cancel
Showing results for 
Search instead for 
Did you mean: 

What's best 2 x 16 GB or 4 x8GB sticks of 3600 RAM

XYchromosone
Level 10
As per the title, can anyone tell me which is best out of the 2 kits of RAM specifically out of these two:
https://www.mwave.com.au/product/gskill-trident-z-rgb-32gb-4x-8gb-ddr4-cl17-3600mhz-memory-ab94984
or
https://www.mwave.com.au/product/gskill-trident-z-rgb-32gb-2x-16gb-ddr4-cl17-3600mhz-memory-ab95245

They both have the same ratings/timings. but I am not clear if there is any advantage between the 2. The only difference is one kit comes a a pair and the other is a kit of 4 sticks of RAM. TIA
13,189 Views
19 REPLIES 19

Arne_Saknussemm
Level 40
Depends....

Look up interleaving..

Also T topology boards will run 4 sticks better often...

Arne Saknussemm wrote:
Depends....

Look up interleaving..

Also T topology boards will run 4 sticks better often...


thanks. I wonder though if the difference is noticeable in real time or everyday use?

cniedzi
Level 8
I would even say in real life the difference between 3200 and 4000 is practically unnoticeable 🙂

cniedzi wrote:
I would even say in real life the difference between 3200 and 4000 is practically unnoticeable 🙂


Of course!
but that's not why people do things lol.

you get a kick out of firing up Aida64, do a **** town of memory benchmarks, comparing between:
- stock Jedec
- 2666, 3000, 3200, 3600, 4000!

then look at graph/ numbers, then you start wondering to yourself, is there a way to squeeze more out of it? hmm lemme tighten the timings again.

THEN OH SHI- IT CRASHED, FROZE, BLEU SCREEN OMGWTFBBQ MOOOOOMM.
then you relax the timings, then run memtest all day, maybe 10 hours when you finally believe you got the timings subtimings down.


our definition of fun is silly looooooooooooooooooooooooooooool
no siggy, saw stuff that made me sad.

haihane wrote:
Of course!
but that's not why people do things lol.

you get a kick out of firing up Aida64, do a **** town of memory benchmarks, comparing between:
- stock Jedec
- 2666, 3000, 3200, 3600, 4000!

then look at graph/ numbers, then you start wondering to yourself, is there a way to squeeze more out of it? hmm lemme tighten the timings again.

THEN OH SHI- IT CRASHED, FROZE, BLEU SCREEN OMGWTFBBQ MOOOOOMM.
then you relax the timings, then run memtest all day, maybe 10 hours when you finally believe you got the timings subtimings down.


our definition of fun is silly looooooooooooooooooooooooooooool


That's true 🙂 🙂 🙂

cniedzi wrote:
I would even say in real life the difference between 3200 and 4000 is practically unnoticeable 🙂


this is not my question, i asked what is better [faster and more stable] 2 x 32GB sticks of RAM or 4 x 8GB sticks rated at the same specs, in this case 3600MHz

XYchromosone wrote:
this is not my question, i asked what is better [faster and more stable] 2 x 32GB sticks of RAM or 4 x 8GB sticks rated at the same specs, in this case 3600MHz


I've answered your question:
XYchromosone wrote:
thanks. I wonder though if the difference is noticeable in real time or everyday use?


And this is an answer to your 1st question: in theory T topology makes it is easier to obtain tighter timmings & higher speed with 4 sticks kit.

cniedzi wrote:
I've answered your question:

And this is an answer to your 1st question: in theory T topology makes it is easier to obtain tighter timmings & higher speed with 4 sticks kit.


Thank you for adding the extra detail to your first reply, I very much appreciate it.

Mappi75
Level 8
lol running G.Skill 32GB 4.133Mhz 17-17-37 Kit the first time on the Asus Maximus XI Extreme (thanks to the 9900KS)!

Now running (and reaching) 24 hours HCI MemTest Pro and no errors (reaching 2000% coverage).

I would not say faster memory wont help (maybe the sweet spot is a 3200mhz) but the key are the CAS Latency and its subtimings.

BTW: 16gb modules usual dont have very good timings like 4x8gb kits.


look youtube videos and you will see a difference in fps in games (more than you will think).