Dell Precision M6700 Owner's Review

From when I sold them on eBay...


I can also dig up my order spec. Regardless of what that PDF may say, they shipped 1333 for quite a while. Several months, at a minimum. I wasn't really active in the owners' community once Throttlegate was resolved.

 

That sucks, and I wasn't doubting that they sold it to you in error. Clearly the guy who set up the configurator didn't read their own spec sheet very well.

But the press announcement pdf here and the "om_en.pdf" manual specs clearly do say 1866MHz support up to 16GB.

The previous Dell laptop I bought was a little over a decade ago, so I have no particular knowledge of the credibility of Dell's spec sheets, but I have no particular reason to doubt them unless you have more info?

 

Intel's processor specs page says they support at most 1600 for every processor that Dell sells on the M6700. I linked it before, but: ARK | Intel® Core

 

Does this render the modular capacity of the system useless?

 

Considering that this is my review thread, and I am running 8gb of 1866 ram, maybe some time should be spent showing you that 1866 ram will work in the M6700 and have a benefit to the processor. I have some 1333 and 1600 around to compare it to. This should be fun.

 

That's the Kingston HyperX, right?

 

That would be interesting. Would also be interesting to see the difference between 8GB of 1866, 16GB and 32 GB.

 

Well I hate to be the naysayer in the group, but I've already seen this test done before. The results I viewed showed a measurable advantage toward 1866 but negligible in real world performance.

None of the slightly higher numbers in the test resulted in any significant real word advantage. Only the double jump from 1333 to 1866 made any noticeable difference and even that wasn't significant. The conclusion in that test was that higher speed RAM had no appreciable effect on overall performance.

I'm going to go out on a limb and guess big spread in test numbers, little or no advantage RW performance. RAM is just no longer an issue anymore. And even then the graph fall off significantly beyond 8GBs. Please feel free to enlighten me if you disagree.

In a related comparison, last week I watched the battle between eSATA and USB 3.0. The Benchmark showed a 15 pint spread advantage 3.0. That resulted in a yielded of 2~ second during a 15GB file copy time. Not exactly something to write home about is it? But it does serve to underscore the significance in these test and how they should be viewed:

What good is a 500hp car if the roadway is restricted to 65 mph? Yes, I know 0-60 (access time) is faster. I grant that. Anyway, if you wanted to argue, the USB 3.0 was faster.

 

I've seen the same tests that you've done I think. Depends what precisely you do.

Sure, the RAM frequency is up 16.7%, but the timings for 1600 can be had at CL9 and Kingston HyperX is CL11. That means that the first word timings are 13.75ns for both, but 8th word is 18.125ns with 1866 and 19ns for 1600 (Ivy bridge pulls 8 words at once). That means that in a purely RAM limited scenario, at worst case of random RAM accessing you should get a 4.6% speed boost.

You're quite right to suggest that for the vast majority of people, a) applications aren't ram limited and b) that part which is ram limited does not run for very long. Most synthetic benchmarks make some complete assumption of a scenario that has little bearing on real life. For an average-ish user of this sort of laptop's demands, maybe a 2-3% speed boost or so? Whether that's worth the £ is entirely subjective. However, if what you do is very heavily CPU & RAM based (e.g. statistical modelling), you'll see slightly higher gains. Again, whether that's worth the £ is entirely subjective.

 

I would indeed be very interested to see that. I'm going purely by Intel's spec sheet, so if the CPU somehow supports a higher speed memory than is listed, that information would be very useful.