Can anyone help me explain this here? look at the fsb, both same but multi is different like Intel put a 3.0 core and added the second as ONLY a 2.8ghz core? WTF? and yes it is the same when I put it all back to default speeds.
:slapass:
Printable View
Can anyone help me explain this here? look at the fsb, both same but multi is different like Intel put a 3.0 core and added the second as ONLY a 2.8ghz core? WTF? and yes it is the same when I put it all back to default speeds.
:slapass:
could it be because of the EIST? i thought that the cores could underclock individually on a Pentium D. Correct me if im wrong :)
I was under the impression that the new EEs were going to be unlocked, have you tried changing the multi?
Whoops
You mean C1E, I have that disabled in bios, and unless it is un-disableable then it should both be the same. Either cpu-z is screwed or Intel screwed up, I dunno which. i need to find a way to benchmark the cpu's individually then I can see if the second core is actually underclocked. Any ideas how I can bench the cores individually. I have cpumark 99 but there is no option for having it bench 1 core or a second core.Quote:
Originally Posted by K.I.T.T.
There is a option in bios to allow lower the multiplier but only to enable or disable it, I have it disabled now since enableing it gives me no options or anything anyways. this is also not a Expensive edition :p: it is a normal dual core.Quote:
Originally Posted by craig588
Think you can choose which cpu you use in prime95, use its inbuilt benchmark function and see what that says
Is BIOS up to date?
Maybe since they are recent, there are some BIOS problems
That did it prime shows CPU 0 is faster then cpu 1Quote:
Originally Posted by Highland3r
wonder what the deal here is. I will call Intel and raise Hell monday.
[Sat Jun 25 12:54:14 2005]
Compare your results to other computers at http://www.mersenne.org/bench.htm
That web page also contains instructions on how your results can be included.
Intel(R) Pentium(R) D CPU 3.00GHz
CPU speed: 3786.77 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: unknown
L2 cache size: 1024 KB
L1 cache line size: unknown
L2 cache line size: 128 bytes
TLBS: 64
Prime95 version 23.8, RdtscTiming=1
Best time for 384K FFT length: 10.344 ms.
Best time for 448K FFT length: 12.600 ms.
Best time for 512K FFT length: 14.191 ms.
Best time for 640K FFT length: 16.806 ms.
Best time for 768K FFT length: 20.341 ms.
Best time for 896K FFT length: 24.368 ms.
Best time for 1024K FFT length: 27.144 ms.
Best time for 1280K FFT length: 36.098 ms.
Best time for 1536K FFT length: 43.507 ms.
Best time for 1792K FFT length: 51.973 ms.
Best time for 2048K FFT length: 58.455 ms.
Compare your results to other computers at http://www.mersenne.org/bench.htm
That web page also contains instructions on how your results can be included.
Intel(R) Pentium(R) D CPU 3.00GHz
CPU speed: 3786.75 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: unknown
L2 cache size: 1024 KB
L1 cache line size: unknown
L2 cache line size: 128 bytes
TLBS: 64
Prime95 version 23.8, RdtscTiming=1
Best time for 384K FFT length: 11.085 ms.
Best time for 448K FFT length: 13.496 ms.
Best time for 512K FFT length: 15.230 ms.
Best time for 640K FFT length: 17.997 ms.
Best time for 768K FFT length: 21.832 ms.
Best time for 896K FFT length: 26.100 ms.
Best time for 1024K FFT length: 29.093 ms.
Best time for 1280K FFT length: 38.681 ms.
Best time for 1536K FFT length: 46.591 ms.
Best time for 1792K FFT length: 55.701 ms.
Best time for 2048K FFT length: 62.690 ms.
Woah thats really wierd! Dont get rid of the chip, could be worth a pretty penny!
i wonder what Intel's responce is going to be? When i inform them they are either lying to the masses, or producing faulty chips.
On the other hand i wonder what reviewers will say if they found out about this and confirmed this with other chips, wow could put a hurting on intel sales real quick. I should tell them to send me a dual core expensive edition for free as hush money :D
any updates :D
They are closed right now :(Quote:
Originally Posted by ibby
***THIS IS AN ELECTRONICALLY GENERATED ACKNOWLEDGEMENT. PLEASE DO NOT REPLY TO THIS EMAIL.***
Thank you for contacting Intel Corporation. This is an acknowledgement that we have received your e-mail and will respond to your inquiry as soon as possible. If you haven't already done so, you may wish to visit our technical support web site at http://support.intel.com and our corporate web site at http://www.intel.com for any product, service or company information that may be of value to you.
Again, thank you for your interest in Intel.
I guess this could explain as to why it happens:Quote:
Originally Posted by SlimySquid
Whole review: http://www.xbitlabs.com/articles/cpu...umd-820_5.html.Quote:
Originally Posted by X-bit Labs
I don't think you should call Intel on that one.
umm, so you wouldn't complain if your cpu at stock speed that was sold to you as a dual core 3ghz was actuall a 3.0ghz and a 2.8 ghz? That is how it is, the multiplier is locked at 15 for core 0 and 14 for core 1, at stock speed with watercooling your not gonna try and tell me it is thermal throttling, exspecially when it works the same way all the way to 3.9ghz. Hell if it is supposedly thermal throttling at stock speed at 3.9ghz is should be smoldering and melting according to your idea. If you think that is the case you are missing marbles. This is pure and simple Intel screw up, I just wonder how many other dual core chips are the same screwed up way.Quote:
Originally Posted by Nasgul
oh one part of that story you missed........(The strange thing is however, that the temperature still remained within acceptable range of 75-78C.)
I haven't even broken 40c yet.
After reviewing your screen shots, see how both cores have the exact same FSB and BUS Speed, your CPU supports EIST. Just like K.I.T.T. pointed out earlier.
I for once thought it was what X-bit Labs said: "Pentium D processors do not allow monitoring the temperature of both cores, this feature is implemented only in dual-core server processors". Since that you're getting the "readings" from ONE core ONLY, seemed like the second core was throttling but then again jus like X-bit Labs says: water cooling is a good idea.
Anyway, here's how EIST works: (BTW, I'm not calling Intel on this one)
*
http://i6.photobucket.com/albums/y218/Nasgul/NoEIST.pnghttp://i6.photobucket.com/albums/y21...ul/EISTx14.png
Pentium D 830, 840 and 840XE support EIST, except the 2.8 core because it's at 14X already.
no matter how much I load up eithe cpu 0 or cpu 1 it NEVER EVER leaves 14 as a multiplier. So again what good is enhanced speed step technology when no matter what load the cpu cores are under it never adjusts itself? I have disabled speed step in bios now, and it is still the same. So either A Intel did screw up or B the motherboard is not operating properly by disabling it, or C it can't be disabled. I dunno which but by calling Intel on it, and possibly asus.Quote:
Originally Posted by Nasgul
what EIST really does from Intel and Asus documents........ Setting EIST to auto then saving bios, and going into power settings on screen saver in windows will allow you to adjust it so the cpu will save power. Select any power scheme EXCEPT, Home/office desktop or always on. After you adjust the power scheme you may notice the CPU Internal frequency lowers slightly when CPU LOADING IS LOW.
So what you are trying to tell me my cpu is doing goes against EVERYTHING on both intel's documents explaining it and ASUS documents explaining it. Since running various benchmarks and playing games and running prime95 has yet to get the cpu 1 to go back to where is should be.
let us know dude :confused2
Just thought I would let you guys nknow that I am having exactly the same problem the only difference is that I have an ABIT AW8-MAX and there is no option in there for EIST here is a screenie I have it overclocked 290mhz but it does the same at stock setting
http://img152.echo.cx/img152/9809/screenie0yr.th.jpg
in your pic, the vcore on the second proc is higher then on the first...
yeh I dont know why that happens sometimes I cant work i out
How are you loading it? mabye it only thinks it needs 1 cpu to do the work it has. Just like with HT you need to run dual prime to use more than 50% of the cpu,you might need 4 prime95,s running to use all of both cpus.The 14 multi looks like the cpu at idle not throttle.What does throttle watch show? :)
Bench both cores separatly (through Windows Task Manager) and see if there is a difference between performance of those cores.
Run 2x superpi together (two installs) so that both cores are 100% loaded. Then compare the two results :)
no yours is the C1e enabled, if you go into bios and disable that you should see both multipliers at 15. But atleast your second core is showing 15 for multiplier where as mine never shows anything but 14, that is why I still say this proc I got is hosed from Intel.Quote:
Originally Posted by nando19
Your Prime95 benchmark run confirms that the cores are running different speeds.
(58.5s*3787MHz)/3534MHz=62.7s (Result is identical with the actual result in Prime95 for the core 2.)
I have disabled C1EQuote:
Originally Posted by SlimySquid
Is it possible that windows is doing this or some windows program/driver is downclocking one core?
Can someone test this under, lets say, server 2003 or linux possibly? I just don't want to jump on the "Intel screwed it up" wagon until this is verified.
EDIT: Even 98 if thats even possible...
well there may be another setting in bios then cause core 1 being lowered is a easy spot of that. if I enable it in my bios then both my cores will be running multiplier of 14. But having all those disabled my core ) shows 15 for multi and core 1 shows 14 a almost clear sign something is not right with cpu. Yours is reversed, your core 0 is 14 multi and your core 1 is 15 almost a easy spot the Intel power saving are kicked in on yours.Quote:
Originally Posted by nando19
maybe cpuz read it un-correct ?
can you try with other app ?
He already said that Prime95 confirmed that they're running at different speeds.Quote:
Originally Posted by Andrewv
Here is another link but this time it is throttling both cores with it under load from super pi and it is definatly set to 15x multiplier in the bios I think that it is bios pains from this Abit Motherboard
http://img22.echo.cx/img22/1821/bob9jo.th.jpg
Yep! the core 2 is running slower then the core 1.
Core1
http://venes.org/bilder/dualcore/core1pentiumd.JPG
Core2
http://venes.org/bilder/dualcore/core2pentiumd.JPG
What 2 do?
How to fix?
They shuld run at the same speed both i guess?
This is what Intel had to say about the whole ordeal.
(((Intel(R) does not use nor recommend benchmarks. I need you try the processor on another motherboard. If the problem persists please contact an Intel(R) Technical Support Center via telephone for a replacement. The North America Technical Support Center is 916-377-7000. If you live in another country, you will need to go to:
<http://support.intel.com/support/9089.htm> to find the Support Center for your country.)))
I really want to get atleast 1 good review site to look into this, then maybe it will catch the eye of intel and they might tell us what is going on.
Send e-mail to every review site that there is, maybe some of them make a fuzz about this and then we might know.
Email theinquirer, they like to put up stories which show Intel in negative light :)
I loaded up 2k pro and it is same, can't do 98 since it is not a multi core enabled OS.Quote:
Originally Posted by ahmad
I have emailed, hardocp, anandtech, xbitlabs, theinquirer, and no responce.Quote:
Originally Posted by Hans.Gruber
I guess Intel gets to get away with this then, there is onviously something happening in the chip that they are not disclosing.
Why else would every result ran so far show core 2 being clower then core 1 at every benchmark thrown at it? I am at a loss to what is happening and I wish we could get to the bottom of it.
:cheer2:
Have seen the same over on OCforums, someone posted about their new chip, one core shows 15* the other 14*
Maybe its an issue with a certain batch of chips?
Cornel
I am still confused.The screeshot shows core one under load in task manager and core 1 at 15x then the second screen shows core 2 under load in task manager but you are still showing core 1 on cpu-z and as it is not under load it is at idle of 14x.looks right to me.
I think you need to run 4 apps to get full load.There is also a rumor that windows xp might not have good support for dual core.
nm, that's weird
You think that's rare? many prescotts throttle stock, inside a hot case, very unlike a reviewers station wide open, now you got two!!! It's true they should have advertised it: Intel Pentium D 830 up to 3ghz processor, sometimes® but it's not "faulty" it's a "feature".Quote:
so you wouldn't complain if your cpu at stock speed that was sold to you as a dual core 3ghz was actuall a 3.0ghz and a 2.8 ghz?
Get better cooling should fix most issue's at stock speeds.. Thermalright XP90, Xp 120 with a high throw, high pressure, 32mm wide fan like panaflo H series will fix you right up.:)
Overclocking like that? And really want to stress and use both cores at that true speed ~3.7Ghz? better get phase.
you are incorrect, obviously you have not read the entire thread and do not know the architecture of the cpu, if you think that is a feature and it is normal.Quote:
Originally Posted by Zebo
Ok so I should get some air cooling cause you said so, because it would be better then my water cooling I am using now? Oh so I should now get phase cooling cause my watercooling keeping it at 48c fully loaded is not good enough for you?Quote:
Originally Posted by Zebo
I think you need to do little more research.
:slapass:
That would be a ok explanation, but how do you explain all the benchmarks showing core 2 being slower at everything then core 1? You can't even begin to tell me that prime 95 and super pi are not fully loading the cores up.Quote:
Originally Posted by fordf250
Load up Sisoft Sandra...it'll tell you want the Minimum and Maximum multiplier of the chip is.;)
The results should be the same for each core. 15x Maximum and 14x Minimum. If not, I'd say there there is definitely an issue with the chip.
My $0.02...;)
My cpuz gives me the same readings with a 830D.
Now the odd things are .. if i put 2 folding clients on it, the readings in all programs give me 2x3ghz, snm, throtlewatch,didn't tried anything else. Throtlewatch, if i stpo one it reads the correct 2800 on one and the one that is on load 3000. If i load both clients it reads both at 3000.
Cpuz never gives me the second at 3ghz. Maybe its a problem with this program. I Tried superpi and the diference is 1sec, i dont consider that very significant.
Anyone got any more sugestions of programs to check the speeds?
I've managed to put the right clock spees in cpuz and every other aplications ... just run systool and go to the Intel Enhanced Speedstep tag and click on 15x and write ..
I think this may be a bios related problem. I'm using 0408.
It does sound like a BIOS issue to me as well. It's as if you have EIST enabled even when you disabled it in the BIOS. Geesh, something is wrong and it should show two 15X processors unless at idle or very light loads. Compare single thread performance to a 2.8 or lower.Quote:
Originally Posted by SlimySquid
Donnie27
roflQuote:
Originally Posted by IvanAndreevich