In order to offer the “Wintel” world computers that could be as small and as powerful as the latest Mac mini systems from Apple, Intel Corp. started to promote its mobile processors in the desktop market segment. Such a move may mean better adoption rates for small form-factor (SFF) systems, which are not something new, but which still have not got significant market share.
“By taking an Intel mobile processor and combining it with a microATX, picoBTX, or proprietary form factor board, Mobile on Desktop systems combine the quiet, cool efficiency of an Intel mobile processor with the full-featured functionality of a desktop. The result is a sleek system that packs great performance into an attractive package,” one of Intel Corp.’s statements read.
Currently the company promotes Intel Core Duo, Intel Core Solo and Intel Pentium M processors for SFF computers. Additionally, the company advertises certified barebones and small form-factor mainboards that support the mobile processors from the company.
It is not the first time when Intel offers its mobile processors to markets that historically used other chips. For example, last year Intel certified the Pentium M for servers, whereas this year released an Intel Xeon processors based on the Core Duo design. Given that the company’s current Intel Pentium 4 and Intel Pentium D consume a lot of power and dissipate a lot of heat (up to 130W), it is impossible to use them in truly small, quiet and efficient computers.
While mobile central processing units (CPUs) solve the problem of excessive power consumption and temperatures with their thermal envelopes of about 30W – 35W, they typically cost more than their desktop counterparts, which makes computers running them more expensive. Furthermore, current mobile processors from the company do not support 64-bit technology and some other capabilities the desktop have.
Later this year Intel is expected to release its processors code-named Conroe, the chips that will have much lower power consumption (65W) compared to current chips, yet, they are expected to sport all the advantages the desktop processors have, such as 64-bit technology.
Comments currently:
31
Discussion started: 03/31/06 08:25:30 AM
Latest comment: 04/06/06 09:14:18 AM
Expand all threads |
Collapse all threads
[1-5]
1.
Intel is about three years late on this one. I had been looking for Pentium M based desktops the day they came out for the laptops, as were many other people. They would have been perfectly suited for servers and they could have kept grossly overcharging people, which is what they do with these processors. For the desktop, the price is terrible for the performance, but some people would have tolerated it anyway. But, instead they did not offer them even to the hobbyists. I could understand not wanting to confuse the general public with two lines that overlapped, but the well-informed always wanted these chips on the desktop and they could have made them available in that limited segment. While it would not make that much money because the volumes would not be very high, it would keep brand loyalty to Intel and keep them from jumping ship to AMD because the Pentium 4 sucked bad.
As it is, they lost a large portion of the server market because these chips were not made available to it, and would have been a great fit if they were. The hobbyist market continues to revile Intel's Pentium 4 and Intel has developed a negative reputation in many circles.
Why did it take them so long to figure this out? It is almost pointless now with the Conroe coming out anyway. Most people will wait, if they can, and not waste money on this when a clearly superior processor is only a few months away.
It never fails to amaze me just how stupid even big companies can be.
[
Posted by: TA152H
|
Date: 03/31/06 08:25:31 AM]
+ expand thread (17 answers)
- collapse thread
Pentium M has been used in servers for quite a while now already. Pentium M has also been an option on the desktop. The price for the processors wasn't that bad either: the motherboards that hobbyists could buy tended to be expensive though. You seem to think that Intel should control and push everything, but that's not the way it works.
[
Posted by: Ranting about nothing
|
Date: 04/01/06 12:23:23 AM]
You are basing this on, what? Up until recently there were no motherboards that supported this processor. Certainly no server motherboards, and no configurations that supported these processors in dual configurations.
To my knowledge, Aopen started this several months back, I looked for at least a year for a motherboard that supported these processors and none were available or supported. So, what are you basing your post on?
[
Posted by: TA152H
|
Date: 04/01/06 08:43:31 AM]
Considering an example on the server side: blades using Pentium M were introduced 2-2.5 years ago. That's one segment where lower power made up for lower performance, at the time.
For desktop, ATX-type boards have definately been available for more than a few months. Just search.
Last sentence was most important in the reply.
[
Posted by: Ranting about nothing
|
Date: 04/02/06 05:41:11 PM]
Sorry but he is right,
the difference is that Aopen and DFi (I think) have been releasing intel855 based desktop boards unofficially, and now Intel is releasing it for the "general" public, that makes a huge difference.
[
Posted by: Kaz
|
Date: 04/03/06 02:53:02 AM]
What do you mean "he is right"? Pentium M applications other than notebooks have been around for ages. This is a fact.
You think that a Pentium M blade is "unofficial"? You think that a Mac Mini is "unofficial". These are just applications. There is no huge difference, except in your mind(s).
[
Posted by: Ranting about nothing
|
Date: 04/03/06 03:49:25 AM]
Maybe this further comment will make it even clearer: it is OEM's like HP/IBM/Fujitsu that made Pentium M blades "official", two or more years ago. In fact, that has been and gone already. You are both just confusing marketing with product.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 04:05:51 AM]
Tell me where are they?
http://www.fujitsu-siemens.com/products/standard_servers/bl ade/index.html
[
Posted by: Kaz
|
Date: 04/03/06 04:24:42 AM]
You have yet to show one example of where they were selling these boards for people.
Aopen even stated they were the first ones to do this. So, if you could be more specific, it would be helpful.
[
Posted by: TA152H
|
Date: 04/03/06 10:05:27 AM]
AOpen even stated that there later motherboards were the first ATX-motherboards, whilst they were already selling ATX-motherboards for Pentium M. None of this helps support your statements about what Intel should have done.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 11:54:07 AM]
To make it clear, although the discussion has moved on: any OEM capable of making a notebook motherboard could have made an ATX-motherboard at any time since the introduction of Pentium M. Intel didn't prevent this in any way, so what is your point about Intel? As it happens, AOpen wasn't the first in anything, but was perhaps the first that you know about. That's perhaps interesting for historians of motherboards.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 12:08:03 PM]
Ranting,
You really don't understand the market at all, that's what you are showing me.
Intel sells microprocessors, and to that extent other companies can make what they want. Except, that's not generally how the market works, and you clearly do not understand this. It's the difference between theory and reality.
Other than really big companies, like IBM and HP, companies do not generally go against the direction Intel sets. Intel sells processors, but also helps in very substantial ways with design and other supporting components. This includes chipsets, motherboards, validation, as well as talking to customers of the company. For example, Dell will contact Intel to meet with customers.
If Intel doesn't support a platform, it does not take off. This proves it, since the Pentium M would have done much better than the Pentium 4 for servers, particularly low power ones. But, did they?
So, Intel simply selling processors is not what I am talking about. They need to support the platform and help other companies deliver products with it. That's what they do. This proves my point exactly, they sold the processors, and no one besides a monster like HP used them in any way other than the way Intel told them to.
Your remarks about Aopen are unclear, so I will not even comment on them. I am not sure what point you are trying to make.
[
Posted by: TA152H
|
Date: 04/03/06 12:35:08 PM]
Copy of my reply to the same kind of post below (I don't enjoy typing the same thing twice):
Firstly, just look at hobbyist-motherboards for Pentium M, which is seems is really all that interests you, although you make broad statements about markets in your original comment on the article. There is nothing special about them. There was no extra support from Intel necessary, over and above what was already there. It had really nothing to do with Intel. Viiv (on Core Duo etc.), which is a market-move by Intel and partners, waited for the right market-timing for Viiv, for example; that certainly wasn't 3 years ago.
Seondly, just read your own comments if you want to to see comments from someone that doesn't have a clue about markets. If Pentium M was so good outside of its target-market, it would have seen widespread adoption outside of its target-market and Intel would have been happy to sell chips regardless. Your enthusiasm has blinded you to the limitations of Pentium M outside of its target-market. Even in blades, where the fit should have been better than in other parts of the server-market, Pentium M had but limited success.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 01:38:21 PM]
As usual, you're wrong again. I was looking for Supermicro server motherboards for the Pentium M. But guess what, they didn't make it.
You are equally wrong about the suitability of the Pentium M for the server market, since Intel has moved it there. You don't know anything about servers, so I'll explain. Most people running servers care a lot about power, and power/performance. This is where the Pentium M did especially well. In fact, the Pentium III is still a popular choice for server, and Supermicro still makes motherboards for them.
You apparently know very little about motherboards. Even small differences between the Tualatin 512K Pentium III-S and the Tualatin 256K Pentium III were important enough that you should not use one in a motherboard not made for it (due to differing load line levels). Tossing a Pentium M in a motherboard not made for it would be sheer stupidity, and Intel did not support it, help with development of it, offer reference designs, or offer their own motherboards.
So, no one did it. If you don't get it by now, you're just not worth the effort to educate. The fact is, the market proved my point, whereas you have none but fluff and air. Learn the market, and maybe in six months I'll try to have a discussion with you. Until then, you're not informed enough to be interesting.
[
Posted by: TA152H
|
Date: 04/03/06 04:13:21 PM]
It doesn't really matter what you were interested in, although your comments seem to jump all over the place. Let's take what you say here. The facts are much better explained by the realisation that there simply was no market for the kind of servers you wanted, based on Pentium M, or it was so small as to be uninteresting. Major OEM's did, in fact, design servers based on Pentium M, but not the kind of servers you wanted. They made the effort, but not for you. Smaller players also saw a limited market, that they also made the effort to move into with actual products, but not in servers and also not for you then. It seems that your perception of the market did not match reality, as seen by the players in the market (including Intel's major customers). This all adds up to indicate that Intel was wrong in its actions and proves your point? How? It seems rather more likely that you just don't have a clue. What is happening now has no relation to the past prospects for Pentium M. Do you even realise that Core Duo is not Pentium M? Sossamon is not Pentium M? Merom/Conroe/Woodcrest are not Pentium M? It's not my facts and comments that consists of fluff and air: your wishful thinking is nothing but.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 05:57:02 PM]
Sorry for the spelling, although it would have been more consistent to use 'Xeon LV' anyway.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 06:35:25 PM]
"In fact, the Pentium III is still a popular choice for server"
Priceless :)
[
Posted by: Ranting about nothing
|
Date: 04/03/06 08:01:12 PM]
MAC mini?
That isnt an example. That is not even a computer. Besides it was a POWER PC machine not Pentium M.
[
Posted by: Kaz
|
Date: 04/03/06 04:26:47 AM]
2.
As I said, been and gone. As an example, check out HP BP10e G2 (retired).
[
Posted by: Ranting about nothing
|
Date: 04/03/06 04:50:56 AM]
+ expand thread (9 answers)
- collapse thread
Sorry, BL10e G2
[
Posted by: Ranting about nothing
|
Date: 04/03/06 05:02:04 AM]
Ok.
But its still blade server.
You never see that on a regular tower/rack server.
Because if Fujitsu had them on RX100/RX200 I would certainly buy one.
[
Posted by: Kaz
|
Date: 04/03/06 05:19:19 AM]
Perhaps you should write to Fujitsu and ask why it didn't want to use Pentium M, except in blades and mobile applications? The answer is easy enough to guess, but you have no point worth making anymore anyway.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 05:27:27 AM]
If that is the best you can come up with, that illustrates my point exactly.
A big company like HP can make these things with or without Intel's help or direction. And this is a lousy example, since it is a tiny blade server sold by a massive company with huge developmental resources. If they were so common, you'd have been able to find something more mainstream.
I kept trying to find motherboards that would work with the processors, and no company was supporting them. Sure, HP can build something that was super expensive, because they are excellent products for that market, but that's not what I was talking about. Intel did not make these processors available to that market. When they do, they come out with reference guides, official support, supporting products, etc... That way smaller companies that can afford the development work can make products, not just monsters like HP that do not require, but still like, help from Intel.
[
Posted by: TA152H
|
Date: 04/03/06 10:14:26 AM]
You just don't understand the point. Not only are you factually wrong, you are just ranting about nothing. If you can't see it, too bad.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 10:18:04 AM]
As a final comment, it's the wishful thinking of enthusiasts, regardless of the facts, that really never ceases to amaze :)
[
Posted by: Ranting about nothing
|
Date: 04/03/06 11:35:37 AM]
Who is the one that is ranting?
You use generalities to obfuscate that you have no real knowledge and your position is weak.
I will put it in simple terms you will understand. Until recently, one could not easily build a supported desktop or server from available components based on the Pentium M. I know this, I kept looking.
That is what my complaint is about. If Intel had supported this type of use, there would have been a lot of support out there for using the Pentium M on the desktop. There wasn't, because Intel didn't want to overlap their product lines, and instead pushed their inferior Pentium 4 line.
So, while big companies could have done the R & D to build something Intel didn't support, it wasn't generally practical and smaller companies like motherboard makers were not willing or able to, until very recently. Intel support would have changed this immediately. Do you understand this at all?
Net result is, the products many of us wanted were not available because of Intel's position on the Pentium M and its platform. Does this help?
[
Posted by: TA152H
|
Date: 04/03/06 12:40:44 PM]
Firstly, just look at hobbyist-motherboards for Pentium M, which is seems is really all that interests you, although you make broad statements about markets in your original comment on the article. There is nothing special about them. There was no extra support from Intel necessary, over and above what was already there. It had really nothing to do with Intel. Viiv (on Core Duo etc.), which is a market-move by Intel and partners, waited for the right market-timing for Viiv, for example; that certainly wasn't 3 years ago.
Seondly, just read your own comments if you want to to see comments from someone that doesn't have a clue about markets. If Pentium M was so good outside of its target-market, it would have seen widespread adoption outside of its target-market and Intel would have been happy to sell chips regardless. Your enthusiasm has blinded you to the limitations of Pentium M outside of its target-market. Even in blades, where the fit should have been better than in other parts of the server-market, Pentium M had but limited success.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 01:05:32 PM]
Reflecting briefly on this, what I think you really fail to understand is that Pentium M was/is an actual processor with an actual TDP and an actual performance in benchmarks that made it particularly suitable for some things and not for others.The only way Intel could have changed that would have been to design it otherwise. Your 'complaint' is really only that those with a deeper understanding than yourself, ie. Intel and it's partners, understood this (and acted accordingly), whereas you did not.
[
Posted by: Ranting about nothing
|
Date: 04/04/06 12:12:00 AM]
3.
I guess you didn't understand the point. It is Apple that makes Mac Mini (Intel) an "official" application of Core Solo/Duo, not Intel.
[
Posted by: Ranting about nothing
|
Date: 04/03/06 04:52:49 AM]
4.
good
[
Posted by: nboy
|
Date: 04/05/06 12:36:42 PM]
5.
lol
[
Posted by: TOVi
|
Date: 04/06/06 09:14:18 AM]
[1-5]
Make Home Page
Add to Favorites
XbitLabs is using Twitter
Windows Live Alerts
