Overclocking
12 Mar 2005 22:21:14 UTC
Topic 188493
(moderation:
What success have people had running E@H on overclocked machines? Aware that overclocking could cause invalid results, I set my FSB to what I thought was not a particularly aggressive overclock. I got a string of invalid results for my trouble. I'm now running a mere 5% overclock in the hopes that at least that will be fine.
AMD Athlon64 3500+ (11 x 210MHz)
Corsair XMS PC3200XL (2 x 512MB)
Overclocking
)
Running A64 3000+, 9 x 250 => ~3500+. Not a single invalid result so far.
Try raising your core voltage, it should help.
Running a¨P4 on an Asus
)
Running a¨P4 on an Asus P4P800 MB.
My Pentium 2.4 is overclocked at 20% using the internal build-in overclocking from the Bios.
Runing Boinc 24/7, no problems at all untill now, CPU at 42°C or 107°F using a tower with one inlet fan and one outlet fan and a good CPU cooler and heatsink.
Greetings from Belgium
Thierry
> What success have people
)
> What success have people had running E@H on overclocked machines?
I'm now running 7 overclocked boxes. They're all AMD socket A (Athlon, Duron, Sempron) and the average overclock is around 30%. They've done around 135 valid results, another 50 pending and zero invalid. However, before being added to E@H, they were well tested for prime stability so I wasn't really expecting problems.
All boxes are using moderately elevated Vcore and stock HSF. I reckon the three most important things I did to improve the overclock were cooling, cooling and cooling, in that order :).
1. Lowered case ambient temp by adding extraction fans.
2. Changed the stock cpu fan to 80mm 5400rpm to give higher airflow.
1. Lapped each heatsink and applied arctic silver (extra 1-2C reduction).
Please note that #2 doesn't do much without #1. Fast hot air doesn't really cool any better than slow hot air. Both #1 and #2 in combination give about 12C+ in full load CPU temp reduction over the stock AMD fan on its own.
Cheers,
Gary.
My main development box is
)
My main development box is the only overclocked one I posess and it is only slighty so.
Athlon XP 2600+ bumped to 2.25 GHz
Has yet to garble anything. I'd go faster, but, the processor fan is running at 7000+ RPM to keep the temp in the low-mid 50c range while crunching and compiling. Add that drone to the other three systems in here and you get a case of tinitus just waiting to happen.
I've seen some other comments in here regarding garbled results on overclocked machines and I concur with the CPU voltage increase advice.
Thanks, everyone so
)
Thanks, everyone so far.
Cooling is a bit tricky for me, since my absolute priority is a near-silent machine. I use an expensive fanless PSU in an aluminum case, a passively cooled video card and Northbridge, and a hard drive enclosure. The CPU is cooled with a Thermalright XP-120, a SilenX 120mm fan, and OCZ Ultra 5+ (same as Arctic Silver). There is another SilenX 120mm for rear exhaust. Both fans are run at very low rpm. The CPU is a Winchester core (90nm process), which runs a bit cooler than the 130nm chips. All this seems to allow me to run the CPU at a 210MHz, with the voltage bumped up a notch, in the mid-40s (at least until summer). There is a separate voltage control for the HyperTransport, which serves as the memory controller on the Athlon64, and that I've left at Normal so far. I also have the DDR voltage bumped up one notch. I'm nervous about cranking any of the voltages any higher. The temperature rose several degrees when I pushed up the HT voltage as well.
I see I've now received credit for a few results returned at 11 x 220MHz (10% overclock), so maybe it's not as bad as I thought, but I dare say I'll never achieve 30%!
hm, because I can only
)
hm, because I can only overclock by raising the FSB on my Mainboard i can only overclock 100 MHz (FSB 133 -> FSB 141) so i have 1900 instead of 1800 MHz... that works without problems, although thats not the world its a bit better than nothing...
I'm running a mobile xp 2200
)
I'm running a mobile xp 2200 on a desktop it has a 512 L2 at 13x159 for 2078, tomorrow I'll try to get it up to 13.5
My other box is a 2500 barton at 11x200 with voltage at 1.70v
both run at under 40 degrees. Very stable all validate good.
I'm quite happy with AMD's
I can't really do to much
)
I can't really do to much with my machine, as all overclocking is mananged by MSI's Core Center and AMD's cool'n'quiet. Right now with stock HSF with ambient room temp of about 18C, i'm running 50C on the CPU and 29 on the board. 210 x 11 @1.49v for a 5% OC.
All my rigs are overclocked.
)
All my rigs are overclocked. My slowest machine is a tbred B 1700 running at 2.06ghz. The horse is an A64 winchester 3000 running at 2.93ghz. All boxes have one side removed for extra ventilation. Will probaly have to detune some of them when summer arrives. Have no problems with Einstein. See below,
>The horse is an A64
)
>The horse is an A64 winchester 3000 running at 2.93ghz.
Do you know how high it's safe to raise the voltage? I don't know if your 3000+ is a Winchester core, but the normal voltage on these is 1.4. I've set mine to 1.45. I've read about people setting it all the way to 1.5. Also, have you altered the HyperTransport voltage at all? Thanks.