Tuning your AMD CPU for Gaming

[Frakk]

Frakk, please show proof. "Seems to run smoother" has nothing to do with "can help stability", ...because placebo can go a long way.

I can show proof.

Thanks.

I'm not engaging in silly (prove it) games with you. Just because you think one thing does not make it a fact, you can go on repeating yourself a thousand times and it still will not make it true, and its certainly not going to make me stoop so low as to posting prove it links all over the place in some game of one-up-manship of who has the best proof.

The IMC's are rated for 1333Mhz, its on AMD's website and i will amend my post to simply state that, i hope that makes you happy because i'm not doing anything more than that.

 

[BeepBeep2]

I'm not engaging in silly (prove it) games with you. Just because you think one thing does not make it a fact, you can go on repeating yourself a thousand times and it still will not make it true.

The IMC's are rated for 1333Mhz, its on AMD's website and i will emend my post to simply state that, i hope that makes you happy because i'm not doing anything more then that.

lol...silly prove it games? You made a statement about stability based on the fact that you "think it feels smoother"...do you know that there are things that you read on the internet that are not true?

How do you come to these conclusions, and have these conclusions been cross checked? I don't mean "Have other people tried it too, without issue", I mean, have you all actually proven the results, that it does what you say it does?

The scientific theory comes in to play. I don't post anything based on my own opinions or results, what advice I give and the things I say have been repeatedly confirmed through many people's observation and experiment.

The CPUs are rated for 3.2/3.4/3.6/whatever too, which is on AMD's website...better not go above that or you could damage your CPU.

I hope you realize that every CPU has a compatibility rating, if it doesn't make that rating then the part is defective. For example, Ivy Bridge CPUs are only rated for "1600", however most of them can push DDR3-2400 to 2800 easy.

 

[Frakk]

lol...silly prove it games? You made a statement about stability based on the fact that you "think it feels smoother"...do you know that there are things that you read on the internet that are not true?

How do you come to these conclusions, and have these conclusions been cross checked? I don't mean "Have other people tried it too, without issue", I mean, have you all actually proven the results, that it does what you say it does?

The scientific theory comes in to play. I don't post anything based on my own opinions or results, what advice I give and the things I say have been repeatedly confirmed through many people's observation and experiment.

The CPUs are rated for 3.2/3.4/3.6/whatever too, which is on AMD's website...better not go above that or you could damage your CPU.

I hope you realize that every CPU has a compatibility rating, if it doesn't make that rating then the part is defective. For example, Ivy Bridge CPUs are only rated for "1600", however most of them can push DDR3-2400 to 2800 easy.

I think in your words is not saying its a fact, its saying exactly what the words mean I think

Are you drunk?

You may disagree with what i think and if you do its good you should say so, but don't try and force what you think down my throat, it does not work.

 

[BeepBeep2]

Alright, I think I'm done helping out average joes, I must just only cater to the HWBOT, extreme overclocking crowd.

Those kind of people are the clueless type, where I belong. You know, the guys that know how to overclock the furthest? Yep, the dumbest of us all.

No, I'm not drunk. Apparently, just stupid.

 

[Automata]

I think in your words is not saying its a fact, its saying exactly what the words mean I think

Are you drunk?

Hey now, let's not resort to personal attacks nor let this thread get out of hand.

 

[Frakk]

Hey now, let's not resort to personal attacks nor let this thread get out of hand.

Out of hand? yes lets get back to what this topic is actually here for.

 

[BeepBeep2]

Out of hand? yes lets get back to what this topic is actually here for.

Okay, so can I post my Maxmem score now?
Can I start my own topic with the same title, or is that not allowed due to possible trolling?

I don't think you understand that I am not your average joe. Maybe I don't work for AMD, or ASUS, Gigabyte, etc, but I am a reputable and well-known extreme overclocker which I have learned from other reputable overclockers like IMOG, Bobnova, Janus67, chew*, Macci, Hondacity and people and/or engineers within companies like 64NOMIS (Simon Solotko formerly @amd), Raja@ASUS, bingo13@ASUS, The Stilt etc. and other top guys around other forums while having validated behaviors and results with many other members throughout multiple online overclocking communities.

I am here to help you guys, I'm not here for my own benefit.

EDIT:
Let me make it clear, I learn something new every day, and people teach me without ever knowing they do so. However I would like a little respect or credibility, because I feel I have proven myself to be a reliable source of knowledge.

 

[I.M.O.G.]

Frakk, I think the point BeepBeep2 is trying to make has some importance.

Essentially, while it could come across as a challenge to you, what he is trying to make clear is that some of your input/advice is just patently false, and has been dis-proven by many people if you do further reading. That doesn't make you a bad person, it just some of the things you are saying are incorrect, and he is trying to share some experience to maintain some degree of technical accuracy.

For example, your comment about HTLink. Increasing it to 2200 is in NO WAY necessary. Everyone, pretty much everywhere, suggests keeping it at 1800-2000 because increasing it offers NO benefit. This has been tested, thoroughly, by world class overclockers... People who hold top global benchmarking rankings. If increasing HTLink beyond 2000MHz doesn't help a 6GHz+ PhenomII running a 4GHz CPU-NB clock get a better score in Superpi32M, then increasing HTLink beyond 2000MHz doesn't make anything on a daily driver "feel smoother". It is technically inaccurate to suggest anything otherwise.

I haven't read completely, so I apologize for "sniping" this specific point... But when inaccurate things are said, others will come in to make corrections in the honest hope at keeping this forum accurate. That is a good thing, and there is no reason to be defensive about it.

Better to be corrected, than to go on being wrong. This doesn't need to be a challenge or one upmanship between anyone, just accept some input, do some further reading if you aren't sure about what you are being told, then make an informed opinion based off that.

The IMC part that was in the first post is also inaccurate, FWIW. It can't be damaged by running higher frequency. It is so completely inaccurate in fact, that it can make people want to jump out of their chairs to correct it... I think that is part of the problem that is leading this thread off the intended path.

 

[Frakk]

Okay, so can I post my Maxmem score now?
Can I start my own topic with the same title, or is that not allowed due to possible trolling?

I don't think you understand that I am not your average joe. Maybe I don't work for AMD, or ASUS, Gigabyte, etc, but I am a reputable and well-known extreme overclocker which I have learned from other reputable overclockers like IMOG, Bobnova, Janus67, chew*, Macci, Hondacity and people and/or engineers within companies like 64NOMIS (Simon Solotko formerly @AMD), Raja@ASUS, bingo13@ASUS, The Stilt etc. and other top guys around other forums while having validated behaviors and results with many other members throughout multiple online overclocking communities.

I am here to help you guys, I'm not here for my own benefit.

EDIT:
Let me make it clear, I learn something new every day, and people teach me without ever knowing they do so. However I would like a little respect or credibility, because I feel I have proven myself to be a reliable source of knowledge.

We all spend our free time here because we want help people, or give something back.

Just because i don't agree with you is not me saying your wrong. and me not saying "oh yes your absolutely right" is not me disrespecting you.

I have my own opinions, just accept that.

 

[BeepBeep2]

lol, your opinions go against the other 99% of the experienced OC community, should I start a silly petition or silly poll at 5 different forums to get a consensus for you to see that?

Tomorrow, I think I'm going to start a thread about tuning AMD platforms for gaming. I feel one is needed. Maybe other things too, like energy efficiency, and office productivity.

 

[Bobnova]

Frakk, if you don't want to be asked for proof, don't make statements like it's a proven fact.

1.Example:

The only reason you need to turn up the HT Link to 2200Mhz (and no more) is because it can help with stability when running a high CPU-NB.

This is a statement of fact. To make a statement of fact honestly it needs to be two things. 1) Fact. 2) Provable to be fact. If it is not provable, it is not fact.

2.Example:

I have found that for me the only thing raising HT can do is make things run smoother.

This is a statement of opinion, people can disagree with your opinion (and will, as "smoother" is not generally considered a viable benchmark), but since you didn't make any statements of fact you cannot be called upon to cough of proof of those facts.

Beepbeep is entirely correct to ask you for proof on your factual statement, not asking for proof is how we end up with ridicules beliefs that hold back the entire community. Had I happened upon that statement of fact first, I would have demanded proof.
An example of this can be found in the belief that the order of components in a water loop makes a difference. It doesn't, and yet because people who don't actually know the facts make a statement of fact that it does and are not called on it, it is generally believed that loop order matters.
Check it out, ask for proof, check the physics and Hey look at that! Loop order doesn't matter!
Another example is LLC, one guy made a statement of fact without proof that LLC would kill your CPU. Despite that hundreds (that I know of personally) of CPUs have been run with LLC and overvolted without issue for years. (Worse in this case, he fabricated some "evidence", but that is beside the point).


The long and short of it is, the burden of proof lies with he who makes the statement.
If you want people to listen to you, you'd best have the facts and proof that they are, in fact, facts.

 

[Frakk]

Again also @ Bobnova, i stated no facts there. only my own findings.

I think you are confusing me stating my own findings with stating facts.

Yes, people are welcome to disagree with me. not a problem, never has been, never will be, just don't expect me to go against what worked for me.

 

[Bobnova]

We aren't expecting you to go against it. We're expecting you to have something better to back up a claim of "stability" with than "smoother".
If you mean smoother, say smoother. Stability is something else entirely.
I strongly recommend you change modes from statement of fact to statement of opinion, as well.

 

[Frakk]

Thats not what he was talking about, he was going on about the IMC's

But i have changed it to "i found" has that settled it?

 

[BeepBeep2]

Frakk, if you don't want to be asked for proof, don't make statements like it's a proven fact.

1.Example:

This is a statement of fact. To make a statement of fact honestly it needs to be two things. 1) Fact. 2) Provable to be fact. If it is not provable, it is not fact.

2.Example:

This is a statement of opinion, people can disagree with your opinion (and will, as "smoother" is not generally considered a viable benchmark), but since you didn't make any statements of fact you cannot be called upon to cough of proof of those facts.

Beepbeep is entirely correct to ask you for proof on your factual statement, not asking for proof is how we end up with ridicules beliefs that hold back the entire community. Had I happened upon that statement of fact first, I would have demanded proof.
An example of this can be found in the belief that the order of components in a water loop makes a difference. It doesn't, and yet because people who don't actually know the facts make a statement of fact that it does and are not called on it, it is generally believed that loop order matters.
Check it out, ask for proof, check the physics and Hey look at that! Loop order doesn't matter!

The long and short of it is, the burden of proof lies with he who makes the statement.
If you want people to listen to you, you'd best have the facts and proof that they are, in fact, facts.

+1.

Another example is LLC, one guy made a statement of fact without proof that LLC would kill your CPU. Despite that hundreds (that I know of personally) of CPUs have been run with LLC and overvolted without issue for years. (Worse in this case, he fabricated some "evidence", but that is beside the point).

While completely off topic, hopefully I can chime in on this without being rude...

vdroop has been part of both intel and AMD VRM specification to reduce the chance of damaging transient overshoot, which is a temporary voltage spike that can push outside the normal working boundaries of electronic components...

To combat this, you need to run higher PWM switching frequencies (lessens the peak time over the overshoot between transistor on/off switching) which is why such options to adjust PWM switching frequencies are on high end boards, however increasing PWM switching frequencies increases heat due to a loss in power and thermal efficiency.

Due to the nature and timeframe of the voltage spike, a decent o-scope is needed to measure the spikes.

Normally though, the spike is rather minimal and is so short it shouldn't have a noticeable effect on the longevity of parts...but could explain why some chips just tend to instantly pop when put under a load at high voltages, etc.

Here is a small read about transient response
http://en.wikipedia.org/wiki/Transient_response

Thats not what he was talking about, he was going on about the IMC's

But i have changed it to "i found" has that settled it?

Can you please post an example of your instability at HT Link 2000 MHz vs 2200? Thanks.

Oh, and who told you that you can damage the IMCs by overclocking them? Yourself? If you would just state your source (which is extremely simple) I'd be satisfied.

 

[Frakk]

Its gone far enough off subject.

Lets not drag it to LLC, this thread has nothing to do with any of that. the whole thread is already a complete mess, this would have been far better in PM's, in fact anymore please send me a PM.

 

[BeepBeep2]

Frakk, the discussion on LLC and the possibility of it being harmful should be over now...
I assume Bobnova will chime in with something like "that's cool!" and then we move on...

That was not to begin a lengthy discussion, just explain why somebody may have said what was said in his example before.

Due to the nature of PMs (1 person to 1 other, "believe me if you want, or just ignore me,") and the misconceptions involved, I'm gonna keep my statements public. Sorry.

 

[Bubba-Hotepp]

Does ED need to put you guys in the corner for time-out?

 

[I.M.O.G.]

Thats not what he was talking about, he was going on about the IMC's

But i have changed it to "i found" has that settled it?

Were you going to change the wording on the part about the IMC in the first post? That is the second sentence of the first post, it is written as though its a fact, and its inaccurate.

Taking things to PM doesn't make this a more informative, accurate thread. We can always do a Tuning your AMD CPU for Gaming Revision 2.0, where there is less disagreement, if everyone's input is taken under consideration... Otherwise if its an opinion piece you are sharing, its going to be subject to public peer review. Thats what forums are all about - the benefit of community knowledge.

 

[Frakk]

Frakk, the discussion on LLC and the possibility of it being harmful should be over now...
I assume Bobnova will chime in with something like "that's cool!" and then we move on...

That was not to begin a lengthy discussion, just explain why somebody may have said what was said in his example before.

Due to the nature of PMs (1 person to 1 other, "believe me if you want, or just ignore me,") and the misconceptions involved, I'm gonna keep my statements public. Sorry.

Move on and thats fine. Any more goings on about the IMC's ecte... in the thread will get ignored.