It might still run a couple of degrees warmer than before. Since I just applied fresh thermal paste, that could be why it's sitting a degree or two higher. Also, one of the pins wasn't seated quite right earlier.
It used to be totally normal—around 131°F under load. Is it possible I botched the thermal paste application? I doubt it, though, since the GPU isn't acting up. Before this, it would freeze at about 163°F, but now it runs fine, which makes these load temps feel a bit off. But specifically right after swapping the paste? Anyone have an idea what's going on?
Had a little fun playing around, but it was a bust. I can't even get things to spin up... it crashes once it hits 72-73°C.
Motherboard: P35-DS3 rev 1.0 - the voltage regulation is pretty rough; after 1.60V, the next step jumps all the way to 1.80V (though maybe a tiny bit lower actually). Cooler: stock
The motherboard is a bigger culprit here than the stock cooler, but hey... it is what it is. At least the clock speed is decent considering the hardware I'm working with.
EDIT: I could probably squeeze out a slightly higher clock, but without some serious cooling, there's zero stability. No point in chasing it.
Look, just drop the multiplier to 6 and bump the Front Side Bus up by 5MHz at a time. You'll see pretty quickly when you finally hit that wall. Honestly, hitting 400MHz is a killer result.
Check out that link above from Kyle Sanchez4 or Edward Peterson70. If you finish reading it and still feel a bit lost, just go through it one more time. Seriously, if it still doesn't click after a second pass, read it a third time and I promise everything will finally make sense.☕
So they've got him clocked at 2.7GHz. Honestly, that motherboard isn't anything to write home about for a Duo, either—it's pretty basic. You'd be much better off with an older 965P chipset if you want any actual performance.
Edward Peterson70 said:And before Peak Performance came along, everyone used to talk trash about Colorama cards, even when they had better cooling than the stock models. Look where we are now...
Do you run a Colorama yourself? Their iPhone 15 Pro models used to bundle with either the X1 (or maybe it was the iPhone X, I can't quite remember if one was for ATI or NVIDIA) and the S1, and they were always cheaper than the Microsoft ones. What's the move here?...
As long as it’s Microsoft and I can afford it, I'm happy.👍 Is a 250mm gap actually caused by the GPU, or is the processor just hitting its limit? (Though, let's be real, it's probably both)😁
It looks like I’m starting to see artifacts at 683MHz when I check things in ATI Tool, and I’m honestly not even sure about hitting 67xMHz in AT&T. To be safe, I’m just keeping the core at 655MHz. Powercolor 256 mb
How does everyone feel about using ATI Tool to overclock an X1950Pro? The renderer looks a bit smoother than what I get with AT&T. Personally, I think AT&T performs better, and there’s zero difference in my max clock speeds. I’m just starting to play around with RivaTuner now to see if it actually lives up to the hype (it looks pretty beefy).
EDIT: My max core clock hit 675MHz, though I kept it steady at 655MHz. mema - I had mine running at 805MHz.
darkowl72 said:The readout says 1.43125V, but the actual reading is stuck at 1.392V. Drake is sitting at 1.360V.
It really should be set to 1.36V in the load line. You’d see the Drake drop from 1.392V down to 1.360V—but honestly, that’s splitting hairs. It doesn't really matter.👍
Is it actually possible for the default voltage to be sitting at 1.45V🤷? That feels way too high to me. Also, when you're talking (or rather, when you all are talking—sorry if I'm being too informal here👍 ) about voltages, my understanding is that Drake (at least as far as I'm concerned) isn't the actual voltage under load, but the difference between idle and load. For example:
At 1.43125V, it shows 1.392V and a Drake of 1.360V -> Drake = 1.392V - 1.360V
