| Clawhammer |
 |
 |
Joined: Fri Sep 23, 2005 6:22 pm
Posts: 4406
Location: In the closet
|
You're such a dick! How dare you ask!! IMOK
Benchmarking absolutely. Real world everyday application, hardly.
All in, it's all about bragging rights, benchmarking and some benefits in zip programs and also video/photo editing engines. Real world it's negligible. There's less than 2% difference between 2133MHz and 1600MHz RAM with Sandy Bridges, so paying five times more for the faster memory is totally wasted money. I think as you've accurately stated on this forum too, 8GB is probably going to be enough for anyone, but since RAM is dirt cheap, 4x4GB even 2x8GB set of 1600MHz RAM is good purchase and opens up a brave new RamDisk world not to mention nothing screams poser than peering into the side-panel window and seeing all DIMM slots full.
In general a 2% increase in the theoretical Sandra benchmark would translate to no real world performance improvement. Where you might see one is in any application that is strictly memory bandwidth limited - they're hard to find because generally to pass data through memory you have to create it or load it up from somewhere else, which is usually slower than the transferring data through memory. If you had an application that cached all of it's data and then performed relatively simple but numerous operations on it you might notice memory speed - I would guess that cached database manipulations, spreadsheet calculations etc. might fall unto that category.
After playing around a bit just now with the frequencies here is what I get.
Increased RAM clock from 800 to 852:
- 3DMark gave 1.5 % better results.
- 7zip had 4% faster compression.
Increased RAM clock from 800 to 1066 and set timings to 6-6-6 (to save the time searching for lowest working timing):
- 3DMark 1.5 % higher score.
- CPU Score : 1.5 % higher.
- 7zip had 7% faster compression.
Conclusion is that it is (mostly) a waste of time. 7zip compression gains from clock frequency, but is mostly independent of timing. 3DMark gives almost negligible improvement. With lower timings maybe up to 5%. So now back to these particualr sticks...at $38 for 8GB or $76 for 16GB, this memory is among the cheapest sets available anywhere, while also allowing those who are interested to get some of the best performance possible. There is no way you lose buying these DIMMS. Not only is it very good for what you get in speed and capacity for the price, its actually some of the least expensive memory out there period. On top of that its low profile, which can come in very handy for working with larger heat-sinks and/or SFF rigs. I actually had to remove heat spreaders on normal sized ram when working on a shoebox HTPC rig I built for my sister and even then that ram just barely fit, these sticks would have made things so much easier if I had them then. Hell, they almost don't rise above the slot clips.
I think the working sweetspot is 1866-9-9-9-24-1T @ 1.42v. I had them stable at 2133, but with slower timings and more voltage the performance was almost exactly the same. I also had them stable at 1866-8-9-9-24-1T @1.5v, but the performance improvement was so miniscule that it didn't justify the voltage bump in my opinion.
|
|
Login
Register
FAQ
Search
© 2010 Future US, Inc. All rights reserved.