1 person found this helpful
Basically you want to raise the BCLK
There are guidelines to voltage maxes you want to stay within. Google that
Memtest86 is a decent tester but I recommend http://hcidesign.com/memtest/download.html for memory testing & LinX http://www.xtremesystems.org/forums/showthread.php?t=201670 for CPU stability testing. Those two are the best testers I've come across and I've been Overclocking for a while.
Thanx to both of you.
Very helpful and informative ...but it all looks kind of too scary for me to mess with.
I have a saying that I should apply to myself in this instance:
"Dont f*ck with somethin' you know nuthin about"
My sytem does have ASSUS Turbo V installed.
Allegedly this will do an OC on the fly (not saved in BIOS). Still needs a little tech knowledge to set it up and save it as a Profile though.
Think I will take my own advice.
If you are having no problems why do you want to Overclock?
You make it sound like: If you have problems, overclock and your problems go away.
That is not the case, quite the contrary. If you have problems at stock speed, solve them. Only try to overclock if you have no problems whatsoever and if the effort warrants the performance benefits.
Craig, before you bother overclocking make sure you have your OS well tuned. I do not recall seeing your results on the posted PPBM4 results although many times forum names and posted names have no connection. You can run PPBM4 on CS5, currently there are 11 CS5 results posted. With that information I can see how efficient your CPU is running and how good your disk system is setup. Of course you are such a frequent forum contributor you may already have the data.
I was to nervous to OC my self so paid £50 to get it professionally tuned.
make sure you get a decent Cooler, Noctua NH-14D are Very Good.
I have never benchmarked any of my systems.
I have just worked on the basis that they all functioned absolutely fine and why "fix it if it aint broke"
As regards OCing.
My CS5 works brilliantly but I was just curious if there was an additional "easy to achieve" performance gain .
Yes . I can easily get my puter dudes who built the machine to do it but probably I cant be bothered lugging this massively heavy box over to their workshop though.
I can easily get my puter dudes who built the machine to do it but probably I cant be bothered lugging this massively heavy box over to their workshop though.
Just bribe them with some Cloudy Bay, and they will flock to your edit-suite to do the work there!
Bribes are good. Going back many years, I needed to shoot from the construction site of the Telluride, CO airport. The construction was not going well, and we were about to be denied access, to do what we needed. My partner was starting to reach for his checkbook, when I told the foreman that I would show up at the appointed time with a cold case of Corona cerveza and limes, and the deal was struck. When we arrived, just as they were shutting down, the foreman called for a Caterpillar D-10 to grade a road to get the RV (our subject) out to the edge of the runway site, so we could shoot it against Mt. Sneffels. If we had paid cash, it would have cost US$1000 at the least, but a cold case of beer, and limes, and we were done.
PS - I'm watching and learning, as I hope to follow in your footsteps in a few mos.
I could probably just drink the wine (or the beer, or both) and do the OCing myself under the illusion that I knew what I was doing.
Shooternz, I have similar setup to yours and similar depth of knowledge about OC.
Although I've built my system myself and it is running the CS5 just perfectly, the OC option is there and it is tempting - I'm just looking for reasons to explore it. The only thing I've done so far is to activate the XMP ('extreme memory profile') of my RAM modules (Corsair XMS3 1600) in the BIOS. If you haven’t done this already, probably your RAM is running at slower speeds predetermined by the 'auto' mode in the BIOS. The activation of the XMP also overclocks a little the CPU - the net result in my case was a bit snappier performance and 2C to 4C increase of the CPU temps. Surprisingly to me this 'slight' OC did not have any noticible effect on my work with CS5. I timed the rendering of the same 9 layer video clip 'before and after' and the results were identical. The playback also. Sure more decisive OC-ing will affect positively the net performance of the system but will also rise its temps and shorten the life of the components according to articles I read on the topic (unless you go for advanced CPU coolers, etc.). I'd love to see some 'before/after' stats of some basic and advanced editing tasks performed by the same system before and after OC. This might give reason to OC-ing.
Is the PPBM4 file used in a CS5 project.
I am getting tempted to try this Computer Racing Game
Yes, PPBM4 runs fime in CS5. If you look at the Benchmark Results page you can see a dozen or more examples of results which Harm has recently highlighted in yellow, some with MPE turned on and some with MPE turned off. We are developing the next generation PPBM5 and debugging it right now.
Some people are of course interested in it for bragging rights (I have become a victim myself I have to admit) but the basic purpose that I originally designed it for was to develop a real application benchmark that could evaluate what elements of hardware and software contribute to a well designed and tuned Premiere Pro environment. When I first started this back in 2005 (despite the fact that my BIO says I started in the forum in 2006) there were a lot of opinions of what was necessary for success but no test data to back up opinions. One of the first rumors that I proved wrong is that the graphics card was an important item (PreCS5 of course). But with the submitted results and comparing to other results with the CPU intensive test we can easily see if your OS is tuned for maximum performance, also if your disk system is not setup properly it will show in the disk intensive test. I have dozens of emails that with the first try I may not publish until the person has the opportunity to make a few adjustments.
Message was edited by: Bill Gehrke
I have actually looked at this table few times. It contains a number of variables which is hard to isolate and prove causality that would be linked to any single factor. Probably it is my reading, but I cannot identify the 'before and after' OC-effect comparison. To prove causality the only independent variable that should be let to vary is the OC. No doubt OC helps, I'm just not able to decipher how much from that table.
Looking at the above Benchmark Results I noticed the user compared to the other users he is useing 6gigs of ram and standard Sata and some of the readings are very respectable. As for OC I feel it is best not done as there must be some down side in the long run, I must say this thread to me has been the most useful thanks for starting it.
Nice results, Baz.
You bested my scores on Cinebench with the HD4870. Must be due to the GTX-480. And your cooling is very good. Congrats.
its due to your recomendations for the CPU Cooler, its a very good peice of kit. its huge. the graphics card did not make any differnce to the cpu scores. will try it later with the quadro fx5800.
And what cooler is that? Your temps are really
I having been doing a few tests on a i7-920 and see how far i could push it on my setup. I got it to 4.35 GHZ running stable for testing purposes. ON AIR. Temps never went beyond 65c. Thats a 63% Gain.
Graph of Cinebench R10.
cinebench R10 GTX480 4.35ghz.jpg 318.3 K