AusDog Posted February 5, 2009 Share Posted February 5, 2009 With time on my hands I tend to mess. Started bumping the CPU up 2 mhz at a time and running a new stress test called Burn In 2008 and predictably as I added the temps climbed slowly.Finally added 4 mhz and suddenly the temps dropped by 6-7C. The E5200 which started at 2.5 is now quite comfy at 11x320 or 3520. As we noted in the past there are idiots who do much worse things to their rig but usually they are on an ASUS board as opposed to my cheapo Biostar. This TG31-A7 is amazing and if it goes I'll switch to ASUS. Derek. Is there a faster CPU that is equally OC able and what board would you match it with...while keeping the price reasonable.Heard INTEL now has 8 cores packed into one CPU.Sounds like it would ably replace a microwave at generating heat. BTW. Did get a 22" widescreen Samsung refurbished with 3 year warranty and it blows me away. No bad pixels. Link to comment Share on other sites More sharing options...
DerekT Posted February 5, 2009 Share Posted February 5, 2009 When you are at the level you are now at, the difference in time and work done is not a great difference. In other words, an E5200 @ 3.5Ghz vs an E8500 @ 4.3Ghz will show little difference in many normal computer functions. There will be a difference, but nowhere near as great as 2.5 --> 3.5 so it becomes a point of diminishing returns where the enthusiasm for greater numbers is the key, not greater performance. Canned benchmarks show a difference in numbers but I liken it to a 2 kilo grapefruit in one hand and a 2.3 kilo grapefruit in the other. Since you already have a new system, moving to different parts for greater numbers is fine if you generate income and do this for enjoyment, hence the term enthusiast. Link to comment Share on other sites More sharing options...
AusDog Posted February 5, 2009 Author Share Posted February 5, 2009 I just noticed you are from BC. When I lived in Montreal all the Quebecois could talk about is Vancouver. Someday. There seems to be fairly wide variance in temp measuring progs. Using RealTemp, RivaTuner, and a Real Temp plug-in for Rivatuner. About 6-7c diff. Since they are all measuring the same thing I find this curioser. And I'm ready for my reaming. I'm now at 11x325 and using that new stress test Burn-In 2008 it stays at 48-49C. Also which 2 2gig sticks of Corsair would give me the most bang for the buck. Link to comment Share on other sites More sharing options...
DerekT Posted February 5, 2009 Share Posted February 5, 2009 Be careful you don't end up with getting nothing for something. :D: I've seen this happen many times before and sadly it is often with those on the more fixed incomes, those who can not easily afford the losses. I just noticed you are from BC. When I lived in Montreal all the Quebecois could talk about is Vancouver. Someday. There seems to be fairly wide variance in temp measuring progs. Using RealTemp, RivaTuner, and a Real Temp plug-in for Rivatuner. About 6-7c diff. Since they are all measuring the same thing I find this curioser. And I'm ready for my reaming. I'm now at 11x325 and using that new stress test Burn-In 2008 it stays at 48-49C. Both cores report working at 100%. And it is like getting something for nothing. Fun but probably silly. Poor computer. Link to comment Share on other sites More sharing options...
AusDog Posted February 5, 2009 Author Share Posted February 5, 2009 I am certainly pushing the ragged edge. You didn't think much of the 8800 GS and thru sheer dumb luck that's the card that is being pushed past the expensive 8800 cards. I tried the insane values once and 3D Mark 06 is the best way to stress your CPU if you wanna see it heat up. Took the video core from 580 to 680, shader to 1740, and memory to 1039. Jumped my 3D mark by 1100 points but I held my breath. The CPU got to 54C which worried me more than the vid card. Wasn't there long but backin everything back to sanity.btw. His reason that it was able to whip the GT was because lower core voltage kept it cooler. Link to comment Share on other sites More sharing options...
DerekT Posted February 6, 2009 Share Posted February 6, 2009 You didn't think much of the 8800 GS I still don't think much of it. Compare it clocked up to a stock 8800GT and you can get some decent benchmark numbers, but a higher res will show the bandwidth gasping of the GS. Not enough RAM and not enough bitrate. Yes, on 1280 X 1024 benchmarks you will gain with the overclock, but you can't create more DRAM and you can't create more bitrate. 512MB castrated to 384MB and a 256 bitrate castrated to 192 bitrate will certainly show in higher resolutions. I've seen people clock up the GS to 750/2000 but it matters not when you are running high eyecandy in 1600 X 1200 or higher resolutions. Sorry, I still don't like it. Yes, with a lower res it can compete even though the drop in memory will make for more loads and higher load times but higher res and you need to drop eye-candy and the load times will still be bad on high DRAM games. Clocking up your E5200 to 4.0Ghz will not compare it to a E8400 clocked to 3.6Ghz. You CAN'T make hardware through the overclock. A 2MB cache vs 6MB cache will show out in the wash. 3D Mark is NOT a great CPU tester. Uuse the Intel Linpack CPU/Memory test. http://67.90.82.13/forums/showthread.php?t=197835 Download Realtemp from Here Run Linpack for full memory (1) and use a 5 pass iteration. Run Real Temp and watch the temperatures. Let the Linpack program run for the full 5 passes. The output in the screen will give you five sets of values. They should be identical for a stable CPU and DRAM. Be sure to turn off your screensavers, display settings that turn the monitor off for this test. Link to comment Share on other sites More sharing options...
AusDog Posted February 8, 2009 Author Share Posted February 8, 2009 I thought I recognized the guys name that wrote the web page touting the 8800 GS. It's the guy who also wrote the program RivaTuner which is respected in many circles. Unless I didn't read carefully enough he only used the lack of heat as showing a lack of problems. Luckily I ran Ati Tool which showed artifacting at a much lower setting than he was pushing his cards to. Course if I got the cards for free from the manufacturer I'd flog the heck out of them as well. I didn't think the video benchmark was a good test but merely amusing that having the vid card that hot beside the CPU and pushing the CPU a bit produced much higher temps than progs designed to do that to a CPU. I don't know how much is due to raised ambient temps and how much to the workload. Already had Real Temp but will sure see what the Intel prog has to say. Oh...and it was Real Temp that was showing lower temps than RivaTuner so if Real Temp is more trustworthy then that's indeed good to hear. I still lust after a GeForce 260 but I don't game THAT much :laughing: Link to comment Share on other sites More sharing options...
AusDog Posted February 8, 2009 Author Share Posted February 8, 2009 The guy scared me from using it. 22C hotter than other similar programs... Link to comment Share on other sites More sharing options...
DerekT Posted February 8, 2009 Share Posted February 8, 2009 LOL :D: It's called "Burn Test" for a good reason. The guy scared me from using it. 22C hotter than other similar programs... Link to comment Share on other sites More sharing options...
AusDog Posted February 9, 2009 Author Share Posted February 9, 2009 ThermalTake Tribe is a very reasonably priced external liquid cooler for this CPU but if I getting cues of the CPU being pushed about as hard as it can be, freezing hell out of the CPU doesn't change the archetecture of the CPU nor does it protect the RAM or Board. Right? Geez. You might have taught a few things to this old dog. Oh. And as I'm going to try this Intel "HotBox" what would be a smart bail-out temp for the CPU test? Link to comment Share on other sites More sharing options...
DerekT Posted February 9, 2009 Share Posted February 9, 2009 If you are not crashing or finding faults, then there is no need to test. Testing for the sake of testing is never a good thing. Such tests stress the CPU greatly. Why stress it? I know there are some who benchmark for the sake of benchmarking but they usually do so to compare it with others rather than show stability. Most who benchmark don't stress test much at all. I'd rather spend the time using my system than testing it. :D: Link to comment Share on other sites More sharing options...
AusDog Posted February 11, 2009 Author Share Posted February 11, 2009 Sounds good to me. I wanted to put an after-market cooler on my 8800 but can't get Arctic, or XFX, or Nvidia to tell me if either the passive Arctic or the neat lookin 3 fan model will fit. So they're worried about liability but I told all three just give me your best guess. No answers. Link to comment Share on other sites More sharing options...
DerekT Posted February 11, 2009 Share Posted February 11, 2009 If it isn't broke... My advice is just enjoy your system. You bought a GS because of cost and now you are going to spend money on a cooler? You could have bought the 8800GT? Just use the system. Have you checked overclock.net? They will give you better answers than I will. Link to comment Share on other sites More sharing options...
AusDog Posted February 11, 2009 Author Share Posted February 11, 2009 Too much time on my hands. The other thing I was thinking of was taking my RAM from 2gb to 4gb but the Dominator 1066 has settings of 5555+15. My current is of course 4444+12 but only 800. By getting the 1066 I could probably avoid backing anything off. Right? (and then I'll let this thread die before it becomes another never-ending dialog). Thanks for your always incisive advice. Link to comment Share on other sites More sharing options...
DerekT Posted February 11, 2009 Share Posted February 11, 2009 The difference with your system will be at best 3 - 5% and often no difference at all. Too much time on my hands. The other thing I was thinking of was taking my RAM from 2gb to 4gb but the Dominator 1066 has settings of 5555+15. My current is of course 4444+12 but only 800. By getting the 1066 I could probably avoid backing anything off. Right? (and then I'll let this thread die before it becomes another never-ending dialog). Thanks for your always incisive advice. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.