64gb 1600 vs 1333

N

note235

Supreme [H]ardness
Joined
May 23, 2005
Messages
5,055
Im building an editing rig and am planning on getting a full set of 64gb but was wondering if there would be any difference between 1333 and 1600?
its about $60-70 more which i can spend towards another hdd
 
If you are using applications that will utilize that much memory, then the volume of available memory will outweigh the differences in speed. Besides, there no real-world noticeable difference between 1333, 1600, or even higher speed memory with a SB, IB, or SB-E processor since the IMC is amazingly efficient.
 
Last edited:
You aren't going to have good luck getting the CPU to run @ 1600mhz with 64GB. You'd be better off with 1333mhz and tighter timings.
 
I just want to throw my 2 cents in as well and say that I agree with both Monkey God and DejaWiz, when you are talking about that much memory there won't be much of a difference.
 
GreenMachine said:
I just want to throw my 2 cents in as well and say that I agree with both Monkey God and DejaWiz, when you are talking about that much memory there won't be much of a difference.
Click to expand...

To clarify...there won't be any noticeable difference between memory speeds with this particular platform. Regardless of RAM quantity.
 
I got the g.skill 1600mhz kit and got it to run on a sabertooth and 3930k. Right now im testing the stability at 4.4ghz, and seems to be working relatively good so far. Temps may be higher and are uneven between the cores but thats because i had problems setting the h100 and probably messed up the thermal paste.
 
Lucky you. Sounds like a sweet rig.

I have a different opinion. I'd get the 64GB kit @ 1600MHz. As a non-electrical engineer my impression of faster memory is that it's simply better binned than the lower speeds. This means that @ 1333MHz, the 1600MHz-capable RAM should run at tighter timings.

Enjoy and have a good weekend.
 
got g skill ripjaw x when they were on sale
they're here but my ssd is coming tomorrow
and the rest of the memory set should be here next week
but ill post my oc on my gigabyte ud5 x79
 
Can anyone even point to a real life application benchmark that shows meaningful difference between 1333 and 1600 RAM?
 
I've seen benchmarks of premiere where it's off by up to 5-10% (but average of 5ish I think ) but but cost wise it's like 20-25%
 
got my 64gb
my rig is at 4.2 idling around 31 C for all cores

tried 4.4 and idling at 34 C
 
Last edited:
What about a ramdrive situation, higher bandwidth/lower latency has got to help if you're running a high IOPS dataset, right?
 
Machupo said:
What about a ramdrive situation, higher bandwidth/lower latency has got to help if you're running a high IOPS dataset, right?
Click to expand...

You would not notice the difference. Not at all. Remember you will not be loading GBs at a time.
 
drescherjm said:
You would not notice the difference. Not at all. Remember you will not be loading GBs at a time.
Click to expand...

I understand that for total bandwidth (you would only see that on loading / unloading the environment/set), but what about the decreased latency? I guess my question is: how well does latency scale (inversely) with increasing memory clockrate?

edited to add: talking about clockrate latency, not adjusting the timings (assume they are constant)
 
Last edited:
You will not tell the difference between a few nano seconds of latency. Remember that memory latency is in nano seconds.
 
I'd go simply for stability with that kind of data need.

Forget about timing tweaking etc. Nothing more time wasting than a big project that goes tits up half way through due to unstable ram.

1600 running at 1333 with nice loose timings.
 
I got 2 kits of this: http://www.amazon.com/gp/product/B006JO2KXQ/ref=noref?ie=UTF8&psc=1&s=pc

It took me 3 kits to get 8 working modules @ 1600Mhz 9-9-924, but I saved like $200+ vs a "name brand"

Its a no name brand but excellent performance / price. Make sure the kit is fulfilled / sold by Amazon, if you get a bad module its much faster to deal with them vs the company.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
as has been pointed out, the difference in speed would be negligible, so save yourself $60
 
You must log in or register to reply here.
Back
Top