cancel
Showing results for 
Search instead for 
Did you mean: 

VCCSA above 1V causes instability

WillyK
Level 10
R4E, i7-4960X, 32G RAM (DDR3-1600 XMP), BIOS 4901. Running 4.7G OC.

Auto VCCSA varies between 0.921-0.940V by default.
Too low, I thought. Bumping VCCSA a little should increase the IMC stability has been said.

However, If I try to bump VCCSA to values between 1.0-1.2V, the system gets unstable (Prime95 errors in 3-5 minutes).

How come? This is the opposite of what was expected and I don't understand that. Any ideas?
14,851 Views
35 REPLIES 35

Raja
Level 13

WillyK
Level 10
Raja, funny you should mention this because I went thru this exact guide once again (have read it before), all of it, just 1 hour ago and I still don't get it (stupid me). I know, I have a lot to learn yet.

Well, I'll give it another try though. Probably I've missed something since you are suggesting that reading for me. At the end of the day it would have been nice with a few words of wisdom though. What are we looking at here? Apparently this is not the expected VCCSA behavior.

Raja
Level 13
Posts 3-5 in that thread cover why certain voltages are sensitive to changes. "Analogy number two" should help you understand the process of signal related voltages a bit better and how related stages are delicate.

I did not include an analogy for every type of "signal nasty", but the part about using more strength to throw a ball and how it can decrease one's accuracy or ability to maintain required timing, is part of the reason why more voltage is not the answer in certain loading conditions.

WillyK
Level 10
Thanks Raja, but I'm still stuck here. I’ve read your guide again (all of it) and it basically tells me that anything is possible. Btw I forgot to congratulate you on the brilliant guide before. Your analogies are very good making very complex issues understandable for the broader public.

However, I was looking for some practical advice here. Most of our learning and decision making in life is based on pattern recognition. If it looks like a duck, walks like a duck, talks like a duck… then it’s probably a duck, you know. Yes, systems are different, parts are different, etc. but there are many common patterns and lessons learned across. Naturally. Otherwise it would have been meaningless to have this forum here sharing experiences and discussing favorable settings etc.

That said, I’ve never heard about lower VCCSA helping stability. On the contrary. Especially in your guides you often argue for the necessity to increase VCCSA in order to achieve OC stability. Values around 1.2V have been mentioned repeatedly. That’s a pattern in my book.

Hence my original question. Why are my default (Auto) VCCSA values so low, and why increasing VCCSA introduces instability? The timings and the frequencies are the same so the CPU/RAM should be able to “catch the ball” just as before. Higher voltage should make sure the “ball is bouncing” fast and firmly enough in order for that to happen. I just don’t get it.

In other words. Have you experienced such anomaly before and if you managed to resolve it, how did you do that? Would you care for sharing what was causing this unusual behavior? That’s all I’m asking for.

Raja
Level 13
I certainly have experienced times where increasing voltage makes things worse.

I think you're taking things too literally. Signal voltages need to be tuned - in most cases higher = better but there are cases where it is not. This depends on the CPU sample itself and the parts used.

Let's try and dumb this down a bit more. If I ask you to throw a ball at a tiny target, and you throw it as hard as you can what will happen to your accuracy? Recall the part in the guide about throwing accuracy if voltage is excessive. We have to take into account sample variance - arms have a different strength and ability to throw. Thinking about things this way do you still feel anything is possible? What if your CPU cannot swing voltage comfortably over 1V?


The majority of samples need more voltage. 0.95V is stock for VCCSA most of the time - in 99% of cases CPUs that are clocked up to DDR3-2400 need upwards of 1.10V (some a lot higher than 1.25V) - especially with a semblance of good timings.

-Raja

WillyK
Level 10
Ok. Let me give you a bit more specific description, simply laying down the facts of life.

Before: The RAM (KHX1600C9D3K8/32GX) is XMP DDR3-1600 rated, and XMP was working fine on my original R4E rig with SB (3960X) @ 4.6G and BIOS 4206, 24/7 (4.7-4.8G was Ok too, but I couldn't use it because of insufficient cooling).

Upgrade: New 4960X CPU, BIOS 4901 rig moved "as is" in a new case with a very powerful cooling (I only use water cooling on all my rigs). I've pushed the CPU up to 4.8G (on DDR3-1333) without much hassle. It gets slightly unstable on 4.8G but nothing serious. I'm confident there's even more headroom in this CPU sample and that it could hit 5G+ if I have had even better cooling (closing 85 C degrees @ 4.8G).

After the upgrade, the rig is 100% stable @ 4.7G on DDR3-1333, but not completely stable with XMP DDR3-1600. Nothing major, just rare Prime95 glitches within a specific FFT interval: FFTs between 32K-36K (i.e. smaller FFT are Ok and larger FFT are Ok too). This should be significant.

One of the common approaches to stabilizing the CPU-RAM collaboration is to try adjusting the VCCSA. It puzzled me that it was very low by default (0.921-0940V), and it puzzled me even more that increasing VCCSA would cause instability. Hence my original question.

Needless to say that the RAM passes all Memtest86 etc. testing (after all it's been working fine for the past 3-4 years), and so does the CPU pushed to the limits (various testing). I haven't been able to provoke errors with e.g. OCCT, AIDA64, etc. but I trust Prime95 is the ultimate stability test and as long as one of its workers crashes, I know this will happen in real life too. I'm using this rig for heavy video/music editing and a lot of image processing, so I'm not comfortable with any degree of instability (e.g. crashing a rendering job after 4 hours is not fun at all). On the other hand using DDR3-1333 instead of DDR3-1600 makes a difference for my usage of this rig, so that's why I'm so eager to find a solution.

So what do you think?

Raja
Level 13
You asked initially why this happens as you didn't understand it - we've covered that in posts 2-6 above.

What is the question now? How to make it stable? Memory instability is dealt with in the guide with a hint to which timings to adjust also.

Also, don't be puzzled why VCCSA stays in that range at DDR3-1600 - it's because the processors support up-to DDR3-1866 at stock VCCSA. VCCSA does not ramp until one runs higher DRAM frequencies for obvious reasons. If you had a CPU that needs upwards of 1.20V just to run DDR3-1600 you'd probably want to contact Intel (assuming one is using the correct DRAM).:)

WillyK
Level 10
Never mind. Thanks for your time anyway.

Obviously, I'm not asking out of pure curiosity. I need to solve that, and believe me I've read every single guide here (and beyond). Apparently, I haven't been able to make it work. Relaxing the RAM timings is basically no better than using DDR3-1333, so I don't consider this a usable solution.

What I was trying to tell you in my previous message is that nothing indicates that the CPU or the RAM have reached their limits in itself, but their collaboration is poor @ higher OC compared to my previous CPU. That's all. I know R4E was not made for Ivy-Bridge-E, but I hear people are getting decent results with 4960X and newer BIOSes so it should be possible. Anyway, thanks for trying to help me out.

Raja
Level 13
Well as you overclock the CPU you increase the amount of data transferred over the associated buses in a given timeframe. There can be some give and take as a result of this. Especially if the memory controller is not comfortable with a given memory IC or density.