I think it's about time we got some Intellers over here...

[Lithan]

http://www.xtremesystems.org/forums/showthread.php?t=64950&page=1&pp=25

Long story short. They had the idea of doing a stability/performance compairison between p4d 840ee and x2 4800+. P4d 840ee crashed about half a dozen times and was losing three of four benches. So They decided to make unknown and undocumented "changes", erasing the records of the crashes and AMD's 100+ hr uptime as well as the performance records, and rebooting both systems. So now Intel is winning three of four and both systems have only 4 hrs uptime. Tom's seems much more pleased with (and willing to not destroy the evidence of) these results.

I don't know why Intel sends Tom $1200 cpus... they could send him a crumpled five and he'd bend over just as quickly.

[caater]

problem was with NF4i chipset... guess it isn't ready after all..
i don't recall if they tried only gigabyte, but they finally ended up with 955, removed one gfx card from athlon system and started all over again..

as for perfomance numbers.. it seems odd indeed. i'd be surprised if the 4800+ didn't win all of tests..

[saaya]

moved to wampa :P

[jayjaya29]

I've never liked or relied on Tom's Hardware.

[Aphex_Tom_9]

I don't know why Intel sends Tom $1200 cpus... they could send him a crumpled five and he'd bend over just as quickly.
hahaha, yeah i dunno what's up with that test. but we all know about the AMD's previous uptime, so i say let them continue the tests, but we'll remember what happened

[Mr. Tinker]

The tests have convinced me not to pair a Pentium-D with that Gigabyte board, but your title seems like you're saying "Ha! See! I TOLD you Intel sucks!" when really this is just one processor line. All that can really be demonstrated is that the Gigabyte board sucks with a Pentium-D. It would be a stretch to say that it proves that Pentium-D with NF4 is a bad combo, but I might give it that.

[Lithan]

I dont think it can be blamed on the chipset entirely. It was always crashing as temp maxxed out, and to try to stabilize it, Toms has lowered vcore below stock vcore (or so the people who have watched it closer then I have said). Right now Tom's is getting 90*C temp readings off it. It's too hard to follow anymore when they delete/edit stuff every hour or so. Really all it serves is a demonstration of the P4D's heat issues and tom's whoring. I wouldn't take anything away from the bench results. The first ones showing amd Dominating were skewed by the Intel's massive downtime, and the new ones are complete bull that tom probably just makes up.

Moving on, I've seen plenty of reviews running Single core P4's without any problems on NF4, so at the very least this can't be written off as an instable chipset. The most blame that can be shifted to Nvidia is calling their support of the P4D poor, which at this point would be very rare, seeing as they've had months to work on support and the chipset was very likely designed with dual core support in mind. I'm much more inclined to believe that stock cooling simply wont cut it for the Dual core prescotts, considering the heat output numbers we've seen, it's pretty easy to understand.


Tinker, I never said Intel Sucks. Although I do think that "Ha, I told you P4D sucks!" would be a pretty good title as well. Personally, I think Dothan is perhaps the best chip available right now for my uses. I do think that Netburst was a mistake for intel, like a balloon. With Northwood, netburst wasn't really being pushed, so it's flaws were well hidden, and AMD wasn't offering much competition since it was so deep in R&D for A64 that the XP line was basically neglected except minor speed/fsb/cache bumps. Then prescott came out. Balloon got inflated, stress marks showed. Right around then AMD released A64 and pretty solidly beat Prescott in all aspects except performance of many encoding apps, which I figure is a concern to a very small number of people relatively. Then with the P4D, they basically took all the problems with prescott, multiplied them by two and since they had already had HT, they didn't even gain much in the HT/SMP enabled encoding apps that had been their previous strongpoint. I'll ignore the X2 beating them in damn near everything because I agree that the X2 is priced at such a point it's a fairly pointless buy unless you absolutely need the best encoding performance available (ie: the same situation the old EE chips were in).

But Im rambling and that's another thread.

[Mr. Tinker]

I dont think it can be blamed on the chipset entirely. It was always crashing as temp maxxed out, and to try to stabilize it, Toms has lowered vcore below stock vcore (or so the people who have watched it closer then I have said). Right now Tom's is getting 90*C temp readings off it. It's too hard to follow anymore when they delete/edit stuff every hour or so. Really all it serves is a demonstration of the P4D's heat issues and tom's whoring. I wouldn't take anything away from the bench results. The first ones showing amd Dominating were skewed by the Intel's massive downtime, and the new ones are complete bull that tom probably just makes up.

Moving on, I've seen plenty of reviews running Single core P4's without any problems on NF4, so at the very least this can't be written off as an instable chipset. The most blame that can be shifted to Nvidia is calling their support of the P4D poor, which at this point would be very rare, seeing as they've had months to work on support and the chipset was very likely designed with dual core support in mind. I'm much more inclined to believe that stock cooling simply wont cut it for the Dual core prescotts, considering the heat output numbers we've seen, it's pretty easy to understand.


Tinker, I never said Intel Sucks. Although I do think that "Ha, I told you P4D sucks!" would be a pretty good title as well.

Agreed. Completely. Here's to X2 and dual core Dothan on the desktop. Cheers? :toast: ?

[Lithan]

Cheers man. Though Im not a fan of either in the near future. Come summer 2006-spring 2007 when i expect to start seeing "good" Multithread optimized games, I'll be right up front debating which one to buy. If COD 2 supported Multithread and actually gained anything from it, I'd buy an X2/Dual Dothan the second I saw one under $300.

[MaxxxRacer]

Well after this massive display of whoring its safe to say i will no longer trust a single word THG says.. though this is not to say i trusted them before.. it was 1 in 20 words i could believe before..

they should truly be ashamed of themselves for this. if it was switched and amd was having the problems intel is having id say the same thing. i may not like intel but i give credit where credit is due.

as mr. tinker said cheers to X2 and hopefully Dual core dothan.

[Mr. Tinker]

Cheers man. Though Im not a fan of either in the near future. Come summer 2006-spring 2007 when i expect to start seeing "good" Multithread optimized games, I'll be right up front debating which one to buy. If COD 2 supported Multithread and actually gained anything from it, I'd buy an X2/Dual Dothan the second I saw one under $300.
Being an animator, I'm a fan of both about 5 years ago. It just don't get fast enough for us. However, there is hope:

http://www.worley.com/fprime.html

http://www.saarcor.de/

[Lithan]

Ahh True dat. You will like Dual core. I wouldn't even notice it except in my checkbook.


Are those programs you work in, or programs meant to design dual core optimized games?

[MaxxxRacer]

mm i seriously want a daul core cpu.. would be great for Brazil RS.. that sucker makes renders take forever.. i rememer a 17hour single frame render with my AXP 2.3ghz... X2 at 2.8ghz (a little OC) would bring that down to around 6 hours prolly.

[[XC] leviathan18]

i think they killed their reputation they should say sorry for all the comunity because that is a joke

[Lithan]

Update for anyone currently blaming NVidia. The rig was switched to an i955x early yesterday or late monday. The problems continued unabated. So it was NOT the chipset at fault.

[trance565]

3rd mobo, and still having troubles

[Holst]

I think this goes to show that THG are at best imcompitent, at worst just plain liars.

Im pretty sure if we gave 2 systems to most of the senior overclockers here they would get them running stable for days no problem... some of our folders manage this with loads of overclocked systems.

Seems like a totaly pointless exercice to me if the hardware they are using is untested and badly installed/badly chosen.

[MaxxxRacer]

ive noticed one of the biggest issues toms has is installing the hardware improperly, thus cuasing it to crash. yes i know this is sad.. but it IS true..

[bxa121]

huh?!!
THG is intel biased?

i had no idea!! - no really :(

where do you guys suggest i subscribe to from now on???

[[XC] leviathan18]

^^ yes and a lot they have like 4 days trying to make intel rig looks better than the amd system they even restarted the whole thest after amd had 100hrs+ and intel like 50hrs