to resource the programmer to try to make sense out f the dual core CPU. It takes up to a year to optimise the game for more threads and even if you make the perfect job you can count on twenty to thirty percent performance increase, and this is the best case scenario.   Once you start making the multi threaded game you end up making and having the multithreaded bugs as well.

You can keep one core busy with the physics and collision detection, second core will have to wait for the score to move on with the Artificial intelligence while the third core could possible calculate the graphic data. In this best case scenario you have to realise that the core number two and three would always have to wait for the core number one to finish its job and pass the job to the cores two and three. In this concept there is absolutely no place for quad core as games are non parallel applications. A game developer expert said that you can use the core number four to stream and load the data in the game and this is what the guys at Remedy did at IDF quad core demonstration. But this takes time and money and it is not commonly embraced by developers.

Game developers are in the dawn of dual core programming and now all the sudden AMD and Intel wants them to go quad core. For the time being Quad cores are good for rendering and serves but not for games.   So if you want to play games, you can forget about quad cores, you simply don’t need them and can gain just marginal performance out of them. Give the developers some time and this might change, but we are talking quarters not months.  My feeling is that game developers will come around and optimize for these new super face quads.

Further expanding its quad-core processor family, Intel has announced today two highly energy-efficient 50-watt server processors that represent a 35- to nearly 60-percent decrease in power from Intel’s existing 80- and 120-watt quad-core server products.

As companies increasingly focus on reducing electricity bills and cooling costs associated with their computing needs, these new processors, requiring just 12.5 watts of power for each of the four cores or processing engines, deliver similar performance yet set a new standard in energy efficiency.

Bragging rights for dual-core chip supremacy seems to be dying down and Intel gave us a glimpse into the quad-core chip coming this year.

Clovertown, a four-core processor, will start shipping to computer manufacturers this year and hit the market in early 2007. Clovertown will be made for dual-processor servers, which means that these servers will essentially be eight-processor servers (two processors x four cores each).

The company will also come out with a previously announced version called Tigerton around the same time for servers with four or more processors.

Core expansion will be the major theme for Intel over the next few years, says Chief Technology Officer Justin Rattner. “By the end of the decade, chips with tens of cores will be possible, while in 10 years, it’s theoretically possible that chips with hundreds of cores will come out, he added.”

Multiplying the number of cores brings distinct advantages. First, it cuts down overall energy consumption for equivalent levels of performance. If the recent Core Duo chips released for notebooks from Intel had only one core, the chips would consume far more power, he said.

Integrating processor cores into the same piece of silicon or same processor package also increases performance by reducing the data pathways

“To go from core to core can be a matter of nanoseconds,” Rattner said. “As soon as you move cores together you get an automatic improvement in available bandwidth.”

Advanced Micro Devices will also come out with chips with four cores sometime this year.

Nevertheless, adding cores requires careful planning. Energy efficiency, data input-output and memory latency (the time it takes data to go back and forth from memory and the processor) will be major issues with each level of core expansion.

To get around some of these issues, Intel is conducting research into circuit re-design and chip architecture as it has in the past. In addition, the company is working with application developers to determine how the architecture of its chips can be optimized for the end-user.

By working with one server application developer, Intel determined that it needed to make three small changes to the architecture of one of its future server chips. Before the changes, the application only ran well in simulations on chips with 16 cores. After that, performance began to decline, Rattner said.

Two pieces of silicon in a single package seems more likely. At around the same time, after all, Intel will release Woodcrest, a dual core server chip based around the same Merom-Conroe-Tigerton-Clovertown architecture. It will contain only two cores and consume 80 watts of power, less than the 165-watt server chips Intel sells now.

A test is currently running at a large financial institution (secret) on an experimental basis with Woodcrest chips, Rattner said.

Intel has already released one dual core processor that contained two pieces of silicon. While using two pieces of silicon can be cheaper to design and manufacture, some have said dual silicon chips don’t provide the same level of performance.