I don't know what it is I have installed 48 gigs of ram and the bios shows only 40 gigs I have run tests on my ram and have checked threw cpuz all the slots to see if the ram is being recognized wich it is so im kinda confused as to why only the bios and windows only show it being 40 gigs not 48gigs any ideas?

Robbanswe
working 3,9-4,3ghz watercooled is optimal
often ít is memorys mhz and latency settings in bios that matter .
 

So I shouldn't worry about the 8 gigs not showing up?

Robbanswe
No problem with ram.
You drive your SR-2 to fast with so much memory.
48gb require higher timing settings and a Little more volt.
 
 
 
 

Ah ok ill just leave it then thanks alot for the quick reply.

No doubt your memory amount is not showing up because of a BIOs setting not being correct or a very slight overclock instability. At BIOs default check to see if all memory is being read. If not try a slight bump in DIMM voltage.

Thx for all the replies ill try the coms reset to see if that works ill also try adding another psu.

I had same problem and followed this guideline:
1. disable 1st CPU (check manual p.38)
2. start computer, you should see 24GB memory
shutdown..
3. disable 2nd CPU
4. start computer, you should see 24GB memory
This is how you can see which CPU has the problem.
I had to remove CPU and clean CPU-contacts (with water and paper) twice before SR-2 saw all the memory! Of course you can have a problem with one dimm, so you need to check, add/remove, dimms one by one to find faulty one. But a bad CPU-contact may lead you in wrong direction.
I'm running 48GB at1700MHz (with zero problems, knock knock)
I have set timings manually (9,9,9,27) and VTT to 1,35. I have same settings as with 24GB. I have a mixed set of Kingston and G-skill 1600Mhz memory.
 
 

Bit of an old thread but I just got myself the SR-2 and have the same issue when using 12*4 Corsair Dominator GT 2000 MHz ram with dual x5670's overclocked to 4.1 GHz and stable. Latest BIOS (A58).
 
During post it only sees a maximum of 40 GB. Same in windows. However, other tools like CPU-z and Evga e-leet software see all 48 GB so it's not a ram issue.
 
I tried all suggestions mentioned here. Increased dimm voltage (tried as high as 2V). I set the timings manually to 9-10-9-27 @ 1.65V or auto. Same result. Even if I press the CMOS button and clear bios config....I can boot up and it still sees only 40 GB with no overclocking done. I also tried disabling cpu0 then cpu1. No difference.
 
What else can I try??

48 gig posting is not the same as 96, the board with no special tricks will post 48 if the board, the memory, and cpu's are all working right.

stingr9137
I was able to reliably post 96GB using the following Kingston memory:
KVR1333D3D4R9S/8G
 
The product code stands for:
KVR = Kingston Value Ram
1333 = 1333mhz
D3 = DDR3
D = Dual Rank
4 = 4x DRAM chip
R = Registered memory
9 = CAS latency
S = with thermal sensor
/8G = single 8GB stick (sold in other sizes and packagings like 16G as two sticks or 32G as 4 sticks)
 
This was the only way I was able to get it to reliably post 96GB.  It is a bios implementation problem that has no chance of EVGA addressing as its not implemented by them (3rd party).  Also,  16GB sticks have no chance of posting.  I tried tried tried tried.  :-(
 
Hope this helps.

I am running 96GB with 12x8GB Crucial ECC Dual Rank RDIMMs without any major issues. The trick is to hard-set all the memory settings, and for 96GB make sure you set command rate to 2T rather than 1T manually in the BIOS.
 
Also, somebody reported successfully using 6x16GB a while back, but not 12x16GB. You may have better luck with Quad Rank DIMMs, but I never tried it and that would require down-clocking the memory multiplier, which can in turn, counter-intuitively, make the stability and POST-ing reliability worse.

Given what you just said, unless you were joking, any answer I could give would be lost on you.

If you are running with the settings you have in your sig, first place to start looking for the problem is your bclk. 187 is way too high to guarantee you'll see all your RAM. Reset everything to defaults, and then start testing again. Beyond 180 bclk, things start to get wobbly in all sorts of subtle ways.
 
If you reset everything to defaults and you are still getting a missing DIMM issue (or, in your case, 2 DIMMs, which means you are probably missing a whole memory channel on one of the CPUs), start pulling DIMMs one by one, to identify exactly which memory bank you are missing. In linux this would be relatively easy (but labels tend to be wrong, just to confuse things). It would probably be easiest to disable the CPUs in turn and see which one is showing 32GB instead of 48GB. In my experience missing banks often don't show up consistently, you may need to reboot a large number of times to get some statistical sense on which CPU it happens.
 
The other thing to look at is whether your RAM is single-rank or dual-rank. I find on my backup SR-2, with single-rank DIMMs, (6x8GB), one of them sometimes doesn't show up during the POST. This doesn't happen when I use dual-rank DIMMs.
 
Last time this happened to me, it first started happening intermittently, but after a while it became consistent. When it became consistent I was able to reliably track it down to a particular DIMM, which turned out to have duffed out completely and never showed up even on it's own. Since I replaced it the machine hasn't skipped a beat.

Now, if a memory module is not appearing in the BIOS, that is a different story. I was referring to the fact that memory companies count 1GB as 1000MB, when it is really 1024MB.

Well, I just know that sometimes something is so obvious, it is easy to miss.
 
I, obviously, didn't fully understand his question. I was just trying to help. And 8GB shows in BIOS, windows says 7.8GB usable.
 
Also, this forum popped up as a, "You might be interested in."
 
(I have only been building computers for a little while)
 
I am very sorry if I caused confusion.

Actually I'm now running 198 blck @ 4.35 GHz and it doesn't matter whether I run stock blck or 200 blck...the issue remains the same. I've already tried swapping the cpu's and disabling each one. I narrowed it down to 2 specific dimm slots on one of the banks which is why I believe it's the motherboard and I have a replacement on the way. Should be here any day now so we'll see. I also tried moving the dims from one bank to the other and even then the same 2 dimms wouldn't work, however, they do show up with evga e-leet as total memory and older versions of cpu-z (latest one doesn't see it). e-leet sees 48 GB...always but bios and OS NEVER see more than 40 GB. I can clear cmos completely and it wont make a difference, even after I set the timings manually and regardless of what voltage I use.
 
But I also have yet to come across anyone else using the two full banks with the same ram and cpu's as I do to confirm whether they have the same issue or not. I suspect it's the motherboard but it may also be a combination of bios settings (timings, voltages, etc.) together...and finding the exact specific combo is almost like looking for a needle in the hay stack and I don't have time or the will to spend days at a time looking for the perfect combo. 
 
If the replacement board shows the same symptoms then I'll be surprised and pissed off at the same time. 
 
As for removing a dim one at a time....I found that they have to be removed (or inserted) in specific locations one by one for the pc to even boot. And it's not as shown in the manual (it said something like for 8 GB use slot 1 and 3 where on my board I had to use 2 and 4). And it doesn't work the same way on the other bank. It's quite confusing when what they label and refer to doesn't match up with the motherboard. Same goes for the software temps when they refer to cpu0 when it's really cpu1. 
 
 

gordan79
If you are running with the settings you have in your sig, first place to start looking for the problem is your bclk. 187 is way too high to guarantee you'll see all your RAM. Reset everything to defaults, and then start testing again. Beyond 180 bclk, things start to get wobbly in all sorts of subtle ways.
 
If you reset everything to defaults and you are still getting a missing DIMM issue (or, in your case, 2 DIMMs, which means you are probably missing a whole memory channel on one of the CPUs), start pulling DIMMs one by one, to identify exactly which memory bank you are missing. In linux this would be relatively easy (but labels tend to be wrong, just to confuse things). It would probably be easiest to disable the CPUs in turn and see which one is showing 32GB instead of 48GB. In my experience missing banks often don't show up consistently, you may need to reboot a large number of times to get some statistical sense on which CPU it happens.
 
The other thing to look at is whether your RAM is single-rank or dual-rank. I find on my backup SR-2, with single-rank DIMMs, (6x8GB), one of them sometimes doesn't show up during the POST. This doesn't happen when I use dual-rank DIMMs.
 
Last time this happened to me, it first started happening intermittently, but after a while it became consistent. When it became consistent I was able to reliably track it down to a particular DIMM, which turned out to have duffed out completely and never showed up even on it's own. Since I replaced it the machine hasn't skipped a beat.