Did I make a mistake? DDR3-1333

[VirtualLarry]

I stocked up on Gskill 8GB (2x4GB) kits of DDR3-1333 CAS9 1.5v RAM when Newegg had them on sale a while back for $60. (Since, it's been as low as $50!)

Did I make a mistake? Did I fall for Newegg's advertising.

Do most motherboards and CPUs prefer DDR3-1600? That seems to be what I see most people buying, when they list their specs and build lists.

I thought that 1333 was spec, and that 1600 was beyond spec, although many platforms can be "tweaked" to utilize 1600.

Does it make much of a difference?

 

[hans007]

well amd 1600 and 1866 are spec.

your 1333 is 1.5 v and most of the cheap 1600 is 1.65v... which is basically the esame ram. so you can always set your 1333 to 1.65 and 1600

 

[abekl]

On the Intel side 1333 is spec. Many people on the Intel side buy 1600 simply because it costs about the same, is a little faster, and most motherboards support it.

 

[Shmee]

should make very, very, little difference. What system will this be going into?

 

[Absolution75]

About the only thing you'll notice are the numbers on the package the ram came in.


I can't remember the last time RAM actually made a difference. I wonder at what speeds you actually would start to notice (as in, what is the lowest speed you can run ram at acceptable speeds - probably in the < DDR2 666Mhz range.

 

[Patrick Wolf]

I say yes, you made a mistake. 1600 is optimal if you do anything at least somewhat demanding with your PC. 1333 is fine if you want to save a few bucks or overall speed isn't the main concern.

 

[Shmee]

I say yes, you made a mistake. 1600 is optimal if you do anything at least somewhat demanding with your PC. 1333 is fine if you want to save a few bucks or overall speed isn't the main concern.

I sure hope this was a joke D:

Memory timings and frequency has very little to do with overall speed of a system. As long as you have at least 4 GB of average speed ddr2 or ddr3, you will be fine in most cases.

I don't want to make assumptions here, but both of you have been around for a while, I would think you both would know that a few hundred MHz on ddr3 doesn't make much difference in overall system performance.

Anyways larry, you bought it, looks like a good deal, enjoy and don't worry about it

 

[hans007]

i would say the only reason it would matter ever , is if your pc has an IGP . like on the llano it will matter.

if you are overclocking by bus speed it might matter , though you cant really do that on sandy bridge and up, so that also doesnt matter.

it likely wont matter at all even on a bulldozer am3+ system. so in reality it probably only reall matters on a llano system.

 

[BTRY B 529th FA BN]

Naw, its fine.

Me on the other hand, I would of bought low timings @ 1.5v 1333 or 1600

 

[MarkLuvsCS]

http://www.bit-tech.net/hardware/memory/2011/01/11/the-best-memory-for-sandy-bridge/9

One caveat is the website doesn't list raw numbers we'd like to see like FPS, MB/s, or a more tangible basis. It does show the difference of performance across the boundaries lie within a few &#37;.

Memory speed isn't as important as some may think. In fact I would say memory speed really doesn't matter unless it's revolving around using an integrated GPU for gaming like llano.

 

[Patrick Wolf]

I sure hope this was a joke D:

Memory timings and frequency has very little to do with overall speed of a system. As long as you have at least 4 GB of average speed ddr2 or ddr3, you will be fine in most cases.

I don't want to make assumptions here, but both of you have been around for a while, I would think you both would know that a few hundred MHz on ddr3 doesn't make much difference in overall system performance.

Anyways larry, you bought it, looks like a good deal, enjoy and don't worry about it

1600 is a marginal improvement for a few tasks, but still the optimal choice considering the little to no cost increase from 1333. Refer to ^bit-tech article.

To answer the OP's question does it make a difference. No, not really. But newegg (and other sites) has had and still has many great deals on 1600 stuff, so it's hard not to recommend it for a SB build.

 

[VirtualLarry]

Wasn't bit-tech the outfit that tested Crossfire performance, with x16/x16 versus x8/x8, and determined that there was like a 20% performance difference, which was later debunked by nearly every other site that tested that?

It's what caused me to waste $130 per motherboard on a pair of X48 mobos. I could have saved a bundle by just going with EP45-UD3P mobos instead.

I blame Bit-Tech for exaggerating performance differences, and costing me a bundle of money. I don't put much stock in their articles these days because of that.

 

[Patrick Wolf]

Wasn't bit-tech the outfit that tested Crossfire performance, with x16/x16 versus x8/x8, and determined that there was like a 20% performance difference, which was later debunked by nearly every other site that tested that?

Don't know, but you shouldn't blacklist a site for one bad article. Their memory findings aren't exclusive.

There's also this article for 1156. Basically the same conclusion.
"Go for a faster product if you find DDR3-1600 memory that provides the same timings as your preferred DDR3-1333 RAM kit at only a little price premium."
http://www.tomshardware.com/reviews/core-i7-870-1156,2482-9.html

Better yet.
"Therefore, we believe that inexpensive DDR3-1600 SDRAM with not very aggressive timings would be the most reasonable choice for contemporary LGA1155 systems: in our opinion, memory like that offers the best price-to-performance ratio today."
http://www.xbitlabs.com/articles/memory/display/sandy-bridge-ddr3_8.html#sect0

 

[StrangerGuy]

Wasn't bit-tech the outfit that tested Crossfire performance, with x16/x16 versus x8/x8, and determined that there was like a 20&#37; performance difference, which was later debunked by nearly every other site that tested that?

It's what caused me to waste $130 per motherboard on a pair of X48 mobos. I could have saved a bundle by just going with EP45-UD3P mobos instead.

I blame Bit-Tech for exaggerating performance differences, and costing me a bundle of money. I don't put much stock in their articles these days because of that.

Ever noticed big name reviewers these days are only reviewing top-end boards, memory, PSUs? But the same thing doesn't happen for CPUs and GPUs because performance differences between those two are real, but mobos/ram/PSUs aren't and they are not going to piss off vendors when they show a $50 board performs the same as a $300 board.

For example, the Antec Neo Ecos are fantastic for their price YET I couldn't find a big-name review on them when I did my research on google.

The whole scene has become a mockery of itself; from becoming finding value/$ years back to how much can you overspend on stuff that doesn't really matter.

 

[XLNC]

Ever noticed big name reviewers these days are only reviewing top-end boards, memory, PSUs? But the same thing doesn't happen for CPUs and GPUs because performance differences between those two are real, but mobos/ram/PSUs aren't and they are not going to piss off vendors when they show a $50 board performs the same as a $300 board.

For example, the Antec Neo Ecos are fantastic for their price YET I couldn't find a big-name review on them when I did my research on google.

The whole scene has become a mockery of itself; from becoming finding value/$ years back to how much can you overspend on stuff that doesn't really matter.

I think there's a problem with the review process in general. Sites like Anandtech wait for the manufacturer to send products, and they usually come with "review guidelines" on how a product should be reviewed. If the manufacturer doesn't send out products for review, you won't see it. Case in point, OCZ only sent out 128GB Vertex 3s for reviews, not the smaller ones which more people bought. Anyways, that's a discussion for another thread.

To the OP, no you did not make a mistake. DDR3 will be the standard for another few years. If nothing else, you maybe able to sell it for a profit at the end of its life.