Question Does the version number matter when upgrading RAM?

[PC-Mac-User]

I bought 16 GBs more memory for this older system I have. The timings, speed, brand, and all the other numbers are identical to what was in the system. Except the version on my original sticks is 8.22. The memory I bought at eBay it is something like version 3.52. I wonder if version number has to do with just the batch they came from, or could there be physical differences internally too?

On a side note, do company's still do memory matching when you buy everything at once? Like in the past, if you bought two sticks for duel channel, the better brands would match them which was a step above just considering the timings. I'm hopping the memory version is not going to be an issue.

 

[bill001g]

Pretty much all those number mean nothing when it comes to predicting if it will really work.
Even when they take 2 sticks off the same manufacturing line they may not function together. What they do when you buy kits of memory is the factory measures the timings and matches ones that test with the same timings.

Anything else may or may not work. It is made even harder if you plan to run XMP profiles and in your case it sounds like you are going to try to go from 2 to 4 sticks. 4 sticks of memory can sometimes be a challenge on some motherboards/cpu even when you buy kits.

Since you purchased it all you can do is try it. Make sure you leave something like memtest86+ run overnight to be sure your machine is stable. You might have to adjust the timing, it is likely with enough messing around you can get something that will work.

 

[PC-Mac-User]

Yes, I've been using Memtest86 for years. Great program
Just curious. What would be a good "free to the point" program for getting some kind of score on my memory speed. I would like something simple to see what my memory runs at now and how fast it runs after I add the two other sticks. I know you can play with timings too, but it would be nice to see a score and then compare it to what's normal across many system. I know PC mark used to do stuff like that I think it was called, but is there something more basic and to the point?

 

[bill001g]

There are memory benchmarking programs but I seldom use them so I don't know off hand.
You best benchmark would be some game you actually play that has a benchmark option.

In general memory is the very last thing you mess with trying to get better numbers on some benchmark. If I remember right it mostly affects benchmarks that have stuff like file compression and decompression It also has some impact on the IGPU performance but most people chasing performance numbers have a actual GPU.

You in most real life cases will not be able to detect difference in memory speed settings. You might get 1 or 2 FPS in some games which is so small you will never see it unless you constantly have a frame counter sitting on the screen.

In your case getting it to work at all at any speed will likely be more improvement if your applications can actually use the added memory.

 

[geofelt]

CPU-Z will tell you what your ram speed is.
It will also identify the specs for each stick. Hopefully they are identical.
If you fail memtest, try increasing the ram voltage higher than the spec.

 

[PC-Mac-User]

CPUID CPU-Z MSI is showing different MHZ speeds. It's jumps from 799.8 to 800, 1400 and sometime 1500 MHz while sitting idle at desktop with nothing running. Not sure what to make of this.

I know about the voltae settings, I know it was something you had to do when over clocking. Hopefully the regulators on the board are working solid. I run everything standard, Never over clock, although I know from the past that sometimes just changing the clock speed a tinny bit can improve or make timings worse.

​