All results were obtained at 4600MHz s on a normal install of Windows 7 64-bit. These are our results, yours may vary
so only use these as a guideline for a similarly configured system.
Memory
AIDA 64
Read
AIDA 64
Write
AIDA 64
Copy
AIDA 64
Latency
SuperPi 8M
Frequency
Timings
CORSAIR DOMINATOR
PLATINUM 2666C10
25303
24938
28805
28,7
1.24.110
2400
8-10-10-21-1N
26416
25149
29918
28,2
1.23.937
2600
9-11-11-28-1N
26818
25244
30188
27,9
1.23.828
2666
9-12-12-28-1N
27032
25347
30823
27,8
1.23.688
2800
10-12-12-31-1N
27438
25510
30386
27,9
N/A
2933
11-13-13-35-1N
25149
37724
29631
40,6
1.26.237
2400
9-11-10-28-1N
27271
40831
32148
40,3
1.26.268
2600
10-12-11-31-1N
27929
41522
32851
40,2
1.25.691
2666
10-12-12-31-1N
28914
35251
31245
38,8
1.25.769
2800
11-13-13-35-1N
30161
30902
29903
38,9
1.26.003
2933
11-14-13-31-2N
30735
31557
30232
40,3
1.25.706
3000
12-14-14-35-2N
34993
37573
34790
43,3
1.26.564
2400
10-11-11-31-1T
39050
40435
37542
45,4
1.26.814
2600
12-13-13-35-2T
30820
24141
26487
44
1.27.361
2666
12-13-13-35-2T
33998
43638
38005
44,7
1.29.513
2800
12-14-14-35-2T
CORSAIR VENGEANCE
PRO 2666 C11 2x8GB
CORSAIR VENGEANCE
PRO 2666 C12 4x8GB
you intend to do with it?
We would have loved to tell
you that the performance of
this set closely matched that
of our previous Vengeance
Pro memory, but that would
be untrue. What we ended
up with is memory which will
need you to tune it before you
can get decent performance
from it. Simply loading the XMP
profile will give you the desired
2,666MHz but the numbers
will be quite poor. With such
memory densities you should
expect some slightly lower
performance (Command rate
drops to 2N for example) but in
this particular case and as you
can see in the benchmarks, the
memory was underwhelming.
Tried as we may have, we
just couldn’t get it to perform
as expected and in fact the
standard 2,666MHz test scores
are the lowest.
We then took to decreasing
the speed and tightening the
timings and it is then we started
recording numbers we were
expecting. At 2400MHz, the only
frequency we were able to set
1N , the memory performed
admirably and as you can tell
efficiency was quite high. If you
are noting anything odd about
the results compared to the last
Vengeance Pro kit, it is because
we used an updated version of
AIDA 64 for this review which
yields higher results than the
previous one.
When we moved on to
2,600MHz, we had to give up on
the 1N command rate and use
the DRAM’s specified timings.
However this was the memory’s
sweet spot where we recorded
the best results with our setup.
We tried this on both the Z87
and X79 platforms and found
that the 2,600MHz setting was
indeed the right setting for
this memory. Anything higher
required too great a sacrifice
in the secondary and tertiary
timings which led to dramatic
drop offs in performance.
In reality you’re looking at a
2,600MHz set here. Sure it says
2,666MHz on the box and in the
SPD, but you’re going to want to
set this up at 2,600MHz to get
maximum performance. It’s
not a set designed for memory
frequency records, thus you
should not expect it to do such.
Should you be in need of lots of
memory though and are willing
to spend some time tweaking
this set, there are worse ways
to spend $600. As far as we are
concerned, if you can’t find the
C11 kits (which is most likely the
case as they were sold out when
we checked) then this kit may
be a valid alternative provided
you need the capacity more
than you do sheer frequency or
performance.
[ The Overclocker ]
Summary
The Corsair C12 32GB
C12 2,666 Kit is perhaps
better suited for memory
intensive applications
than it is for overclocking.
By current memory
pricing standards, it is
competitively priced
and gets the job done.
For those who need
performance above
everything else, do
consider some other
Corsair kits before
settling on this one.
Would you buy it?
Only if performance needs
are secondary to capacity.
Issue 28 | 2014 The OverClocker 27