Even though Intel is very bullish with revenue shipments of quad-core processors in late 2006 and their introduction in early 2007 for both servers and desktops, the company’s mainstream product line will still utilize dual-core processors. Apparently, the mobile and desktop chips that should come in early 2008 will have two processing engines.
Intel’s code-named Penryn and Wolfdale processors, which are to be manufactured using 45nm process technology, due to arrive in early 2008 will be incarnations or Merom and Conroe processors, respectively. They will only have two processing engines, which means that “wide-spread” quad-core chip era is not around the corner. On the other hand, this means that the epoch of the single-core chips comes to its end: Intel does not seem to plan more single-core central processing units.
Penryn will be a 45nm derivative of the Merom, whereas Wolfdale will be 45nm incarnation of Conroe, reports Cnet News.com. It is unclear what is different in the new chips, most likely, it is larger caches and/or higher clock-speed, however, it is known that all the four aforementioned processors are based on the micro-architecture known as Intel Core 2. It was reported earlier that Intel Corp. is planning single-chip quad-core processor Bloomsfield (both Kentsfield and Clovertown use two chip per single substrate design), which is likely to be made using 45nm process technology.
An Intel spokesperson reportedly confirmed that the Penryn and Wolfdale code-names symbolize projects that are under development, but declined to mention the details of the products other than to note that they are subject to change.
Also in 2008 Intel Corp. is set to introduce the successor of the Core and Core 2 micro-architecture along with the new processor code-named Nehalem, which should deliver architectural advantages in performance, power consumption and so on. Currently it is unclear whether Nehalem processors will have two, four or more cores. Nehalem processor will be built using 45nm production technology.
Intel Corp., the world’s largest chipmaker, for years introduced new processor micro-architectures on proven process technologies and then delivered their “shrunk” versions using thinner fabrications processes. Intel Corp. will continue to follow that strategy, however, it recently announced that micro-architectural shifts are going to happen much more frequently – every two years – compared to earlier 5-year micro-architecture cycle.
Comments currently:
4
Discussion started: 05/12/06 10:00:23 AM
Latest comment: 07/20/06 02:33:26 AM
Expand all threads |
Collapse all threads
[1-2]
1.
"micro-architectural shifts are going to happen much more frequently – every two years – compared to earlier 5-year micro-architecture cycle"
What a joke, what they actually mean is instead of pushing a certain marchitecture until it breaks they will now spend some money on R&D
[
Posted by: alpha0ne
|
Date: 05/12/06 10:00:24 AM]
+ expand thread (1 answer)
- collapse thread
You're joking, right?
Have you seen the amount of money they spend on R & D? It is enormous.
Keep it in perspective, AMD is still making a K7 and has no end in sight. The K8 is just a K7 in drag, much less of a design change than the Prescott was from the Northwood. On top of this, Intel has the Pentium M line and the Itanium line.
I think Intel is being a little disingenous about all this stuff. If you look closely at what they've done, it is not much more than two year intervals. You could argue the Pentium MMX could be called a microarchitectural change, the Pentium Pro, of course. Then the Pentium II, then the Pentium 4, then the Prescott, then the P8. So, from around 1995 or so until now, they've had six.
I think a lot of this is just going to be how Intel defines what a micro-architectural change is. Look at the K8, it is less of a change than a lot of those mentioned above, but AMD defined it as the next microarchitecture so it is. But, if you look at it, it is about the same type of change as the Pentium MMX (which had a longer pipeline, better superscalar performance, and added instructions), much less than the Pentium Pro vis-a-vis the Pentium, about the same as the Pentium II compared to the Pentium Pro, far less than Pentium 4 compared to the Pentium III, far less than the Prescott compared to the Northwood, and far less than the P8 compared to the Pentium 4.
So, I think it is all fluff. We've been seeing iterative updates every couple of years, or major changes, for a long time. Intel is probably just going change the emphasis more. Of course, I could be wrong, but considering how long it takes to design a processor, a two year cycle is impossible unless a lot is borrowed from the predecessor for at least some of the iterations.
[
Posted by: TA152H
|
Date: 05/12/06 03:50:07 PM]
2.
why intel using weird name on chip it hard for to know which want are more better to buy by now i think AMD are more easier to recognize. alphaOne i think it is good for intel do more R&D for better performance and less heat
[
Posted by: Zymath
|
Date: 05/12/06 11:22:23 AM]
+ expand thread (1 answer)
- collapse thread
I find both rating easy to understand, AMD's can be quite difficult when the ratings are similar, where the difference in some software are more significant than what the rating says, but other than that, both Intel's and AMD's ratings are good enough.
[
Posted by: Some random nerd
|
Date: 07/20/06 02:33:26 AM]
[1-2]
Make Home Page
Add to Favorites
Print article
Send link to friend
Digg this
Post to del.icio.us
Post to Slashdot
