Asus p/i-p6np5 and pp200mhz - FastVid needed?
John L. Hardy IV
This is motherboard and chip are arriving tomorrow at my vendor, so no bios
info yet. Do you think I would need to run VastVID on a mobo this new?
Also - one more time, <g> anyone have the url to the FastVid page? Our
proxy server is a little wiggy right now so I couldn't search for it..
Thanks -
--
John IV
JHa...@Adobe.com / Jo...@MSN.com / Jo...@GTE.net
Adobe Systems Internet/Persuasion SQA Team, Seattle, WA
Juan Tuboneya
If I were to run NT, would I still need to run fastvid? What exactly is
fastvid? TSR? A BIOS patch (i.e., you run it once and it patches the
BIOS and you wont need to run it again)? Would I need it if I am going
to run NT and Windows 95?
Thanks
John L. Hardy IV
I got this combo mobo/chip two days ago.. and noted in the bios that there
are options for write combining and write posting along w/ some other pci
functions.. all are enabled by default. I haven't pulled down fastvid yet,
but it was my understanding that it enabled these functions for bioses that
couldn't toggle them.
--
John IV
JHa...@Adobe.com / Jo...@GTE.net
Adobe Systems Internet/Persuasion SQA Team, Seattle, WA
Juan Tuboneya <tubo...@westside.exisinc.com> wrote in article
<321eb...@SLIDER.EXISINC.COM>...
Kenneth Johansson
"John L. Hardy IV" <jha...@adobe.com> wrote:
>I got this combo mobo/chip two days ago.. and noted in the bios that there
>are options for write combining and write posting along w/ some other pci
>functions.. all are enabled by default. I haven't pulled down fastvid yet,
>but it was my understanding that it enabled these functions for bioses that
>couldn't toggle them.
you still need fastvid for write combining on the linear frame buffer.
Per Allansson
The P6NP5 mobo does not enable everything that FastVid does.
Get Fastvid - you'll get a big improvement in video throughput.
(like from 30 to 70 MB/s ... )
/Per
John L. Hardy IV wrote:
>
> I got this combo mobo/chip two days ago.. and noted in the bios that there
> are options for write combining and write posting along w/ some other pci
> functions.. all are enabled by default. I haven't pulled down fastvid yet,
> but it was my understanding that it enabled these functions for bioses that
> couldn't toggle them.
David S. Livingston
k...@canit.se (Kenneth Johansson) wrote:
>"John L. Hardy IV" <jha...@adobe.com> wrote:
>>I got this combo mobo/chip two days ago.. and noted in the bios that there
>>are options for write combining and write posting along w/ some other pci
>>functions.. all are enabled by default. I haven't pulled down fastvid yet,
>>but it was my understanding that it enabled these functions for bioses that
>>couldn't toggle them.
>you still need fastvid for write combining on the linear frame buffer.
No you dont....Install new 103 bios and just turn it on... works
great..
david
Yang Hong
In article <508cf2$9...@molokini.conterra.com>, DSLivi...@conterra.com
(David S. Livingston) wrote:
Are you sure? I have 1.03 bios on my board and all the results with and
without fastvid remain no changed --- you still need fastvid!
Yang
Juan Tuboneya
Two different answers. Can someone else add to this? Is FASTVID needed
or not?
Thanks
Per Allansson
> > you still need fastvid for write combining on the linear frame buffer.
> >
> >
> > No you dont....Install new 103 bios and just turn it on... works
> > great..
> >
> > david
>
> without fastvid remain no changed --- you still need fastvid!
>
> Yang
If I understand this correctly(which I probably don't... :( )
the 1.03 (or higher) BIOS is able to turn on (in Fastvid terms):
- Write Posting
- VGA (banked) Write Combining
Fastvid is also able to turn on "SVGA (linear frame buffer) write
combining" - it does indeed make a big difference on my system -
I have a Diamond Stealth 968 2Mb VRAM card. Worth noting is that
I need to load a VESA 2.0 driver before running fastvid to get
everything working. Is it possible that systems which have gfxcards
which have onboard support for VESA 2.0 don't need Fastvid - that
this SVGA lfb WC is indeed turned on by the BIOS???
/per
Andreas Schildbach
tubo...@westside.exisinc.com (Juan Tuboneya) hat geschrieben:
> Two different answers. Can someone else add to this? Is FASTVID needed
> or not?
>
> Thanks
>
> David S. Livingston (DSLivi...@conterra.com) wrote:
> : k...@canit.se (Kenneth Johansson) wrote:
>
> : >"John L. Hardy IV" <jha...@adobe.com> wrote:
>
> : >>I got this combo mobo/chip two days ago.. and noted in the bios that there
> : >>are options for write combining and write posting along w/ some other pci
> : >>functions.. all are enabled by default. I haven't pulled down fastvid yet,
> : >>but it was my understanding that it enabled these functions for bioses that
> : >>couldn't toggle them.
>
> : >you still need fastvid for write combining on the linear frame buffer.
>
>
> : No you dont....Install new 103 bios and just turn it on... works
> : great..
With BIOS 0103 you don't need Fastvid on the ASUS (X)P6NP5 for Write
Combining on the LFB (Linear Frame Buffer).
For some reason, EMM386.EXE disables WC on LFB in a way that it can't
be re-enabled using Fastvid!
- Andreas
---
Andreas "Goonie" Schildbach
schi...@informatik.tu-muenchen.de
Ride more roller-coasters!
Per Allansson
Per Allansson wrote:
>
> Yang Hong wrote:
> >
> > In article <508cf2$9...@molokini.conterra.com>, DSLivi...@conterra.com
> > > you still need fastvid for write combining on the linear frame buffer.
> > >
> > >
> > > No you dont....Install new 103 bios and just turn it on... works
> > > great..
> > >
> >
> > Are you sure? I have 1.03 bios on my board and all the results with and
> > without fastvid remain no changed --- you still need fastvid!
> >
> > Yang
>
> If I understand this correctly(which I probably don't... :( )
> the 1.03 (or higher) BIOS is able to turn on (in Fastvid terms):
>
> - Write Posting
> - VGA (banked) Write Combining
>
> Fastvid is also able to turn on "SVGA (linear frame buffer) write
> combining" - it does indeed make a big difference on my system -
> I have a Diamond Stealth 968 2Mb VRAM card. Worth noting is that
> I need to load a VESA 2.0 driver before running fastvid to get
> everything working. Is it possible that systems which have gfxcards
> which have onboard support for VESA 2.0 don't need Fastvid - that
> this SVGA lfb WC is indeed turned on by the BIOS???
reported crappy values in the LFB mode test, and good values
in banked mode - but I still need to run Fastvid to get good
values in banked mode. I'm getting really confused...
Robert B. McGehee
Per Allansson <al...@tripnet.se> wrote:
:The P6NP5 mobo does not enable everything that FastVid does.
:Get Fastvid - you'll get a big improvement in video throughput.
:(like from 30 to 70 MB/s ... )
:
:/Per
:
Where can one find fastvid???
-------------------------------------------------------------------
Robert B. McGehee - University of Mississippi - RM Racing, Inc.
r...@sunset.backbone.olemiss.edu
http://sunset.backbone.olemiss.edu/~rbm
http://www.inlink.com/~rmracing
-------------------------------------------------------------------
Kenneth Johansson
tubo...@westside.exisinc.com (Juan Tuboneya) wrote:
>Two different answers. Can someone else add to this? Is FASTVID needed
>or not?
>Thanks
I need fastvid for linear mode. I also have to turn of USWC write
posting in BIOS otherwise sound locks my computer.
VSPEED also reports wrong values for linear mode. I always get 13 MEG
with fastvid regardless of what i set in BIOS or FASTVID.
John L. Hardy IV
It's somewhat convoluted.. here the results I have obtained, with
assistance from DLivingston too. First you must enable all the pci-write
combining settings in the chipset screen. Also in this screen, change your
video cache mode to 'USWC' (this is very important and can skew vspeed if
it's not on).
Now if you run vspeed you should see something like this:
copy DRAM to banked VGA: 87.72 million yada yada
copy DRAM to linear framebuffer: 89.97 millon
I thought all was well at this point.. but decided to run fastvid just to
see what was reported. It said:
Write posting 'enabled' already
VGA Write Combining 'enabled' already
the final option
SVGA write combining wasn't enabled. but Fast vid enabled it at a hex
address fb00000 for my Matrox Mill.
So from this it appears that 1.03 will take you 2/3 of the way, but it may
be unable to set the final toggle for SVGA write combining unless it knows
the linear frame buffer's address. I'm trying to see who's responsibility
this last one is.. bios or Matrox maybe.. anyone?
--
John IV
jha...@adobe.com / jo...@gte.net
Adobe Systems Internet/Persuasion SQA, Seattle, Wa
Per Allansson <al...@tripnet.se> wrote in article
<322E10B0...@tripnet.se>...
:
:
Andreas Schildbach
k...@canit.se (Kenneth Johansson) hat geschrieben:
Yes, this is because you have to throw out EMM386.EXE out of your
CONFIG.SYS in order for LFB Write Combining to work.
Andreas Schildbach
"John L. Hardy IV" <jha...@adobe.com> hat geschrieben:
> It's somewhat convoluted.. here the results I have obtained, with
> assistance from DLivingston too. First you must enable all the pci-write
> combining settings in the chipset screen. Also in this screen, change your
> video cache mode to 'USWC' (this is very important and can skew vspeed if
> it's not on).
>
> Now if you run vspeed you should see something like this:
> copy DRAM to banked VGA: 87.72 million yada yada
> copy DRAM to linear framebuffer: 89.97 millon
>
> I thought all was well at this point.. but decided to run fastvid just to
> see what was reported. It said:
>
> Write posting 'enabled' already
> VGA Write Combining 'enabled' already
> the final option
> SVGA write combining wasn't enabled. but Fast vid enabled it at a hex
> address fb00000 for my Matrox Mill.
>
> So from this it appears that 1.03 will take you 2/3 of the way, but it may
> be unable to set the final toggle for SVGA write combining unless it knows
> the linear frame buffer's address. I'm trying to see who's responsibility
> this last one is.. bios or Matrox maybe.. anyone?
Again, you have to throw out EMM386.EXE of your CONFIG.SYS. It will
shut off WC on LFB otherwise.
Per Allansson
I have a Diamond Stealth 64 Video (2M VRAM) gfxcard - and I don't get
the same results as you do. If (ahem... when) I turn on all pci-write
combining settings and USWC in bios, and start up the computer without
loading any drivers, I get:
copy DRAM to banked VGA: 19
copy DRAM to LFB: not enabled
then I load Fastvid - now I get an even lower result in banked mode
(LFB still not working)
[ Fastvid does report that WP & VGA WC is (already) enabled, and
SVGA LFB disabled (--> enabled) ]
Now I reboot(cold) to disable whatever fastvid did - and load
univbe - now I get
copy DRAM to banked VGA: 19
copy DRAM to LFB: 19
then I load Fastvid - now I get
copy DRAM to banked VGA: 82
copy DRAM to LFB: 19
Why do I need to load Univbe + Fastvid to get fast banked mode?
Why isn't LFB mode enabled?
Must the gfxcard support VESA 2.0 onboard (like the Matrox?)?
John L. Hardy IV wrote:
>
> It's somewhat convoluted.. here the results I have obtained, with
> assistance from DLivingston too. First you must enable all the pci-write
> combining settings in the chipset screen. Also in this screen, change your
> video cache mode to 'USWC' (this is very important and can skew vspeed if
> it's not on).
>
> Now if you run vspeed you should see something like this:
> copy DRAM to banked VGA: 87.72 million yada yada
> copy DRAM to linear framebuffer: 89.97 millon
>
> I thought all was well at this point.. but decided to run fastvid just to
> see what was reported. It said:
>
> Write posting 'enabled' already
> VGA Write Combining 'enabled' already
> the final option
> SVGA write combining wasn't enabled. but Fast vid enabled it at a hex
> address fb00000 for my Matrox Mill.
>
> So from this it appears that 1.03 will take you 2/3 of the way, but it may
> be unable to set the final toggle for SVGA write combining unless it knows
> the linear frame buffer's address. I'm trying to see who's responsibility
> this last one is.. bios or Matrox maybe.. anyone?
>
Andreas Schildbach
p...@netcom.com (Philip M. Burton) hat geschrieben:
> Andreas Schildbach (schi...@informatik.tu-muenchen.de) wrote:
> : k...@canit.se (Kenneth Johansson) hat geschrieben:
>
> : > tubo...@westside.exisinc.com (Juan Tuboneya) wrote:
> : >
> : > >Two different answers. Can someone else add to this? Is FASTVID needed
> : > >or not?
> : >
> : > >Thanks
> : >
> : > I need fastvid for linear mode. I also have to turn of USWC write
> : > posting in BIOS otherwise sound locks my computer.
> : >
> : > VSPEED also reports wrong values for linear mode. I always get 13 MEG
> : > with fastvid regardless of what i set in BIOS or FASTVID.
>
> : Yes, this is because you have to throw out EMM386.EXE out of your
> : CONFIG.SYS in order for LFB Write Combining to work.
>
> Is this an issue if you are running Win 95 or Win NT?
DOS: Yes.
Win'95: I think so.
Win/NT: No.
Neal Westfall
John L. Hardy IV (jha...@adobe.com) wrote:
:
: Write posting 'enabled' already
: VGA Write Combining 'enabled' already
: the final option
: SVGA write combining wasn't enabled. but Fast vid enabled it at a hex
: address fb00000 for my Matrox Mill.
:
: So from this it appears that 1.03 will take you 2/3 of the way, but it may
: be unable to set the final toggle for SVGA write combining unless it knows
: the linear frame buffer's address. I'm trying to see who's responsibility
: this last one is.. bios or Matrox maybe.. anyone?
FWIW, I'm using an ATI 3D Xpression and all three come up enabled when I
turn everything on in bios. However, if USWC is enabled, running quake
results in a lockup, unless I boot Safe Mode Command Prompt Only.
Neal
John L. Hardy IV
Per -
Appears the only difference would be the video card, perhaps it does have
to do w/ VESA 2 compliance being built in.... but it's possible that
fastvid is not detecting the base hex address for your lfb properly.
Should also note that for what it's worth, bios 1.04 is avail. at Asus.
It's pretty sad that we have to go through these contortions in the first
place. <g>
I just found this quote as part of my message down at the bottom <g>.. I
don't know if it was you who originally wrote it..
: > : >
: > : > Fastvid is also able to turn on "SVGA (linear frame buffer) write
: > : > combining" - it does indeed make a big difference on my system -
: > : > I have a Diamond Stealth 968 2Mb VRAM card. Worth noting is that
: > : > I need to load a VESA 2.0 driver before running fastvid to get
: > : > everything working. Is it possible that systems which have gfxcards
: > : > which have onboard support for VESA 2.0 don't need Fastvid - that
: > : > this SVGA lfb WC is indeed turned on by the BIOS???
--
John IV
jha...@adobe.com / jo...@gte.net
Adobe Systems: Internet/Persuasion SQA, Seattle, Wa
Per Allansson <al...@tripnet.se> wrote in article
<3232CDEF...@tripnet.se>...
: I have a Diamond Stealth 64 Video (2M VRAM) gfxcard - and I don't get
: >
: > Write posting 'enabled' already
: > VGA Write Combining 'enabled' already
: > the final option
: > SVGA write combining wasn't enabled. but Fast vid enabled it at a hex
: > address fb00000 for my Matrox Mill.
: >
: > So from this it appears that 1.03 will take you 2/3 of the way, but it
may
: > be unable to set the final toggle for SVGA write combining unless it
knows
: > the linear frame buffer's address. I'm trying to see who's
responsibility
: > this last one is.. bios or Matrox maybe.. anyone?
: >
: > --
: > :
:
:
John L. Hardy IV
Jan Walter - Centurion Services
Andreas Schildbach (schi...@informatik.tu-muenchen.de) wrote:
: > Is this an issue if you are running Win 95 or Win NT?
:
: DOS: Yes.
: Win'95: I think so.
: Win/NT: No.
We have had some problems with write combining when running NT 4 Preview,
as well as problems with well-proven drivers under the OS/2 4 Beta. We
were using a Hercules Terminator 64 (S3 Trio64 V+) video card. (Yes, we
know that these are silly to use on a PPro since they are relatively slow.)
This seems to have something to do with true 32 bit S3 specific video
drivers.
Cheers,
Jan Walter
--
// Centurion Services
// "Quality from the Start"
// Fax:(604) 279-1800 | www.infomatch.com/~centurion
// cent...@infomatch.com
// WE ARE HIRING: www.infomatch.com/~centurion/jobs
Dan Butterfield
In article <50tul0$r...@holocron.odc.net>, nwes...@odc.net (Neal Westfall)
wrote:
>FWIW, I'm using an ATI 3D Xpression and all three come up enabled when I
>turn everything on in bios. However, if USWC is enabled, running quake
>results in a lockup, unless I boot Safe Mode Command Prompt Only.
Win95 also hangs during bootup on my system if I have USWC enabled in BIOS,
in addition to the VGA or linear frame buffer write combining options of
either FASTVID or the BIOS. If I turn off USWC, everything works fine and I
get high values reported from VSPEED using either the BIOS VGA/LFB options or
FASTVID.
My question is, what is this USWC option doing? I did not measure any
performance difference with or without it, and it does not seem to be
equivalent to any of the 3 FASTVID options:
1) PCI Write posting
2) VGA write combining
3) linear frame buffer write combining
Of course the BIOS manual explains nothing other than what bit setting it
corresponds to:
This option is for USWC Write Posting to PMC register 53h, bit 5. The
settings are: Disabled or Enabled. The Optimal and Fail-Safe default
settings are Enabled.
I've even looked at the Natoma technical documentation for this bit setting,
and unfortunately it did not shed much more light. Does anyone know what this
option does exactly and how it relates (or *if* it relates) to the things that
FASTVID does?
Dan Butterfield -Keller, TX- http://ourworld.compuserve.com/homepages/danielb
"Never taunt the Happy Fun Ball"
Dan Butterfield
In article <516o8j$l...@rtpnews.raleigh.ibm.com>, dan...@vnet.ibm.com wrote:
>I've even looked at the Natoma technical documentation for this bit setting,
>and unfortunately it did not shed much more light. Does anyone know what this
>option does exactly and how it relates (or *if* it relates) to the things that
>FASTVID does?
Here's what the Natoma technical documentation says about this bit setting:
USWC Write Post During I/O Bridge Access Enable (UWPIO). 1=Enable.
<
allows posting of CPU-to-PCI cycles destined for a USWC (Uncacheable,
speculative, write combining; essentially PCI adapter memory for which
writes to can be combined/cached, like an SVGA adapter) region, even during
a passive release cycle.
It says nothing specific about what passive release is, other than:
To comply with PCI Specification, revision 2.1 latency requirements, the
PMC supports passive release. The PMC disables CPU-to-PCI posting during
the passive release, except for transactions to the USWC region (resumably
only if the above bit is turned on).
What's odd is I turn this option OFF in BIOS, but still get high write speeds
from VSPEED on writing to the LFB and VGA buffer if I enable write combining
(via FASTVID or BIOS) to those areas. Presumably, then, this bit setting only
applies during a passive release cycle (which means ???).
I saw another post which seemed to imply this conflict is only caused on
machines with a Matrox Millenium and a Sound Blaster, and that the actual hang
occurs when the Sound Blaster is initialized / used. This would be consistent
withit hanging on my system as Win95 starts (initializing the SB), and on your
system when running Quake in DOS (using the SB).
Perhaps SB DMA uses a passive release cycle and somehow triggers this
conflict? Don't know, just grasping at straws...
Dan Butterfield
>Perhaps SB DMA uses a passive release cycle and somehow triggers this
>conflict? Don't know, just grasping at straws...
I finally found a technical summary document
(http://www.ocean.aust.com/tech.html) that had some information about what
passive release is. Supposedly "This mechanism enables more efficient ISA Bus
Master cycles to PCI by absorbing ISA overhead before requesting PCI. On
multiple data cycles, the PCI bus is free for other agents' use in between
each ISA bus data phase." It goes on to say "As an example, audio cards place
less load on the system, enabling more CPU and PCI concurrency."
I guess this explains why the SB is the triggering agent for the failure,
since it would be the only thing (in my system, it is the only ISA card) which
could cause a passive release cycle. It also explains why FASTVID (or
the BIOS setting to enable write combining) reveals the problem, since
without it the LFB or VGA buffer are not treated as USWC regions, but are
instead just plain uncacheable memory. I imagine it means the only
performance aspect affected by this setting is writing to the SVGA card
concurrent with using the SB on a system with FASTVID write combining options
enabled. This would probably be a factor for most games (how much is
anybody's guess; only someone w/o this bug could tell us), but explains why
the setting does not seem to affect VSPEED, and why the setting seems to do
nothing for a system without FASTVID write combining turned on.
Of course, why it fails is anybody's guess. Probably some bug in the Matrox
card handling writes during a passive release cycle, or perhaps the
440FX chipset itself.