dfx said:
Guys this was achieved via phase change
Here's the break down i had tested on this chip
3000+ @ 2.45ghz (Stock HSF)
3000+ @ 2.6ghz (Watercool)
3000+ @ 2.8ghz (Phase Change)
All tested prime stable
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
dfx said:
BioVader said:What mobo you using?
i'm kinda surprised that you didn't get a higher cpu score at 2.8 ghz..dfx said:
(cf)Eclipse said:i'm kinda surprised that you didn't get a higher cpu score at 2.8 ghz..
at 2.4 with my clawhammer, my overall score was 4897, but my cpu score was 5237..
Heh, my limit seems to be 2480MHz with a 3200 "CG" CH. My stupid board undervolts, though, if only I could get the 1.85vcore I am entitled to by my birthright, I would overclock to 2500, and beyond!Circuitbreaker8 said:2.5ghz seems to be my max on my A64 3400 CG CH. So right now i'm
@
250x10 ( 2500mhz )
4x HTT
1.7v CPU
2.8v Mem
ram is running at a lovely 250mhz 2.5-3-3-6 I love my OCZ EL
(cf)Eclipse said:depends on your cooling. put the voltage up until it starts going over 60ºc under full load. that's when i would start getting mildly worried about my temps. (usually that means like 1.7v or so. i don't like to run everyday with more anyhow)
that was answered earlier in the thread. check page 9 or 10. i forget the make, but i know it's 3.6vnotoriousformula said:DFX: i think you missed my post.. what ram are you using and @Vcore?
thanks.
notoriousformula said:DFX: i think you missed my post.. what ram are you using and @Vcore?
thanks.
(cf)Eclipse said:yeah, the temp sensors to tend to be a bit of a problem. hence why i don't suggest going over that, cause the a64 dies are rated to 90ºc
that was answered earlier in the thread. check page 9 or 10. i forget the make, but i know it's 3.6v
What's the vdimm on that BH-5? My BH-6 can do that at 3.8...dfx said:got my BH5 burned in for a few days
here's my new update
Darth_Fluffy said:Who cares, those are synthetic benchmarks anyway. You could hack your Sandra into thinking you have a 10THz Pentium III, and you will get astronomical scores.
(cf)Eclipse said:dfx.. something looks wrong with that unbuffered score. with the speed and timings you're running your ram at, it should be more in the order of 2700-3000mb/s
(cf)Eclipse said:then you're just looking for big numbers then. you'd probably be better off at 1t at a slower speed like 250mhz or so, your bandwidth should go up a bunch.
hell, i just tested unbuffered with my ram running at 227mhz, 2.5-3-3-7 1t, and it got about 2550mb/s, just something to consider
Darth_Fluffy said:Who cares, those are synthetic benchmarks anyway. You could hack your Sandra into thinking you have a 10THz Pentium III, and you will get astronomical scores.
and which mobo you willbe going with.. let me guess.. Mr. Troublesome MSI Neo2? wait for nice stuff like DFI Ultra-D.. I'm planning on getting a Winchester myselfDarth_Fluffy said:Beats mine ... I'm moving to Winchester, though
Darth_Fluffy said:Is there something bad about Neo2? The original is just peachy..
Wait, what's the maximum vcore you can get on that DFI board you use right now? This stupid board undervolts. And I need about 1.8vcore. For great justice .
Darth_Fluffy said:Is there something bad about Neo2? The original is just peachy..
Wait, what's the maximum vcore you can get on that DFI board you use right now? This stupid board undervolts. And I need about 1.8vcore. For great justice .
HTT alone means nothing whatsoever. It's just an input frequency that's used as a base for some of the devices... such as the CPU, and the HT bus, memory. If you're lowering multipliers to tone down the CPU frequency, bus frequency, and memory, you're actually going to have worse performance. Overclocking HTT only makes sense if at least your CPU and HT can run at native multipliers with it.notoriousformula said: