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

Mp Vs Xp?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

verbatim

Member
Joined
Mar 12, 2003
Location
Western Australia Perth
I know u can unlock the amd xp to b reconised as a mp but does the mp actually do anything better in a dualy config than xp's?

I have been reading the amd propaganda about "advanched cache coherency" etc.

What do u think?
 
They are both exactly the same CPU's regardless of what you may here. In AMD SMP systems, SMP ability is determined by chipset implementation and not the CPU. All AMD CPU's have been able to run SMP, just no chipset has been around to support them.

Saying that though, the MP's are cherry picked XP's.
 
Please read the stickies in this forum.

This issue is addressed in the sticky called "The MP, XP, Thunderbird, Duron SMP answer."
 
Some people have problems with modded XP's not being recognised as MP from time to time. For this reason I say go with MP's. Not only that, MP's hold there value when you come to sell them on. Nobody really wants to buy modded XP's.

But if you are looking for a cheap way in to the world of SMP, then mod away!!

Hit the search icon for more information, as this has been debated before.
 
This issue is addressed in the sticky called "The MP, XP, Thunderbird, Duron SMP answer."

No its not. I read that b4.

This is not a question of compatability but of function.

I was asking weather the mp's actually act together better than xp's because they were made for smp
 
This is down to the individuals modding ability and not the CPU

i dont think this is entirely true. mp chips are tested before being packaged. xp chips are not tested for smp. this is one of the reasons why mp chips are more expensive.

since the mp chips get tested, then you can conclude that some xp chips will not run/function properly under smp environments, no matter the skills of the person performing the modification.
 
True, but the vast majority of the time, the issue is not that XP's are MP rejects. It's just that the demand for XP's is higher and that apparently this extra "MP Verfication" testing drives the cost up $50 per chip. They only perform the MP verification on enough CPU's to satisfy the demand for MP's. So the chance is there that you will get XP chips that fail the "MP test". Very slim though.

If so you can always sell them, they should still work fine in single configuration.

In any situation where you may need a warranty, I suggest MP's.
 
i wonder why some xp chips fail under the smp environment when both chips are supposed to be identicle. is it because of the error in assembly by the robots, like an error in bridge gaps, bridge connections, or is it in the silicon itself?

if such errors, then mp chip testing at the factory will spot these out, and discard them. but under xp production, there are no check-ups.

or there may very well be a difference in design (within the silicon architecture iteslf) that AMD is not openly saying.
 
The design is the same. The only difference is that some chips are tested for SMP use. If they pass, then they are labled MP's and sold for more money, due to testing costs and market demand. If they don't pass SMP tests, but do pass operational tests, then they are sold as XP's. Most XP's, however, are never tested, at all.

I have five AMD duallies, all running modified XP's and I have never encountered an XP or Morgan Duron that hasn't worked flawlessly in SMP mode. (I have run 12 different XP's and 4 Morgan Durons.)
 
cmcquistion said:
I have five AMD duallies, all running modified XP's and I have never encountered an XP or Morgan Duron that hasn't worked flawlessly in SMP mode. (I have run 12 different XP's and 4 Morgan Durons.)

Another vote for XP's.
I've modded 6 XP's to MP's, no problems, ever.
 
verbatim said:
This issue is addressed in the sticky called "The MP, XP, Thunderbird, Duron SMP answer."

No its not. I read that b4.

This is not a question of compatability but of function.

I was asking weather the mp's actually act together better than xp's because they were made for smp
It is hidden down in the links in Frank's post. Easy to miss.

http://www.hardwarezone.com/articles/articles.hwz?cid=2&aid=393
http://www.hardwarezone.com/articles/articles.hwz?cid=2&aid=395


Mine is 1.3 Morgan Durons that run quite nicely on a Tyan Tiger MPX
 
I guess the real question is what are you willing to pay for a peice of mind? i have used both modded xp and true mps in my asus 266-d board, and i will never get another xp to mod like this, i dont think the price diff is too high, so i went with the MP's
 
i never tried modding my xp1800 for smp use. but i did try to mod it in order to unlock it's multipliers, without much success. i think the rear window defogger kit that i purchased at pep boys was to blame. i dont think the defogger acted like it is supposed to act. i could not increase the multi, but i could decrease them, at certain stages only.

but i view the mp vs the xp as buying a quality psu. everyone stresses quality psu, and i think the same should be said about the mp chips in the smp world.

most people have gotten their xp to work under smp environments, but with some trickery, like multiple re-boots to bios after each setting when overlcocking in order to set the xp chips at the desired settings.

here are a few problems that i found to be the main cause of my avoidance with the xp chips running under smp environments.

Now, I have an MSI K7DMaster-L mobo flashed to the 1.5 BIOS. On the PCB it says VER:1. I have tried clearing CMOS etc, but the multiplier adjustment wont work. The only mod I have made to the chips is to close the last L5 of each to enable SMP operation. This seems to work fine except they still show up as XP1700 in the BIOS bootup screen, not MP1700s. Also, Windows sees two XP1700s, not two MP1700s.

I've had that issue creep up before. For me, the solution was to set the multiplier to the default (in your case, 11,) then reboot. Watch it recognize them as MP1700's, go into the BIOS again, change the multi to 11.5, repeat. Keep doing this, until you're up to 12.5.

Multiplier adjustment takes two reboots to take hold on my board but it's consistent - always been that way. I'm surprised that you're not seeing at least a consistent behavior pattern

the thread can be located here.

http://forum.oc-forums.com/vb/showthread.php?s=&threadid=178397

so it comes down to this; if you like experimenting or do not have a problem with adjusting to the way your rig works, then the xp trick is not such a bad thing (actually a good thing:eek:) to undertake, and will save you considerable amount of money.

if you dont like trial and error, and piece of mind is more important to you in the smp world, then i would spend the extra $100 dollars or so to get the mp chips. you spend more money, but save time and effort on two things, 1) you dont have to perform the multiplier unlock trick and 2) you dont have to perform the smp operation mode trick either.
 
cmcquistion said:
MP's will run the same temperature as XP's at the same voltage.


The MP's do actually run cooler than the equivalent XP:

The XP2000's have a maximum thermal power of 60.3W and a typical thermal power of 54.7W.

The MP2000's have a maximum thermal power of 58.2W and a typical thermal power of 52.8W.

As stated, the MP's are cherry picked and are basically downgraded (underclocked) XP's. This ensures that the MP's are super stable at their rated speed.
 
theotherphil said:



The MP's do actually run cooler than the equivalent XP:

The XP2000's have a maximum thermal power of 60.3W and a typical thermal power of 54.7W.

The MP2000's have a maximum thermal power of 58.2W and a typical thermal power of 52.8W.

As stated, the MP's are cherry picked and are basically downgraded (underclocked) XP's. This ensures that the MP's are super stable at their rated speed.

I stand corrected.
 
Back