GA-5ax and half cpu speed
kotor
M.B. Gigabyte 5ax p.c.b. 3.0 (last bios F3) chipset Ali
c.p.u amd k6III 450
suond blaster Live! Value (liveware 3.0)
matrox mga g200 agp (last driver 5.33)
last driver for agp and bus master chpset
S.O. win2k (linux, win98)
I have two installations of win2k in the two differents partitions of my hd,
one in the acpi mode, other in the non acpi (i have forced non acpi
mode -standard computer -in the 2nd installation of win2k) with apm module.
The apm activation in the second installation of win2k cause blue screen
after shutdown or suspend mode, and is not possible (not disponible) the
standby mode.
The acpi installation have normal boot and normal powersaving mode, but is
very half cpu performance: the system information and directx diagnostic
tools display half cpu frequency (266mhz) and is very bad (confusioned music
play...) the result of directmusic test in the directx diagnostic tool
(normal test result in the non acpi installation).
Is not possible, for example, in the acpi installation, to work with
microsoft windows media encode for videocapture with my webcam, and the
error is "...computer not sufficientely performant.." (very confortable and
performant work in the non acpi installation ...).
In the acpi installation the software Sisoft Sandra 2001 display in the
mainbord information "cpu speed 266 mhz", "front side bus speed = 50 mhz",
(is normally 100 mhz...), "memory bus speed 50 mhz" , "pci bus speed 17
mhz..." and the cpu benchmark is half respect to the normal (correct bus
speeds and benchmarks is displayed in the same software in the non
acpi installation...). Others diagnostics softwares display same half bus
speed.
The half performance of cpu in the acpi installation is visible with the
velocity of windows open and other work performance.
Thankful for suggestions....
gbt-tech
with Ali without getting any response so far, and frankly I'm not expecting
one now.
Mike Mullen
CST
Gigabyte UK
"kotor" <ko...@kotor.com> wrote in message
news:9au87a$qhl$1...@nreadA.inwind.it...
Shazam
based board (such as the Asus P5A) with this weird problem.
gbt-tech <gbt...@btclick.com> wrote in message
news:0uzA6.5680$Rj2.19478@NewsReader...
Colin Bocking
Mike Mullen...Your reply is less than helpful!
You work for Gigabyte tech support and you should have something a little
more constructive to say about this problem.
W2k is an important OS worldwide, yet this problem seems to preclude its use
with the GA5 AX?
Surely Gigabyte owe us a proper technical explanation?
Your further comments will be appreciated
Colin
"Shazam" <gm...@toad.net> wrote in message
news:T5NA6.1649$%r6.1...@news.abs.net...
Jeff Cochrane
The P5A uses a different revision of the ALi chipset I believe.
That may have something to do with it.
In reality though, the point is probably moot, For two reasons, Firstly,
Win2k was not really designed for operation on Socket 7 processors (Even
though it will run on them albeit rather slowly compared to a PII or III) as
Socket 7 was dead long before Win2k was released (According to Intel at
least and therefore by definition, Microsoft as part of the 'Wintel'
alliance) and secondly, The 5AX has been discontinued and has not been
available for some time now. (At least, that is the case here in Australia)
Jeff C.
PS,
Upon reflection, maybe the problem is not only with the chipset as GB say,
but also maybe there is a problem with Win2k and the clock generator chip
used in the GB boards?
That would explain the half speed problems with the 5AX. AA's but not with
the P5A's wouldn't it??
Your thoughts?
Jeff C.
"Colin Bocking" <millenn...@crouchvale.freeserve.co.uk> wrote in message
news:9b2n50$ums$1...@news5.svr.pol.co.uk...
gbt-tech
a long time. We did identify the problem and spent a lot of time
investigating it looking for a solution, but in the end we had to turn to
Ali looking for one, and since they don't make the chipset anymore, well its
not a high priority with them. I'm sorry for the problems with Win2k but it
is an operating system designed around certain expectations of the system it
is installed on, and if your system doesn't meet those specs you will have
problems.
Mike Mullen
CST
Gigabyte UK
"Jeff Cochrane" <jeff_c...@no.damn.spam.jeffscomputers.com> wrote in
message news:3ad4ea20$0$25505$7f31...@news01.syd.optusnet.com.au...
Colin Bocking
I am not a W2K user myself but many people DO use this OS and your
explanation is appreciated. I just wonder whether Microsoft make any mention
of this incompatibility in their advertising blurb for W2K.....but I guess
we all know the answer to that one!
ps GA5AX still available here in the UK!
Colin
"gbt-tech" <gbt...@btclick.com> wrote in message
news:2aeB6.970$211.2111@NewsReader...
Trinity
AMD K6-3/450 as I am. For office applications (word processing, databases,
etc) it feels as fast as my Celeron 800.
The only time the Super7 lags is with games and that is because the K6-2/3
family has a relatively weak floating unit. And since Win2K is not a gaming
OS, I don't buy the argument that Win2K is not designed for operation on
Socket7 processors.
The ASUS P5A has used the following Ali Alladin V revisions: E, G
The Gigabyte 5AX has used the following Ali Alladin V revisions: E, and H.
They both used the E revision. I honestly cannot believe that it is a
problem with the chipset.
"Jeff Cochrane" <jeff_c...@no.damn.spam.jeffscomputers.com> wrote in
message news:3ad4ea20$0$25505$7f31...@news01.syd.optusnet.com.au...
Colin Bocking
On reflection, it seems unlikely that Microsoft would launch an OS (Win 2k)
which is inadequate on such an important architecture as SS7, outdated
though it now is.
Any further comment, Mike or Jeff??
Colin
"Trinity" <gm...@toad.net> wrote in message
news:YXqB6.2128$%r6.2...@news.abs.net...
Jeff Cochrane
at least) long before Win2k came out, therefore, Microsoft didn't (and
rightly so I believe) plan specific support for Socket 7 at all.
The mere fact that it runs on those low end machines is a bonus that was (I
beleive) unintended.
SS7 WAS an important architecture (Note the 'WAS'), but in Intels (and
almost everyone else's) view, it was dead and buried and had been for some
time when Win2k was released.
Jeff C.
"Colin Bocking" <millenn...@crouchvale.freeserve.co.uk> wrote in message
news:9b7ql2$o70$1...@news7.svr.pol.co.uk...
Colin Bocking
But SS7 hardware is still on sale here in the UK. Win2K was released over a
year ago! One might reasonably assume that the OS would support the hardware
especially as it is a business OS and businesses do not tend to be bothered
about the highest end machines. In any case, the GA5AX and Win2K DO work
well together in some cases!
As I have said before, it does not matter to me because I am not a Win2K
user, but it would just be nice to know where the problem lies
Colin
"Jeff Cochrane" <jeff_c...@no.damn.spam.jeffscomputers.com> wrote in
message news:3ad7d720$0$25508$7f31...@news01.syd.optusnet.com.au...
Jeff Cochrane
Now you have my curiosity raised.
I'm wondering who is still making these SS7 boards??
(Methinks they need shooting) :)
Jeff C.
"Colin Bocking" <millenn...@crouchvale.freeserve.co.uk> wrote in message
news:9b97fn$bdt$1...@news5.svr.pol.co.uk...
ji...@vcn.bc.ca
>Hmmmm,
>Now you have my curiosity raised.
>I'm wondering who is still making these SS7 boards??
Gigabyte was until at least Feb 2000 (Manufacture date of my GA-5AA board)
Colin Bocking
Do you know them?
Available this very day from insight.com together with AMD K6 2 500.
We have a very long thread here but I still think we deserve an explanation
from either Gigabyte or Microsoft (the latter might be out of the question).
Colin
"Jeff Cochrane" <jeff_c...@no.damn.spam.jeffscomputers.com> wrote in
message news:3ad82601$0$25519$7f31...@news01.syd.optusnet.com.au...
Shazam
I completely disagree. If Win2K doesn't "support" the Socket7/Super7
platform then how come it automatically recognizes the chipsets for the
following Socket 7 chipsets (Intel TX, HX etc) and Super7 chipsets (Via MVP3
and Ali Aladdin V)?
When I installed Win2K on my old Supermicro Intel TX based motherboard and
my newer Tyan S1598C2 VIA MVP3 based motherboard, Win2K automatically
recognized the chipsets on their motherboards. It runs great!
I have not tried it on my Gigabyte GA-5AX yet but I have heard of many
people running Win2K successfully on other Ali Aladdin V based motherboards.
Win2K runs on these motherboards because MS purposely supplied support for
the older platforms. It was not "unintended". Are you trying to imply that
MS "accidently" put drivers for those chipsets in Win2K and "accidently"
wrote the installation procedure to "accidently" detect which chipset it is
being installed on so it can "accidently" load the correct motherboard
drivers on bootup?
Jeff Cochrane <jeff_c...@no.damn.spam.jeffscomputers.com> wrote in
message news:3ad7d720$0$25508$7f31...@news01.syd.optusnet.com.au...
Shazam
I think I found the solution to the 1/2 CPU speed problem. According to the
AMD Zone forums, you have to disable ACPI and/or APM and then Win2K will run
at full speed on the GA-5AX. It's been verified by a user to work.
To me, the solution makes sense. On the Gigabyte GA-5AX, it has a thermal
sensor that "slows down" the cpu when it overheats. Maybe on Win2K, with
APM/ACPI enabled, it thinks the cpu is overheating and slows it down
half-speed.
So this is not an Ali Aladdin V chipset issue. And it is not a problem of
Win2K not supporting the Super7 platform.
This is a problem with the Gigabyte's handling of ACPI/APM and its secretive
thermal sensor (why doesn't Gigabyte publish the specs to this thermal
sensor so someone can write a temperature monitoring utility that shows us
the CPU temperature while in Windows?).
I now have a low opinion of Gigabyte's technical staff if they could not
have figured out the solution to this problem. How could they blame it on
the Ali Aladdin V chipset when no other Aladdin V based motherboard has this
problem? It's just plain common sense that the problem is with the Gigabyte
GA-5AX.
kotor <ko...@kotor.com> wrote in message
news:9au87a$qhl$1...@nreadA.inwind.it...
Colin Bocking
I must say, I have always had problems with ACPI/APM on this board (I have 3
of them!)
Whaddya say Gigabyte?
Whaddya think Jeff?
Colin
"Shazam" <gm...@toad.net> wrote in message
news:a0kC6.174$Z74....@news.abs.net...
Jeff Cochrane
It does makes sense to me.
(I am SO glad I have never sold or used one of those pieces of crap called
the 5A? series :)(By choice I might add), seems as though they are a f^%##^g
nightmare to support! <G>)
All I can say is kudo's to the person[s] who nutted out the problem and came
up with a viable answer.
Jeff C.
"Colin Bocking" <millenn...@crouchvale.freeserve.co.uk> wrote in message
news:9bd0qc$vam$1...@newsg4.svr.pol.co.uk...
Nikolaos Tampakis
this motherboard has. Both the F3 and F4b BIOSes (the ACPI ones) will
simply not allow win2000 to install on a SCSI drive. Since I own an
all-SCSI system I can't know if it's only a matter with the
install-drive being SCSI or with SCSI presence in general, but it simply
didn't work no matter what configuration changes I made. Installing with
an older (F2 or F1) BIOS was fine. Also, the F4b BIOS will NOT allow a
floppy seek to be performed at startup if both IDE channels are
disabled. And disabling floppy seek in the BIOS doesn't help either,
cause a seek is attempted anyway.
I will have to agree that it has nothing to do with the chipset but
rather with the motherboard (either the BIOS or the hardware layer). My
board is a 4.1 revision, I recall older posts suggesting the situation
might be OK with the 5.2 revision.
Furthermore, I don't really think the 'secretive' thermal sensor has
anything to do with APM or ACPI. If I remember well, Gigabyte calls this
system 'ACOPS' or something like that, and it rather works on its own,
independently of any OS. This isn't to rule it out as a possible cause,
of course, but according to Gigabyte's word it shouldn't be (it
shouldn't also be possible to write any utility that displays CPU
temperature from the 'secretive' sensor, since it appears to be a
self-contained circuit).
I have switched over to a PIII platform recently, and moved my SCSI
stuff there. The GA-5AX + K6-III 400 found home in an all-IDE
environment now, so I guess win2000 will eventually install in ACPI
mode. I'll post anything interesting on this issue I may find (I doubt I
will, of course).
Regards
Nikos Tampakis
Shazam <gm...@toad.net> wrote in message
news:a0kC6.174$Z74....@news.abs.net...
Shazam
the problem with installing Win2K on SCSI drives is fixed. I do have a SCSI
based CDRW and CDROM but no SCSI hard drives so I guess it doesn't matter
with me.
Question, has anybody actually verified that the Gigabyte's thermal sensor
works? How does verify that it is working? I remember one guy posting that
he has verified that it works but I've never seen anyone else.
Oh well, dispite its problems I still feel the Gigabyte GA-5AX is one of the
better Super7 motherboards out there...
Nikolaos Tampakis <ntab...@hotmail.com> wrote in message
news:98738871...@athnrd02.forthnet.gr...
gbt-tech
happy about the tech support. I can only apolgize, but a certain point we
have to refer problems to Taiwan, especially when it looks like a bios or
chipset issue, unfortunately we are kind of stuck with thier response as we
don't have the resources to work on these things here, though we have
identified and resolved things they've missed befor it's usually been in the
line of turning on bios functions Taiwan have turned off.
Mike Mullen
CST
Gigabyte UK
kotor
DISABLE ACPI/APM in the bios (disable "power management" and "...apm" in the
bios power screen??) ? after o before win2k installation?
Thanks..
Shazam
happen is that Win2K does not find ACPI in your system and installs with APM
mode.
It's also possible to just disable ACPI from the Control Panel in Win2K.
Check out the forums in www.amdzone.com. There was some threads about this.
Hope this helps.
kotor <ko...@kotor.com> wrote in message
news:9boutg$v3t$1...@nreadA.inwind.it...