I noticed a few things wiht my kit. When on auto my RTL (round trip latench) is 54 on one stick and 52 on another stick. Should I match these up? And to what number? Also B2B Casdelay is disabled. Is this correct or should it be enabled?
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
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.
RTL and B2B CasDelay.
- Thread starter Copyright
- Start date
RTL: Do NOT match them up - they're offset for a reason. For triple channel, you'd notice something like 52, 54, 56, etc. You'll also notice that the numbers automatically adjust upward with higher memory speeds and higher uncore speeds as well.
B2B: Disabled is the default, which is also usually equivalent to a value of 4. Anything under 4 (3,2,1,0) also acts as if it's at the default setting. There's no benefit to adjusting the value higher except in the instance of extreme memory overclocks, and there's a fine line between getting things to run optimally, and taking a big performance hit.
B2B: Disabled is the default, which is also usually equivalent to a value of 4. Anything under 4 (3,2,1,0) also acts as if it's at the default setting. There's no benefit to adjusting the value higher except in the instance of extreme memory overclocks, and there's a fine line between getting things to run optimally, and taking a big performance hit.
Thank you! Looks like im just playing the wait game. Been having crashes once or twice a week that started with this kit. I can pass every stability test out there but still got crashes. Just did a bios update and backed my Command Rate down from 1T to 2T to see if this resolves the issue. Crossing my fingers.
If you can pass every stability test, but still crash, then something's still unstable. Could be a software thing, or could still be hardware-related.
MemTest, with one instance running for each core/thread & the memory divided evenly among them, will hammer the hell out of your memory controller. It will find errors that Memtest86+ does not, and that other stability tests will not. If that happens, it could be your memory controller rather than a problem with your actual sticks - in which case speeds/voltages would need to be adjusted.
http://hcidesign.com/memtest/
On another note, your tRRD and tRANKIDLE values look very low. Could just be a reporting issue though - I wouldn't manually adjust them unless you find out it's a real problem.
MemTest, with one instance running for each core/thread & the memory divided evenly among them, will hammer the hell out of your memory controller. It will find errors that Memtest86+ does not, and that other stability tests will not. If that happens, it could be your memory controller rather than a problem with your actual sticks - in which case speeds/voltages would need to be adjusted.
http://hcidesign.com/memtest/
On another note, your tRRD and tRANKIDLE values look very low. Could just be a reporting issue though - I wouldn't manually adjust them unless you find out it's a real problem.
Thanks for the info.. I am holding on to see if the 2T will fix the issue. If not I will start checking other places. I don't want to make to many changes at once. Looking at the values above it shows them as ZERO for the two you mentioned so I am thinking this is a reporting problem but will look into it if I find I am still unstable.
Sgt.Steiger
n00b
- Joined
- Dec 1, 2009
- Messages
- 11
Sgt.Steiger
n00b
- Joined
- Dec 1, 2009
- Messages
- 11
When I picked up this memory I installed it, cleared CMOS, loaded optimized defaults, and enabled XMP Profile 1:
XMP Profile 1 sets:
It was completely stable from there.
For my OC @ 4.0Ghz I left XMP Profile enabled and changed the following:
I would have posted images from my overclock results, but I think the level of privelages I have here right now are making posting images a real PITA.
So I will give you my Mods Rigs thread for my machine:
http://www.evga.com/forums/tm.aspx?m=120266
XMP Profile 1 sets:
- CPU MULTI = 18x EDIT: originally entered 14x here
- BCLK = 166
- VTT = 1.35V
- VDIMM = 1.65V
- Memory Divider = 2:12
- Timings as you see them in my previous post
It was completely stable from there.
For my OC @ 4.0Ghz I left XMP Profile enabled and changed the following:
- CPU MULTI 20x
- BCLK = 200
- CPU VCORE = 1.350 (As set in BIOS, HT ON, LLC ON, ~1.365 IDLE and ~1.385V LOAD in CPUZ)
- VTT = 1.4V
- VDIMM = 1.65V (Set Manually)
- Memory Divider = 2:10
- Still leaving memory timings as you see them in my previous post.
- PCH VOLTS = 1.100V
- PLL VOLTS = 1.800V
I would have posted images from my overclock results, but I think the level of privelages I have here right now are making posting images a real PITA.
So I will give you my Mods Rigs thread for my machine:
http://www.evga.com/forums/tm.aspx?m=120266
Last edited:
Sgt.Steiger
n00b
- Joined
- Dec 1, 2009
- Messages
- 11
When I first was trying for my Max Static OC (all turbo and power saving features disabled), I started by just raising the BCLK to 200 and dropping the multi to 14x while keeping everything else the same.
This setup passed 12+ hours of Prime95 Blend stability and 25 passes LinX.
As soon as I started raising the CPU Multiplier, around 3.8Ghz I had to start increasing VTT, nothing below 1.4V would work.
When I started erroring @ 3.8Ghz I thought it was VCORE, but it was then that it hit me.
If Corsair XMP equipped memory is Intel branded RAM, and 1.35V VTT is set by the XMP Profile and that is with the BCLK at 166, then an increase would probably be necessary in VTT to go from 166 to 200 BCLK.
Increasing VTT was what made my OC stable, and I dropped VCORE from above 1.4V down to 1.35V and stayed there.
I don't think you'll need 1.4V VTT for ~1600Mhz and below DDR3 speeds, but 200 BCLK and 2000Mhz DDR3 def need that much VTT for my CPU and motherboard.
Good luck sir. Have fun.
This setup passed 12+ hours of Prime95 Blend stability and 25 passes LinX.
As soon as I started raising the CPU Multiplier, around 3.8Ghz I had to start increasing VTT, nothing below 1.4V would work.
When I started erroring @ 3.8Ghz I thought it was VCORE, but it was then that it hit me.
If Corsair XMP equipped memory is Intel branded RAM, and 1.35V VTT is set by the XMP Profile and that is with the BCLK at 166, then an increase would probably be necessary in VTT to go from 166 to 200 BCLK.
Increasing VTT was what made my OC stable, and I dropped VCORE from above 1.4V down to 1.35V and stayed there.
I don't think you'll need 1.4V VTT for ~1600Mhz and below DDR3 speeds, but 200 BCLK and 2000Mhz DDR3 def need that much VTT for my CPU and motherboard.
Good luck sir. Have fun.
Last edited:
Sgt.Steiger
n00b
- Joined
- Dec 1, 2009
- Messages
- 11
There has been a lot of discussion about the maximum safe VTT for Lynnfield CPU's. Most refer to an Intel spec sheet that shows ~1.21V Max for VTT.
At best the information available is confusing. I had a chance to ask some questions over at EVGA forums chat, a couple of Intel reps showed up. Was on the side of the road changing an alternator on my mom's Prelude when the chat started.
I found this page:
http://www.intel.com/Consumer/Game/extreme-memory.htm
Now, if you look at the right side of that page, there are PDF's listing supported XMP equipped memory. These are "Self-Certified". One of them is my kit.
My XMP profile sets VTT to 1.35V VTT with the profile enabled.
My reasoning is this, and I have posted this at several sites, if Intel knows about these kits that clearly set VTT above 1.21V, and they list them as supported on their own motherboards and processors, then clearly 1.21V is not the maximum safe VTT voltage.
Just my two cents.
At best the information available is confusing. I had a chance to ask some questions over at EVGA forums chat, a couple of Intel reps showed up. Was on the side of the road changing an alternator on my mom's Prelude when the chat started.
I found this page:
http://www.intel.com/Consumer/Game/extreme-memory.htm
Now, if you look at the right side of that page, there are PDF's listing supported XMP equipped memory. These are "Self-Certified". One of them is my kit.
My XMP profile sets VTT to 1.35V VTT with the profile enabled.
My reasoning is this, and I have posted this at several sites, if Intel knows about these kits that clearly set VTT above 1.21V, and they list them as supported on their own motherboards and processors, then clearly 1.21V is not the maximum safe VTT voltage.
Just my two cents.
Makes sense to me. I have it at 1.3 and will increase it if it crashes again until I can verify if it stable or not. Whats wierd are my crashes are sometimes a week apart and I can pass linx, intel burn in, and prime for 8 hours no problem.
I started thinking i never left my VTT on auto and checked my voltage in windows. So I did it and sure enough it set the VTT to 1.35V. Im gonna let that stay there and see how it acts. I still had to go back and set my timing manually as it set it to 8-8-8-26 3T probably because I pushed the system to 4ghz I assume. Im still trying to get used to this XMP stuff lol.