Would an Opteron 165 reach ~2.5GHz as easily as an Opteron 170?

[Tomash]

Running on 2 gigs of OCZ Platinium DDR500 RAM, and a DFI LanParty nF4 Ultra-D.
Is the 170 worth the extra $70?

 

[maaoouud]

Not really any difference between them I'd say. The 170 gives you a maximum multiplier of 10 vs. the 165 at 9. However the maximum multiplier isn't really important unless you're going to overclock it to the extremes, you could just run the HTT at 277MHz (should have no problems with that on the ultra-d) and get 2.5 GHz on the 165 with multiplier x9 set and then use a divider to get roughly the right memory clock. You don't loose performance if you use a divider for the memory.

 

[Tomash]

Not really any difference between them I'd say. The 170 gives you a maximum multiplier of 10 vs. the 165 at 9. However the maximum multiplier isn't really important unless you're going to overclock it to the extremes, you could just run the HTT at 277MHz (should have no problems with that on the ultra-d) and get 2.5 GHz on the 165 with multiplier x9 set and then use a divider to get roughly the right memory clock. You don't loose performance if you use a divider for the memory.

Jesus. That sounds a lot more complicated that back in 2004 when I quit OCing.
Thanks though.

 

[maaoouud]

mad shrimps athlon 64 general overclocking guide

Covers the basics for it. Page 4 seems to be the one with the important stuff.

 

[Tomash]

mad shrimps athlon 64 general overclocking guide

Covers the basics for it. Page 4 seems to be the one with the important stuff.

Huh. Thanks.

 

[swaaye]

I hit 2.6GHz with my 165 at 1.46v. It has no headroom left and only big voltage boosts would help. 2.8GHz is probably the best you can really expect out of air cooling with these chips, and that is somewhat rare.

 

[zsouthboy]

My 165 did (and would still do, except the box is in pieces right now) 2.5 at stock voltage. Right out of the box, stock retail heatsink.

I didn't bother pushing it or anything. Seriously.

And yes, it does rock.

 

[swaaye]

I think I could do 2.5GHz with stock voltage. If not, it would be very close. At most 1.40v. The thing is, once you start to need more voltage, the need quickly spikes upward. You begin to hit a wall where you need tons more voltage and cooling to get anywhere.

3.0Ghz, say, is never achieved without extra voltage and excellent cooling. Basically no one hits 3GHz with air cooling.

I think it's safe to say though that 2.5GHz is nearly guaranteed. Though I have heard of some people who have especially poor chips, or an especially poor core. Core 1 and Core 0 are never equal overclockers. My Core 0 is definitely the wimp of the two.