|
Page 4 of 4
ASUS
and their new entry from VIA have come up with a good board, but failed
to surpass the Nforce chipset. This is quite easily explained by the
small onboard cache that the Nforce2 Ultra 400 sports. With the speed
gained with IDE load times, it’s a fine line on whether this board is a
worthy contender in the market or not. Unfortunately, I personally
don’t think it’s quite there yet.
I have had a month with the
board, and have tried every possible BIOS I could find. Not one of them
gave me the lockable AGP bus that this board claimed to have. Having
only the AGP, and not AGP/5th PCI like the nForce, is not a big deal,
as the onboard sound is fairly decent, so there is no need of a PCI
soundcard. I like a lot of bass, as those that have seen my gaming
setup already know, and it was far from disappointing.
I cannot forgive being unable to use the full potential of my memory, however.
The
maximum FSB you can select is 227, and the board is more than stable at
that speed, but because my video card does not like anything out of
spec, anything I tried in 3D, no matter what multiplier I used,
resulted in VPU recoveries, or hard crashes. In the GA-7N400 Pro2, the
same FSB was easily attainable with the same remaining hardware, but
not as rock stable as the A7V880. Looking through the forums, there
seems to be a few people who have working locks, but to me it seems to
be BIOS related. With no updates since early August, I’m beginning to
lose faith.
In the compatibility area, I could not get any
SoundBlaster Audigy to work properly. I actually fried 3 before I
figured the local FutureShop had had enough of my returns. However,
this is the only board I could get my video card to work with both 8x
AGP and fastwrites enabled. Most people are screaming “You don’t need
fastwrites!” Unfortunately for me, this ASUS A9800PRO requires it, to
be stable with an AGP aperture of 64MB, which I find gives me the most
stable overclocking performance. This is partly due to the temperature
sensors embedded in the memory being enabled, but that in itself is a
whole different story. Let’s just say that it nets me another 15 MHz at
the GPU.
I managed to get the board under a waterblock, and
cranked the CPU up to 13x200. At 1.85v (2.0v real), I got a clean boot
and ran Prime95 for 8 hours without any errors. The voltage the board
provided is one of the most stable and accurate to what is in the BIOS,
but the CPU temperature was way off, both in the BIOS, and using ASUS’s
included utility. Using the L-12 mod to force a default 166 MHz FSB
from the mobile chip’s default 133 quickly remedied the problem.
Dropping in a desktop chip instead of a mobile seemed to help; whilst
the temperature change from idle to load was more accurate, it was
still 10C below the real temperature. This is pretty much standard for
ASUS boards though, so is completely expected. 3200+ speeds (11x200)
were easily attained at default voltage, but this can be highly
chip-dependant. As long as no 3d is involved, overclocking was easier
with this board, and more stable, than any other I have tried so far.
I’m still waiting to give Epox a try though.
All
in all the ASUS A7V880 is a good motherboard. At default speeds, it’s
rock stable, and only has a few compatibility issues. Once you begin
overclocking though, expectations are poorly met. Hopefully a BIOS
update will solve the only issue I really have, but the non-locking AGP
bus has been A7V880 user’s complaint from the start. Either they have
not solved the issue, or overclocking is not a priority. Either way, if
you can forgive that lone issue, the higher IDE access times lead to
better overall performance than NVIDIA boards, if you can forgo the
2-3% difference in benchmarks. For me, the faster load time makes more
than up for it. If you have made the transition to SATA, however, and
are into overclocking, this board is not for you. ASUS and VIA have the
tips of their finger on the crown, but have not taken it for their own.
Have a comment? Visit the forum!
|