Confirming Installed RAM Speed

[gadgetfreak98]

All:

I just completed an upgrade of my 2009 Mac Pro (updated to 5,1 BIOS) to Xeon 5680s. That seemingly went fine.

I already had 36GB of RAM installed before the upgrade. I vaguely recalled buying 3 sticks of 1333 speed RAM last time around, just in case I ever wanted to do exactly this upgrade and figured a few extra dollars future-proofed me. I don't recall doing that with the first 3 sticks I had installed, so I thought I might have had a mix of RAM speeds. The labels on some of the RAM clearly say 1333, but the other three are unclear.

So I pulled up System Profiler and lo and behold it shows that all 6 sticks are 1333, as you can see in the screenshot. Is that reliable -- e.g., should I feel comfortable that all 6 are actually 1333? Or is it possible that due to all the hacky upgrading, that it just reports the standard 5,1 speed (or the speed of the fastest sticks) for all?

Would love to know I'm good to go if this is reliable. Thanks!

 

[tsialex]

All:

I just completed an upgrade of my 2009 Mac Pro (updated to 5,1 BIOS) to Xeon 5680s. That seemingly went fine.

I already had 36GB of RAM installed before the upgrade. I vaguely recalled buying 3 sticks of 1333 speed RAM last time around, just in case I ever wanted to do exactly this upgrade and figured a few extra dollars future-proofed me. I don't recall doing that with the first 3 sticks I had installed, so I thought I might have had a mix of RAM speeds. The labels on some of the RAM clearly say 1333, but the other three are unclear.

So I pulled up System Profiler and lo and behold it shows that all 6 sticks are 1333, as you can see in the screenshot. Is that reliable -- e.g., should I feel comfortable that all 6 are actually 1333? Or is it possible that due to all the hacky upgrading, that it just reports the standard 5,1 speed (or the speed of the fastest sticks) for all?

Would love to know I'm good to go if this is reliable. Thanks!

The only thing that you can do to be sure is always clear NVRAM after you change memories, this force the redirection of the memories.

If the memory controller re-detects the DIMMs and if all DIMMs support 1333MHz, your Mac Pro will run at 1333MHz.

 

[gadgetfreak98]

So I did the NVRAM reset. The System Profiler shows the same as what I posted last night... e.g., 6 installed sticks, all at 1333MHz. So should I be confident that's accurately reported then?

Thanks!

 

[tsialex]

So I did the NVRAM reset. The System Profiler shows the same as what I posted last night... e.g., 6 installed sticks, all at 1333MHz. So should I be confident that's accurately reported then?

Thanks!

Yes.

To have peace of mind, run AHT extended test. Google “AHT + github”, you will find a nice article about how to download and install AHT.

 

[gadgetfreak98]

Confirmed on speed, thanks. AHT reported all was 1333... oddly, though, it also showed that the 4GB sticks were 1GB and the 8GB sticks were 2GB. But it still got the total of 36 correct

 

[flowrider]

If I were you, I would try to even out the RAM assigned to each processor as much as possible. Rather than the 12GB assigned to Processor A and 24GB assigned to processor B. I would put two 8GB in slot 1 & 2 and a 4GB in slot 3 = 20GB

For slot 5 put the last 8GGB DIMM and the last two 4GB DIMMS in slots 6 & 7 = 16GB.

Lou

 

[tpivette89]

I wonder if mismatching RAM sticks (in terms of GBs) has an effect on performance. I know it does this with the newest Mac mini (one guy installed a single 16GB stick + a 4 GB stick and got reduced performance (in Geekbench). I know Geekbench is usually a laughing stock around here, but measurably changed values should amount for something.

https://forums.macrumors.com/threads/2018-mini-8gb-vs-20gb-benchmark.2156474/

For instance, would there be a performance penalty if one were to install 2 x 16GB sticks + 1 4GB stick vs just 2 x 16GB sticks (in a single CPU system). I will be in a position to test this out in a few days...