Because in biblical times they didn't have enough steroids to pump up goliath any further, but david was so puny that he could put their small quantity of roids (and the rage) to epic use
ndeed it's all about leverage...what u wanna do with your power... just for show or are you gonna use it...
Originally posted by: Andrew1990
hmm, I kind of remember seeing how all Phenom II X4s were supposed to easily overclock past 4.2GHz+ on air. We all know how that turned out.
Even so, I am most likely picking one of these up to put in my system.
I don't recall any source promising 4.2Ghz+ on air other than some dumb fan boys in forums. I've seen a lot of people claim their Phenom II to reach 3.8Ghz on air and that's all AMD claimed it would do, beyond that it was water cooling to get above 4.0Ghz.
Nemesis 1
Lifer
- Dec 30, 2006
- 11,366
- 2
- 0
Originally posted by: LoneNinja
Originally posted by: Andrew1990
hmm, I kind of remember seeing how all Phenom II X4s were supposed to easily overclock past 4.2GHz+ on air. We all know how that turned out.
Even so, I am most likely picking one of these up to put in my system.
I don't recall any source promising 4.2Ghz+ on air other than some dumb fan boys in forums. I've seen a lot of people claim their Phenom II to reach 3.8Ghz on air and that's all AMD claimed it would do, beyond that it was water cooling to get above 4.0Ghz.
Not true . Most reviewers report getting higher O/C but its not stable. Stable being the key here for me it has to be 24/7 stable.
It is a nice Cpu . Tho . Good luck. I would like to see that link were AMD claimed 3.8 ghz . I just want the Date. When the over clocking thing on LNC started. AMD was claiming 6 GHz . They never really said Air or Water but people were saying more than 4.GHz. Because guys recall I got really upset. That AMD was claiming 4ghz on air. I said they better get it right this time or I would take legeal action . As it is I believe next stepping gets them there and more . So its Cool.
And mark is right....
For the percentage amd is slower then the lower clocked INtel chip it really doesn't matter...because both of them get their arse handed to them by nvidia gpus....
My QX6700@3.46ghz for which I originally paid 600 dollars for got me aliitle more then 2000ppd( 4 instances and was before SMP client and the idea of running 2 SMP clients to take advantage of the idle times)....For 260 dollars of investment in 2 9800gt...I average 9400ppd...
So AMD or INtel aside, dont base cpu performance on folding much anymore....
With old gpu clients and drivers one video card took up one core regardless for polling....So 2 cards in my quad box would generally run me 50% cpu usage on top of the gpus running full blown. NOw with newer drivers and clients there is no hit or minute on the cpu...
So the basically any bargain basement dual core (intel or AMD)(no point of looking at single core though I think mark has a celeron of one of those) on a 3 to 4 pci express slot mobo....Pick up open box buys at newegg of 9800's for around 90 -100 a piece...It would take 3-4 quad cores to equal the performance of a dual core dual gpu set. At total cost for 3-4 total systems???? The money saved from not getting a 300-400+ dollar cpu for a 40 dollar one (like mark did) easily paid for the slightly more expensive motherboard, beefier 700watt minimum power supply, and the a good percentage of the video cards...IN the end he still runs a client on the cpu and can make a nice 13-15k ppd box
Can anyone find that AMD slide that was floating around here that showed how much overhead they were supposed to have?
edit: The line in the graphics is just a tad above 3.9GHz for the upper end of overclock headroom on air. 4.7GHz for water.
Weird, the edit button did a quote of my post instead...
At any rate, I photoshopped the linear y-axis scale onto AMD's slide so you see where the water-cooled overhead claims were extending to. (>4.7GHz)
http://i272.photobucket.com/al...t/PhenomIIHeadroom.jpg
Thank you. Well it looks like AMD did not really over-sell the air-headroom too bad. Good for them. Ill have to look into the water.
Unfortunately reality showed up and told the real story.....
http://www.xtremesystems.org/f...howthread.php?t=213248
prime stable and water is only 4.2ghz range.....air was only getting to 3.7ghz
http://www.xtremesystems.org/f...howthread.php?t=213248
prime stable and water is only 4.2ghz range.....air was only getting to 3.7ghz
That actually lines up pretty well with the center of the ranges provided by AMD's marketing slide on headroom.
The central limit theorem in action...appears to be that AMD was remarkably spot-on.
The 920's didn't even hit 3.7ghz and were much closer to the center of range as HTT overclocking could be limitations...the higher ranges are always the BE 940's with unlocked multis....
I would say we could talk about motherboard limitations on the platform being the factor except for the fact unlocked multis are a good percentage of our real world results.
Just chalk it up to first stepping. They will likely get better. Are they in the same realm as ocing chips as the intel chips? NO....I think they lose that as well.
I would say we could talk about motherboard limitations on the platform being the factor except for the fact unlocked multis are a good percentage of our real world results.
Just chalk it up to first stepping. They will likely get better. Are they in the same realm as ocing chips as the intel chips? NO....I think they lose that as well.
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,352
- 10,050
- 126
Wow, check out the first page of posts. Nothing but comments from people that are AFRAID to run Prime95 for 8 hours, for fear that they will burn out their mobos.
Is this because so many people on XS are afraid to run prime, because they aren't that stable with their overclocks, and using the mobos as a FUD excuse?
Or are there endemic problems with socket AM2+ mobos VRM sections, still, ala the 780G debacle in which Phenom 9850s would burn out mobos, because they couldn't handle 125W TDP CPUs?
I OCed my Q6600 to 3.6Ghz, at 1.33v, and I ran Prime95 25.9 64-bit overnight, no problems at all. (OTOH, sometimes when I run OCCT:linpack, it hits temps so high that my mobo shuts down.)
Crazy talk! I run my machines 24/7 under loads equal to prime95.
I saw all those posts a couple months ago, not sure what happened.
I ran P95 on my 9850/Biostar 790GX combo for a couple of hours, no problems here. The 9850 was @ 3 Ghz, drawing much more juice @ load than any PhII. I do have ramsinks (leftover from a Zalman VF900) on the mosfets near the CPU socket, though.
"only getting to 3.7ghz". i'd be quite happy w/ "only" 3.7ghz for a $145 tri-core w/ 7.5mb cache.
Cookie Monster
Diamond Member
- May 7, 2005
- 5,161
- 32
- 86
But that depends on application no? Like gaming for example where the denebs have the edge, compared to F@H which is clearly faster on intel cpus.
ditto my thoughts as well. Also what's the power-consumption look like for a 3.7GHz 720 vs. 3.3GHz Q6600...performance/watt should favor the AMD rig if my perception on AMD's 45nm power-consumption is correct.
Markfw
Moderator Emeritus, Elite Member
- May 16, 2002
- 25,567
- 14,520
- 136
The $145 is a tri-core, not quad-core, and thus price in this example doesn't count.
The thread Duvie linked to was 940/920 overclocks.
As for the poster, he runs superpi and calls it stable ? I don't think that counts either, I could probably do that on my 940.
The thread Duvie linked to was 940/920 overclocks.
As for the poster, he runs superpi and calls it stable ? I don't think that counts either, I could probably do that on my 940.
Mark doesn't price count in any example of price/performance comparisons being drawn?
I don't care how many cores/threads i7 has when compared to a C2Q or C2D but I do care what the price/performance is for them for the applications I care about.
No disagreement over what is being considered stable by some folks versus what is considered stable by others. That is why the qualifier needs to always be present when using the word stable...i.e. "superPi stable" versus "prime95 stable" or "OCCT stable", etc.
Anyone throwing the word "stable" around without including the qualifier should know beforehand that they will be criticized for not clarifying what they mean by stable. Its the nature of the business of the OC'ing.
TRENDING THREADS
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 10K
-
-
Discussion Speculation: Zen 4 (EPYC 4 "Genoa", Ryzen 7000, etc.)- Started by Vattila
- Replies: 13K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 7K
Facebook
Twitter