I definatly want some sets :D
Printable View
I definatly want some sets :D
A 288 MHz CPU-Z screenshot at ~3.8V? Stability?Quote:
Charlie
Do THIS with 1gB of BH5, hehehehe....
Sorry if I'm not completely impressed, but I'm also interested in the stable results one can obtain from this product.
I am not going to spend that much money on memory, to take CPU-Z screenshots, and run SuperPi 1M through 32M all day. A screenshot isn't enough to call it stable, nor is a half hour benchmark, in my opinion.
How fast can it go, looping memtest86 for 24 hours? Don't take offense, I'm not challenging you guys, but what you need out of a memory stick (excellent 'bencheable', 'screenshotable' results) isn't what everyone needs out of a memory stick. Unfortunately I find that these impressive screenshots are breeding some dissapointment, when people discover that their sticks cannot run such speeds with any sort of long-term useable stability.
Excellent for benching, yes - that's very impressive and good to see :). But would you care to share some "thoroughly tested" stable results with us, for our own education and benefit?
Just my two cents, I'm interested in what you guys think about this.
Quote:
Originally Posted by felinusz
i totally agree with you felinusz, that's the same thing that interests me. Stability is def a key factor...
They don't ship to Europe :(
Thats correct, I don't were you live, but in holland we have to go to bytes @ work in Belgium. I love this shop, 4 min after my pre-order I get the confirmation that I'm by the first batch :woot: .Quote:
Originally Posted by Juliette
Quote:
Originally Posted by felinusz
Well, I actually AGREE with you!!!!
And I "don't" agree at the same time.
I said this before but I'll say it again.
This is BENCHMARKING, it is computer "racing".
In automobiles, there are various levels of competition ranging from 1/8 mile drag races to the 24 hours of Le Mans....
A Le Mans racer doesn't compete at the local 1/8 mile dragstrip, nor vice versa.
Benchmarking is the same thing
I have a computer for "computer stuff", it is a P4-630/P5P800Mushkin PC4400 that runs at 15 x 270 = 4050 like a ROCK, meaning there's NOTHING I can do to make it crash. I can game/stream video/music/prime/sPi 32M, whatever... PLUS it runs at only 1.34v so it will last a while even at OC'ed speeds.
Then I have a "benching computer", I keep it at home and it's laid out on a table. There's nothing valuable on the HDD, and I don't CARE if I have to reformat or reinstall o/s 3 times a day. It is a DFI nF4/San Diego 4000+/XP4000 machine. I don't WANT/NEED it to run sPi 32M or loop PRIME95. I "play" superPi 1M and 3D benching, that's my sport. If I can complete 1 run out of 20 attempts in 3DMark 2001 at a given OC, that is a SUCCESS.
DEFINITION:
STABILITY = The ability to complete an assigned task without fail
When the assigned task is to complete a superPi run 1 time out of 100 attempts, that IS stability!!!!!
When the assigned task is to run D2OL 24/7 in dual instances without failure, that is ALSO a "measure of stability" (see definition above)
24/7 stability is secondary to me, I don't EXPECT top speed from my 24/7 rig, so I'm happy with how it runs, even if it's only at 80% of the hardware's capability.
Do you guys REALIZE that if your computers were at JPL in their LAUNCH TRACKING HQ, they would be considered ridiculously un-reliable and unstable because they have a different expectation of stability than you do and Prime95 and suerPi 32M just won't quite do it...
Just remember the definition of stability...
And FWIW, I can complete superPi 32M at 269mHz using the MUSHKIN Redline XP4000 (the Indy 500 of benchmarking)
lots of propaganda about this ram...:D
might pick up a set
EDIT: any clue to when newegg gets it?
Quote:
Originally Posted by OPPAINTER
if he said bench stable, then there would be no confusion. All my rigs are on 24/7. While some people complain that bench stable is not stable, and they are right in a regard, i don't think the whole "burn-in" process to these winbond chips sinks in...if it boots...you can get it stable. It may take 3 years of burn-in to get it 100% error free over a month...but it can be done. My UTT stick wouldn't even boot past 233mhz when i first got them, and now, a wee bit over 6 months later, 260 is starting to get bootable into windows. That took me over 6 months...
These Muskin sticks are booting @ 275 avg? Do you know how much time and money in powerbills that would have saved me? Only bench stable? Who cares! Burn 'em in! They'll get as stable as you want, after a time.
I want a red pcb.
Charlie & OPP.
Any chance it could run @ 255-260 2-2-2-5 on my Intel setup ?? (P5P800...see sig:P )
I wish it could :p: (i'm getting a P4C800-E + dothan soon too)
Thanks
What's your burn-in procedure? Everyone seems to do something different and I'm trying to find out whats best.Quote:
Originally Posted by cadaveca
too damn expensive ram if you ask me.
OPP which bios did you used and settings :)
Why don't you shoit to us then? Come on Charlie... :) You would have even better set than John has ;)Quote:
Originally Posted by charlie
Thanks!Quote:
Originally Posted by the farmer
Memory configurator was darn expensive, ranking it with OCZ PC5000 modules :(
Well said, and I understand your personal school of thought on the stability issue.Quote:
Charlie
I have a computer for "computer stuff", it is a P4-630/P5P800Mushkin PC4400 that runs at 15 x 270 = 4050 like a ROCK, meaning there's NOTHING I can do to make it crash. I can game/stream video/music/prime/sPi 32M, whatever... PLUS it runs at only 1.34v so it will last a while even at OC'ed speeds.
Then I have a "benching computer", I keep it at home and it's laid out on a table. There's nothing valuable on the HDD, and I don't CARE if I have to reformat or reinstall o/s 3 times a day. It is a DFI nF4/San Diego 4000+/XP4000 machine. I don't WANT/NEED it to run sPi 32M or loop PRIME95. I "play" superPi 1M and 3D benching, that's my sport. If I can complete 1 run out of 20 attempts in 3DMark 2001 at a given OC, that is a SUCCESS.
DEFINITION:
STABILITY = The ability to complete an assigned task without fail
When the assigned task is to complete a superPi run 1 time out of 100 attempts, that IS stability!!!!!
When the assigned task is to run D2OL 24/7 in dual instances without failure, that is ALSO a "measure of stability" (see definition above)
24/7 stability is secondary to me, I don't EXPECT top speed from my 24/7 rig, so I'm happy with how it runs, even if it's only at 80% of the hardware's capability.
Do you guys REALIZE that if your computers were at JPL in their LAUNCH TRACKING HQ, they would be considered ridiculously un-reliable and unstable because they have a different expectation of stability than you do and Prime95 and suerPi 32M just won't quite do it...
However, my school of thought on the stability issue involves thorough testing before declaring an overclock "stable". "Stability is in the eye of the beholder" as I always say :).
And it's simply that, merely an opinion, and one that you can't change :). Likewise, I'm not going to try and change your opinion, because I respect it :).Quote:
felinusz
A screenshot isn't enough to call it stable, nor is a half hour benchmark, in my opinion.
However, it would be really cool of you to drastically increase the value of your testing to our entire community, and perhaps post some results that would appeal to and subscribe to the opinion that myself and many others hold. It'll only confirm the integrity of the product you're testing out, and cause more of us to benefit from it's capabilities, after we're given justification for spending money on it :).
What is the consensus when comparing the pc4000VX against the redline pc4000?
I agree if you've done 32M, you can show it to us :)Quote:
Originally Posted by bachus_anonym
man I gotta tell you guys something. This site was founded on respect and mutual admiration and there has in the PAST always been a "kind of reverence" for the hardcore overclockers and forefathers of our hobby. Lately, I have seen this degrade. Bacchus, I'm not talking to you! You pose some interesting questions in a mostly respectful way, but I see new people with 50-300 post counts coming on the forum and basically dis-respecting people who have 1000x the skill and knowledge that most others have. I find it unsettling to see these posts become degraded by people thread crapping on what someone reports. This thread, and the gentleman who had the Venice vs San Diego thread are prime examples of not showing proper respect. Just b/c you started overclocking a few months ago and consider yourself expert, doesn't qualify you as truely expert.
I'm not saying you have to take my word, charlie's, or Opp's word as gold for that matter, but show some proper respect or the knowledge you have come to take for-granted will slowly migrate away from this climate of hostility and suspicion.
Alright then!
I do not know what Charlie, OPP and you John are going to do with those early sets you got from Mushkin. But as everybody is looking at big boys here then they expect some good stuff from them. I know that OPP or Charlie are not on the same "stability bandwagon" as many other members.
For Charlie all what matters, as he said, is 1M and 3D. For OPP, it looks like only 3D. But it just doesn't meant they cannot deliver more stable results. That's just the only thing I do not understand, especially when you guys got those sticks for review.
I think that true review should contain not only benches from the edge of stability but also some "community-wide" recognized day-to-day stability tests.
My $.02
Now, let's just all be respectfull to eachother... We're just gonna sit tight and wait for OPP's review.
I hear you John. Respect should be given to the people who have taken time out to come up with some findings and views of thier own. Even if you have better results or have seen better I dont see the point of dis-respecting the member about thier results. Overclocking achieves totally different results. Its a gamble. It varies.
The Venice vs. San Diego thread may have to do with what the thread starter, Ben did before in a thread pervious. It was a challenge to s7e9h3n about his Venice. Ben tried to say that his Winnie was better than the Venice and there was nothing much about Venice's and such. s7e9h3n has shown quick results when having the cpu to show what the Venice's may do as like a preview. He didnt do the whole benchmarking thing, just a few tests which was all good with most members. Ben however challenge him and shown different results to the ones that s7e9h3n shown and started to lay into him about that his Venice was crap and his Winnie was great. Everyone didnt like how he went about this and it cause a few problems. s7e9h3n took time out to preview his Venice but Ben tried to be all mouth and it began from there.
I having bought s7e9h3n 3800+ Venice and said to him I would take up the challenge but it ended there and the thread was deleted before anything serious could happen, which was the best thing.
So to his new thread, the members that remembered what happpened in the last thread have lost respect for Ben how he went about his business before.
If you're talking to me, I'm sorry. Since a few months I look around on this side and every time I read interesting stuff and know I have a lot to learn :) , that's why I'm always so happy if bachus_anonym ask for information. Maybe it's better to let him ask.Quote:
Originally Posted by JNav89GT
If my English was better, my reply would be much longer, just want to make clear what I'm doing here.
Full load of different types of data 24/7. You just can't be worried about errors. Staart at the 1stmhz beyond what's you "stable" point, and load away...for days. only give 'em a fan if they are HOT like burn your finger hot to the touch., and then only enough so that they don't error too much.Quote:
Originally Posted by HiJon89
I must be honest..i have killed some sticks this way, but these BP pcb's and winbond chips can take alot.
Now, you might misunderstand me. Burn-in works differently for different chips, from my experience(which, JNav89GT, has nothing to do with post count. Making posts doesn't make you smart, or more educated.). CH-5 won't get much in max MHZ, but burn-in will increase max mhz @ low volts/2.5CAS. BH5 will get small increases of low volt mhz, but can go higher at the lower lat/high volts.
random errors won't "burn away", at least that i have seen. It's will eb that same error in the same spot, that one day, just won't happen as often. Eventually, it's gone. But to burn for 24 hours, and expect an increase... :slapass: doesn't work that way. Over a week... :nono: maybe that last mhz will ge a bit more stable...YOU GOTTA WORK IT! :fact: BURN!!! :flame:
Of course, somebody's gonna take it too far, and fubar thier sticks...you gotta get them hot, but just high volts WON'T do it. ya gotta treat the mright...on the edge, but not too far.
well I was out of loop on that one! ty for clarification Bennah.
Couple of other points. I think you guys are a bit too critical for one. AFAIK, Opp bought his stix, he wasn't given them by mushkin.
As long as there is a context of what the user appreciates in ram and also a context in comparison I feel their comments/posts are legitimate. Furthermore, if they have been a contributer and a pioneer such as Opp, I think they should be given a bit more respect versus Joe Schmuck fresh off of hardocp or anandtech forums. I'm not trying to start a flame war, I'm not trying to bait anyone into an argument that I'm right etc... All I'm requesting is that people respect other peoples posts and in this case a positive experience with a product without critiqueing it like a freaking research article.
In the end each user determines their acceptable stability. If this product is better than his last, then for him it's better, and for MANY that is a solid endorsement. Don't forget Opp has endorsed many a products in the past and has had many people satisified with the products he recommends.
I have dealt and traded with Opp in past and my experiences have always been positive. He has been around a LONG time in our circle of friends, and although I'm not a huge contributer to this site re: pioneering etc... I feel I have made some contributions along the way as well. When someone like Opp is dis-respected, it offends me to a point as well.
I will do some more 32M tonight, summer like started today in SoCal and here inside it's like 78F. But here's a start, 269@2-2-2 and I know for sure it'll do 271+ 32M.
That was kinda' fun, like running the 1 mile race in high school track!!
Sorry for the pathetic time, winny has only 9x multi!
And the sPi window is messed up from me scrolling down late in the test to see what step it wsa on :D
http://www.xtremesystems.org/forums/...chmentid=30323
C
Makes sence. And Im sure a lot frequently visit this forum just to see you guys with niiiiiice wallets "race". I for one do. I love it.Quote:
Originally Posted by charlie
The thing is, a lot a people inhere ground their 24/7 investments on your "race"-results. In that context it would be so damn nice if you extended your racing a bit, so that we might get an idea on the components perfomence in a more 24/7-like environment :)
Other than that, just keep on trucking! Its (partly) what makes this forum an intricate part of my life :D