When we were getting ready for the IDF Fall 2009, we anticipated to hear some exciting details about the upcoming 32 nm processors based on Nehalem microarchitecture. Yes, I am talking about the CPUs known as Westmere.



According to Intel’s “Tick-Tock” concept, that they have been actively promoting for the past few years, Westmere should become Nehalem’s die-shrink manufactured with new 32 nm process. Intel confirmed that they have no problems with 32 nm process, it has been certified and the first semiconductor wafers are already coming off the production lines. Within the fourth quarter they are planning to launch mass production. First mass production solutions using 32 nm cores will start selling in the very beginning of 2010. And this time Intel decided to focus primarily on desktop and mobile mainstream processors codenamed Clarkdale and Arrendale.

For a in-depth review visit.

Yes I understand its how they make money but do they really need to make so many kinds of chips. I mean a new one for laptops desktops and severs per year would be great that way there are big jumps.

well Intel has finally decided to try and sort out the mess they've made with the hundred different processor lines and names they have (exaggeration) by consolidating them in the Core i3/i5/i7 lines (though keeping the Pentium/Centrino brands)... no more of this Core2Duo and Core2Quad and Core2Crap.

I wonder... does AMD have similar/contesting research and development going on to compete with what Intel is doing here with Clarkdale and beyond? I've heard the newest Phenom II processors are competitive with the Core i7...

I too am quite confused with the huge product line Intel has out right now. Not just their latest core i5/i7 1156 socket, but with their upcoming i3 and possible i9. I think they're slimming down the core2 series as they want to move consumers toward the i5/i7. But they've also got their pentium, celeron, atom, and xeon. Then they have a similarly named part, but for a mobile platform such as a modified core2duo, that is nothing like the core2 desktop it was named after. The confusing part is how different people refer to the same chip by 2 different names of the consumer name and the code name.

AMD is trailing in the game again; their Phenom II stuff was designed to compete with the upper end of core2duo and core2quad line which it did really well. By the time they got their 945 and 955 out, they were really close to competing with the core i7 920, but not clock for clock. But as performance per dollar is concerned they held the crown for a long time. Now with the core i5 and core i7 8xx line, I think Intel got that back because the i5 is (or was) priced lower than the PII965 with higher performance at a lower clock.

Last I heard about AMD was their 6-core Phenom II consumer CPU's coming early or mid next year. It wont be much of an architecture change as they will be focusing how to manage the heat of 6 cores while maintaining a high enough clock to compete with similarly priced Intel chips. They'll also be trying to package it into an AM3 socket. After that, Bulldozer is still rumored to be coming as well. It's kind of hard to tell where AMD is headed as they don't have the same future projection as Intel does with their neat little charts.

I guess a lot of it is down to new types of computers emerging in the last few years: the HTPC livingroom mediacenter type, the 'baby-laptop' netbooks, and the handheld stuff.

AMD received similar flak for flooding the market with lots of different cpus: every extra $10 gets you a different, slightly better model, it seems. Maybe that's part of a deliberate marketing strategy by both Intel & AMD, rather than simply being the result of a bad market messing up their original plans?

AnandTech had a writeup about the costs of moving to smaller scaled chips (costs according to AMD's split off GlobalFoundries company):

Click to view attachment

With such enormous costs and a bad market, flooding the customers with choice sort of makes sense, I presume: less demand -> lower prices but only through the addition of pricepoints/models.

While I can understand the efforts to compensate and recover from the hundreds of millions or billions of retooling and production costs, why flood the market with many re-badged products. I can understand wanting to hit different popular price points in order to maximize profit, but it seems like they're trying to meet nearly EVERY price point. Granted, it is pretty hard to impart degrees of separation with CPUs as there are only so many different configurations of a given architecture. There are only so many different clocks, cache levels, core counts, thread counts. But I still argue that having a limited amount of selection can be healthy for income.

For example; the recent downsizing of Detroit's big 3 auto makers cut back on the re-badgeing of the same model over many different names. While this didn't reduce R&D costs, it sure made it easier for a consumer to pick out a product, and start paying for it. For those who aren't totally familiar with the rebadge phenomenon, an automaker would make a single car and market it under multiple brand names with a different bell or whistle for each brand to try and make it stand out a little. I would point this analogy back at the Phenom II 955 and 965. Even AMD admits that they are the same exact thing, except one is binned just a little higher. What people are finding out is that for the extra $40, overclockers can only clock a few megahertz higher. If they had just stayed with the 955, retailers would only have to stock 1 part number and leave consumers an easier choice when deciding on a top AMD processor. I could also direct this to Intel; with the release of their i7 nehalem processor, they had the 920, 940 and 960 which is easy enough, however they clocks were very close, but with price points several hundred dollars apart with very little performance differences. Perhaps the 940 was just an appeasement to give an illusion of more choice. It too was a sneaky marketing since the 920 and 940 are just binned differently but still cost the same to produce.