Despite a bad experience with an old ASUS AM2 board, I must say I'm impressed with their
top-end product range. I managed to obtain a Maximus IV Extreme for very little, fitted a
2700K; obtaining 5GHz didn't take long (valid.canardpc.com/show_oc.php?id=2334534).
I like the UEFI layout, and I've had no problems with weird startup issues after a failed oc
run, etc. The feature set is very impressive, similar to the Rampage described here with
respect to onboard items. I can see why enthusiasts buy this type of board despite the
price points. If anything though, perhaps SB is too easy to oc unless one is going to try
exotic cooling (I'm just using a 2nd-hand Venomous-X I won off eBay and two cheap
Coolermaster fans). Perhaps that's why I've seen comments recently from overclockers
who've done their SB setups and have then sought out older P55/X58 items because
they're more difficult to oc, ie. more of a challenge. I suppose then there are those who
oc to go for benchmarks (so SB/SB-E are king atm) and there are those who oc for its
own sake, the technical challenge, for which SB is a breeze. Thus, for the enthusiast
crowd who would normally buy this class of mbd, have Intel made it too easy to oc?
I reached 4.8 with hardly any changes required, while 5.0 only needed a couple of tweaks
recommended by forums posters and ASUS' own guide. A friend of mine obtained 4.8 in
under a minute just by using the Turbo 50 option on his Asrock board (Z68 Extreme4
with a 2700K).
The performance results for the SB-E used here are interesting; it doesn't seem as if that
many tasks can yet scale nicely from 4 to 6 cores, while of course I get better single-threaded
results for the Lame and WinZip tests given my higher clock (www.sgidepot.co.uk/misc/tests-jj.txt)
One thing I would say though, I was delighted that my Maximus was able to use my Mushkin
8GB DRR3/2133 kit at its rated speed & timings without any trouble at all, even at 5GHz. I
didn't have to change anything; definitely wasn't expecting that (from reading numerous oc
guides, I thought I'd have to lower it to 1866). This echoes the article author's comments
about the easier RAM setup on the Rampage. ASUS have done an excellent job support 2133
speeds.
CaedenV makes a good point about the effect of this class of CPU power on gaming
performance, namely not a lot for these newer games used as tests. Ironically, the most
striking effects occur with much older games, inparticular X3TC and Stalker which see
major speedups when run on such a system, though of course at this level the frame
rates are crazy anyway. Interestingly, even 3DMark11 starts to show the artificial boost
to overall scores from having powerful CPUs, even though the game fps results may be
less interesting. The effect is totally OTT with 3DMark06 of course, but I wasn't expecting
to see it so much with 3DMark11.
I've tested the same two GTX 460s (EVGA 1GB FTW, 850 core) on a wide range of systems;
the Maximus with a 5GHz 2700K gives 8130, vs. 7768 for the same cards on a cheap P55 with
an i7 870 @ 4.27, but the fps results are almost identical (infact the P55 gave a higher score
for the Combined test, and that's with older drivers, etc.) See:
http://3dmark.com/compare/3dm11/2220257/3dm11/3215406
Hence, as Marcus52 says, perhaps a broader range of more CPU-intensive game tests
would be more appropriate for reviews like this? Problem is though, some games hit
the CPU hard mainly for multiplayer scenarios and that's tricky to replicate for a test.
kitsunestarwind, what do you get for your CB 11.529 CPU score on your 3930K at 5.0?
Ian.
Facebook
Twitter
Instagram