| Corsair Vengeance LP 16GB DDR3-1600 |  |
| Reviews - Featured Reviews: Memory | ||||||||
| Written by David Ramsey | ||||||||
| Monday, 12 December 2011 | ||||||||
Page 2 of 6
Closer Look: Vengeance LP DDR3-1600Corsair Vengeance LP CML16GX3M4X1600C8 memory is delivered in a nice box with each DIMM module in an individual closed plastic tray. You can choose red, blue, or black heat spreaders (there's also a special white, low-voltage version). And according to the photo on the box, it can even be used in systems without a CPU!
Corsair does not provide a height specification for the Vengeance LP modules, but if they did, it would probably be about 31mm.
The label shows the XMP profile specifications (1600MHz at 8-8-8-24) rather than the SPD specifications of 1333MHz, 9-9-9-24. But that's OK: you're going to run this memory at its XMP specs, right?
With black heat spreaders and a black label, the Corsair Vengeance LP memory is low-profile in more ways than one. Still, low-profile memory's a good idea for X79 systems, since almost any air cooler is going to overhang one or both of the banks of memory on either side of the CPU socket.
Join me as I explain my test methodology in the next section.
|
||||||||
Comments
Importantly, as the reviewer notes, you do not really see a perceptible difference between high speeds and lower latencies. It is good RAM, and the low profile heatsinks (which is all you need if you are going to run at 1600Mhz, these do not get hot at all, they are warm to the touch) enable them to fit even under large heatsink overhangs. Perfect.
enable C state tech and rebench please, the memory performance suffers greatly on Core i processors when C state tech is disabled.
In benchmarking tests the most important thing is consistency, so that users can judge the relative performance of the products being tested. All the memory in these benchmarks was tested on the same system with the same settings, so even if the C state thing is legitimate (and I will do more research on this), the results are comparable.
these write results are also a good deal less than what i've seen from other benchmarks using the C9 vengeance kit on lesser hardware (the C8 sticks should obviously be a little faster)
see
##pureoverclock.com/review.php?id=1216&page=5
Can you provide an example of a reproducible benchmark I can use to test this? If so I can incorporate it into future memory reviews.
C states enabled: read 18924,, write 14995, copy 16914, latency 48.8ns
C states disabled: read 18897, write 15082, copy 17101, latency 48.9ns
The difference in write performance is a fraction over 1%, well within the margin of error of these tests. I conclude that at least with this motherboard and CPU, whether or not C states are enabled makes no difference in memory performance as measured by AIDA64's tests.
"You can choose red, blue, or black heat spreaders (there's also a special white, low-voltage version)."
In any case I can only review what a vendors sends me. Corsair sent me a black memory kit, so that's what the appearance score was judged on.
Sigh.
#forums.aida64.com/index.php?/topic/292-c-state-tech-disabled-negatively-affects-memory-bandwidth-tests/page__p__1712#entry1712