Curiosity

So i have the setup in my sig currently and have since upgraded to a Nepton 140XL which BTW dropped my temps @ 4GHz @ 1.225v from ~30*C idle 56*C Load to 17-20*C Idle 40*C Load. I have since overclocked it further to 4.5GHz @ 1.30v stable with a max temp of around 42.5*C. Now my curiosity has peaked with this chip, and am wondering how much further that I can push it. I have only had a few hours worth of tinkering around with it thus far and have not had the time to invest into further clocks. Also as a note I have only used the multiplier for OCing with no additional freq added. I intend to find out what that will be limited to when I have the time.

This is a bit of speculation but I feel that i can push to around 4.8GHz no problem which seems to be the average for these chips probably around 1.325-1.35v. I have seen that around 1.42v is the stock voltage for 4GHz in turbo mode, is this true as I have not seen this thus far for my setup. Would anyone care to enlighten me on this subject and possibly speculate on what ones opinion is on how far I can push this? I would love to see this chip hit 5Ghz but that seems to be few and far between based on my research.

As long as your load temps are under 62 which is max stable temp for visheras and your vcore below 1.525 your good to go, try 5ghz.

Damn those UD3 boards are suprising me big time.

I would say how far the cpu might go is dependent on the Rev X.X of your motherboard and roughly when the cpu might have been purchased. Truth is all we can do is guess based on seeing a number of cpus. I just made a post earlier today about "how" it seems the later FX 8 cores cpus are acting. Let me go find the link.
RGone...

I wrote that in response to some of our stick around and help users that are here all the time and have just purchased new FX 8 core processors and are now running them upward thru their paces.

Thats a very interesting read RGone, it seems to somewhat reinforce what I'm thinking with this chip. As for the rev. it it a 4.0 model and seems to like the 8320 I have chosen for it. Also I'm sure that the higher Frequency RAM helps(currently @ ~2133Mhz) a little with the clocking. @Zakarro I would prefer to that under 1.4-1.45v if possible, however this is still only about a 23% OC, of which me best was roughly 40% in years past.(on air none the less)

I believe that if the chip has it in it.. it shall be able to hit all mentioned marks, only time and tinkering shall tell. This is however still interesting as this new cooling solution has definitely flipped my thinking for this chip.

oh and to make a note the vcore reads around 1.225-1.272v still ironically. However I might need a dif voltage readout app.

Mag10 when you say the temps are 40c under load what are you loading it with and also what temp, are you referring to the socket temp or the core temp. You should post a screen shot of Hwmonitor after 20 minutes of Prime blend so we can see what's, what.

Mandrake4565 said:

Mag10 when you say the temps are 40c under load what are you loading it with and also what temp, are you referring to the socket temp or the core temp. You should post a screen shot of Hwmonitor after 20 minutes of Prime blend so we can see what's, what.

Click to expand...

OCCT linpack test should produce the most heat to really see how his cooler does. Prime is so 14 years ago

Zakarro said:

OCCT linpack test should produce the most heat to really see how his cooler does. Prime is so 14 years ago

Click to expand...

Maybe but it still works, another reason why we recommend prime is so we can all be on the same page. This is also the same reason why we recommend Hwmonitor as a temp reading software, it just makes it easier to help someone when we know what all the readings are and again are on the same page.

Mandrake4565 said:

Maybe but it still works, another reason why we recommend prime is so we can all be on the same page. This is also the same reason why we recommend Hwmonitor as a temp reading software, it just makes it easier to help someone when we know what all the readings are and again are on the same page.

Click to expand...

To be honest I seen OCs that will pass prime but give OCCT errors. OCCT I trust 200%

Zakarro said:

To be honest I seen OCs that will pass prime but give OCCT errors. OCCT I trust 200%

Click to expand...

I've seen that too but I think I've seen the opposite phenomenon as well. Both are good. The thing I like about OCCT is you can set the time of the stress test run.

Zakarro said:

To be honest I seen OCs that will pass prime but give OCCT errors. OCCT I trust 200%

Click to expand...

That maybe so and I do know people that use OCCT and they have no issues with it. I have not used it only because Prime Blend works for me, I know if my rig passes 2 hours blend I never have an issue. I use this rig 8 hours a day for work, which will utilize near 7g of ram and have the Cpu running between 50-100% during those hours, add on to that the hours for daily computing. It has also at times has run F@H for 2 weeks straight 24/7, if you are unfamiliar with F@H it is a pretty demanding program and runs the Cpu and/or Gpu at 100% constantly. I have run it at 24/7 OC as seen in my signature and still not a hiccup. I have not had or seen one person cross through this forum that we have helped said that after passing 2 hours prime blend they were having Oc related stability issues. The point though, we have countless people come here looking for guidance and help. Unfortunately, we cannot reach through the internet and put our hands on ones rig and help them tune it. Therefore, it makes our job a lot easier when we all get on the same page. Using software we are all familiar with is one step in getting there.

trents said:

I've seen that too but I think I've seen the opposite phenomenon as well. Both are good. The thing I like about OCCT is you can set the time of the stress test run.

Click to expand...

I always leave infinite

Mandrake4565 said:

That maybe so and I do know people that use OCCT and they have no issues with it. I have not used it only because Prime Blend works for me, I know if my rig passes 2 hours blend I never have an issue. I use this rig 8 hours a day for work, which will utilize near 7g of ram and have the Cpu running between 50-100% during those hours, add on to that the hours for daily computing. It has also at times has run F@H for 2 weeks straight 24/7, if you are unfamiliar with F@H it is a pretty demanding program and runs the Cpu and/or Gpu at 100% constantly. I have run it at 24/7 OC as seen in my signature and still not a hiccup. I have not had or seen one person cross through this forum that we have helped said that after passing 2 hours prime blend they were having Oc related stability issues. The point though, we have countless people come here looking for guidance and help. Unfortunately, we cannot reach through the internet and put our hands on ones rig and help them tune it. Therefore, it makes our job a lot easier when we all get on the same page. Using software we are all familiar with is one step in getting there.

Click to expand...

yes I know what folding is but that means nothing

2 hours of stability testing is not enough, maybe for work but not for gaming.


My general rule for any test is at LEAST 8 hours over night preferibly 12 with 0 errors. whether it be memtest, OCCT cpu test or OCCT gpu test. If this is not done I will find instability.

Zakarro said:

I always leave infinite



yes I know what folding is but that means nothing

2 hours of stability testing is not enough, maybe for work but not for gaming.


My general rule for any test is at LEAST 8 hours over night preferibly 12 with 0 errors. whether it be memtest, OCCT cpu test or OCCT gpu test. If this is not done I will find instability.

Click to expand...

Yes we all have our preferences and what works for some may not be the next persons choice. Either way Zakarro if you want to start a discussion of which is the better stress test by all means go ahead. Just try not to hijack another thread. We're getting away from the original post here and need to focus on why we're here.

Johan45 said:

Yes we all have our preferences and what works for some may not be the next persons choice. Either way Zakarro if you want to start a discussion of which is the better stress test by all means go ahead. Just try not to hijack another thread. We're getting away from the original post here and need to focus on why we're here.

Click to expand...

Sorry didnt even realize I was OT just responding to that guy giving him advice that 2 hours is not enough Ive gotten errors at 2:30 go figure