Worth an extra $30-40 for DDR3 2400 vs 1600?

[Jeff7181]

Is it worth spending an extra $30-40 on a 16 GB kit of DDR3 2400 vs. DDR3 1600? Specifically I'm looking at these:

http://www.newegg.com/Product/Produc...82E16820211775

http://www.newegg.com/Product/Produc...82E16820211814

 

[nickbits]

IMHO no

 

[Jeff7181]

Right after asking, I came across this: http://www.anandtech.com/show/7364/memory-scaling-on-haswell

Looks to me like the difference is almost nothing... 1-3% at most maybe. And it appears the 2400 kits run at 1.65V while the 1600 kits run at 1.35V. Think I'll go with the DDR3 1600.

 

[zCypher]

Definitely not.

 

[bryanl]

Memory dealer Corsair said the best improvement from 1600 to 2400 was 10.5% for a 3.5 GHz Haswell, 11.9% for one overclocked to 4.5 GHz:

http://www.corsair.com/~/media/Corsair/Blog/2014-02/HaswellRealWorld/HRW-StockPercent.jpg

http://www.corsair.com/~/media/Corsair/Blog/2014-02/HaswellRealWorld/HRW-OCPercent.jpg

I would not pay a premium for any memory rated 1.65V, especially 2400 MHz made from 1600 chips, as that Adata is.

 

[NewYorksFinest]

IMO, no.

That said, DDR3-1866 RAM by G.Skill is the same, or ~$5 cheaper than DDR3-1600.

 

[BrightCandle]

It really depends. In some software it makes much more difference than others. Winrar for example shows quite a lot more difference than most software, up to about 10%. Arma 3 is one game that has shown similar gains (10%) with higher speed memory, but almost no other games show any useful gain at all. Not surprisingly it only helps if memory latency and bandwidth is the issue with the software you are using, and that is not true of most software.

 

[Blain]

I would buy This kit or This kit, based on my quest for quality.

 

[Jeff7181]

I would buy This kit or This kit, based on my quest for quality.

Not many reviews on those... I picked the one I picked because it had a decent amount of positive reviews. Even though they're 1.5V and not 1.35V I think they'll be fine for me. I have 32 GB of ADATA in my home server, so I'm comfortable that it's quality stuff, although the Mushkin is a pretty good price and in my experience they've been pretty reliable.

 

[BonzaiDuck]

OP didn't offer details about his hardware. The consensus seems to be that you can't squeeze much more out of performance RAM above DDR3-1866 with a Z68 chipset or my SB-K 2600K processor. Others here provided some stats about performance gains under Z87 and Haswell.

Whatever the rated speed, I wouldn't run the kit at the rated speed if it required goosing the DIMM voltage to 1.60V or higher. I have a nice set of DDR3-1866 2x8GB which runs at 1.5V. The only reason I'd buy faster-rated RAM @ 1.6+V (ignoring price for the moment) would be to downclock from the default [meaning rated] speed and run at 1.5V. And maybe I'd tighten the latency settings as long as the RAM would run at 1.5V.

But why spend the money -- just to do that, when you could simply buy RAM at a chosen target speed requiring only 1.5V?