Could you share the link on OCN? It's a good idea to start up an overclock chart.
well worth the £17 and a weeks wait without it
Can I send my old chip back to Intel after receive the new one? So I know which one to keep?
4930k should it stay or should it go??
- Thread starter Nickolp1974
- Start date
Could you share the link on OCN? It's a good idea to start up an overclock chart.
Can I send my old chip back to Intel after receive the new one? So I know which one to keep?
I only got 4.5Ghz prime 95 stable @ 1.4volts. If we got by the stats on the Asus forum getting 4.7 or higher is very rare and this chips require more volts to be prime 95 or handbrake stable. Asus stable when binning chips only ONE sample out of 95 did 4.8GHz.
They also give this information.
4960X ES samples. All tested at DDR3-2133 with 16GB of memory. Physics test pass voltage (handbrake encode may require more).
Out of 45 tested:
1 sample managed 4.8GHz at 1.40V
9 samples managed 4.7GHz at 1.40V
21 samples managed 4.6GHz at 1.40V
13 samples managed 4.5GHz at 1.40V
1 sample managed 4.4GHz at 1.40V
The 4.8GHz sample would not make it into Windows at 5GHz, even with 1.47V using watercooling with multiple Thermochill PA120.3 and 120.2 rads in series and 2 X Petra top DDC pumps. Full load temps are right on the wall for this type of cooling at 1.45V.
I don't think getting much more than 4.5 or 4.6Ghz is a given. The sample I have will post and is stable in games at 4.7Ghz @ 1.4volts but is prime 95 stable at 4.5Ghz @1.4volts (handbrake 1080p high profile @ 1.33volts and under 1.3volts for games). If I just want to be stable in games I have a very good sample but if I want a stable system that can pass prime95 then I have to accept a lower frequency. If you want to start a cpu chart then you will need to say what is stable in your book?
They also give this information.
4960X ES samples. All tested at DDR3-2133 with 16GB of memory. Physics test pass voltage (handbrake encode may require more).
Out of 45 tested:
1 sample managed 4.8GHz at 1.40V
9 samples managed 4.7GHz at 1.40V
21 samples managed 4.6GHz at 1.40V
13 samples managed 4.5GHz at 1.40V
1 sample managed 4.4GHz at 1.40V
The 4.8GHz sample would not make it into Windows at 5GHz, even with 1.47V using watercooling with multiple Thermochill PA120.3 and 120.2 rads in series and 2 X Petra top DDC pumps. Full load temps are right on the wall for this type of cooling at 1.45V.
I don't think getting much more than 4.5 or 4.6Ghz is a given. The sample I have will post and is stable in games at 4.7Ghz @ 1.4volts but is prime 95 stable at 4.5Ghz @1.4volts (handbrake 1080p high profile @ 1.33volts and under 1.3volts for games). If I just want to be stable in games I have a very good sample but if I want a stable system that can pass prime95 then I have to accept a lower frequency. If you want to start a cpu chart then you will need to say what is stable in your book?
Last edited:
Soldato
I reckon most chips would do at least 4.7 on 125 strap. It's not as rare as some people make out. 4960s generally don't clock as well anyway. Although mine will do 4.7 @ 1.4v on 100 strap, and 4.810 @ 1.42v on 125 strap.
Just took this. Bench stable...Including Prime for 6 hours last time I tried these settings.
Just took this. Bench stable...Including Prime for 6 hours last time I tried these settings.
Last edited:
Depends what you mean by 'do' - game? prime? ibt etc etc.
I pulled a right fast one with intel lol! - Bought a 4960x on MM for £400 in April, had an intel online chat with a rep and he claimed I could buy the ITP for it even though it was second hand (just thought to cover myself)
Testing with the 4960x was pretty poor results - it couldn't handle 32gb (8gb x 4) of 1866 C10 G Skill unless I kept it at 4.4Ghz (and required 1.38v for that!) - I dropped to 16gb (4x4) of 1866 C9 which just about got it stable (IBT, prime etc @ 1.38v / 4.5ghz). Stock 4Ghz was using up to 1.30v when testing! - so I decided to claim on ITP.
Intel just laughed at my first attempt - they said anything up to 1.5v is acceptable - wtf!! lol
I tried again a week later, this time though I emailed them - they kicked up a fuss asking for evidence of who said to me that I could buy ITP on a second hand CPU. I provided evidence (I had a transcript of the chat with their rep) - So they unwillingly agreed to replace it lol
New one arrived batch number 3330, straight away 4.5ghz fully stable ibt and prime blend at 1.24v! mem at stock xmp of 1866 c9 - it dropped my load temps (custom 280mm rad with ek supremacy) by 15 Degrees c! :> really happy with this one.
This one is capable of 4.7 <1.4v - i did a quick cinebench run when i first got it on 100 strap - that ran fine, 4.8 boots but WHEA error when trying to run CB. Not pushed above 1.4v as i have to agree with 8 pack, these ivye cpu's degrade quick with too high volts. Could prob get more than 4.7 but cba. just wanted a stable 4.5 <1.3v and this one exceeded my expectations (think it falls in the top 20% being able to do 4.7 <1.4v anyways!)
currently FULLY stable @ 4.5 with offset +0.010 medium llc. I could do +0.005 with no LLC in all stability tests other than prime95 with 75% ram useage (thats an awesome stability test btw). Happy with only 1.25v for 4.5ghz completely stable :>
remember guys ITP can only be used once - once its been used then that's it - but its the best £16 I've ever spent lol
Last edited:
Soldato
It will do prime all day long at 4.5 1.24v with medium LLC now (had a memory issue previously), but if it's under 1.3v there's no point pussy footing it!
Last edited:
Depends what you mean by 'do' - game? prime? ibt etc etc.
I pulled a right fast one with intel lol! - Bought a 4960x on MM for £400 in April, had an intel online chat with a rep and he claimed I could buy the ITP for it even though it was second hand (just thought to cover myself)
Testing with the 4960x was pretty poor results - it couldn't handle 32gb (8gb x 4) of 1866 C10 G Skill unless I kept it at 4.4Ghz (and required 1.38v for that!) - I dropped to 16gb (4x4) of 1866 C9 which just about got it stable (IBT, prime etc @ 1.38v / 4.5ghz). Stock 4Ghz was using up to 1.30v when testing! - so I decided to claim on ITP.
Intel just laughed at my first attempt - they said anything up to 1.5v is acceptable - wtf!! lol
I tried again a week later, this time though I emailed them - they kicked up a fuss asking for evidence of who said to me that I could buy ITP on a second hand CPU. I provided evidence (I had a transcript of the chat with their rep) - So they unwillingly agreed to replace it lol
New one arrived batch number 3330, straight away 4.5ghz fully stable ibt and prime blend at 1.24v! mem at stock xmp of 1866 c9 - it dropped my load temps (custom 280mm rad with ek supremacy) by 15 Degrees c! :> really happy with this one.
This one is capable of 4.7 <1.4v - i did a quick cinebench run when i first got it on 100 strap - that ran fine, 4.8 boots but WHEA error when trying to run CB. Not pushed above 1.4v as i have to agree with 8 pack, these ivye cpu's degrade quick with too high volts. Could prob get more than 4.7 but cba. just wanted a stable 4.5 <1.3v and this one exceeded my expectations (think it falls in the top 20% being able to do 4.7 <1.4v anyways!)
currently FULLY stable @ 4.5 with offset +0.010 medium llc. I could do +0.005 with no LLC in all stability tests other than prime95 with 75% ram useage (thats an awesome stability test btw). Happy with only 1.25v for 4.5ghz completely stable :>
remember guys ITP can only be used once - once its been used then that's it - but its the best £16 I've ever spent lol
4.5ghz fully stable ibt and prime blend at 1.24v, this is just average level chip.
Some can do 4.5G with less than 1.2V.
Soldato
Between 1.25-1.3v for 4.5 is slightly above average I'd say.
I've yet to see anyone with sufficient evidence showing 4.5 @ 1.2v however.....But that might mainly be due to the fact you wouldn't really want to run those clocks when you could potentially be running 4.7 at 1.3v or under. That's a good chip
I've yet to see anyone with sufficient evidence showing 4.5 @ 1.2v however.....But that might mainly be due to the fact you wouldn't really want to run those clocks when you could potentially be running 4.7 at 1.3v or under. That's a good chip
That maybe the case for 4930's but 4960's are more difficult to clock due to the larger cache. Besides the lower volts what makes this chip above average is the fact it can do 4.7 <1.4v
I've had it as low as 1.22v (and probably could have taken it lower) with IBT and prime95 blend. I'm a stickler for 100% stability so have found one test which pushes the boundaries more so with IVY-E. Would love to see the results for people trying p95: custom,, 5 min per FFT and 12288 ram if you have 16GB.
Soldato
Nothings ever 100% stable
That maybe the case for 4930's but 4960's are more difficult to clock due to the larger cache. Besides the lower volts what makes this chip above average is the fact it can do 4.7 <1.4v
I've had it as low as 1.22v (and probably could have taken it lower) with IBT and prime95 blend. I'm a stickler for 100% stability so have found one test which pushes the boundaries more so with IVY-E. Would love to see the results for people trying p95: custom,, 5 min per FFT and 12288 ram if you have 16GB.
It seems you have tried the intel tuning plan. Just wonder can I send the old chip back to intel after I receive the new one, so I know which one to keep?
Also, with Intel tuning plan, is this exchange available for anyone with no question asked, even if the old chip is still functional?
Doesn't work that way mate.
you have to satisfy them that the chip isn't working as it should be at stock after you've overclocked it. They are not interested in if the chip doesn't perform at overclocking, only if the chip doesn't perform at stock settings.
So for example - my case was that the chip was running at 1.3v under full load at stock which I deemed to be too high and not acceptable for an Ivybridge CPU (which is supposed to be more power efficient than the older architectures).
They only send you a replacement once you have sent them your old CPU.
you have to satisfy them that the chip isn't working as it should be at stock after you've overclocked it. They are not interested in if the chip doesn't perform at overclocking, only if the chip doesn't perform at stock settings.
So for example - my case was that the chip was running at 1.3v under full load at stock which I deemed to be too high and not acceptable for an Ivybridge CPU (which is supposed to be more power efficient than the older architectures).
They only send you a replacement once you have sent them your old CPU.
are they going to test the chip and verify the problem you told them?
I finally managed to find the source of that 1.38v 5G chip. Owner only posted a CPUZ screen snapshot, without any proof.
Even if it's true, it's just a voltage for booting into Windows, definitely not enough for benching.
I asked 8pack in another post, how much voltage does his best 4930, 4960 need to bench-stable @5G. Let's look forward to his answer.
Last edited:
OcUK Staff
I dunno man. I dont bench 5g I bench 6.3. Binning I have one that can bench on air at 1.39v 5g DDR2666 C9-12-12-21 1T quad, one that can do the same at 1.34-1.35 and one the same 1.325. When I say bench I mean all 3D including Cinebench and all physics tests including Vantage HT enabled.
Last edited:
Soldato
1.325 for 5ghz and 2666c9 is a winning chip indeed