How to set up GTL Ref Values for 45nm & 65nm

[BailA]

Hi,i have a Maximus Formula II with Q9650,my voltage is this:

Cpu V=1,3
Cpu PLL=1.5
FSB Termination Voltage=1,1

What'is the ideal setting for GTL?
Thank's and sorry for my english

[CryptiK]

^ BailA - If you aren't overclocking much (less than ~4GHz) you can just leave them on auto if you like. Each CPU likes something different though it seems, so perhaps experiment a little bit to see what works best. Don't worry, at only 1.1v FSB Termination voltage, you won't be doing any damage no matter what you set the GTL Refs to.

You may find you get some added stability if you set +50mv/+10mv/+50mv/+10mv so give that a try first if you don't want to leave them on auto.

Hi.

Ok, so i've been reading this thread and tried to understand it, i could use some help if possible though:

I'm running an e8400 at 4.05, with 1.4 vCore (bios setting, 1.36 CPU-Z report) on an EVGA 790i.

As soon as i lower vCore, Core0 starts failing Prime FFT tests. Core1 can go down to 1.26 vCore and still run stable.

I have all GTLref lanes set to auto and they're all at +00mV according to BIOS.

Is there any way i can help Core0 by messing around with GTL settings? Please help me if you can, 1800FSB 24/7 would be awesome but 1.4 is a bit high.

Yes GTL tuning may help, but I doubt it will close the gap between your cores much. Try setting different GTL Ref's and see if there is an improvement. You wont damage anything so long as your Vtt (vFSB) is not too high. Try setting Vtt (vFSB) to around 1.25v and play around with the GTL's.

[yeahoo]

Here's my settings. Please suggest me for my settings

Q9550 C1, Ramapage Extreme x48
453fsb@8.5@1.296v(windows v) with LLC enable
vtt 1.33
vNB 1.59 (lower than that it won't stable, i run 6gb of RAM)
CPU PLL 1.65x

CPU GTLVref
0. 80mv
1. 70mv
2. 80mv
3. 70mv
NB GTLV = Auto
i ran prime95 for about 45 mins then i stopoped

i tried to set to as people sugguest somewhere in this thread as
0. Auto
1. -40
2. Auto
3. -40
NB GTLV = Auto
this setting is totally unstable. i can't prime95 at all

Will the CPU GTLV i set to 80 and 70 damage the CPU?

what NB GTLV i need to set?

Thanks

[CryptiK]

No, at your Vtt 80 & 70mv wont damage anything.

I set +50 to +70mv for the NB GTL Ref when testing with a Q9650 - this helped lower vNB too.

[Saturn™]

Hello, people!
two days I can not adjust the FSB 533, Prime95 is not stable.... please help me.

Intel Core 2 Duo E8400 [E0] ~4264MHz [8x533]
Asus Maximus II Formula P45 rev.A3 [bios 2202]
Geil Ultra Plus [2x2Gb] DDR2-1066Mhz 5-5-5-15
Corsair TX650W
Microsoft Windows® 7 Ultimate x64

Ratio CMOS Setting: 8
FSB Frequency: 533
CPU Clock Skew: Auto
NB Clock Skew: Auto 
FSB Strap to North Bridge: 400
DRAM Frequency: 1066

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

1st Information:

CAS Latency: 5 DRAM Clocks
DRAM RAS to CAS Delay: 5 DRAM Clocks
DRAM RAS Precharge: 5 DRAM Clocks
DRAM RAS Activate to Precharge: 15 DRAM Clocks
RAS to RAS Delay: AUTO
Row Refresh Recycle Time: AUTO   
Write Recovery Time: AUTO
Read to Precharge Time: AUTO

2nd Information:

Read to Write Delay (S/D): AUTO
Write to Read Delay (S): AUTO
Write to Read Delay (D): AUTO
Read to Read Delay (S): AUTO
Read to Read Delay (D): AUTO
Write to Write Delay (S): AUTO
Write to Write Delay (D): AUTO


3rd Information:

Write to PRE Delay: Auto
Read to PRE Delay: Auto
PRE to PRE Delay: Auto
All PRE to ACT Delay: Auto
All PRE to REF Delay: Auto
DRAM Static Read Control: dis
DRAM Read Training: dis
MEM. OC Charger: enab
Ai Clock Twister: strong
Ai Transaction Booster: Manual

Common Performance Level:10
Pull-in of CHA PH1: Disabled
Pull-in of CHA PH2: Disabled
Pull-in of CHA PH3: Disabled
Pull-in of CHB PH1: Disabled
Pull-in of CHB PH2: Disabled
Pull-in of CHB PH3: Disabled

PCIE Frequency: 100

CPU Voltage: 1.375
CPU PLL Voltage: 1.50
FSB Termination Voltage: 1.21
DRAM Voltage: 2,2
North Bridge Voltage: 1.36
South Bridge 1.5 Voltage: 1.50
South Bridge 1.1 Voltage: 1.10

CPU GTL Reference (0): Auto
CPU GTL Reference (1): Auto
CPU GTL Reference (2): Auto
CPU GTL Reference (3): Auto
NB GTL Reference: Auto
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto


CPU Configuration:

Ratio CMOS Setting: 8
C1E Support: Disabled
Max CPUID Value Limit: Disabled
Intel Virtualization Tech: Disabled
CPU TM Function: enable
Execute Disable Bit: Disabled

Load-Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

[Telo]

FSB Termination Voltage: 1.21

Your VTT might be a bit low for 533FSB.

[CryptiK]

Hello, people!
two days I can not adjust the FSB 533, Prime95 is not stable.... please help me.

Intel Core 2 Duo E8400 [E0] ~4264MHz [8x533]
Asus Maximus II Formula P45 rev.A3 [bios 2202]
Geil Ultra Plus [2x2Gb] DDR2-1066Mhz 5-5-5-15
Corsair TX650W
Microsoft Windows® 7 Ultimate x64

Ratio CMOS Setting: 8
FSB Frequency: 533
CPU Clock Skew: Auto
NB Clock Skew: Auto 
FSB Strap to North Bridge: 400
DRAM Frequency: 1066

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

1st Information:

CAS Latency: 5 DRAM Clocks
DRAM RAS to CAS Delay: 5 DRAM Clocks
DRAM RAS Precharge: 5 DRAM Clocks
DRAM RAS Activate to Precharge: 15 DRAM Clocks
RAS to RAS Delay: AUTO
Row Refresh Recycle Time: AUTO   
Write Recovery Time: AUTO
Read to Precharge Time: AUTO

2nd Information:

Read to Write Delay (S/D): AUTO
Write to Read Delay (S): AUTO
Write to Read Delay (D): AUTO
Read to Read Delay (S): AUTO
Read to Read Delay (D): AUTO
Write to Write Delay (S): AUTO
Write to Write Delay (D): AUTO


3rd Information:

Write to PRE Delay: Auto
Read to PRE Delay: Auto
PRE to PRE Delay: Auto
All PRE to ACT Delay: Auto
All PRE to REF Delay: Auto
DRAM Static Read Control: dis
DRAM Read Training: dis
MEM. OC Charger: enab
Ai Clock Twister: strong
Ai Transaction Booster: Manual

Common Performance Level:10
Pull-in of CHA PH1: Disabled
Pull-in of CHA PH2: Disabled
Pull-in of CHA PH3: Disabled
Pull-in of CHB PH1: Disabled
Pull-in of CHB PH2: Disabled
Pull-in of CHB PH3: Disabled

PCIE Frequency: 100

CPU Voltage: 1.375
CPU PLL Voltage: 1.50
FSB Termination Voltage: 1.21
DRAM Voltage: 2,2
North Bridge Voltage: 1.36
South Bridge 1.5 Voltage: 1.50
South Bridge 1.1 Voltage: 1.10

CPU GTL Reference (0): Auto
CPU GTL Reference (1): Auto
CPU GTL Reference (2): Auto
CPU GTL Reference (3): Auto
NB GTL Reference: Auto
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto


CPU Configuration:

Ratio CMOS Setting: 8
C1E Support: Disabled
Max CPUID Value Limit: Disabled
Intel Virtualization Tech: Disabled
CPU TM Function: enable
Execute Disable Bit: Disabled

Load-Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

This type of non GTL Ref related question would be best made in the Maximus II Formula Thread

Those voltages would have been stable on my old M2F but I wasn't using 2 x 2GB sticks.

You may need to set Vtt to ~1.25 - 1.30v depending on your CPU's needs. vNB should be ok for 533 FSB.

If raising Vtt doesn't help, try lowering your FSB to say 510 FSB and working your way up.

[Saturn™]

I assign values up to 1.38! Prime95 falls after a few minutes.

[CryptiK]

Try setting GTL's to +40/auto/+40/auto and see if that helps. Perhaps also bump up vNB a little.

What are your NB temps like under load? Some M2F's ram very hot, and the washer mod was required.

[Saturn™]

With the above settings, the system is completely stable in Prime95 at FSB = 520 ... 533 has already snowed errors... Maybe it FSBWall?

CryptiK,
ok, I'll try...
MoBo Temp = 57° Max. I have never seen temperatures above 60 °. Cooling is original, no fans ...

[CryptiK]

Could be FSB wall but I would be surprised if this was the case. Try also, CPU clock skew = 200 NB clock skew = normal

57*C is pretty warm, perhaps you need some washers under the mobo heastink screws - i posted a pic of this in the M2F thread.

[AntecX]

Hello
I have a DFI X48 LT T2R liquid cooled and a Q6600 VID 1.2000v
But I have trouble setting the GTL for this CPU ...

I have set:
VTT 1.327v in BIOS, 1.312v in WIN
CPU GTL 0 / 3 @ 115
CPU GTL 1 / 2 @ 105
NB GTL @ 100

but my Q6600 is not stable at 423x9 with 1.400v

[CryptiK]

I'm not familiar with the older X48 DFI GTL system, they don't use multipliers or mv values. You'll have to just use trial an error to see what works best, or look in old threads and see what others used with quads as a starting point.

[sheri]

What are the default GTL multipliers for REX? I heard that 0/2 has different multi than 1/3?
Funny thing is that i managed to push E8400 to 6100 MHz with all GTL's on auto
http://valid.canardpc.com/show_oc.php?id=1150675

edit:
ok, i got it

0/2 multi is 0.635x
1/3 multi is 0.667x

But despite 0.667x multi on 1/3 i should aim into 0.635?

[CryptiK]

I dont know, I assume 0/2 is 0.635x and 1/3 is 0.667x

I have to tweak them for stability when benching at FSB limit of CPU. Can help get that last couple of MHz. Congrats on 680 FSB

[sheri]

I dont know, I assume 0/2 is 0.635x and 1/3 is 0.667x

I have to tweak them for stability when benching at FSB limit of CPU. Can help get that last couple of MHz. Congrats on 680 FSB

Thx , but on 1/3 i have to aim into 0.635x multi right?

[Willhub]

Hello.

CryptiK, would you have any ideas what sort of GTL's I need for my X3350 and Asus P5Q-E?

I posted a thread here: http://www.xtremesystems.org/forums/...02#post4407602

Thanks
Will.

[CryptiK]

Thx , but on 1/3 i have to aim into 0.635x multi right?

Yeah one of my CPU's likes around that value, but another likes lower than that or I get 3D corruption when pushing the max FSB of the CPU. Vtt also needs to be adjusted for FSB, so I do that first then set GTL Refs.

Hello.

CryptiK, would you have any ideas what sort of GTL's I need for my X3350 and Asus P5Q-E?

I posted a thread here: http://www.xtremesystems.org/forums/...02#post4407602

Thanks
Will.

No idea TBH never used one, but I'd say you could use what other people set with Q9650's as a rough guide - the Q9650 thread in the Intel section should be helpful for you. Start at default GTL's, and experiment with Vtt and fine tune using GTL's. Expect it to take awhile to get it dialed in though quads are tricky compared to duals.

[Willhub]

Thanks I'll have a look at that thread.

Hopefully this app in the OP will give me some decent GTL ref, but I am abit confused about the calculator for the GTL's.

Max and min GTL are different in that to what I have in my bios and in the spreadsheet I cannot change them.

Also unsure if I need to change that 0.15v in blue shade?

I notice this " PS2. This calculator is made for MF2 board but will work with all boards that got correction number in equation (like +40 or - 20 and so on). " So does that mean I have to change those things in yellow?

[Willhub]

Anyone know how to use the GTL Calculator?

[Willhub]

Can someone help please?

I don't get this:

If you have a 45nm cpu, you are aiming for the voltage given by the 0.635x multiplier, so to work out what to put for the the 0.667x GTL Ref (Y) values (it will be a negative number to make the resulting voltage the same as the 0.635x GTL Ref voltage), use the following eqn, but substitute the Vtt you use, in place of the example Vtt.

In the examples given, I have rounded off to 3 decimal places throughout for simplicity.

For me, Vtt = 1.14v

Equation:

Vtt x 0.667 + Y = Vtt x 0.635

1.14v x 0.667 + Y = 1.14v x 0.635
0.760v + Y = 0.724v
Y = 0.724v - 0.760v
Y = -0.036v
Y = -36 mv

So to verify it:
1.14v x 0.635 = 0.724v
1.14v x 0.667 = 0.760v

0.760v - 0.036v = 0.724v

or an easier, non algebraic alternative:

1.14v x 0.635 = 0.724v
1.14v x 0.667 = 0.760v
0.724v - 0.760v = -0.036v
= -36mv

So for me it is 1.38 (VTT) x 0.635 = 0.8763

Problem is, in my GTL Ref I canot put 0.8764, or 0.875, that is higher than it will possibly allow.

Please I need help.


Thanks

[MAXMIZ]

Hi, I do not understand your calculations, if he can tell me if my settings are correct.

Asus P5Q-E with E8500

CPU Voltage: 1.42
CPU GTL Voltage Reference (0/2): 0,635
CPU GTL Voltage Reference (1/3): 0,635
CPU PLL Voltage: 1,56
FSB Termination Voltage (VTT): 1,32
NB Voltage: 1,30
NB GTL Reference: 0,635

Thank you in advance.

[h0t]

Hello everybody.
I'm having some trouble understanding that the GTL Ref! : \
Would I could say that I have values ​​for? Here is (I think!) The information they need to help me please:

CPU Ratio Setting: 9
FSB Frequency: 380
DRAM Voltage: 1.88
Timings: 9-9-9-24
CPU Voltage 1.25000
CPU GTL Voltage Reference (0 / 2) Auto
CPU GTL Voltage Reference (1 / 3) Auto
CPU PLL Voltage 1:52
FSB Termination Voltage 1.12
NB Voltage 1.12

My system:

E8500
P5Q3 Deluxe WiFi-AP
Gainward HD4870 Golden Sample
LC-Power Arkangel 850W V2.3 140mm
2GB - Corsair XMS3 CM3X1024-1600C7DHX

Thank you!

[PsySc0rpi0n]

Can you help me setting my values?

I have an Asus P5K Pro, E8400 E0, Gskill pc2-6400 1066 with this settings:

CPU - 4140Mhz
Bus Speed - 460Mhz
Rated FSB - 1840Mhz
FSB Strap to NB - 333Mhz
Multi - 9x

img9877r.jpg

I still didn't manage to understand the GTL reference voltage.

Can those values be correct?

For cpu gtl i have these options 0.57x; 0.59x; 0.61x; 0.63x; Auto
For NB gtl i have these options 0.61x; 0.67x, Auto

As i have vCore and vNB and FSB Termination Voltage a little bit raised i have chosen lower values for CPU and NB GTL reference voltages. Is this correct??

Cheers

[The|Hunter]

First of all sorry for necro bump

You guys have helped a lot of overclockers with this thread



Both on Normal worked up to 444. Probably changed with +.15 vtt raise. Then -3%/Normal was the best combo - passed 37 hours prime95 blend. Putting them both on the same since just bluescreens or fails prime faster. When I went to +.2 vtt the best setting proved to be -6%/Normal. Not what I expected.
-6%/-3% failed within a few minutes too.
The GTLs allowed lowered vtt and mch when first applied at 4.22GHz oc. Didn't notice any further gain when changed GTL again for +.2 vtt.



I changed from 2x2GB Corsair 8500 to 2x2GB G.Skill 8800 SuperPi because the Corsair wasn't happy over 1094. So far the G.Skill only fails if I give it more than 2.0v. (Just testing) Both sets pass Memtest, the bootable cd version, and large fft's in Prime as long as settings are ok. Possible maybe. What stress test is best for vista to push ram in particular?

The last two. Of course. I have not played with them much. Embarrassingly obvious. Will see if they help.

HI
so you could use both at normal up to 445mhz?

Im still at Ga-x48-ds5 and I had the same Corsair dominator before, but mine was faulting at 1080mhz (even 2.25v).. 1075mhz was max stable (448mhz) with these settings fsb +0.15v, mch +0.300v, cpuv 1.3500v, llc off, gtlr1,2 - normal.
Anything higher would report hw error in windows7 mem diagnostic tool - always between 23-28% per round (its testing either plus LRAD or Stride6 -cache enabled), while mem testx86 4.2 passed up to 1090mhz..




Now i bought 4x 2gb Kingston HyperX 1066mhz and i can't run at 1080mhz or 450mhz fsb, mch +0.325, fsb +0.200v, cpuv 1.36250v, gtlr1,2 normal..

5-5-5-15, extra advanced ram timings 5-6-10-74-6-perf.level 7-auto-2t. And it can make a system halt with high pitch sound..

(these are mobo default extra timings at 445mhz+ 5-5-9-70-5-pl7-auto-2t, even at 455mhz)
---
I can pass windows7 mem diagnostic tool up to ~ 1092mhz (455mhz) at those manual adv timings (including perf. level 7), but ingame benchmarks like Metro2033 (high mode, dof off, physx on) raises cpu cores ~75-98% (avg 90%) and when i see the most cpu usage (i monitor via g15 display) it can freeze with this high pitch sound - almost every time at 3rd repeat (or sometimes on 1st round)..
----

The next game is Resident evil 5 dx10 (all max, 4xaa), 2nd fixed benchmark. Also cpu/mem bandwidth intense and same symptom as Metro2033, anything beyond 1080mhz or lower adv timings like 5-5-10-74-6-pl7-auto-2t causes system freeze -high pitch sound. At the most cpu/ram bandwidth intense parts (raising cpuv didnt help)
-------

Now my question is, should i change those GTLR1,2, will it smooth things out or help eliminate this halting?


i use both GTLR1,2 at normal up to ~ 455mhz and always ended up with this high pitch sound. Even at mch +0.350v and cpu way higher then i use to with 4gb, but i didnt raise adv timings so much, yet.. (memory enhance : Turbo; LLC -Off in all cases)

It's because with 45nm chips they use a lower Vtt of 1.10v as default, same goes with P45 NB it uses a lower default voltage than the older X48/X38/P35 chipsets.

The GTL0/3 and GTL1/2 aren't cores, they are mid/end GTL reference pins.

0/3 are data bus pins. 1/2 are address bus pins. They are logic sampling voltages, taken mid swing and end swing.

...

Hi

Could one of those GTLR affect memory read/copy somehow as well? And what is what in Gigabyte's case

GTLR1 data bus
GLTR2 address bus
?


Tnx both