• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

200 bclock on i7 920?

86 5.0L

86 5.0L

Supreme [H]ardness
2FA
Joined
Nov 13, 2006
Messages
7,118
I have had my 920 since Sept. but recently got a H50 cooler, now I can really stretch out my 920s legs, right now Im sitting at 195 bclock @1.255v, should I up the voltage? whats a good upper limit for voltage? also, I lowered the multi for the uncore and memory to x12 and x6 respectively, any advantages to having this as low as possible? or that I should keep the uncore multi twice what the memory multi is?

 
your image is WAY too small to tell anything, i think you copied the wrong link.
 
past 1.35v is not recommended but doesn't mean you shouldn't. So i suggest increasing vcore to 1.325v and bclk to 200 (x21) for 4.2GHz and stress test it with Prime95 or IBT. If stable for an hour then nock the voltage down a notch and stress test for another hour and if still stable then drop the voltage another notch and repeat.
Obviously if it's not stable the increase a notch and stress test and repeat.

Arkuatika said:
your image is WAY too small to tell anything, i think you copied the wrong link.
Click to expand...

Lol click on it?
 
1. I personally would not feel safe exceeding 1.4v, but I've seen some exceed even 1.5v for benchmarking purposes. I wouldn't recommend this for 24/7 use.

2. There is no advantage from having the uncore and memory multipliers as low as possible. Yes, the uncore multiplier should be twice the RAM multiplier.
 
PC_User said:
2. There is no advantage from having the uncore and memory multipliers as low as possible. Yes, the uncore multiplier should be twice the RAM multiplier.
Click to expand...

Quite right, infact most people have 2xRam + 1
 
so I shouldnt worry about keeping the uncore(thus the memory) speed close to the default 2133MHz?
 
Say your memory is running at 1333Mhz, your uncore should be at least 2666. Most people (like myself) have their uncore higher than 2x your memory speed by one notch.
Your memory (depending on it model) should be able to go above it's rated speed. So you have some Crucial 1333mhz stuff, what model is it? Their bottom budget stuff or Crucial Ballistix?
 
Ooo running mixed speed rates. Ok you're probably looking at what around 1400mhz for the ram? The slower sticks should handle that fine. Stick the RAM voltage @ 1.66v
 
Something like this:

bclk - 200
cpu multi - 21 (turbo)
qpi - 3600
uncore - 3000
memory - 1400 (with stock timings of the 1333 stick 7-7-7-24)

vcore - 1.325v
pll - 1.90v
qpi - 1.350v
dram - 1.66v (ignore warning)
cpu differential amplitude - 800mV

If this is stable (20 runs of IBT on max or 1 hour prime95) then start to lower the vcore a notch and stress to find a sweet spot. If not as said before increase a notch a stress etc
 
I upped the memory multi a notch, stock voltages minus 1.25 for the CPU

 
Very nice temperatures!
You say 'stock voltages' so they're all on stock and not 'auto' as auto has been known to increase quite a lot. Push for 200bclk now mate :)
 
as far as I know, voltage settings on a flaming blade are not auto-capable, anyway to check? in the BIOS I assume.

I am having trouble getting to 200bclock, I am adjusting voltages .1mV at a time, trying out different combos but always ending up with BSODs

brb, gonna try 1.3V+ on the core

EDIT: got it, thanks everybody, now to enjoy my 920 :D

 
Last edited:
quick question regarding "stock" voltage. where does one determine what's stock? is this info found on the intel data sheets?

i have a xeon x3440. i'd like to know what the stock vcore is supposed to be.
 
Usually have a look at Intel spec sheets but your Xeon says Voltage: 0.70 - 1.400 V. As it's in the same family as such as the i7 lot i would hazard a guess 1.2v
 
Rossi~ said:
Usually have a look at Intel spec sheets but your Xeon says Voltage: 0.70 - 1.400 V. As it's in the same family as such as the i7 lot i would hazard a guess 1.2v
Click to expand...

Diiferent batch numbers will usually have different "stock" voltages.

For example, take the Q6600... "stock" voltages ranged anywhere from around 1.25 to about 1.325.. maybe even higher depending on what stepping and what batch number you got.

The easiest way to tell is to load up CPU-Z and look at what it says the CPU voltage is when running at stock speed and voltages in the BIOS.
 
Last edited:
Cpu-z isn't all that accurate though im not sure what is these days? I just go on what's in the BIOS hw monitor as it's .25v out from cpuz so i'd trust that.
cpu-z has been better though past couple of years, i remember when it used to waaaaay out lol
 
well, if i disable speedstep and all c-states, but leave turbo boost on, my mobo is reporting 1.26875v at 2.66GHz.

mystery solved!

edit: okay, this is so confusing. under the voltages section of my bios, i have vcore set to auto. the default "normal" value (which i believe is just a guideline and it's not a value you can change) is showing 1.26875v.

if i fire up gigabyte's easytune6 utility, it's reporting an actual value of 1.28125v. cpu-z shows 1.24800v.

also, in my bios, there is a section that shows all the current settings as far as bclk, cpu and mem multipliers, vcore and dram voltages, etc. on that page, the bios is reporting 1.24800v.

which number do i trust? i'm inclined to trust the 1.248v since both cpu-z and my bios are reporting that as a real-time number. the 1.26875 isn't a real time number, and i think this easytune6 program is kind of a piece of junk.

thoughts?
 
Last edited:
You must log in or register to reply here.
Back
Top