• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Intel : Prescott successor canned? Pentium-M is the future?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
I don't think anything FORCED them to, I think they just looked ahead into the future at where their design was leading them, and they decided that they would have to take a new direction to getting where they want to go. They probably could have continued fighting against the inherent flaws in Netburst architecture, but they decided it would be better to change directions.
 
Isnt the Pentium M a p6 morphed together with netburst. If so then netburst isnt really dead, but I reallllllly like what intel is doing, theinquirer was reporting this for weeks that intel is going to can the p4.

Im intereted in seeing what amd does with the dual core opteron, being that its supposed to be compatible with current boards.
 
Wow! What a reponse from the forum! I never thought alot of people feel about the P6 architecture the same way I do!

Now to answer some responses...(from what I know)

goheels said : "i don't think dual cores are really going to be that great. i mean, microsoft definitely doesn't want to write for dual core support, that'd mean practically starting longhorn over. not that i care, since i've gone linux.

and even if microsoft manages to support it, very few apps will be written to utilize it"


The approach to dual-core is to have denser, more efficient computing. Its no longer about brute speed with slapstick features to improve an inefficient architecture. (ie : HT)

MS's Longhorn DOES support Dual-Core CPUs, if you look around, you'll now see information regarding the new power management support for this kind of setup.

AMD obviously encouraged MS about this when they told everyone Opteron was designed with dual-core from the start.

Dual-core will be identical to today's dual-CPU setups...But with a few advantages:

(1) One less heatsink is needed. Physical advantage of having dual core in one package.

(2) In Longhorn, if the PC is idle, One core will be switched OFF, while the other will switch to low-power mode to maintain the system. This dynamic power adjusting (depending if the apps need it) is identical to the power saving features to today's notebook CPUs...Except the response of the system is much better. (like current dual CPU setups).

(3) If the OS can see the dual-core and the apps are multi-threaded, you have no issues. If the app is single threaded, you can multi-task far more than single core. This is based on my experience with dual CPU. (assuming Dual-core is similar)...One special thing...Don't expect 100% faster with two CPUs. It'll depend on OS and apps, and how they're written and optimised. In the most ideal case, you'll see a 80% at best.

AMD has a clear advantage in dual-core CPUs since they designed the Opteron with dual-core in its beginning.

Chowdy says : If the next opterons are to be dual-core with same pin compatability as the old opterons, that's awesome. But remember, the FX is basically an opteron for the desktop. Does that mean dual-core FX?

This will depend on AMD, if they see a market for this, then expect a dual-core FX. If not, you're stuck in "single life".

Gregory_WE : Apparently, there's some inherent issue (power leakage) with the process going from 130nm to 90nm which no one has been able to correct.

That's about right, as mentioned, all CPU manufacturing companies are facing a bit of a brick wall when moving from 130nm to 90nm process. From what I gather, its basically because as you shrink, you need to pump in more current. In the case 90nm, you have to pump in quite a bit to maintain signal integrity compared to 130nm process.

So now, you got a fair amount of wasted heat being produced from the extra current needed to maintain signal integrity...This is what is known as "leakage".

Many will attempt to reduce this as best as possible...But as we all have seen, the Prescott is an example of quite significant leakage. (I hear about 1/3 of heat produced is leakage...Very inefficient).

Obviously, Intel realises it can't keep doing this and still be competitive. (especially against the AMD64 CPUs).

BTW, you can buy desktop Pentium-M mobos from Commell and Lippert. They both supply Mini-ITX mobos for Pentium-M. In the Commell one, there's a Mini-AGP version that allows you to install a ATI Mobility Radeon 9600 Pro. The catch is that its extremely expensive to setup such a system.

Its best to wait until Intel pushes this kind of setup to maintstream, so it is more affordable.
 
Krowa 02 said:
Isnt the Pentium M a p6 morphed together with netburst. If so then netburst isnt really dead, but I reallllllly like what intel is doing, theinquirer was reporting this for weeks that intel is going to can the p4.

Im intereted in seeing what amd does with the dual core opteron, being that its supposed to be compatible with current boards.

A really rough way to see it...

Pentium-M
=> Pentium-III with 400Mhz or 533Mhz (upcoming) FSB
+ Highly optimised Power Saving features
+ Large Cache (slight increase in latency due to large size compared to PIII)
+ Improved Branch Prediction
+ Slightly longer pipeline (but still very short compared to P4).
+ SSE2

Slap this with a chipset and wireless => Centrino Mobile Technology.

Essentially, the Pentium-M is a hybrid design that takes the best of the older Pentium-III Mobile with some of the P4-M features. I don't think it uses the Netburst architecture, but a highly modified PIII design with higher bus speeds and optimizations.

It performs much closer to the Athlon compared to the P4 ever did in a clock-for-clock comparison. (I say, overall, both are reasonably even).
 
Last edited:
So what could we expect to see from a Desktop Banias/Dothan CPU?

400/533/800 FSB? What about implementing HyperTransport?
SSE3?
Improved efficiency Dynamic Branch Predictors (similar to ones installed on Prescott)?
Increased size of Branch Target Buffer (similar to what AMD with with Opteron)?
Increased size of Global History Counter?
Optimized scheduler improvements (A LA Prescott)?
Increased number of execution units (bring it back up to par with AMD's Athlon)?
High speed 512 KB or 1024 KB L2 (shorter pipeline, faster cache needed)?

Could be really interesting to see what the desktop version of Banias/Dothan brings to the desktop CPU market
 
The rumor mill reports that AMD’s .09 micron die shrink will meet more traditional expectations for a die shrink, which is to say that AMD’s .09 micron chip will run cooler, draw less power, and scale well.

I cannot imagine how Intel R&D did not know the P-4 Core would have this scalibility issue when it was deployed and for them to have missed it is a big sign that Intel is in trouble OR just cares nothing about it's customers.

Imagine all the P-4 Boards that are going to be left in the junk drawer becaue the entire chip platform has to be scrapped ? LOL..

I would be so mad at Intel right now if i owned one,that The filters would be having trouble keeping up with me. What is really sad is how DUMB Intel must be to not have been ready to phase out the P-4 smoothy with an introduction of a replacement line. DURRING THIS ANNOUNCMENT..............

Wow and to think people have been brainwashed so well as to not see this as a kick in the butt from a giant manufacture simply amazes me.......................if I was an INTEL OWNER I WOULD JUMP TO AMD SO FAST THE PRESHOTT WOULD HAVE NO TIME TO COOL DOWN...Come on here people, look at what Intel just did ?
Is no one really mad at Intel for this surprise bombshell ?
 
What good would editing my post be now when you sensitive types have quoted it twice now? lol If it bothers someone that much then they need some therapy ;) Anyway, I won't flame anymore.
 
i suggest you find a new forum sunburn, or relax

yeah right now im just a little mad at intel, though not really. I would have to have upgraded my mobo for new xoens newayz so its not really a big issue to me. But still these dual cores is a really bad idea. becuase you need software writen to take advantage of it, it is nothing like hyper threading. People need to do some research on this before they post.

Anyways go AMD if they can totally pwn intel more power to them, they might be my next dualy system.

btw tiamat-vote for love is a really realyl cool song.
 
I think the processors need to go Dual core before any software company would write for it .. And Hyperthreading is almost a dual processor in practice any way. And multi Tasking is for sure going to benifit from a dual cored cpu no matter how the software is written......
 
winxp's processor drivers will probably needs to be
updated to support the new 2-core 4-core 8-core CPUs ;)
 
hey hey, i havent heard much about the k9 lately, whats up with it? all that i know about it is that amd has a shortage of engineers for the k9. anyone know how progress is going with it?
 
I cannot imagine how Intel R&D did not know the P-4 Core would have this scalibility issue when it was deployed and for them to have missed it is a big sign that Intel is in trouble OR just cares nothing about it's customers.
Its called arrogance.......pride always comes before a fall
 
Krowa 02 said:
Isnt the Pentium M a p6 morphed together with netburst. If so then netburst isnt really dead, but I reallllllly like what intel is doing, theinquirer was reporting this for weeks that intel is going to can the p4.

Im intereted in seeing what amd does with the dual core opteron, being that its supposed to be compatible with current boards.

In a way, yes. It shares a very similar execution backend to the P6. However, it's front-end resembles the P4 a bit more. It's more flexible than the P6's original front-end.

The pipeline is also 15-stages. Which is pretty much exactly half-way between the P6 and Netburst. It also has many features that aren't found in either P6 or Netburst such as a dedicated stack manager and a very advanced prediction algorithm.

So ya, it's.....different. It's not really based on P6 and it's not really based on Netburst. I think ultimately, the execution core that resembles that of the P6 will have to go. If it's to be a desktop processor it will need more power than that. Aceshardware recently did a few tests and the FPU results for Banias did not turn out too well.

So, if you beef up the execution back-end, it'll resemble Netburst a lot more than it would resemble P6. Only this time it won't extend the pipeline to an insane 31 stages (past the trace cache), add slower cache on it, not worry about power consumption at all and not even worry about the memory bottleneck that occurs at high clockspeed.
 
Back