Puget Systems print logo
Read this article at https://www.pugetsystems.com/guides/142

Intel Performance Results


To start things off, we used MaxMem to give us our write, read, copy and latency results. Since read speeds and latency are based primarily on the frequency and not greatly affected by the timings, we see a gradual increase in performance as the frequencies become higher. One unexpected result is that the frequency does not affect the write speeds by a significant amount.

These results are great to show that higher frequency RAM is better at sustained read and copy tasks, but in order to answer if it is enough to overcome the increased timings we have to run some real world benchmarks.


Our gaming benchmarks could not be any more clear: on Intel-based systems, you gain little to no performance advantage with the higher frequency RAM. The only benchmark that showed any appreciable variance was DiRt 2 with very low settings. Even then, the variance was very minor with only a 1.5% variance between 1333MHz and 2133MHz.

X3: Terran Conflict DiRt 2 Unigine Heaven





On our application benchmarks, we finally saw some performance variances, although the results are a bit strange. While the 2133MHz RAM did perform the best, the 1600MHz and 1866MHz RAM were both worse than the RAM running at 1333MHz. Due to this, it appears that we cannot make a blanket statement for WinRar when it comes to the performance of higher frequency RAM. 

Euler3d also gave us some mixed results that are almost exactly the opposite of our WinRar benchmark. Again, the 2133MHz RAM is faster than 1333MHz RAM, but 1600MHz and 1866MHz RAM are both faster than the 2133MHz RAM. So once again, while we see some performance differences, they do not line up exactly with the frequency speeds.

All of our other application benchmarks (shown below) showed only minor variations not significant enough for us to specifically comment on.

Cinebench CPU  Cinebench GPU x264 HD Benchmark
Handbrake 0.9.5 Lame 3.98.2 Windows Media Encoder
TrueCrypt - AES TrueCrypt - Various  


< Previous Next >

Switching in to analytical mode, are your part failure reports based on frequencies in general or separated by brand, or perhaps IC manufacturer? Does it include memory overclocked by the end user?

As for the scores, I can validate your results 100% (not that you need it.) :) Games rarely have ever showed a benefit when overclocking the RAM frequency compared to overclocking the CPU. However, lower timings do slightly benefit the system's responsiveness like opening and closing big programs.

I often recommend that users buy good names with the lowest memory timings for best overall performance. Sometimes, buying high frequency CAS9 memory, down clocking it and use lower timings works out really well. However, if the user requires the performance of very large programs, it's best to stick with low latency, lower frequency, higher capacity memory for maximum stability.

Posted on 2012-04-03 01:59:52

The failure rates are for the last 12 months between all of the desktop (not saptop or server ECC) RAM we've sold. The majority of the failures occur in-house, although roughly 10% or so do happen in the field. As far as we know, none of the failures were the result of overclocking.

The brands are almost entirely Kingston and Patriot. We've tried other brands, but the failure rates have always been higher so we've largely settled on those two.

Posted on 2012-04-03 22:01:32

We keep track of reliability data by SKU.  While you can trace that down to a certain IC manufacturer, it doesn't guarantee that the ODM isn't jumping from one IC chip to another.  I can say though that we haven't had a situation in which we felt compelled to track by IC manufacturer.  Kingston, by far, puts out the most stable memory we've seen, and it is no coincidence that they are also the most stable in their IC sourcing, and also the most conservative in their frequency binning.  This is all just overclocked 1333, right?  I've haven't looked for a while, but that was the case last time I checked.  In that case, the reliability of the parts is going to be highly dependent on how aggressively the ODM decides to bin their IC chips.  That's why OCZ has a 8% failure rate, and Kingston a 0.5% failure rate! :)

In the case of the numbers in this article, the reliability data was just an aggregate of all brands we've offered at those speeds.  It pretty clearly shows a trend that you would expect as the memory guys "overclock" these IC chips more and more.

Posted on 2012-04-04 04:44:50
Eric Garay

I was here on the 6th and forgot to say Thanks for the replies and clarification. Agreed on the IC jumping which was very common a few years ago. We would see a very good series launch and then a quiet IC jump to something not as costly which also typically means a drop in quality. Validating your own batches and keeping in contact with the vendor is the way to go. It's always smart to go with the ODM that isn't afraid to keep you aware of any changes and that do their own proven validation. You're doing is the most common sensible way. 

Posted on 2012-04-17 00:05:59

Thanks, your article is very useful and informative. You may like to visit Om Nanotech in case you want to have more information on DDR1 supplier

Posted on 2014-09-11 11:29:53
Cankut Bayhan

yeah whats the best is to wait for the rush to fade for a while and go for the non oc required and yet found its place on the main stream of the product range frequencies which is sorta 2000's on ddr4 rams these days...there is no avoiding from obsolescence the brands would like to benefit so it takes a good observation to find the middle man and go along with it.

Posted on 2015-03-17 18:51:40