They have key notches in different places. Among the pins is a notch. Looking from the top side, the notch on the DDR2 lies slightly to the right of the middle point of the memory stick. On the DDR3, the notch is slightly to the left of the middle. :Desktop_DDR_Memory_Comparison.svg
Thank you all. Found that ddr1 has written 184 at the end of the pin. 240 is written in ddr2 and ddr3. The way to differentiate ddr2 from ddr3 is the position of notches with ddr2 around the middle and ddr3 with notch more towards one end.
The thing is: "the guy" at the repair shop said the ram was probably the cause, because using a low-voltage ram on the system originally designed to use DDR3 ram caused the system to have "extra energy running around" and it fried the logic board.
Shorting is a thing, but yeah, good idea to investigate idomatic explanations like "extra energy running around." The only time I putt DDRL ram into a my DDR machine (this same Macbook), my computer simply wouldn't boot up until I put the old RAM back in.
For the unfortunate folks coming here after the fact looking for answers, DDR3L is dual-voltage compatible. It'll run safely at 1.5v or 1.35v if you're on a newer system that supports the lower voltage. I've got 16GB of DDR3L memory in a Lenovo IdeaPad Y500(third gen, i7 3630qm) and it's running fine. CPU-Z reports it as 1.5v.
I just asked a slightly related question. My system came with the exact same memory and when my new RAM arrived (2 x 8) from Ifixit, I noticed that it was DDR3L and no matter what I did, my system would not boot. I suspect this is the issue - or the RAM sticks suffered from water damage?
@rjones - The older 2009 & 2010 had issues as the standard what not tight enough with the width of the SO-DIMM's some where just too thick! Unless you got some real odd SO-DIMM's the newer DDR3 based systems were much better. Even still people love forcing things which I can tell you will kill the slots! In addition a good bang on the cover in the 15" models right where the slots are could damage them too.
I would say that, it seems your laptop may not support an 8GB module. Or, your laptop may not support asynchronous RAM configurations, (e.g., 4GB & 8GB or, 2GB & 4GB RAM module combinations.) What model/make is your device?
The systems limits listed are not always correct as these where what Apple tested during the development of the system. As Apple did stick with the DDR3 & DDR4 standards larger SO-DIMM's that came out later in time often still work.
There is nothing that can go wrong by using DDR3L instead of DDR3 standard RAM. The DDR3L RAM is "built-in" backward compatible with the standard DDR3 voltage profile as well. It is simply ABLE to operate at a lower voltage than the standard DDR3 RAM does.
Dan, apparently, knows what he's talking about here. I would only add that Over Clockable (AKA: O.C. or XMP) profile RAM may NOT be compatible with older laptops or desktops designed to useDDR3-8500 (1066 MHz) as, the lowest backward compatible frequency for the O.C.'d RAM is most likely going to be 1333 MHz or 1600 MHz, which are not supported by those older DDR3 standard capable devices.
I upgraded my old laptops ddr3 to a 8gb ddr3L stick but I'm getting blue death screen with a bad pool header.what could be the possible reason and any solution for it?.my total capacity is 16gb also I have tried another ddr3L 8gb stick.
I have been stuck with this problem for a month now and I am not able to understand it.I have xilinx kc705 that have a 1GB memory organized as organized as (128 MB x 8) that is one [R]ank of 8, 8-bit devices. so the memory interface is 64 bits. Each device or chip is (128 MB) and has 8 banks (16 MB) each has 14 [R]ows and 10 [C]olumns. as the picture below: so there is 8 of this chip.
My understanding is that each address correspond to 1-byte even though the ddr3 data width is 64-bit and that is because there are 8 chips and all get the same address but each get 8-bit part of the 64-bit. and when each of the chip perform the 8 burst, a 512-bit transfer is done. so I am confused about how to increase the address.
The burst length 8 is a property of the DDR3 protocol. The way it works is that you request a single address, and get 8 words of data back. So you're correct that one burst will give you 64*8=512 bits, but the MIG core will only send one command to the DDR chips (aside from stuff like opening/closing the row).
From the way you describe the situation, I'm guessing that you're using the native MIG UI and not AXI. In that case, the addresses you send via the MIG UI are the actual address bits that go to the DDR chips. Since you have 128M words, you should have 27 address bits. But as you mentioned, MIG is giving you 28 bits. The extra one is the rank bit. This is unnecessary because you only have one rank, so that bit should always be 0. (That's just one of MIG's bugs/annoyances.)
With DDR3 and burst length 8, you always need to request blocks that are 8-word aligned. The lower three column address bits have an effect on the order that the 8 words come back, which is useful if you want one of the words back before the others. However, since MIG ultimately gives you all of the words at the same time, I'm not sure what will happen if you set those to anything but 0.
With the rank and lower 3 column bits always zero, that leaves you 24 bits of usable address space. Those 24 bits give you access to the full space of 512-bit blocks inside your 1Gbyte RAM. So yes, you should increment by 8.
c80f0f1006