Sandy Bridge Memory Scaling: Choosing the Best DDR3
by Jared Bell on July 25, 2011 1:55 AM ESTFinal Words
I think we confirmed what we pretty much knew all along: Sandy Bridge's improved memory controller has all but eliminated the need for extreme memory bandwidth, at least for this architecture. It's only when you get down to DDR3-1333 that you see a minor performance penalty. The sweet spot appears to be at DDR3-1600, where you will see a minor performance increase over DDR3-1333 with only a slight increase in cost. The performance increase gained by going up to DDR3-1866 or DDR3-2133 isn't nearly as pronounced.
As a corollary, we've seen that some applications do react differently to higher memory speeds than others. The compression and video encoding tests benefited the most from the increased memory bandwidth while the overall synthetic benchmark and 3D rendering test did not. If your primary concern is gaming, you’ll want to consider investing in more GPU power instead of a faster system memory; likewise, a faster CPU will be far more useful than more memory performance for most applications. Outside of chasing ORB chart placement, memory is one of the components least likely to play a significant role in performance.
We also found that memory bandwidth does scale with CPU clock speed; however, it still doesn't translate into any meaningful real-world performance. The sweet spot still appears to be DDR3-1600. All of the extra performance gained by overclocking almost certainly comes from the CPU overclock itself and not from the extra memory bandwidth.
Finally, although the effects of low latency memory can be seen in our bandwidth tests, they don't show any real world advantage over their higher latency (ahem, cheaper) counterparts. None of the real-world tests performed showed any reason to prefer low latency over raw speed.
Even though there's merely a $34 price difference between the fastest and slowest memory tested today, I still don't believe there's any value in the more expensive memory kits on the Sandy Bridge platform. Once you have enough bandwidth (DDR3-1600 at a small $9-$10 price premium), there's just not enough of a performance increase beyond that to justify the additional cost, even when it's only $34 between 4GB kits. Once you jump to the 8GB kits, the price difference for CL9 DDR3-1600 is a mere $8, but it becomes much more pronounced at $92 to move to DDR3-2133. We simply can’t justify such a price difference based on our testing.
Of course, testing with Sandy Bridge doesn't necessarily say anything about other platforms. It's possible that AMD's Llano and Bulldozer platforms will benefit more from higher bandwidth and/or better latency memory, but we'll save that article for another day. Also, we've shown that performance scaling on integrated graphics solutions can benefit, particularly higher performance IGPs like Llano. Ultimately, it's up to you to choose what's best for your particular situation, and we hope this article will help you make better-informed decisions.
Facebook
Twitter
RSS
76 Comments
View All Comments
Rick83 - Monday, July 25, 2011 - link
Fancy heat spreaders are the worst that has ever happened to RAM.It gets worse when you have to pay more to get rid of it, as with the new low profile vengeance series from corsair.
Memory doesn't usually get that hot anyway, and the large heat spreaders impede airflow between the modules in fully populated setups, as well as limit what size your cooler can be, occasionally forcing you to get one of those water-cooler-in-a-box things which incur massive extra costs.
The only reason I don't want to have completely naked memory, is that the heat spreader gives the RAM some ESD protection, which is actually useful.
JoJoman88 - Monday, July 25, 2011 - link
The review just made your post the truest of them all jabber!Spacecomber - Monday, July 25, 2011 - link
In the past, one reason to get faster rated memory is that you eventually would see a migration of what was the standard memory module to something running on a faster bus speed. I'm not sure if that really holds true, anymore. It seems that these days you are more likely to see the adoption of a completely new type of memory, rather than an existing standard sticking around long enough for the minimum required memory speeds it is based on to go up.geofelt - Monday, July 25, 2011 - link
One of the price differentiators is the heat spreaders.Apart from the aesthetics, where is the value of fancy heat spreaders? Can it be measured?
Seems to me that they are mostly marketing gimmicks, excepting perhaps for those used on overclocking competitions.
I would like to see some sort of a study to determine the value of heat spreaders.
MrSpadge - Wednesday, July 27, 2011 - link
Short answer: nothing.MrS
BobDavid - Monday, July 25, 2011 - link
see subjectJarredWalton - Monday, July 25, 2011 - link
See the conclusion; we already did a look at that (with HD 3000 and Llano).http://www.anandtech.com/show/4476/amd-a83850-revi...
LoneWolf15 - Monday, July 25, 2011 - link
Ivy Bridge will be out next year. There is a reasonable chance it could have a bump in memory bandwidth. Buy RAM at one or two multipliers above what you need now, and when the upgrade comes along, you won't be wishing for new RAM.DDR3 is so cheap right now, it's worth planning ahead.
dman - Monday, July 25, 2011 - link
I've been looking for a review like this for a while, was a good read even if it didn't come as a huge surprise. I'm definitely interested in the AMD platform results if/when those are available.SteveSweetz - Monday, July 25, 2011 - link
I was disappointed to see this article lacked the detail (and quantity) of the gaming tests versus it's predecessor on AnandTech: http://www.anandtech.com/show/2792/10That article showed that memory frequency and latency changes had a greater impact in some games than others, and that in most cases the memory also had a greater impact on the minimum framerate (an important consideration) than average framerate.
Also disappointing to see no CAS6 sticks tested here. Particularly because 2GB 1600MHz CAS6 were relatively common at one point, but now, for whatever reason, G.Skill is the only company that still makes them. It'd be interesting to see if that's a meaningful exclusive. The previous article showed CAS latency being more important than frequency in some cases.