Socket 939 Sempron found........

[SlitheryDee]

Looks like there's some sort of additional window on the amd system. Maybe an error notification?

Edit: I think the amd box is still up, but it's having problems with one or more the programs running on it.

 

[Duvie]

Its 7:48 in germany those stupid fvckers should have long fixed it......Unless they are not in a hurry....

 

[Duvie]

Originally posted by: SlitheryDee
Looks like there's some sort of additional window on the amd system. Maybe an error notification?

Edit: I think the amd box is still up, but it's having problems with one or more the programs running on it.

Dont usually windows error boxes have a red "X" or something??? I think it could be like a network connectivity warning popup.....I think the network card is f'ing it up and Toms is not in a hurry to fix it.......They probably will reboot it and count it against the AMD system....

It may be an error after further look....Like one of those send report screens.....

 

[AnandThenMan]

Both systems had the same white box a few minutes ago. You'd think they could use something better than a 50 pixel webcam for crying out loud so we could actually SEE the screens.

 

[Sinker]

Oh dear!

I see Porkster found another forum to spout his pig sh*t all over.

 

[boran]

On HT:
now I'd say even more, if the divX thread is actually set to lower priority than the rest, it shows a windows sheduler problem with the Intel processor. because the lower priority thread takes cycles away from the higher priority one (a big no no, it's lower for a reason). but this is all speculation. but when the divX thread is on a lower priority level it shows that windows XP is not dual processor HT aware and just thinks there are 4 physical processors.

imho something to watch out for. (and thg should confirm/deny that the DivX thread is running in lower priority than the rest)

 

[mehmetmunur]

I think that the only thing TH has proven from this stress test is that the Intel platform is not ready and the AMD?s choice to use Socket 939 was very wise. This is more of a comparison of the platforms than the CPUs, and to concoct this ?live stress test? to prove that point is very roundabout, if not a waste of time. I think that we are just encouraging TH by visiting their website. Also, I feel very bad for Duvie -- and the rest of us -- who had to endure porkster, though at times it was quite hilarious.

 

[imported_Opteron]

Hi.

Did anyone notice that stats jumped for AMD rig ?
But during that 'AMD stats not updating' period Intel still managed to close the cap in most benches.
Divx is now almost 10% done what it is in Intel..
Maybe they try to make AMD fans silent by giving them little hope.

About that error message in AMD.
I think that it was "send error report to M$" popup, but now it has disappeared.
I wonder what those other popups where ?
Intel and AMD had similar boxes, and those did look like a taskmanager.
Maybe they were changing priorities so Intel could catch up on Winrar ?

Originally posted by: Sinker
Oh dear!

I see Porkster found another forum to spout his pig sh*t all over.

LOL.. I wonder if Porkster is FUGGER undercover, it would explain many things..

-> http://www.xtremesystems.org/forums/showthread.php?p=904137#post904137


[edit] Man i had trouble posting this, never used this kind of popup forum..
But quite nice system, i just have to get used to use it.

 

[Continuity27]

I don't care who Porkster was/is I just want him to take his lies and misdirection elsewhere...

 

[Sinker]

Originally posted by: Continuity27
I don't care who Porkster was/is I just want him to take his lies and misdirection elsewhere...

My preference would be for him to take it nowhere.
No one should be subject to such blind ignorance.

He has got an appropriate handle, though, he's very thick skinned...

Anyway, back to the test, looks like the AMD machine's settled down again. It probably just wanted to be noticed so it threw a bit of a hissy fit.

 

[imported_Frallan]

Originally posted by: porkster

Originally posted by: michaelpatrick33Wouldn't a fair stress test take into account downtime of failed parts?

I agree, but the starting line isn't fair anyway. AMD boards have been out for sometime when you compare them with just hitting i955x based boards. Also Tom didn't get the Intel system setup properly before the test.

The analogy is like having a race between a existing track winner and a car just rebuilt from spare parts.

Tom Hardware crew failed to test the systems out before the stress test race. They are real lamers.
.

Which is exactly the point some of us has tried to make. X2 vs. P4D is just ....

And as for the Lurkerstatus blown for nothing in my prior post U just had to look at the temp lines last day I did not get a screnshout of it. And That the Intel MoBo has 2 nics and it measures on the other I know but that doesnt explain why the traffic shows up on the week summary And there is still no info on the apps.

 

[imported_Frallan]

Originally posted by: PetNorth
oooohhhhh here he is, the indescribable Porkster!!!! flooding this one too!!!!!! go away troll!!!!!!

Guys, read and laugh (by Porkster):

http://www.community.tomshardware.com/f...mber=569253&page=4&view=collapsed&sb=5

http://www.community.tomshardware.com/f...mber=569349&page=2&view=collapsed&sb=5

http://www.community.tomshardware.com/f...mber=570108&page=2&view=collapsed&sb=5

http://www.community.tomshardware.com/f...mber=570102&page=2&view=collapsed&sb=5

ROFL it seems tom me (reading that and whats in this Forum) as if INTEL has found a really cheap way of doing marketing. It does not matter that most of us know better - a number of noobs will read his posts and decide. 1 Troll on the paylist doing damagecontroll on-line all day would be a marvelous investment orINTEL and this guy is pretty good at it .

 

[Duvie]

[/quote]


Well lets see how the pace is going...did it speed up or slow since last look...for me i has been 11:45 hours...so I was expecting the Intel cpu to have at this point 2 per 1-1/2 hours for a lead now of 16 but yet it only has a lead of 9...so the pace appeared to slow a bit....Did it raise elswehere???


..................CD's .............RAR...............FC.............DIVX

AMD...........200.................1648...........988..............40
Intel............200................1082...............697............930
difference....0..................-566.............-291...........+890

at 21:00 uptime for INtel

..................CD's .............RAR...............FC.............DIVX

AMD...........253.................2050...........1227..............60
Intel............262................1422..............881............1210
difference....+9..................-628.............-346...........+1150
..............gained 9............lost 62.........lost 55..........gained 260

notice how caculated winrar should have been 104 in 12 hour day yet only 62 of that was realized...HMMMM....interestng how HT seems to refocus its efforts elsewhere no consistency...It is like it knows it has the lead in cd encoding ad puts it efforst elsewhere.....Gaming was estimated to be 64 more down and is clse to that at 55.....I dont know if Ht is that intuitive, sporadic or what but its inetersting nonetheless at its randomness....

DIVX is running more like a 200:10 pace now....

overall current lead by percantage

.................CD's .............RAR...............FC.............DIVX

...........INT 3.6%.........AMD 44.2%.....AMD 39.3%.......INT 2016%

[/quote]

..................CD's .............RAR...............FC.............DIVX

AMD...........323................2420.............1359..............130
Intel............317................1720............1060............1450
difference....-6..................-700.............-299...........+1320
..............LOST 15...........lost 72.........lGAIN 47..........gained 170


OK this is interesting...Network problem solved and evidently the AMD was still working all that time...

They took back the CD test...Gained big in winrar...lost ground in Far Cry overall....But somehow gained a Divx backbone...How the fvck did that happen with minimal loss all around...What the fvck are those morons doing??? It was a 200:10 ratio and now it is like 110:10...That is a big change

 

[Duvie]

Originally posted by: boran
On HT:
now I'd say even more, if the divX thread is actually set to lower priority than the rest, it shows a windows sheduler problem with the Intel processor. because the lower priority thread takes cycles away from the higher priority one (a big no no, it's lower for a reason). but this is all speculation. but when the divX thread is on a lower priority level it shows that windows XP is not dual processor HT aware and just thinks there are 4 physical processors.

imho something to watch out for. (and thg should confirm/deny that the DivX thread is running in lower priority than the rest)

I believe that was confirmed with a statement at Toms....

It is also confirmed that in that GKNOT app it is low priority...

I can also confirm as an EX P4 owner and tester of HT that priority process does not work right with HT...I could change priorities and t had little to know efect.....

Any P4 can test this by running FH which is a low priority app and the any other thing like gaming...Run gaming first and record the fps with fraps...then launch FH and do it again...The fact is in a normal single core non HT system the game will take all cycles and with FH running you shold have less the 1-3% drop in fps if that....With the HT on my system I oculd have a drop as much as 10%+...IN some other apps (oter then gaming) combined with FH the drops were in the 30-40% range and the FH was churning away and the thread scheduler appeared to ignore and process settings I was using or trying to change...

 

[Sinker]

Originally posted by: Duvie
I can also confirm as an EX P4 owner and tester of HT that priority process does not work right with HT...I could change priorities and t had little to know efect.....

I guess Microsoft didn't want to (couldn't be bothered to) change the scheduler in Windows to use HT properly. Windows just sees the extra logical core as a physical core, therefore assigning a low priority task to it's own 'core', when doing that slows down the other 'core' as they are in fact the same. Perhaps the scheduler should be more intelligent and see the extra core as a lesser core and only assign things to it if it doesn't foobar the first, true, core.

Mmm, that made sense in my head...

 

[imported_Frallan]

Ok on the German Toms there have been an up date.

First of all its no longer a Stability test...

The Title is now (on last update) Competition of performance.

Secondly Its now starting all over again when they have finally gotten the INTEL system stable - saying basically that it was all a heat problem and mentioning only the 955 chipset (allthogh they do say that SLI was removed from the testing).

3rd:lyThe HT will be disabled on the INTEL...

I acctually think thay are trying to save some data (and face here) from this totally failed test. Its going to be interesting what they come up with next.

On the conclusino of previous runs the say that AMD and INTEL tie 2 AMD wins games and Intel DivX - which I must say is a VERY interestin conclusion on a STABILITY test LoL

 

[boran]

Well it is a fact winXP is more HT aware than Win2K which really saw both virtual cores as real processors, but I tought winXP had a good dual virtual core HT awareness, apparently not so according to your statements. and if the encoding thread is set to idle, also not so with 4 virtual cores. I think HT makes things really complicated for the OS scheduler.

after all you have to take into account what core1A is running in order to see what core1B can do (I think you cant just ask a core "how high is your load ?" from the OS point, might be wrong on that).

 

[imported_Frallan]

Now... I know of the dificulties and of the issues with a test like this. But I just don't trust Toms any more and specially not this test. To much undocumented wierdnes nad to many changes under way.

Now to the issue of this post:

PLEASE PLEASE PLEASE Anand figure out a way of making one of these test. I think there is about 50 ppl here that would love to help designing the test if any help would be needed (but I trust U to do on your own just as well). The test at toms just cant be allowed to be the last word. There are a huge number of ppl reading Toms who still belive everything thats in there and I would really love to have a comprehensive test that I trust to show them.

 

[Viditor]

I have to say that this has been an embarassing test for Intel (not that it was necessarily their fault...).
The real market for these chips will be workstations and low-end servers, where reliability is paramount!
With all the mobo changes and reboots for the Intel (and Nforce) platforms, IT guys are gonna think twice about adopting...

Edit:I forgot to mention that the power issues were highlighted as well...

 

[Topweasel]

I think Farcry Just Crashed on the Intel. Either that or they minimized it.

Edit: Looks like it was the end of a loop (never seeen it yet) or they minimized it.

 

[Sinker]

Originally posted by: Weaselboy
I think Farcry Just Crashed on the Intel. Either that or they minimized it.

Edit: Looks like it was the end of a loop (never seeen it yet) or they minimized it.

I think they minimised it to switch NICs or something. Check the chart, NIC activity has started.

 

[GuitarDaddy]

Although this online test has been a miserable failure up to this point, I think it has opened a lot of peoples eyes about the methods or lack of that THG uses during their review process. And the total lack of credibility in there motives and conclusions. I understand better now why some of Toms reviews seem so far off the mark and don't agree with other credible reviewers results. They have absolutely no clue about control procedures, and their tests are obviously designed with a particular result in mind. If they don't get the result they expected, they just change the test and start over:Q

 

[dug777]

Originally posted by: GuitarDaddy
Although this online test has been a miserable failure up to this point, I think it has opened a lot of peoples eyes about the methods or lack of that THG uses during their review process. And the total lack of credibility in there motives and conclusions. I understand better now why some of Toms reviews seem so far off the mark and don't agree with other credible reviewers results. They have absolutely no clue about control procedures, and their tests are obviously designed with a particular result in mind. If they don't get the result they expected, they just change the test and start over:Q

to be fair to them they had to start over to keep the test fair

Although the constant hardware failures haven't been their fault...i come out of this rather worried at the stability of the intel platform. I would love anand to do a similar test, to see if he can confirm these severe stability problems, it really can't be that hard to do it right can it?

 

[boran]

well imho the stability issues are there because of the massive amounts of currents the mobo has to supply, intel really slapped two presshots together, whereas amd really did quite some re-engeneering with their cores (check some reviews, the power usage for a dual core lies not that much over the high-end single cores) on top of that, being socket compatible means more mature chipsets. but imho socket compatibility is a big plus for AMD, unlike intel they seem to design their sockets with some future plan in mind (remember how long socket A lasted (and even still does to some degree))

 

[SlitheryDee]

Anyone else think that intels "heartbeat" is somewhat more pronounced since NIC activity started?