Just Putzen Round

This is where all those notes you've been keeping come in real handy.

I like to use my highest overclock combination with the lowest voltage and lowest temps for my 24/7. It will likely be in the 4.5 - 4.8 range depending on you're personal comforts.

That my friend is the beauty of AMD.

This would be where you test.
Run some 3D benches at different FSB's and see how it affects the score.

See that's why I like the FX-9590. At stock it runs some really decent speeeeedz and really with decent voltage with all the green on. My hottest temp on the package is only 45c for gaming and heavy use and load testing has me hovering at just 50c. All this running a mere 4.7ghz and 5ghz turbo, there's actually no need to OC past this point. Most games I'm lucky to see 10FPS gain with as little as 200mhz OC to 4.9ghz. Even benching the chip, ALL the tweaks need apply in order to get some gains on performance numbers.

Also Silver_P, you may want to take into consideration all the other stuff that's running besides the game and mods. Just the operating system itself with bells and whistles like Ultimate version in example, takes a toll on performance. So you want to bench PiMod, you'll be better off with W2000 or WXP either NLited for a super stripped OS.

Saying all this, I'm trying to point you in the direction od OS system choice and uses. Would Minecraft game better on 32bit XP and gain some decent frame rates over using W7 Ultimate? Well I'm not really sure, but it IS something you could make a partition or use a spare drive and find out. Call it the MC gaming OS or something.

On a personal note, W10 does no justice for most any benchmarks and for gaming frame rates are very similar to running W7.

ShrimpBrime said:

Is there a difference with heat between Rosetta or F@H? I was thinking to run some F@H on load for a bit as well for testing, but I'm not home enough to monitor it as it should be.

Click to expand...

I mostly "crunch" Rosetta or WCG but I've also done some Folding. I think they are all rather similar as far as running long-term with regards to load/stress/heat? Rosetta is CPU only while Folding can run on both cpu and graphics cards.

PolRoger said:

I mostly "crunch" Rosetta or WCG but I've also done some Folding. I think they are all rather similar as far as running long-term with regards to load/stress/heat? Rosetta is CPU only while Folding can run on both cpu and graphics cards.

Click to expand...

In the past on Phenom chips, I noticed that there are "peak times" for heat with F@H. It's crazy to watch for two hours or more and never see temps go over like 45c for example, then leave it up for a day or two, check temps and see temps another 10c hotter or more where for some reason the Cpu soaked the loop, but then returned to a lower running temp. To be safe, I'd run F@H at 3.8ghz even on cpus that where stable at 4.4ghz. It's non productive to have stability problems for sure.

On that note, I may get back into some folding and see how the FX-9590 does. Just leave it stock meh, cause I fear the beast can melt my coolant lines.

Mr.Scott said:

This would be where you test.
Run some 3D benches at different FSB's and see how it affects the score.

Click to expand...

I'll install Heaven again

Maybe for maximum daily performance... Run the three busses overclocked and sync'd along with the fastest/tightest memory speed/timings that you can for your particular chip/motherboard/ram combo?

I dropped back from 300 to 266 sync'd CPU_NB and HT and bumped up my memory speed to DDR3-2133:

Mr.Scott said:

This would be where you test.
Run some 3D benches at different FSB's and see how it affects the score.

Click to expand...

THIS^^^^^^^^^^

ShrimpBrime said:

On that note, I may get back into some folding and see how the FX-9590 does. Just leave it stock meh, cause I fear the beast can melt my coolant lines.

Click to expand...

Too funny Shrimpy! Sad but true though, these darn Fx 8xxx/9xxx make really good space heaters in the winter.

Your NB is jacked!!!
Mine gets flaky over 2400.

I also can't run over 250 FSB.


So far, I found that 240 FSB is best in heaven.
~6 FPS better than 231 FSB and ~3 FPS better than 250 FSB.

Bet if you had a better board all that would change.

Mr.Scott said:

Bet if you had a better board all that would change.

Click to expand...

Probably.

I can't get 4.9 Ghz stable....
I enabled "One core per compute unit" in the BIOS, which shuts down 3 cores!
Temps dropped like a rock

Pumping 1.63 Vcore still can't make 4.9 Ghz happy. WHEA errors and a hard freeze minutes into Prime95.
NB HT link and RAM are running at 2100, 2300 and 1600 respectively.

Not sure if I have enough CPU/NB volts... (1.3)
Windows won't boot up unless I'm giving it 1.56 Vcore @ 4.9 Ghz

Silver you should look for 4.6ghz or maybe 4.7ghz and Run the HT up 2.6 - 3.0ghz, NB up to around 2400mhz and Get the ram flying as fast as you can.

Don't know why in the world you want to stress test at 4.9ghz. You know it won't be stable while hot.

ShrimpBrime said:

Silver you should look for 4.6ghz or maybe 4.7ghz and Run the HT up 2.6 - 3.0ghz, NB up to around 2400mhz and Get the ram flying as fast as you can.

Don't know why in the world you want to stress test at 4.9ghz. You know it won't be stable while hot.

Click to expand...

That's just it.
It's not hot. Below 62 package while stressing...
It gets hot if I enable one more core though..

NB is already at 2400
Wait. Why high HT link? I thought it had no impact on anything?

Silver_Pharaoh said:

That's just it.
It's not hot. Below 62 package while stressing...
It gets hot if I enable one more core though..

NB is already at 2400
Wait. Why high HT link? I thought it had no impact on anything?

Click to expand...

SP every chip I have, at some point, even if it is below what we consider recommended max temps, will not take any more volts. The only solution is getting it cooler. Shrimpy is making a valid point about the heat. Even when I'm benching with cold water I have had instances where the water is a bit warmer on a particular day but still at -1c as opposed to -6 c. I remember specifically on my 8350 trying to bench Cinebench at a known previous setting, because the water was 5c warmer and even though I was only getting 35c package temps the chip would just net take the volts it needed to run there. Once I got the water back down to -6c I was able to run it.

Mandrake4565 said:

SP every chip I have, at some point, even if it is below what we consider recommended max temps, will not take any more volts. The only solution is getting it cooler. Shrimpy is making a valid point about the heat. Even when I'm benching with cold water I have had instances where the water is a bit warmer on a particular day but still at -1c as opposed to -6 c. I remember specifically on my 8350 trying to bench Cinebench at a known previous setting, because the water was 5c warmer and even though I was only getting 35c package temps the chip would just net take the volts it needed to run there. Once I got the water back down to -6c I was able to run it.

Click to expand...

Oh my gosh that makes sense!

This chip was cold, but no matter what volts I feed it, it just won't take to 4.9 Ghz+.
So for 4.9 Ghz and high I'll have to wait for my loop to come in. Hopefully it will take to that.

I'm not sure why I didn't remember this.. It was mentioned to me before too

Having the loop should help a lot. Like I said it isn't just one of my chips. I've seen it on all of them, when really pushing the volts at some point they just say no maas because of the heat. Even though the temps are below what we consider "max safe".

Mandrake4565 said:

Having the loop should help a lot. Like I said it isn't just one of my chips. I've seen it on all of them, when really pushing the volts at some point they just say no maas because of the heat. Even though the temps are below what we consider "max safe".

Click to expand...

With luck, the Block and res will come tomorrow.
Cleared customs at 3:30 today. Lets hope I didn't get dinged

Even with a water loop, Your still not promised 5ghz with a FX-6300. Remember dude you got it from clearly stated he was benchmarking at 5ghz and higher, no stability involved there.

ShrimpBrime said:

Even with a water loop, Your still not promised 5ghz with a FX-6300. Remember dude you got it from clearly stated he was benchmarking at 5ghz and higher, no stability involved there.

Click to expand...

Yes of course

Running right now at 4.85 Ghz again.
Testing for stability on 3 cores.

I Intend to start fresh and overclock from scratch once my loop arrives.
If temps are cool enough pushing 1.6 Vcore wouldn't hurt the chip would it?

I would not run 1.6v for 24/7. But that's me. I like to stay at 1.45v for 24/7 simply for longevity sake. I'm sure there are many opinions on this though.