Silicon Lottery - worth it?

[JamesIceland 14]

I paid for $469 for it delidded and 5.1Ghz - I didn't see that as a dreadful premium and definitely cheaper than me attempting and failing at a delid myself. I'm very pleased with the results both performance and temperature wise.

I am not sure about the percentage of chips that are able hit 5Ghz but I have built alongside a new friend of mine in the UK, exactly the same system, and he is having stability issues at 5Ghz. So between us it's a 50% hit rate ;-) With Martin it's 66% on here. Maybe we can do a quick poll?

 

 

 

 

 

[martin-w 3,234]

21 hours ago, TechguyMaxC said:

Can you show me what you read the other day that made you think this?  Asus still has the 80% figure up on their overclocking guide.   https://rog.asus.com/articles/guides/the-kaby-lake-overclocking-guide/

 

Nothing definitive Max, just talk on a few hardware forums.

Asus did indeed state 80% based on the samples they tested, and indeed, I'm sure they test many. Would have been initial batches though, mostly pre release, so who knows how they fair now.

I'm referring to not delidded of course.

[Alan_A 1,650]

11 hours ago, joemiller said:

...can hit 5.1 However P3D cares little to nothing about this. So, I left mine at 4.9Ghz . Other games might take advantage of a 5.1 or 5.2; but our beloved sim? pfffff-  Zero!  

Came to a similar conclusion about my Haswell - fine at 4.6, wouldn't get to 4.7, not a temperature or voltage issue, just wouldn't stabilize there.  So I settled at 4.6 and had perfectly good performance in FSX and then P3D.  Would be fine with something similar on the next build - not interested in pushing the clock to the limit.

Will be keeping an eye on the percentages (how many getting to 5.0?)  I agree with James that the premium for a pre-tested 5.1, or even 5.0 isn't huge if you consider a two- or three-year service life.  5.2 is another story.  But a DIY (and delidded) 5.0 seems like a decently safe bet to take. 

 

[martin-w 3,234]

18 minutes ago, Alan_A said:

Came to a similar conclusion about my Haswell - fine at 4.6, wouldn't get to 4.7, not a temperature or voltage issue, just wouldn't stabilize there.  So I settled at 4.6 and had perfectly good performance in FSX and then P3D.  Would be fine with something similar on the next build - not interested in pushing the clock to the limit.

Will be keeping an eye on the percentages (how many getting to 5.0?)  I agree with James that the premium for a pre-tested 5.1, or even 5.0 isn't huge if you consider a two- or three-year service life.  5.2 is another story.  But a DIY (and delidded) 5.0 seems like a decently safe bet to take. 

 

 

 

Yep, my 6700K does 4.6 at a nice temp. Will do 4.7, but why on Earth should I worry about 100 MHz when the increase in frame rate as a result is so low in percentage terms it's not measurable. 

If we are honest with ourselves, we don't need that extra few hundred megahertz. Again, if we were honest with ourselves, it's really just about fun, and sense of achievement.

[joemiller 269]

10 hours ago, martin-w said:

 

 

Yep, my 6700K does 4.6 at a nice temp. Will do 4.7, but why on Earth should I worry about 100 MHz when the increase in frame rate as a result is so low in percentage terms it's not measurable. 

If we are honest with ourselves, we don't need that extra few hundred megahertz. Again, if we were honest with ourselves, it's really just about fun, and sense of achievement.

Correct.... However, most of us got carried away by the urge to " upgrade, OC and tweak,"  then realized.... "Oh heck, I forgot I need to be flying."