Core i7 Vdimm/VTT Graveyard :D

[dinos22]

lol for those settings you can run it at 1.4 i'm sure

[hetsaq]

1.368 CPU
1.4 QPI
1.70 DRAM 7-7-7-21 @ 1640
1.80 PLL

Should be acceptable for 24:7 use? ;]

[Alexxx650]

Alex,

Impressive! How long has it been running at 1.72v ?

May, 1st 2009

Works in day of hours 12, 24/7

[saaya]

equant, might be mem failing and degrading as well...

[leonarderman]

I usually use these info below on EVGA Classifield 760
vQPI 1,45V
VTT 1,65V for Uncore 4,5GHz on Water
VDimm 1,85V for 2200MHz cas 8 Kingston HyperX (2000cas9)

I thing VTT 1,65V and VDimm 2.0V are safety

Now I'm running 24/7:
vQPI 1,35V
VTT 1,55V
VDimm 1,75V

[JarrettM]

I'm wondering if this will work for i7 920, as I notice that the voltage seems to be much higher than recommend for EVGA SLI LE X58, or any X58 for that matter:

Muskin DDR3 1600 MHZ

http://www.newegg.com/Product/Produc...82E16820146871

They run at 1.7-1.9v. Running these at 1.8 shouldn't be a problem, should it? I have this in my current build of X48 and I want to carry it over to i7 to reduce cost.

(off topic question - on the EVGA cards, can you run Crossfire on them?)

[Amurtigress]

I'm wondering if this will work for i7 920, as I notice that the voltage seems to be much higher than recommend for EVGA SLI LE X58, or any X58 for that matter:

Muskin DDR3 1600 MHZ

http://www.newegg.com/Product/Produc...82E16820146871

They run at 1.7-1.9v. Running these at 1.8 shouldn't be a problem, should it? I have this in my current build of X48 and I want to carry it over to i7 to reduce cost.

(off topic question - on the EVGA cards, can you run Crossfire on them?)

Hi Jarret,

any kind of memory specified for 1.8V will work on an i7 system but them running at 1.8 at 1600 indicates that they are oldish. Normal DDR3 RAM is running at 1.5V and more than that means higher clock speeds achieved through that increase in voltage. 1.8V modules should be considered to be technology of late 2008.

Just choose RAM that's running at 1.65V at 1600-a pretty safe hint that it is designed with a special eye on i7s and i5/i3. The less voltage the better, the more headspace for overvolting and overclocking. Either to go easy on the clock and use tighter timings instead, or for higher clock speeds.

The official JEDEC specification for DDR3 RAM up to 1333 is 1.5V.

However, in GENERAL, almost any kind of DDR3 memory should work on i7s...I bought Corsair XMS3 1600 CL9/1.8V sticks before I got my i7, no problems. Aside from not hitting 1600 stably because I used 4 of them...but 1200 was just fine with just about the same performance with only 1.5~1.6V

Hope that helps...

[Fritz_the_germ]

Yoh guys,
I read the last four pages of this thread and I am still a little confused about the safe qpi/vtt voltage.
Setting the qpi/vtt on "auto" I will receive a qpi/vtt of 1,54v. As I told that fact on this page : http://www.xtremesystems.org/forums/...60#post4298760 some guy told me (post#933) a qpi/vtt is not a good idea so I reduced it to 1,36v (the system is still stable, vdimm is 1.62v real).

But is a high qpi/vtt the early grave for a i7 like it is for an 45nm build QX9650
http://www.anandtech.com/show/3381

I understood this thread here so far that:
- the vdimm and the qpi/vtt should not be more apart like 0.5v
- a high qpi/vtt (high is <1,54v) wouldn´t cause a degradation of an i7

Here are my bios settings:

Advanced CPU Features:
CPU Clock Ratio ................................ [21 ]
Intel(R) Turbo Boost Tech ...................... [enabled]
CPU Cores Enabled .............................. [All]
CPU Multi Threading .............................[enabled]
CPU Enhanced Halt (C1E) ........................ [disabled]
C3/C6/C7 State Support ......................... [enabled]
CPU Thermal Monitor ............................ [enabled]
CPU EIST Function .............................. [enabled]
Virtualization Technology ...................... [disabled]
Bi-Directional PROCHOT ......................... [enabled]

Uncore & QPI Features:
QPI-Multi………………………..[36]
QPI Link Speed ..............................[6,91]
Uncore Frequency ............................[3840]
Isonchronous Frequency ......................[enabled]

Standard Clock Control:
Base Clock (BCLK) Control .......................... [192]
BCLK Frequency (MHz) ................................[enabled]
PCI Express Frequency (MHz) ........................ [Auto]
C.I.A.2........................................... .. [Auto]

Advanced Clock Control:
CPU Clock Drive ..............................[800mV]
PCI Express Clock Drive ................... [900mV]
CPU Clock Skew ............................. [0 ps]
IOH Clock Skew ............................. 0 ps]

Advanced DRAM Features:
Performance Enhance ...................... [Turbo]
Extreme Memory Profile (X.M.P) ........... [disabled]
System Memory Multiplier (SPD) ........... [10x ]
DRAM Timing Selectable (SPD) ............. [Quick]

Channel A + B + C

Channel A Timing Settings:
##Channel A Standard Timing Control##
CAS Latency Time ......................[9]
tRCD ..................................[9]
tRP ...................................[9]
tRAS ..................................[24]
CR.....................................[Auto]
B2B.....................................[Auto]


Advanced Voltage Control:
CPU Intel Xeon W3565 D0 Stepping
Load Line Calibration ................. [Level 1]
CPU Vcore ..............................[1,26875V]
QPI/VTT Voltage 1.150v .................[1,36V]
CPU PLL 1.800v .........................[1,6V]
PCIE 1.500v ...........................[Auto]
QPI PLL 1.100v ........................[Auto]
IOH Core 1.100v .......................[1,2]
ICH I/O 1.500v ........................[Auto]
ICH Core 1.1v .........................[Auto]

DRAM
DRAM Voltage 1.500v .............. [1,64]
DRAM Termination 0.750v.......... [AUTO]
Ch-A Data VRef. 0.750v ...........[AUTO]
Ch-B Data VRef. 0.750v ...........[AUTO]
Ch-C Data VRef. 0.750v ...........[AUTO]
Ch-A Address VRef. 0.750v ........[AUTO]
Ch-B Address VRef. 0.750v ........[AUTO]
Ch-C Address VRef. 0.750v ........[AUTO]

The temps. are under load 69° C.

[eagzor]

Asus P6T SE crossflashed to P6T, overvolt jumper for QPI used,
Currently running 1.725v VTT/ 1.72v DRAM to get 4000 mhz uncore stable for about 4 weeks now, i'd like to run the maxmimum 1.9 to reach higher uncore if someone else dare to try before me =P
Everest showing some nice numbers: 20k read, 21k write , 24k copy

[eagzor]

Attempted 1.9v QPI to get 4400 Uncore but failed(hard reboot), instead took a shoot for 4200 Unc @ 1.825v. Intended for 24/7 but hey lets see how that works out.

Edit: had to do 1.875v to attain stability

[Xanth]

24/7 Overclock, OK for 4 weeks now. Prime and LinX stable. I went ahead and raised the vDIMM after the Corsair Dominator decided not to play nice at 2000 (v3.2 Micron). I managed to tighten up the timings, seem to get flaky below 1.73 so I left it at 1.74. I am going to politely ignore the 1.65v warnings for now.

i7 930 @ 4.2 Ghz (200 x 21)
vCore: 1.3375 BIOS
QPI 3600mhz: 1.350 BIOS
PLL: 1.8
Corsair Dominator C8 @1.74 DRAM 7-7-7-17 @ 1603

[zsamz_]

24/7 Overclock, OK for 4 weeks now. Prime and LinX stable. I went ahead and raised the vDIMM after the Corsair Dominator decided not to play nice at 2000 (v3.2 Micron). I managed to tighten up the timings, seem to get flaky below 1.73 so I left it at 1.74. I am going to politely ignore the 1.65v warnings for now.

i7 930 @ 4.2 Ghz (200 x 21)
vCore: 1.3375 BIOS
QPI 3600mhz: 1.350 BIOS
PLL: 1.8
Corsair Dominator C8 @1.74 DRAM 7-7-7-17 @ 1603

you still gotta have qpi vtt .05 close to vdimm
so bump qpi to 1.7v

[quake6]

no, you need to have it in 0.5v range...

[Xanth]

I got bored, figured I would tighten up the timings, so far:


i7 930 @ 4.2 Ghz (200 x 21)
vCore: 1.3500 BIOS
QPI 3600mhz: 1.400 BIOS
PLL: 1.9
Corsair Dominator C8 @1.8 DRAM 7-7-6-16-60 @ 1603

Chip seems to hate the 22X multi. Also, the ram hates running any faster. I might be stuck here for a while.

[saaya]

fyi, ive been running my 920 D0 at
1.15v vtt and 1.65v vdimm, which is a .5v gap, for around 6 months
and
1.40v vtt and 1.8v vdimm, which is a .4v gap, for around 10 months...

its has always been a bad clocker, but i didnt notice any degradation...

[AbsoluteZero]

After benching for hours with i7 980 on 1.6v vtt, 1.9v vcore i left the system for cooling down. An editor on a gaming site saw the system cooled down and strip the system for air cooling, didnt reset the bios, opened the system with Ln2 bios config... Instant death.

Cpu was ES so "Bye bye happiness..."

[BenchZowner]

After benching for hours with i7 980 on 1.6v vtt, 1.9v vcore i left the system for cooling down. An editor on a gaming site saw the system cooled down and strip the system for air cooling, didnt reset the bios, opened the system with Ln2 bios config... Instant death.

Cpu was ES so "Bye bye happiness..."

I wouldn't say so.
You can't kill an Intel processor of our era like that.
Just put the CPU on another mobo and it'll work

[BenchZowner]

After benching for hours with i7 980 on 1.6v vtt, 1.9v vcore i left the system for cooling down. An editor on a gaming site saw the system cooled down and strip the system for air cooling, didnt reset the bios, opened the system with Ln2 bios config... Instant death.

Cpu was ES so "Bye bye happiness..."

I wouldn't say so.
You can't kill an Intel processor of our era like that.
Just put the CPU on another mobo and it'll work

[Kain665]

Hey guys, i'm running a 860, what's the general consensus on the max VTT and degregation for these CPUs? I've run all my CPUs mercilessly into the 80s and 90s with decent voltages and never seen any degradation, I've seen 1.21v, 1.4v, and within .5v of the Vdimm. Can anyone clear this up? Why do ram XMP profiles set VTT to something like 1.65v?

[AbsoluteZero]

I wouldn't say so.
You can't kill an Intel processor of our era like that.
Just put the CPU on another mobo and it'll work

Tried 3 different mobos. It was killed on MSI X58 Pro. I tried on Giga X58A UD7 / MSI X58 Eclipse / Giga X58A UD5.. Nothing is changed. I also tried on Giga X58A UD7 after one month passed but it is still dead. Late autopsy.

[BenchZowner]

That's weird.
There's a protection "mechanism" on-die or on the mobos for sure though.
Try giving 1.9Vcore 1.9Vtt on a Rampage III Extreme or any Classified ( E759/E760/E761/E762 ) on air and you'll find out that while it doesn't post ( gets stuck @ FF ), clear CMOS = CPU is alive and not damaged at all ( no degradation, etc )

[christpunchers]

I've been running my 920 c0 at 200x19 for awhile now.

Ram is Corsair 1600 3x2GB Kit CAS 8 timing.

The Vcore is 1.39, VTT at 1.475, VDIMM at 1.65.

Is this safe in the long run with only air cooling? I find the VTT to be quite high for only 1600mhz memory.

Wish there was some way to lower the Vcore but that's about as low as I can go without screwing up in P95 and Linpack.

I'm in the process of rma'ing the ram... I think it's not 100% fully working.

[moky33]

hi

Vtt 1.66/1.70v
Vcore 1.50/1.55v
Vddr: 1.65v
(water)

Since too months 24/24 (boinc) on i5 750 with sucks maximus III (poor mb,no possible to desynchronize ram on high ,pi 17600..) and at this instant no problem

wait and see

[Erklat]

Since it is massively overvolted out of every spec, please report when something changes due to prolonged usage in such conditions.
Hope it doesn't, but 1.7 Vtt is damn a lot for 24/7.

[steponz]

I predict a dead cpu... that way too much vtt... not likely needed...