Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

1.35 V & 1.5 V RAM for late 2013 iMac's

My decision was that I returned the iMac.

It's a wonderful machine, but 8GB it's not enough for me.

(the test was with 32Gb)

Geekbench 2.0 32bits score was - 14134

Geekbench 2.0 64bits score was - 15988

Geekbench 3.0 32bits score was - 13376

I might buy another when the 3rd party memory issue is resolved.

Ti'll then I want the money in my pocket and might choose to buy the nMacPro.

Regards,
UnFraGile
 
Did you check with Crucial for memory replacement as they said to a user that there was something wrong with a batch ?

Btw, ola :) eu tambem sou portugues :)
 
silvetti said:
Did you check with Crucial for memory replacement as they said to a user that there was something wrong with a batch ?

Btw, ola :) eu tambem sou portugues :)
Click to expand...

All the memory sticks are OK

The iMac really needs some kind of firmware update to resolve this issue with 3rd party memories. Apple don't confirm anything of when or if it will be resolved, for that I decide to return this "beast" it fit my needs, but 8GB it's not enough.

I've tested all memories. terminal memtest, memory test from diglloydtools (8 hours of stress test). Everything is ok.

But sometimes the computer reboot by itself and give me a report with a Kernel_panic error.

Olá Silvetti :) !
 
Last edited:
unfragile said:
All the memory sticks are OK

The iMac really needs some kind of firmware update to resolve this issue with 3rd party memories. Apple don't confirm anything of when or if it will be resolved, for that I decide to return this "beast" it fit my needs, but 8GB it's not enough.

I've tested all memories. terminal memtest, memory test from diglloydtools (8 hours of stress test). Everything is ok.

But sometimes the computer reboot by itself and give me a report with a Kernel_panic error.

Olá Silvetti :) !
Click to expand...
Have you tried in 10.8, maybe it's a bug in 10.9? Clean installed OS? I just don't understand why the iMac is having so much trouble with aftermarket ram from respectable sources like crucial.
 
2Turbo said:
Have you tried in 10.8, maybe it's a bug in 10.9? Clean installed OS? I just don't understand why the iMac is having so much trouble with aftermarket ram from respectable sources like crucial.
Click to expand...

Yes, I've tried 10.8.4 / 10.8.5 and Mavericks.

Friday they will pick up the iMac. It will be a sad moment, but it as to be.

Regards

Marco
 
unfragile said:
Yes, I've tried 10.8.4 / 10.8.5 and Mavericks.

Friday they will pick up the iMac. It will be a sad moment, but it as to be.

Regards

Marco
Click to expand...

WTF :confused: Is there any aftermarket 16GB ram kit thats compatible/works flawlessly? Why must this be so complicated!!! :mad:
 
2Turbo said:
WTF :confused: Is there any aftermarket 16GB ram kit thats compatible/works flawlessly? Why must this be so complicated!!! :mad:
Click to expand...

I'm almost sure Apple will resolve this issue. But I don't want to invest in this iMac with this "problem", mayb later I'll try again, or jump to the nMacPro.
 
So some one mentioned getting 1866mhz ram to work on there late 2013 iMac,

Can any one confirm that there also running it fine with no issues?

Would I be better with ram at 1866mhz cl11
Or 1600 cl 9

I'm looking for the best performing ram for my late 2013 i7 27" iMac

Cheers!
 
CrAzYLiFe said:
So some one mentioned getting 1866mhz ram to work on there late 2013 iMac,

Can any one confirm that there also running it fine with no issues?

Would I be better with ram at 1866mhz cl11
Or 1600 cl 9

I'm looking for the best performing ram for my late 2013 i7 27" iMac

Cheers!
Click to expand...

In a few days I'll be able to tell if works fine.
Just ordered these memories: G.Skill Ripjaws 1866Mhz 16gb CL10
 
Kernel Crashes/System Restarts on Late 2013 iMac

elithrar said:
This is definitely not true.

I have 1.5V G.Skill (this model) sticks in my iMac, alongside the factory sticks (2x 4GB + 2x 4GB) and I have zero issues. This includes both compiling software, running VMs and some gaming: the machine is being "stressed".

Given all of these posts where people keep getting RAM stuck, I'd be more tempted to say it's just user error (static damage, not seated correctly, etc) than a voltage issue.
Click to expand...

Elithrar,

Coming from just under 20 years of Data Center/Server and Network Support/Engineering, I can tell you my memory wasn't damaged by "User Error." I've never damaged a memory module, processor, component, or expansion card.

I have watched this thing restart itself for days now after running for about 12 hours...

Your system may be running well with aftermarket and Apple supplied RAM but I'm willing to guess the supplied Apple RAM is 1.35v and you're not approaching whatever limit we are.

I just got off the phone with Kingston and the correct part number is now KTA-MB1600L/8G. The "L" means Low Voltage... Not mixed Voltage, Not 1.35v/1.5v. A day of reading blogs seems to confirm that the low voltage RAM fixes the problem in many or all cases.

Apple obviously changed the specs and if you think about it, it makes sense. The system is not supplying enough voltage to power all the modules properly when maxed to 32GB. Maybe at a lower amount of memory or with some of the modules not gobbling up a full 1.5v, but not at 32GB X 1.5v.
 
ted071571 said:
Apple obviously changed the specs and if you think about it, it makes sense. The system is not supplying enough voltage to power all the modules properly when maxed to 32GB. Maybe at a lower amount of memory or with some of the modules not gobbling up a full 1.5v, but not at 32GB X 1.5v.
Click to expand...

So far that's the explanation that's make more sense about people trying to use 32Gb modules of 1.5v and getting Kernel Panic and etc...
 
Jaguarxl said:
In a few days I'll be able to tell if works fine.
Just ordered these memories: G.Skill Ripjaws 1866Mhz 16gb CL10
Click to expand...

Actually a CL9 1600 MHz could score better than a 1866Mhz CL10 in some benchmarks. Wait for my 32GB Kingston HyperX PC12800 CL9 ( khx16ls9p1k2/16 ) These got a lot of excellent performance review.

Will post benchmark soon :cool:
 
IA64 said:
Actually a CL9 1600 MHz could score better than a 1866Mhz CL10 in some benchmarks. Wait for my 32GB Kingston HyperX PC12800 CL9 ( khx16ls9p1k2/16 ) These got a lot of excellent performance review.

Will post benchmark soon :cool:
Click to expand...

Don't suppose you can give a quick explanation?? about to buy the G.skill
 
ted071571 said:
Elithrar,

Coming from just under 20 years of Data Center/Server and Network Support/Engineering, I can tell you my memory wasn't damaged by "User Error." I've never damaged a memory module, processor, component, or expansion card.

I have watched this thing restart itself for days now after running for about 12 hours...

Your system may be running well with aftermarket and Apple supplied RAM but I'm willing to guess the supplied Apple RAM is 1.35v and you're not approaching whatever limit we are.

I just got off the phone with Kingston and the correct part number is now KTA-MB1600L/8G. The "L" means Low Voltage... Not mixed Voltage, Not 1.35v/1.5v. A day of reading blogs seems to confirm that the low voltage RAM fixes the problem in many or all cases.

Apple obviously changed the specs and if you think about it, it makes sense. The system is not supplying enough voltage to power all the modules properly when maxed to 32GB. Maybe at a lower amount of memory or with some of the modules not gobbling up a full 1.5v, but not at 32GB X 1.5v.
Click to expand...

The problem with the KTA-MB1600L/8G is it's dang expensive, looks like over $100 for 8GB where as I can get 16GB Crucial for $140.

Is there anything compatible for cheaper?
 
CrAzYLiFe said:
Don't suppose you can give a quick explanation?? about to buy the G.skill
Click to expand...

Yeah actually I read a an article about this subject really long time ago. I found a similar one. Check the 6 pages thread.

http://www.pcstats.com/articleview.cfm?articleid=873&page=1

It tests RAM modules with different bus speed and latency. Gotta find a recent one although I don't expect a lot of difference to be honest. I'd go for lower latency anyway although no major improvement will be noticeable either case.

Quote :

On average, the system with the memory running at 400 MHz (5:4 memory divider enabled) with aggressive memory timings performed 2-3% faster than the system using high speed memory with loose timings.

While that may not seem like a lot to most people, it can make a world of a difference to the enthusiast, especially if you're gunning for that high score in a clan match where every FPS counts.

It seems as if all the large memory manufacturers/suppliers are afraid to lose face by not pumping out high speed memory modules with lax memory timings just so they can list them in their product lines. Many enthusiasts I know, tend to favour slower memory which allows them to run aggressive timings however




----------
2Turbo said:
The problem with the KTA-MB1600L/8G is it's dang expensive, looks like over $100 for 8GB where as I can get 16GB Crucial for $140.

Is there anything compatible for cheaper?
Click to expand...

http://www.amazon.com/Kingston-Technology-16-PC3-12800-KHX16LS9P1K2/dp/B009W0ZFUY
 
As an Amazon Associate, MacRumors earns a commission from qualifying purchases made through links in this post.
CrAzYLiFe said:
Don't suppose you can give a quick explanation?? about to buy the G.skill
Click to expand...

It's pretty simple. The CAS number is a measure of many clock cycles it takes from the processor requesting data, before it actually gets it. So clearly on the face of it CAS 9 (aka CL9 or CAS Latency 9) is better than CL10.

However, this is only for the first word of the transfer. Once the ram has responded, the processor doesn't have to wait so long for the 2nd word or 3rd etc if they are sequential transfers.

Then you add in the fact that the processor can access different memory banks at the same time - reading data from one bank whilst waiting for another to respond. But this only works well of the processor can predict in advance what data it is likely to need next.

And finally, of course if all other things are equal, then the faster the clock speed, the higher the memory throughput.

So adding this all together and then you ask, is 1866 CL10 preferrable to 1600 CL9 and the answer is - predictably - "it depends".

1866 CL10 will outperform 1600 CL9 on large block transfers, so video encoding, mp3 encoding, anything like that where you are moving static, predictable, large chunks of data around. Then the CAS latency will not be so relevant and the higher clock speed wins.

For more random, less predictable memory access - for example FPS gaming - then often the higher random access throughput means lower CL wins.

But to be honest regards 1866 CL10 vs 1600 CL9, there's really not much in it at all and would you ever notice a difference in overall system performance between the two? Nope, I doubt it.

Hope this helps.
 
IA64 said:
Actually a CL9 1600 MHz could score better than a 1866Mhz CL10 in some benchmarks. Wait for my 32GB Kingston HyperX PC12800 CL9 ( khx16ls9p1k2/16 ) These got a lot of excellent performance review.

Will post benchmark soon :cool:
Click to expand...

Can you post a benchmark with 16gb before placing all 32gb?

I think will be better to compare with my benchmark on geekbench. I only have 16gb right now! Hehehe!
 
From my understanding if you mix the gskill with the apple, the gskill will down clock to 1600mhz but have tighter timings the difference in timings may cause problems, I'd recommend not mixing.
 
Only mix modules with same clock and latency.
This way your system won't scale down the speed.
I'm not sure about different latency, but different clock will not harm your system. Only set to the lower one. In that case 1866 would run at 1600.

Need to confirm, but G.Skill can operate on different clock and latency, for overclock purposes.

That´s the memory i bought: G.Skill RipJaws F3-1886C10D-16GRSL
 
Last edited:
CrAzYLiFe said:
From my understanding if you mix the gskill with the apple, the gskill will down clock to 1600mhz but have tighter timings the difference in timings may cause problems, I'd recommend not mixing.
Click to expand...

The system will run at the fastest latency supported by both ram types, and then choose the fastest clock speed that both memories can run at that latency.

so for example,

Mix 1600 CL10 with Apple 1600 CL11 and you will end up running at 1333 CL10. The systems chooses CL10 and then can only run the Apple memory at 1333MHz at that latency, so the whole system is clocked down to 1333.

Mix 1866 CL10 with 1600 CL10 and it will just run 1600 CL10. It won't try 1600 CL9 just because the 1866 memory can run 1600 CL9.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back