AMD's X2 A Failure Over At Tom's

[ProviaFan]

There is something very wrong with the way the benchmark was set up, there is something very wrong with the OS scheduler, and there is something very wrong with your ability to understand these things, Porkster. :roll:

 

[DaveSimmons]

Originally posted by: porkster
For some percentages as of, GMT-10 20050607:05.55am

Thread 1 : AMD leads by 19.12% (lite task)
Thread 2 : AMD leads by 49.01% (lite task)
Thread 2 : AMD leads by 10.52% (moderate-heavy task)(fps, not interrupted session loop count!)(AMD 38fps, Intel 34, usually neck and neck.)
Thread 3 : Intel leads by 94.44% (heavy task)

That makes no sense. AMD completed 1881 runs of the FarCry demo to P4's 1368, leading by 37.5%. Performance here is completed runs, FPS in those runs is secondary.

But as others have said THG is supposedly running a stress test not performance benchmarks. Performance is not directly comparable with this test setup since the P4 is eseentially running 4 applications while the X2 is only running 3.

 

[Sqube]

Can we have some reinforcements for the "For fvck's sake this is not a head-to-head comparison" crowd please?

To be honest, I just find it bizarre that the Intel is so much better that it can remain even remotely close in three tests while being almost twice as good as the AMD chip in a fourth. Wasn't there an extended period when it looked like the Intel chip wasn't doing anything? But all this is besides the point.

You're trying to use information that we can't make sense of due to discrepancies whose origin we can't even guess. porkster, you're counting your chicks before you even bought a goddamn breeding hen.

Let the test end, let Tom come back and tell us what the fvck happened with reboots, Intel CPU usage discrepancies and things like that, and then you can start drawing irrational conclusions about how this test signifies the death knell of AMD. Seriously.

 

[Steelski]

I wannapoint out that people saying that the cheap intel CPU will only perform slightly slower is a bit of a problem to me.

Infact what seperates the top end Intel CPU with the next one along?

Lets point it out....
HYPER THREADING.

The bottom end dual cores do NOT have hyperthreading.

http://www.xbitlabs.com/articles/cpu/display/pentium-840_3.html

check this yourselves and then tell me that the intel cpu will be performing the same if hyperthreading was not on.
I really think that it would be having the same troubles the AMD cpu is having.

A mid range dual cored pentium D840 against a low end 4400+ X2 or 4200+ X2 would have been a far better comparison.

Anyway,
the theme of the article is stability.
noone here can deny that AMD has the more stability here.

 

[The Pentium Guy]

I question the title of this thread though. "AMD's X2 A failure Over At Tom's". Doesn't sound like a failure to me.

 

[Sqube]

Originally posted by: The Pentium Guy
I question the title of this thread though. "AMD's X2 A failure Over At Tom's". Doesn't sound like a failure to me.

When you're trolling, creating a thread title like that just goes without saying. Just remember that whatever legitimate discussion took place in this thread took place in spite of the op.

 

[Whitedog]

Originally posted by: imaheadcase
People still go to Toms Hardware? weird

People with 7 posts shouldn't make such stupid comments.

 

[Concillian]

Originally posted by: porkster
A (Lite Task) is an application that isn't sophisticated in computation, but will try to obtain as much CPU time that is free.

A (Moderate-Heavy Task). The game thread is fairly moderate in CPU use as most of the game work is done in the GFX cards. The demand is on memory being supplied and basic gaming logic.

A (Heavy Task). The divx encoding requires alot of computation, so is classed as a heavy load.

I'm curious as to why you classify sound encoding as a "lite" (actually it's light, but whatever) task, but video encoding as a heavy task?

Similarly why is video encoding heavy, but data encoding (winRAR) is "lite"?

I don't get why one type of encoding is inherently 'heavier' than another.

I'd also tend to disagree with the statement about how FPS of Farcry is more important than number of runs. The FPS does not appear to be time averaged if you watch for a long time. I looked earlier this morning and the Intel was at 28, then later it was 38, then later still you are saying it was 34. These appear to be relatively instantaneous numbers. The number of runs time averages the more immediate frame rate number. If the Intel machine is consistently only 10% behind the AMD machine, then how does the AMD machine have almost 40% more runs than the Intel machine?

To me the appropriate way to compare is to weight each measure by Intel = Intel / (Intel + AMD) and AMD = AMD / (Intel + AMD) this gives each a weighting of score for each mark out of 1 such that when added together they equal 1. I do not see any reason to weight any test independently, as we don't know if any of the threads have been given different priorities in Windows. Applying weightings would highly skew results if the schedulers are not identical.

 

[DaveSimmons]

Originally posted by: Concillian
I'm curious as to why you classify sound encoding as a "lite" (actually it's light, but whatever) task, but video encoding as a heavy task?

Similarly why is video encoding heavy, but data encoding (winRAR) is "lite"?

I don't get why one type of encoding is inherently 'heavier' than another.

The distinction doesn't make sense in this "not a benchmark" test anyway, because 3 of the charts show cycles completed while the fourth shows minutes encoded, but there is no way to calculate how they should be weighted if you want to compute an overall score.

How many "work units" should one cycle of RARing a MB CD count for in computing a performance score? How many "work units" should one minute of divx encoding count for?

 

[UNCjigga]

Did he forget to download AMD's new driver for Windows XP SP2 AND update the BIOS as requested? Tomshardware is so moronic I bet they forgot.

 

[classy]

Well here's the link to ongoing results, here. The AMD is out in front in 3 out of 4 but the encoding is a bit strange. AMD is easily compressing more cds and is completely far more stable. I would asume there is an explanation for this anamoly in the divx test because its one a just small number of apps it outperforms the X2, but its a small small advantage.

 

[mechBgon]

*drops by, prods the trolls with a stick, and goes back to work*

 

[CTho9305]

So does this mean that if you have HT enabled, and are playing a game, while running folding@home in the background (at idle priority), it will actually cost you performance in the game? With n real CPUs, if you have n real tasks plus SETI/genome/folding, you'd presumably get better performance on the real tasks with AMD?

 

[mircea]

Since I own and AMD system and like them a bit much then I should I think (for the moment) I am an AMD fan(boy).
I didn't have time to read trough all but I wanted top point something out.

Originally posted by: porkster
It's amazing how much AMD fans will overlook the faults or base their purchases on generalisations.

* AMD X2 4800+ is not cheaper than Intel 840EE
* The AMD X2 is to be sold as a multitasking CPU, so yes it is expected to multitask properly, which it is not doing!
* It is unfair to expect the competing CPU to have features disabled. You can?t say disable Hyper Threading so as to stop Intel being shown as the better CPU.
* Both systems started with the same settings and the AMD failed the multitasking expectation in the test. Even if the test had a stability theme, the test still was to place a multitasking load on the systems.

.

Since this totaly unrealistic scenario is targeted as a real world scenario then AMD wins on all 4 categories.

See when I want to do all those things simultaneoslly (pardon my english) I want to be completing them too. What I am more interestes is the DivX chart and the minutes completed since last reboot and not total. See I think that while you argue of the "faulty X2" you don't realize that it's actually going to complete the encoding, but the Intel might never do it since it reboots before it gives me the finshed file so it has to start from the begining. As a real world result that's really bad. That's why this is a stability test primarely. I don't care that Intel encodes more minutes of film than AMD if it never finishes in such a scenario.

 

[thehstrybean]

Originally posted by: Son of a N00b
EDIT:: can we have a filter for all these trolls that keep registering just to bring BS into these forums and trying to disprove something that is not true just to get attention/give a reason for fanboyism.....look how he keeps posting incoherently just to back himself...got something to prove g00ber?

I second that...

 

[mjrpes3]

Originally posted by: porkster

The real question is the AMD X2's lack to do the expected tasks! The AMD X2 has refused to go near the difficult thread of divx encoding which gives Warning signs to all those thinking of buying a Work-Horse computer.

Isn't it only the OS that manages threads? The processor has nothing to do with it. So it isn't an issue of the AMD X2 that has "refused to go near the difficult thread of divx encoding," since the processor has no control over threads. So if Windows managed the threads differently, we would see more equality among the four benchmarks.

 

[Darth Farter]

anyone notice thg decided to replace the intel (gigabyte) motherboard?

 

[Duvie]

I hear GKnot is set to low priority by default..can someone confirm this???? If so this may have some effect that the AMD does not even want to deal with it...Perhaps a normal setting would force the amd to deal with i more...Pure speculation but it may explain a few things...At low prioriy it is like FH which plays nice with everything...

 

[bigboxes]

and yet you all keep feeding the troll...

 

[Zebo]

Good lord what kinda slop troth did you crawl out of Porkster? Starting this flame w/o a clue of what you're talking about.

"AMD's X2 A Failure Over At Tom's"

The intel box has crashed at least 5x, AMD zero crashes up 93 hours straight since i post this. Who's failing again? They even swapped MOBOs on the Intel setup to Asus, crashed again, this is because Prescott has 31 documented hard lock bugs. You did'nt know that did ya?

What is showing is that the AMD multitasking is faulty. The AMD X2 CPU's push aside tasks that they find difficult and result in low priority.

No it's showing windows task scheduler is faulty and a stability test which intel is getting trounced. If you want to see X2's dominate display look at some of the reviews, where P4 gets trounced in all the benchmarks including multitasking, including DivX. Toms is there too.. Calls the X2 "the king" for good reason.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2410

http://techreport.com/reviews/2005q2/athlon64-x2/index.x?pg=1

http://www.hothardware.com/printarticle.cfm?articleid=682 Overclockers special

http://www.xbitlabs.com/articles/cpu/display/athlon64-x2.html Overclcokers special

http://www.amdzone.com/modules.php?op=m...index&req=viewarticle&artid=124&page=1

http://www.lostcircuits.com/cpu/amd_x2/17.shtml

http://www.tomshardware.com/hardnews/20050509_170947.html

http://hardware.fr/art/imprimer/571/

http://www.hardware.no/print.php?artikkelid=15703

http://pclab.pl/art14865-7.html

http://www.tweak.pl/artykul/athlon64_x2_4800_4

http://www.sudhian.com/showdocs.cfm?aid=672

http://www.motherboards.org/reviews/hardware/1513_1.html

http://firingsquad.com/hardware/athlon_64_x2_4800/

http://www.simhq.com/_technology/technology_042a.html

http://www.techwarelabs.com/reviews/processors/amd_x2/

http://www.pcworld.com/news/article/0,aid,120750,00.asp

http://www.extremetech.com/article2/0,1558,1813757,00.asp

http://hexus.net/content/reviews/review.php?dXJsX3Jldmlld19JRD0xMTcw

http://www.bit-tech.net/hardware/2005/05/09/amd_a64x2_4800/1.html

We all know AMD CPU's are poor at floating point operations and also heavy work like Divx encoding but in the test it's showing that in one of the tasks of doing a Divx encoding loop, the AMD is struggling to equally dedicate CPU operations for the task.

I know the pig pin is smelly but pull you head out from your arse. You don't have enough knowledge about hardware or software to make you even worth talking too, let alone making proclamations about FPU's and what apps use it. I would suggest a little research before making comments...

AMD's A64's FPU is the best in the business. Totally dominating, other have posted links here's another: Whoops who's intel's daddy: http://www.anandtech.com/mac/showdoc.aspx?i=2436&p=5

Also there are issues with business use reliability regarding AMD's X2 fan running at crazy speeds of 6000RPM.

Bet you wish you would'nt have said this huh After intel crashing 5x and AMD is rock solid 93 hours straight who's having "reliability issues"?!?! Anyway Intel is using a 92x25mm fan which has more throw, of course it can spin at less RPM than a 70x15mm fan sitting ontop of dual core X2, but what does this has to do with stability? Nothing. just more "guessing" from Porky. I can tell you which is quieter dispite the higher RPM...

The only think you're right about in this whole thread is your username, good job. I would suggest a little research before making further comments...

 

[classy]

WEll the Intel box is completely offline now, crashed again.

 

[Sunner]

I didn't read the whole thread, but I just figured I'd mention one little thing.
WinXP will NOT see the EE as 4 cores, it will be able to tell the difference between the physical and virtual cores in an SMT enabled processor, and schedule work accordingly.
Seems like a few people are a bit confused about that.

 

[Azsen]

What I want to know is how come they are rebooting? Why can't they run for a good amount of time without crashing?

 

[boran]

Originally posted by: Azsen
What I want to know is how come they are rebooting? Why can't they run for a good amount of time without crashing?

well that's the whole point of the test, to stress the systems out and see which one fails the most.

 

[Regs]

I think Tom was talking to you, porkster, when he stated this in his X2 review and quote, "Future Applications Require Intelligence".