I7-3930k & 2133 MHz memory

S

StainlessSteel

n00b
Joined
Mar 13, 2012
Messages
10
I'm thinking of building a new system, I already have 16 gb of Mushkin redline 2133 MHz memory (4x4gb) and I'm thinking of getting the asus Rampage IV extreme with an intel i7-3930k processor. I have seen many examples online of people running 2133 MHz memory on this setup, including a video by newegg & asus. But one reviewer on newegg killed his processor and was told by intel that running his ram above 1600 MHz was the cause. Should I run this memory at a slower speed, say 1866 MHz, or at full 2133 MHz speed?
 
I read that same review on newegg and if I remember - he was using memory rated at 1.65volts instead of 1.5 for sandy bridge platforms. I'm building a new sandy bridge e system and will be using a corsair 16gb kit rated at 2133\1.5.

I'd say try it at 1600mhz first at 1.5volts and if you're stable and temps look good, try the memory at whatever its rated for. But don't go over 1.5volts!
 
Most important is to run the memory at 1.5V and below. But read around both on 2011 and 1155 SB there is very little to nearly no improvement going with faster memory and lower latency seems to be the important option.
 
jpmeaney said:
I read that same review on newegg and if I remember - he was using memory rated at 1.65volts instead of 1.5 for sandy bridge platforms. I'm building a new sandy bridge e system and will be using a corsair 16gb kit rated at 2133\1.5.

I'd say try it at 1600mhz first at 1.5volts and if you're stable and temps look good, try the memory at whatever its rated for. But don't go over 1.5volts!
Click to expand...

+1 jpmeaney

With the 2nd generation Intel® Core™ processors only support DDR 3 at 1066/1333 and now on some 1600; that isnt what I am hearing as causing the issue in almost every case is that they are running the memory at 1.65v +. The 2nd generation Intel Core processor only support memory at 1.5v ±5% (1.425v to 1.575v) anything over this has a chance of damaging the processor memory controller.

Personally the performance increase of anything over 1600 doesnt seems to be worth the extra cost.
 
The Socket 2011 platform can handle 1.65v ram all day long. It's not as weak as your standard issue socket 1155 when it comes to ram performance.

The socket 2011 processors all natively support a baseline of 1600mhz ddr3. However the ram past that speed will not typically make a noticeable difference in performance.
 
I have sandy bridge with RAMs set on 1600 and using 1.65V and everything works fine, should I turn it to 1.5/1333?
 
the memory in question is rated at 1.65v but I can use the settings for their 1833 Mhz memory which runs at 1.5v.

I asked this same question on the Mushkin forum and they asked me where did I hear this from, so I qouted an intel rep who posted on AnandTech and that customer from Newegg. Waiting back for a reply.
 
IntelEnthusiast said:
+1 jpmeaney

With the 2nd generation Intel® Core™ processors only support DDR 3 at 1066/1333 and now on some 1600; that isnt what I am hearing as causing the issue in almost every case is that they are running the memory at 1.65v +. The 2nd generation Intel Core processor only support memory at 1.5v ±5% (1.425v to 1.575v) anything over this has a chance of damaging the processor memory controller.

Personally the performance increase of anything over 1600 doesnt seems to be worth the extra cost.
Click to expand...

I've increased the font size on the key area's that folks don't seem to understand.
IntelEnthusiast is the guy from Intel

If you can run your ram within those voltages your in spec and it will work just fine. Go over those and you could damage the CPU (that's where the controller for the ram is at)
 
Ok, the Mushkin guy fired back with official intel documents that show memory running at 1.65v so here is his reponse in full...

I see.
They have said you will burn up your CPU with more than 1.5V since 1366

There are millions of 1366 and 1155 systems out there running 1.65V without issues.
I have run my own 2500K at 1.65V without issues for a year+

Have a look at this, http://www.intel.com/content/www/us/en/ ... sheet.html
and this http://www.intel.com/content/www/us/en/ ... sheet.html

Those pages are right from Intel, multiple memory manufacturers, with voltages ranging from 1.25V to 1.65V
Multiple mobo manufacturers, including Intel.
You'll notice that several modules run at 1.65V even on Intel branded mobos.

I ask you as a logical person if 1.65V is dangerous to the CPU, why is 1.65V memory certified with Intel mobos on their own page...

However to you original question, if you decide to run at 1.5V you should be able to run 1866 CL9
Click to expand...

The second pdf file from intel does show memory running at 1.65v on their i7-3960X processor and board. If they have changed their minds they better remove or update that pdf.
 
IntelEnthusiast said:
+1 jpmeaney

With the 2nd generation Intel® Core™ processors only support DDR 3 at 1066/1333 and now on some 1600; that isnt what I am hearing as causing the issue in almost every case is that they are running the memory at 1.65v +. The 2nd generation Intel Core processor only support memory at 1.5v ±5% (1.425v to 1.575v) anything over this has a chance of damaging the processor memory controller.

Personally the performance increase of anything over 1600 doesnt seems to be worth the extra cost.
Click to expand...

He has a 6 core. Totally different chip. Your thing doesnt apply much.
 
Hrmm, I am running a 3930K with 1.65v 2133 memory... I always thought 1.65v was in spec, hell my 1366 system ran 1.65v for ages. In any case I thought that voltage limit was really not a hard cap at 1.65, but it was no more than .5v over vcore or vtt or something..

can you post proper links to those 2 pdfs? Thanks!
 
Maybe those d-bags OC'd their CPU Vcore with crazy volts, and then tried to blame it on the memory which they also were OCing past 2133?

Which is more likely?? ;)

Overvolted memory blowing a chip or overvolted CPU Vcore?...
 
extide said:
Hrmm, I am running a 3930K with 1.65v 2133 memory... I always thought 1.65v was in spec, hell my 1366 system ran 1.65v for ages. In any case I thought that voltage limit was really not a hard cap at 1.65, but it was no more than .5v over vcore or vtt or something..

can you post proper links to those 2 pdfs? Thanks!
Click to expand...

Here they are...

http://www.intel.com/content/www/us/en/gaming/gaming-computers/core-i7-memory-suppliers-1111-datasheet.html

http://www.intel.com/content/www/us/en/gaming/gaming-computers/core-i7-extreme-memory-suppliers-1111-datasheet.html
 
StainlessSteel said:
Here they are...

http://www.intel.com/content/www/us.../core-i7-memory-suppliers-1111-datasheet.html

http://www.intel.com/content/www/us...-extreme-memory-suppliers-1111-datasheet.html
Click to expand...

Interesting, the EXACT KIT I am using IS certified for use on the 3960X, which should be the EXACT SAME silicon as my 3930k. There is only one kit and its 1.5v certified for 3930k, but plenty of 1.65v kits certified for 2600k.

I am using G.Skill F3-17000CL9Q-16GBZH, 4x4gb, 2133 @ 9-11-10-28 1.65v which is certified by intel for 3960x. SO, IMHO I think I am fine.
 
StainlessSteel said:
Here they are...

http://www.intel.com/content/www/us.../core-i7-memory-suppliers-1111-datasheet.html

http://www.intel.com/content/www/us...-extreme-memory-suppliers-1111-datasheet.html
Click to expand...

Wow, that's interesting. So Intel has either (1) been wrong this whole time, or (2) there is some sort of disconnect between the division that maintains this gaming-oriented Intel site and the guys who write the processor datasheets.

It's one thing to get ram running at 1.65v, and another thing entirely for it to be "safe" for long-term use, and it's not clear which one we have here.

IntelEnthusiast, care to give us some clarification here? I've questioned your insistence on 1.5v modules before based on the fact that only Corsair has bothered to release a high-end line targeted at Sandy Bridge with 1.5v-only modules. All other major memory module makers released TARGETED Sandy Bridge products at launch a year back that included high-end 1.65v modules.

Can you give us a clear answer?
 
Last edited:
jQcjQ.png


This is using IBT Extreme Max.

I'll pass it on 5, 4.8, and 4.6. The only issue about 5 Ghz is heat and an H100 just isn't good enough for it.

I had to adjust my ram voltage though from 1.5 to 1.525 (for 5 Ghz). That's about it.

EDIT: I just did this today at 8:30 AM. Just got done 12:30 PM.
 
Last edited:
defaultluser said:
*** snip ***
Click to expand...

Why is it so hard to believe that intel may have had to change the memory voltage spec after SB's release? There's a reason for the disclaimer "Specifications subject to change at any time with or without notice" that just about every single manufacturer uses.

It's also apparent that the 1.5V spec if viable, since every single memory manufacturer out there that cares about product compatibility is making 1.5V memory, and has been, for quite some time now.

:)
 
DejaWiz said:
Why is it so hard to believe that intel may have had to change the memory voltage spec after SB's release? There's a reason for the disclaimer "Specifications subject to change at any time with or without notice" that just about every single manufacturer uses.

It's also apparent that the 1.5V spec if viable, since every single memory manufacturer out there that cares about product compatibility is making 1.5V memory, and has been, for quite some time now.

:)
Click to expand...

First off, the Sandy Bridge memory spec of 1.5v +/- 5% has not changed since launch.

Second off, did you READ THE LINK? There are a few entries on that list made on Z68 motherboards, entries which could not have been certified until May/June time frame. So even if your pipe dream of last-minute spec changes existed, it wouldn't hold because we have the following out-of-spec memory added to the list MONTHS after Sandy was released:

2133Mhz [email protected]
16GB 4DIMM
G.Skill
F3-17000CL11Q-16GBXL
MSI Z68-GD65(B3)
4.0
v22.3
Core™i7-2600K

1600Mhz [email protected]
8GB 2DIMM
G.Skill
F3-12800CL7D-8GBSR
MSI Z68A-GD80(B3)
3.0
V17.1
Core™i7-2600K
 
Last edited:
^ dude, settle down. Specs can change at any given time with no notice. Those may have been added to the list months after SB was released, but maybe the spec has changed only very recently. We'l likely never know for sure, either.

If you want to start the leg work on getting a definitive answer, try here.
 
DejaWiz said:
^ dude, settle down. Specs can change at any given time with no notice. Those may have been added to the list months after SB was released, but maybe the spec has changed only very recently. We'l likely never know for sure, either.

If you want to start the leg work on getting a definitive answer, try here.
Click to expand...

That's what I'm doing, getting clarification from a trusted source because we have a conflict here. For almost a year now IntelEnthusiast has been preaching the gospel of 1.5v memory, and now we see a page that conflicts with that gospel. I don't feel it's out-of-sorts to ask for an explanation.

And then you decided to post in this thread about spec changes with no proof whatsoever, derailing the thread. I want confirmation from IntelEnthusiast, NOT SOME GUY ON A FORUM. He's the only man who has a chance in hell of knowing what's going on.

If you insist on derailing the thread to focus on you, Super Genius, why don't you do your own detective work and get back to us?
 
defaultluser said:
That's what I'm doing, getting clarification from a trusted source because we have a conflict here. For almost a year now IntelEnthusiast has been preaching the gospel of 1.5v memory, and now we see a page that conflicts with that gospel. I don't feel it's out-of-sorts to ask for an explanation.

And then you decided to post in this thread about spec changes with no proof whatsoever, derailing the thread. I want confirmation from IntelEnthusiast, NOT SOME GUY ON A FORUM. He's the only man who has a chance in hell of knowing what's going on.

If you insist on derailing the thread to focus on you, Super Genius, why don't you do your own detective work and get back to us?
Click to expand...

Because I don't honestly think it's that big of a deal. I've read the statements of 1.5v +/- 10%, now it's 1.5v +/- 5%... Honestly, I just don't give two shits about it. With the abundance of very inexpensive 1.5v modules flooding the market, neither should anyone else.
 
defaultluser said:
That's what I'm doing, getting clarification from a trusted source because we have a conflict here. For almost a year now IntelEnthusiast has been preaching the gospel of 1.5v memory, and now we see a page that conflicts with that gospel. I don't feel it's out-of-sorts to ask for an explanation.

And then you decided to post in this thread about spec changes with no proof whatsoever, derailing the thread. I want confirmation from IntelEnthusiast, NOT SOME GUY ON A FORUM. He's the only man who has a chance in hell of knowing what's going on.

If you insist on derailing the thread to focus on you, Super Genius, why don't you do your own detective work and get back to us?
Click to expand...

Dude put him on ignore , that's what I did :) best move ever!
 
defaultluser said:
That's what I'm doing, getting clarification from a trusted source because we have a conflict here. For almost a year now IntelEnthusiast has been preaching the gospel of 1.5v memory, and now we see a page that conflicts with that gospel. I don't feel it's out-of-sorts to ask for an explanation.

And then you decided to post in this thread about spec changes with no proof whatsoever, derailing the thread. I want confirmation from IntelEnthusiast, NOT SOME GUY ON A FORUM. He's the only man who has a chance in hell of knowing what's going on.

If you insist on derailing the thread to focus on you, Super Genius, why don't you do your own detective work and get back to us?
Click to expand...
The Intel representative on this forum is likely to be an office person caring for orders or public relations.
He is not a designer or engineer and may not have detailed technical knowledge. However he also says here stuff that may he was told to say. If you are enthusiast yourself do a technical research and bring yourself a result not troll the thread over and over this crap everyone can do.
 
Acer_Sheep said:
The Intel representative on this forum is likely to be an office person caring for orders or public relations.
He is not a designer or engineer and may not have detailed technical knowledge. However he also says here stuff that may he was told to say. If you are enthusiast yourself do a technical research and bring yourself a result not troll the thread over and over this crap everyone can do.
Click to expand...

You don't have to be an engineer to send an email to another department to get more information, but you DO need to be an employee of Intel.

Why is everyone hijacking my effort to get some information about this issue? It's not like Intel makes an easy-to-use webpage with all the information made clear, so our only option is to have our helpful local contact send some emails to the right people.
 
The guys over at xtreemsystems seem to agree that 1.65v is fine and even going up to 1.85v is tolerable. My ram is advertised as being specifically for the X79 platform and it is 1.65v ram, infact it's ON THAT PDF.

If the specs have changed and they revoke the certification of 1.65v rams, then I'd like to see some documentation.
 
Another question, does Memtest+ 4.20 detect and test properly the 2133 Mhz memory on the X79 mainboard with the IMC of the 3930K? I heard a 5.0 ver is coming out soon but no ETA as yet.

BTW, the parts I have so far for my new build are:

CM HAF932 Adv case
XFX 1000w ProSeries Platinum PSU
Corsair H100 cooler
Mushkin 16gb 2133 Mhz 1.65v Redline Ridgeback memory
Mushkin Chronos Enhanced Deluxe 240gb SSD
Creative SB Core3D Fatal1ty Champion PCI-E1X sound card
LG BlueRay writer

Looking at buying the ASUS Rampage IV Extreme BF3 mainboard and intel i7-3930k processor. The rest is being recycled from my current system.
 
StainlessSteel said:
I'm thinking of building a new system, I already have 16 gb of Mushkin redline 2133 MHz memory (4x4gb) and I'm thinking of getting the asus Rampage IV extreme with an intel i7-3930k processor. I have seen many examples online of people running 2133 MHz memory on this setup, including a video by newegg & asus. But one reviewer on newegg killed his processor and was told by intel that running his ram above 1600 MHz was the cause. Should I run this memory at a slower speed, say 1866 MHz, or at full 2133 MHz speed?
Click to expand...

I'm pretty sure it's going over 1.5V that caused the processor to die. I know Corsair has 2133 Mhz RAM at 1.5V and it's running pretty good on my friends 3960K.
 
Actually I bet that dude wayyy overvolted his proc or just got a bad one (and in all my years of building pc DOA intel cpu's are VERY VERY rare)
 
Probably overvolted too much. Hell when I was testing my 3960x at 5Ghz I was at 1.5XX+. And I left it for more than 8 hours of IBT. Didn't die. extide is right too. Intel procs are very hard to kill.
 
Read the whole thread. This got me worry because I bought a gskill 1.65v Memory from newegg. The spec said it is designed for x79 platform which is what I building. Cross my finger
 
gigiba said:
Read the whole thread. This got me worry because I bought a gskill 1.65v Memory from newegg. The spec said it is designed for x79 platform which is what I building. Cross my finger
Click to expand...

I'm going to spend the extra $35 and get the intel OC insurance just in case, If the processor doesn't go BOOM! before the end of the 3 years, the chances it will die afterwards is very low and if it does, newer better chips will be out. I, like you, assumed that the memory manufacturers knew what they were doing releasing these kits as X79 certified.
 
You must log in or register to reply here.
Back
Top