Sign in with
Sign up | Sign in
Your question

Stress Test MK II

Last response: in CPUs
Share
June 3, 2005 11:38:29 AM

<A HREF="http://www.tomshardware.com/cpu/20050603/index.html" target="_new">Here we go again!</A>

______________
Bow Down (and pray to your God)

More about : stress test

June 3, 2005 12:11:47 PM

Let the controversy begin!

Some people are like slinkies....
Not really good for anything but you cant help smile when you see one tumble down the stairs.
June 3, 2005 12:27:39 PM

BTW is the DivX encoder running on X2 ? still shows 00 mins run which doesnt look right !!
Related resources
June 3, 2005 12:36:18 PM

Yeah, I was just about to post that!

In all the tests which appear to be working correctly, AMD is in the lead / equal to Intel.

______________
Bow Down (and pray to your God)
June 3, 2005 1:46:35 PM

i think they realised that divX encoding was not running on X2 system.. and they started it a few mins back... the counters now showing it as running for 10 mins :) 

i guess X2's arch is better... so i dont think intel dual proc is gonna be pulling on a fast one... (the only one i think it can is video encoding)


Regards,

Hermit Dave
June 3, 2005 5:09:02 PM

Quote:
Just looking at the clock speed suggests that Intel's dual core processor offers greater performance. But even Intel has stopped talking about clock speed and now uses code designations.

This is THG lingo for "Yes we are Intel Fanboys, but we want to appear, I mean, remain unbiased"
June 3, 2005 7:23:24 PM

Ok, as everyone here knows I have a strong bias towards AMD and have built exclusively AMD systems for at least a year now.

That said, it is somewhat embarassing how much intel kicks our ass at video editing. I know that this is the only place on the stress test that intel is currently winning, but they are not just winning, they are raping.
June 3, 2005 8:19:35 PM

Why do benchmarks continue to use utilities and applications that suit AMD CPU's, or Intel for that matter.

I would have thought it more general a test if they used Windows XP Zip, rather than WinRar.

Any programs that are developed on Linux and have a windows joint build are most likely programmed/compiled from a AMD perspective, so they should also be avoided.

3DMark2005 Graphics test in a loop would have also been more fair a test than Farcry. Least 3DMark is testing latest graphics formulas and based on general coding for Direct X, than a sales pitch for particular branding like 3DRealms or the makers of Farcry.

.
June 3, 2005 11:03:38 PM

Quote:
That said, it is somewhat embarassing how much intel kicks our ass at video editing.


lol, "our" ass?

Anyway, that benchmark is obviously malfunctioning - it has beenat 10 min for hours.

______________
Bow Down (and pray to your God)
June 3, 2005 11:05:27 PM

Quote:
Why do benchmarks continue to use utilities and applications that suit AMD CPU's, or Intel for that matter.


They aren't built to suit a particular processor, each architechure has its own pros and cons.

Quote:
Any programs that are developed on Linux and have a windows joint build are most likely programmed/compiled from a AMD perspective, so they should also be avoided.


LOL - wtf?

Quote:
3DMark2005 Graphics test in a loop would have also been more fair a test than Farcry


Why?

______________
Bow Down (and pray to your God)
June 3, 2005 11:17:06 PM

Anyone know why the Amd's 12v rail is so low?
June 3, 2005 11:17:13 PM

Looks like the first hicup. Find it.

Dichromatic for your viewing plesure...
June 3, 2005 11:23:09 PM

11.69v is fine esp for a 12v rail.

Dichromatic for your viewing plesure...
June 3, 2005 11:38:45 PM

lmao: INTEL REBOOTS!

______________
Bow Down (and pray to your God)
June 3, 2005 11:56:06 PM

Even with Intel idle, I don't think AMD is going to catch up in DIVX. LOL.

Dichromatic for your viewing plesure...
June 4, 2005 1:10:04 AM

11.69 is a little low. 11.49 is worse. It may still be within the 5%, but only just. I sure wouldn't keep a psu that dropped that far. These are supposed to be high end units.
June 4, 2005 2:36:48 AM

If it were a sensor, I'd agree, but since it is a voltmeter cct, it should be very accurate.
June 4, 2005 3:14:30 AM

The voltmeter cct is on the mobo.
June 4, 2005 5:00:08 AM

WHAT THE [-peep-]

WHY DIVX???

90% of all Good encoders are Using XVID instead, Beside its has Better Quality results it is not Otimized for Pentium4 Macro Architeture.


Shame on you THG :D 
June 4, 2005 7:11:42 AM

Quote:
it is not Otimized for Pentium4

Well duh, you answered your own question. If the Intel chip actually appears to loose, heads will role.
June 4, 2005 7:54:16 AM

Woot Intel is crunching away again, It seems to be for the past hour or so.

AMD seemed to have a few bumps at 6am and 8am but still looks to be crunching away.

Although I thought the DIVX was not running on AMD, it has moved up to 20min done.

Intel can definatly handle 4 major processes better due to HyperThreading. I think amd would win a 2 legged race though.

Dichromatic for your viewing plesure...
June 4, 2005 10:27:57 AM

Quote:
it is somewhat embarassing how much intel kicks our ass at video editing. I know that this is the only place on the stress test that intel is currently winning, but they are not just winning, they are raping.

It's windows that gives less priority to background tasks as default.
Since AMD has 2 CPUs instead of Intels 2+2 virtual cpu's, windows is able to share loads more evenly for Intel.

So it is a windows issue, not "intel winning" becouse it's better.
Guess why they did choose 4 tasks ?

I believe that AMD system has defective PSU or something, it's not normal to so low voltages.. (11.2v)

ps.i wonder why there is no power consumption separately for Intel and AMD...
June 4, 2005 10:42:25 AM

This test is pointless. Without setting the process of the Divx thread as a high priority, the test is flawed. It's clear the AMD CPU is working on the other tasks alot more than it's Divx thread.

So if you compare the AMD lack of ability to perform a general work load, the AMD is failing big time. The Intel 840 is doing all the tasks very well, equally.

I would also like to see test applications/utilities that are standard, like Windows XP Zip-Folders, 3DMark2005 graphcis loop, etc.

.
June 4, 2005 10:51:18 AM

It seems AMD is failing the workload.

If you examine the inabilities of the AMD to do all four tasks equally, it appears that the AMD is getting congested with it's instruction queue for the difficult task of Divx encoding and the instructions for the other tasks are taking its command.

The Intel is equally doing all tasks, but the AMD is failing to balance the load of the multitasking. This could indicate that AMD maybe the worse in a server environment as one task maybe abandoned whilst others get somewhat faster attention. This is a major hit on AMD's poor design.

.
June 4, 2005 10:53:45 AM

No, it's because AMD is failing to equally do all tasks at once. The AMD is congesting it's instruction queue for the Divx encoding processing thread.

The AMD X2 may need a redesign to fix this major fault.

.
June 4, 2005 12:53:02 PM

I take it you have no idea what you are talking about, no? Why don't you reply to my post, where I quoted you and answer the questions?

Stop flame baiting! :lol: 

______________
Bow Down (and pray to your God)
June 4, 2005 1:23:59 PM

Hey MR. <b>Porkster</b> Plz stop kissing INTEL's ASS

Thank You
June 4, 2005 1:28:26 PM

Quote:
No, it's because AMD is failing to equally do all tasks at once. The AMD is congesting it's instruction queue for the Divx encoding processing thread.

The AMD X2 may need a redesign to fix this major fault.

NO IT'S NOT WHAT IS HAPPENING

If you install auto Gordian knot you 'll realise that when you starting encoding a movie it sets <b>LOW PRIORITY</b> to their process - That is what is going on

If you want to see Real World Divx/Xvid encoding benchs, plz reffers to: <A HREF="http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=241..." target="_new">http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=241...;/A>
June 4, 2005 5:40:59 PM

Oh lordy lordy, X2s are faulty! oh who will save us! lol

this is pretty funny. I see the fanboys are still out in force. This is just liek the last time, amd gets no respect no matter what the result. and fo course when there is any sign of problem we hear some crying for a recall of all things lol. lets all take a deep breath before making such wild claims.

there are more thena few problems with this stress test, just like there were last time. many inaccuracies and strange occurances that are unexplained. its nice for them that they can just start a 'test' and not have to explain themselves.

but im not going to get worked up over it, i think ill prefer to trust the body of testing results from the majority of reputible sites out there and not this one test that is questionable as usual.
June 4, 2005 6:42:26 PM

Quote:
but im not going to get worked up over it, i think ill prefer to trust the body of testing results from the majority of reputible sites out there and not this one test that is questionable as usual.


So ignore the negative and only see the positives?

I also agree it wil be intersting to heard the conclusion of the tests, but I would also like the see some human comments from the test team during it.

.
June 4, 2005 9:27:24 PM

Lol, Tom's Hardware Bribe at it's best.

I'm just wondering, why cant they run useful applications, like Prime95 and that sort of applications instead? Then all that power wouldn't go to waste...
June 4, 2005 9:37:30 PM

Quote:
It seems AMD is failing the workload.

If you examine the inabilities of the AMD to do all four tasks equally, it appears that the AMD is getting congested with it's instruction queue for the difficult task of Divx encoding and the instructions for the other tasks are taking its command.

The Intel is equally doing all tasks, but the AMD is failing to balance the load of the multitasking. This could indicate that AMD maybe the worse in a server environment as one task maybe abandoned whilst others get somewhat faster attention. This is a major hit on AMD's poor design.


Lol, poor design? AMD is the one that has properly designed a dual-core cpu, not just stitched two cores together...
June 5, 2005 3:22:28 AM

For one thing, how can you conclude anything general about the X2 or 840 from this single test using one chip of each? that is the absolute worst scientific method ive ever heard of. Before I start running around cliaming the X2s are faulty as you say or to say anything about the 840s, I gather data from more then a singel source. show me any other sources that agree with your conclusion...


I actually work in a business that can make use of the added power of dual core and luckily they are open enough to not be hooked on dell, so we have several opteron systems running. Do you really think any business IT is going to look at one 'stress test' from a shaky source as evidence that a product is faulty when so many other seperate sources contridict that?

this is nothing conclusive, there are sooo many varibles that you have to admit you cant be sure any of this is a general observation of X2s or 840s. dont jump to conclusions, otherwise you just sound like you WANT to find fault with X2s. How about examaning the test method instead.
June 5, 2005 3:36:54 AM

Quote:
For one thing, how can you conclude anything general about the X2 or 840 from this single test using one chip of each? that is the absolute worst scientific method ive ever heard of. Before I start running around cliaming the X2s are faulty as you say or to say anything about the 840s, I gather data from more then a singel source. show me any other sources that agree with your conclusion...


It better to take into account all test and methods. But it's wrong to ignore the negative results and only see the positive.

One thing we all can say, this test is showing a major issue with the AMD X2 and I'm correct to put into question if the AMD X2 is FAULTY.

.
June 5, 2005 3:47:04 AM

you are ignoring the evidence provided to you. it is not the x2 that is faulty, it is windows. windows sees four processors for intel, and thus it assigns tasks equally. it only sees two processors for amd and thus is not able to manage the tasks properly. if it wasnt for microsoft's lame os, intel would never even have used ht cause it only corrects microsoft's inability to properly multi thread. but please, youre gonna ignore this and go on rambling.
June 5, 2005 4:06:49 AM

Quote:
you are ignoring the evidence provided to you. it is not the x2 that is faulty, it is windows. windows sees four processors for intel, and thus it assigns tasks equally. it only sees two processors for amd and thus is not able to manage the tasks properly.


So you're advocating that customers of AMD's X2 avoid using Windows?

Sorry but to me this is a fault. Even if they can repair it with software, the X2 is not doing to well, sorry.

.
June 5, 2005 4:14:10 AM

Intel Reboots!!!!!! this would be the third reboot for intel. (I think the counter is wrong)
June 5, 2005 4:18:41 AM

The x2 is not doing well? How about checking out the reviews where the x2 beats intel in almost every single benchmark, and then lets compare the amount of power the intel uses. nobody is going to run 4 threads at a time, and the few who might arent going to be using x2s or windows. you just dont get what multi threading is. ht doesnt add performance to intel, it just tricks windows into spreading the tasks equally. since windows cant spread tasks equally with less processors available, it will put processes on hold.

<P ID="edit"><FONT SIZE=-1><EM>Edited by julius on 06/05/05 00:22 AM.</EM></FONT></P>
June 5, 2005 4:56:26 AM

Quote:
it will put processes on hold.

Actually, what Windows does is give secondary, tertiary, etc tasks a lower priority. That way, the task you are focusing on gets the most cpu usage. For those of us who now how to set our own priorities, it doesn't matter. For joe Q, he wants to see what he is doing, move quickly.
For gaming, it helps, because things like virus scan etc get less cpu time.
As far as I can tell, it looks like HT is taking frames from the game to use for Divx. That is bad for me.
The Intel system is doing quite well since for almost 25% of it's up time, it was idling. Too bad it cant stay running for more than 16 hrs at max usage.
June 5, 2005 4:57:43 AM

I know, I have become so tired of all these bastards with there, "Intel is more stable than AMD." Yeah, that hasn't been true since at least the K7. (Thats about the time I started to care about computer hardware.)
June 5, 2005 5:27:30 AM

Quote:
So you're advocating that customers of AMD's X2 avoid using Windows?

Sorry but to me this is a fault. Even if they can repair it with software, the X2 is not doing to well, sorry.


Really, of all your posts in this topic, you seem to completely ignore the point.

The reason for the DivX problems lies in the OS, which is having trouble dividing the load of four processes into the two cores of the AMD.

Now, the only reason Intel doesnt have this "fault" as you call it, is because of the HT.

Now really, you don't need any software patches or kernel fixes to solve this, is all about opening the task manager and raising the process priority of VirtualDub from "lower" to normal, it's that simple. Why they haven't done this already is a mystery, but someone, please hack into that AMD system and fix the priority, since THG doesn't seem to know how do it it...
June 5, 2005 10:10:28 AM

Quote:
Strange, it looks like FarCry on X2 stopped running a while ago.

Now they can start yelling left and right that Farcry stopped running in X2, and ignoring why it stopped (they turned it off)... :wink:

Intel has booted 4 times already ? :lol: 

Looks like AMD voltages are now ok, also temps are lower but so is fan speed.. i wonder are any tasks running, altough CPU usage is still high.
June 5, 2005 10:49:42 AM

Looks like the test is at a dead end road with tasks not running and figures being skewed from inconsistent runtimes.

.
June 5, 2005 1:40:11 PM

Yeah, this has to be THG's worst test ever, with machines running idle for hours.
June 5, 2005 1:52:28 PM

In case that someone believes that Intel 840XE has quiet fan..

<A HREF="http://koti.welho.com/pnystro2/Hauskat_videot/PXE840_Co..." target="_new">Intel XE840 Extreme cooling. (boxed cooler)</A>

Btw. That 3500rpm in THG test makes me suspicious..
..also those temps, since others got +85-90°c @ full load with boxed cooler.
June 5, 2005 1:56:34 PM

OMG, now that's a noisy boxed fan.

I hope they have som kind f noise protection there, that thing sounded like a jet plane.

<P ID="edit"><FONT SIZE=-1><EM>Edited by BePe86 on 06/05/05 10:01 AM.</EM></FONT></P>
June 5, 2005 2:01:17 PM

They must be at a trade show or something.
June 5, 2005 2:05:57 PM

Intel has crashed an app ? <A HREF="http://koti.welho.com/pnystro2/somepics/cam3.jpg" target="_new">Problems at Intel..</A>

[edit] Or is that "System has just recovered from serious hardware error" notification ?
.....it wouldn't surprise me, except that "recovered" part..<P ID="edit"><FONT SIZE=-1><EM>Edited by HansGruber on 06/05/05 12:02 PM.</EM></FONT></P>
    • 1 / 16
    • 2
    • 3
    • 4
    • 5
    • More pages
    • Next
    • Newest
!