Velocity Reviews - Computer Hardware Reviews

Velocity Reviews > Newsgroups > Computing > Computer Information > No room for Soundcard!

Reply
Thread Tools

No room for Soundcard!

 
 
RayL12
Guest
Posts: n/a
 
      02-06-2011

Well, am I upset!

I upgraded my system at the expense of 480. One of the reasons for
doing so was because my Creative Xfi was up against my gfx card fan.

So I decided upon an ATX board(M4A89GTD Pro) with 2 PCI-E slots
thinking I could use the first in order to keep a PCI slot clear.

Comments on-line were all positive.

So, how disappointing to find that it was designed to use the 2nd PCIE
slot in single gfx card mode and a riser card in the first!? The Geforce
GTX 460 covers both remaining PCI slots completely.

What is it with these designers?? Surely they know that a gfx card is
potentially going to cover the PCI slots?

I'm struggling to find a work-around and a flying lead-out is all I
can think of. Ofcourse, this will mean the Xfi card sits outside the PC?

If anyone has any ideas I'd be grateful.

Cheers...

....Ray.






--
Learn why we are suffering..
www.zeitgeistthefilm.com/
 
Reply With Quote
 
 
 
 
Paul
Guest
Posts: n/a
 
      02-06-2011
RayL12 wrote:
>
> Well, am I upset!
>
> I upgraded my system at the expense of 480. One of the reasons for
> doing so was because my Creative Xfi was up against my gfx card fan.
>
> So I decided upon an ATX board(M4A89GTD Pro) with 2 PCI-E slots
> thinking I could use the first in order to keep a PCI slot clear.
>
> Comments on-line were all positive.
>
> So, how disappointing to find that it was designed to use the 2nd PCIE
> slot in single gfx card mode and a riser card in the first!? The Geforce
> GTX 460 covers both remaining PCI slots completely.
>
> What is it with these designers?? Surely they know that a gfx card is
> potentially going to cover the PCI slots?
>
> I'm struggling to find a work-around and a flying lead-out is all I can
> think of. Ofcourse, this will mean the Xfi card sits outside the PC?
>
> If anyone has any ideas I'd be grateful.
>
> Cheers...
>
> ...Ray.


Have you tested the video card in the white slot ?

Paul
 
Reply With Quote
 
 
 
 
RayL12
Guest
Posts: n/a
 
      02-09-2011
On 06/02/2011 20:43, Paul wrote:
> RayL12 wrote:
>>
>> Well, am I upset!
>>
>> I upgraded my system at the expense of 480. One of the reasons for
>> doing so was because my Creative Xfi was up against my gfx card fan.
>>
>> So I decided upon an ATX board(M4A89GTD Pro) with 2 PCI-E slots
>> thinking I could use the first in order to keep a PCI slot clear.
>>
>> Comments on-line were all positive.
>>
>> So, how disappointing to find that it was designed to use the 2nd PCIE
>> slot in single gfx card mode and a riser card in the first!? The
>> Geforce GTX 460 covers both remaining PCI slots completely.
>>
>> What is it with these designers?? Surely they know that a gfx card is
>> potentially going to cover the PCI slots?
>>
>> I'm struggling to find a work-around and a flying lead-out is all I
>> can think of. Ofcourse, this will mean the Xfi card sits outside the PC?
>>
>> If anyone has any ideas I'd be grateful.
>>
>> Cheers...
>>
>> ...Ray.

>
> Have you tested the video card in the white slot ?
>
> Paul


Wow, Paul, this is an unexpected reply as I pulled my post within
minutes of posting and therefore did not expect this to exist? You are
on the ball, mate

No, Paul, I haven't. I simply followed the manual instructions for
single gfx card use. It had crossed my mind. Upon first reading the MB
specs I was under the impression that I had 2 PCI-E x16 slots.
The manual describes it as: "2 x PCI Express 2.0 x16, support ATI
CrossFireX Technology(@ dual x8 speed)" So, is each PCI-E slot x16 or,
x8? I guess the latter.

As it is, I did put my Creative soundcard in the barely visible PCI
slot. While playing Grid the fans howled. This reduced greatly when I
pushed a 10mm wood chip between the two cards.

I will search the Net to see if anyone else has tried the first slot
under single gfx card use.

BTW, as you suggested, after spending 480 on the upgrade I cannot say
that my apps show any speed improvement. Maybe the exception is Google
SketchUp?

Thank you Paul,

....Ray.


 
Reply With Quote
 
Paul
Guest
Posts: n/a
 
      02-10-2011
RayL12 wrote:
> On 06/02/2011 20:43, Paul wrote:
>> RayL12 wrote:
>>>
>>> Well, am I upset!
>>>
>>> I upgraded my system at the expense of 480. One of the reasons for
>>> doing so was because my Creative Xfi was up against my gfx card fan.
>>>
>>> So I decided upon an ATX board(M4A89GTD Pro) with 2 PCI-E slots
>>> thinking I could use the first in order to keep a PCI slot clear.
>>>
>>> Comments on-line were all positive.
>>>
>>> So, how disappointing to find that it was designed to use the 2nd PCIE
>>> slot in single gfx card mode and a riser card in the first!? The
>>> Geforce GTX 460 covers both remaining PCI slots completely.
>>>
>>> What is it with these designers?? Surely they know that a gfx card is
>>> potentially going to cover the PCI slots?
>>>
>>> I'm struggling to find a work-around and a flying lead-out is all I
>>> can think of. Ofcourse, this will mean the Xfi card sits outside the PC?
>>>
>>> If anyone has any ideas I'd be grateful.
>>>
>>> Cheers...
>>>
>>> ...Ray.

>>
>> Have you tested the video card in the white slot ?
>>
>> Paul

>
> Wow, Paul, this is an unexpected reply as I pulled my post within
> minutes of posting and therefore did not expect this to exist? You are
> on the ball, mate
>
> No, Paul, I haven't. I simply followed the manual instructions for
> single gfx card use. It had crossed my mind. Upon first reading the MB
> specs I was under the impression that I had 2 PCI-E x16 slots.
> The manual describes it as: "2 x PCI Express 2.0 x16, support ATI
> CrossFireX Technology(@ dual x8 speed)" So, is each PCI-E slot x16 or,
> x8? I guess the latter.
>
> As it is, I did put my Creative soundcard in the barely visible PCI
> slot. While playing Grid the fans howled. This reduced greatly when I
> pushed a 10mm wood chip between the two cards.
>
> I will search the Net to see if anyone else has tried the first slot
> under single gfx card use.
>
> BTW, as you suggested, after spending 480 on the upgrade I cannot say
> that my apps show any speed improvement. Maybe the exception is Google
> SketchUp?
>
> Thank you Paul,
>
> ...Ray.
>


Well, I can't guarantee it'll work, but my suspicion is, when you
stick your video card in the first slot, it'll run at x8. And if
both the video card and chipset are Revision 2 type, then each lane
is 500MB/sec. So that gives you 4GB/sec of PCI Express bandwidth.
That should work well with your video card.

Some experiments were done years ago, to measure the effect
of restricting PCI Express bandwidth. It was done, by placing cello
tape over the contacts on the card, to simulate different lane wiring.
The video card supports operation all the way down to x1 rate.

SPECviewperf shows a strong dependency.

http://www.tomshardware.com/reviews/...ing,927-9.html

This game didn't have as strong a dependency. But it also
isn't the best game for such a test.

http://www.tomshardware.com/reviews/...ng,927-10.html

I'd love to see those experiments repeated with modern
hardware, but no computer site has been the least curious
about how PCI Express bandwidth helps. Those taping
experiments should be repeated.

If you're gaming, a little narrower bus probably won't hurt that much.
If we use the top end SPECViewPerf as an example, maybe 4%
lower frame rate. It's really hard to guess.

Paul
 
Reply With Quote
 
RayL12
Guest
Posts: n/a
 
      02-10-2011
On 10/02/2011 00:00, Paul wrote:
> RayL12 wrote:
>> On 06/02/2011 20:43, Paul wrote:
>>> RayL12 wrote:
>>>>
>>>> Well, am I upset!
>>>>
>>>> I upgraded my system at the expense of 480. One of the reasons for
>>>> doing so was because my Creative Xfi was up against my gfx card fan.
>>>>
>>>> So I decided upon an ATX board(M4A89GTD Pro) with 2 PCI-E slots
>>>> thinking I could use the first in order to keep a PCI slot clear.
>>>>
>>>> Comments on-line were all positive.
>>>>
>>>> So, how disappointing to find that it was designed to use the 2nd PCIE
>>>> slot in single gfx card mode and a riser card in the first!? The
>>>> Geforce GTX 460 covers both remaining PCI slots completely.
>>>>
>>>> What is it with these designers?? Surely they know that a gfx card is
>>>> potentially going to cover the PCI slots?
>>>>
>>>> I'm struggling to find a work-around and a flying lead-out is all I
>>>> can think of. Ofcourse, this will mean the Xfi card sits outside the
>>>> PC?
>>>>
>>>> If anyone has any ideas I'd be grateful.
>>>>
>>>> Cheers...
>>>>
>>>> ...Ray.
>>>
>>> Have you tested the video card in the white slot ?
>>>
>>> Paul

>>
>> Wow, Paul, this is an unexpected reply as I pulled my post within
>> minutes of posting and therefore did not expect this to exist? You are
>> on the ball, mate
>>
>> No, Paul, I haven't. I simply followed the manual instructions for
>> single gfx card use. It had crossed my mind. Upon first reading the MB
>> specs I was under the impression that I had 2 PCI-E x16 slots.
>> The manual describes it as: "2 x PCI Express 2.0 x16, support ATI
>> CrossFireX Technology(@ dual x8 speed)" So, is each PCI-E slot x16 or,
>> x8? I guess the latter.
>>
>> As it is, I did put my Creative soundcard in the barely visible PCI
>> slot. While playing Grid the fans howled. This reduced greatly when I
>> pushed a 10mm wood chip between the two cards.
>>
>> I will search the Net to see if anyone else has tried the first slot
>> under single gfx card use.
>>
>> BTW, as you suggested, after spending 480 on the upgrade I cannot say
>> that my apps show any speed improvement. Maybe the exception is Google
>> SketchUp?
>>
>> Thank you Paul,
>>
>> ...Ray.
>>

>
> Well, I can't guarantee it'll work, but my suspicion is, when you
> stick your video card in the first slot, it'll run at x8. And if
> both the video card and chipset are Revision 2 type, then each lane
> is 500MB/sec. So that gives you 4GB/sec of PCI Express bandwidth.
> That should work well with your video card.
>
> Some experiments were done years ago, to measure the effect
> of restricting PCI Express bandwidth. It was done, by placing cello
> tape over the contacts on the card, to simulate different lane wiring.
> The video card supports operation all the way down to x1 rate.
>
> SPECviewperf shows a strong dependency.
>
> http://www.tomshardware.com/reviews/...ing,927-9.html
>
> This game didn't have as strong a dependency. But it also
> isn't the best game for such a test.
>
> http://www.tomshardware.com/reviews/...ng,927-10.html
>
> I'd love to see those experiments repeated with modern
> hardware, but no computer site has been the least curious
> about how PCI Express bandwidth helps. Those taping
> experiments should be repeated.
>
> If you're gaming, a little narrower bus probably won't hurt that much.
> If we use the top end SPECViewPerf as an example, maybe 4%
> lower frame rate. It's really hard to guess.
>
> Paul


Once again, Paul, I thank you for your knowledge,research and, time in
answering.

As you stated, it seems that the number of lanes makes little difference
to throughput? I think I will give the No.1 slot a try. Is it worth it
to you if I d/l Viewperf and send you the results for each slot?

....Ray.

--
Learn why we are suffering..
www.zeitgeistthefilm.com/
 
Reply With Quote
 
Paul
Guest
Posts: n/a
 
      02-10-2011
RayL12 wrote:

>
> As you stated, it seems that the number of lanes makes little difference
> to throughput? I think I will give the No.1 slot a try. Is it worth it
> to you if I d/l Viewperf and send you the results for each slot?
>
> ...Ray.
>


Only if you're curious.

The last time I ran that, I wasn't too impressed with it (SPECviewperf).
Mainly, because it seemed to be dominated by CPU activity. I couldn't
convince myself, it was really stressing the video card. Mind you,
I tested on a pretty low end card.

Some benchmarks leave me with a warmer feeling than others. That
one didn't really impress me. So if you run it, it'll mainly be,
just to see what kind of a benchmark it is.

Paul
 
Reply With Quote
 
RayL12
Guest
Posts: n/a
 
      02-10-2011
On 10/02/2011 20:06, Paul wrote:
> RayL12 wrote:
>
>>
>> As you stated, it seems that the number of lanes makes little
>> difference to throughput? I think I will give the No.1 slot a try. Is
>> it worth it to you if I d/l Viewperf and send you the results for each
>> slot?
>>
>> ...Ray.
>>

>
> Only if you're curious.
>
> The last time I ran that, I wasn't too impressed with it (SPECviewperf).
> Mainly, because it seemed to be dominated by CPU activity. I couldn't
> convince myself, it was really stressing the video card. Mind you,
> I tested on a pretty low end card.
>
> Some benchmarks leave me with a warmer feeling than others. That
> one didn't really impress me. So if you run it, it'll mainly be,
> just to see what kind of a benchmark it is.
>
> Paul


OK, SPECviewperf is out. But if I am to change slots I may as well do
some tests. Have you a benchmarking software suggestion?

....Ray.

--
Learn why we are suffering..
www.zeitgeistthefilm.com/
 
Reply With Quote
 
Paul
Guest
Posts: n/a
 
      02-11-2011
RayL12 wrote:
> On 10/02/2011 20:06, Paul wrote:
>> RayL12 wrote:
>>
>>>
>>> As you stated, it seems that the number of lanes makes little
>>> difference to throughput? I think I will give the No.1 slot a try. Is
>>> it worth it to you if I d/l Viewperf and send you the results for each
>>> slot?
>>>
>>> ...Ray.
>>>

>>
>> Only if you're curious.
>>
>> The last time I ran that, I wasn't too impressed with it (SPECviewperf).
>> Mainly, because it seemed to be dominated by CPU activity. I couldn't
>> convince myself, it was really stressing the video card. Mind you,
>> I tested on a pretty low end card.
>>
>> Some benchmarks leave me with a warmer feeling than others. That
>> one didn't really impress me. So if you run it, it'll mainly be,
>> just to see what kind of a benchmark it is.
>>
>> Paul

>
> OK, SPECviewperf is out. But if I am to change slots I may as well do
> some tests. Have you a benchmarking software suggestion?
>
> ...Ray.
>


3Dmark 2001 SE Build 330 is a 41MB download. I like this for
low end cards, but it doesn't use some of the latest DirectX
features in its benchmark tests.

http://majorgeeks.com/3Dmark_d99.html

3DMark2006 is more up to date with current technology, but it is a
584MB download. That might be a better match to your card
(no "999 frames per second" runs).

http://majorgeeks.com/3DMark06_d4935.html

Those are not perfect benchmarks either. But for the purposes you're
putting them to (ratio comparison, using identical hardware), most
of their sins will not be an issue. That should make a good enough test.

With regard to SPECviewperf, it might be best used on "certified" OpenGL
cards. Those are the cards that Nvidia and ATI charge an arm and leg
for, and the driver is "unleashed" for OpenGL work with hundreds of thousands
of objects. Apparently, you can try to do OpenGL with desktop cards,
but at some small number of objects, they tend to bog down. And
this is presumably a function of something in the driver.

When I was watching SPECviewperf on my system, it almost looked like it
was computing the image on the CPU, and just copying it to the frame
buffer (and there was an OpenGL driver file for the video card). That's
because the jerky rendering it was doing - the long delay between frames,
at least to me, seemed to be a computing delay, rather than a GPU delay.

At work, we had a high end workstation, with just about the most expensive
video card you could put in it, and on complex objects (designs that take
7 hours to load from a network database), the rotation of the final object
was relatively smooth (the frame rate was somewhere between 20Hz and
30Hz or so). You could tell the machine was struggling though...
SPECviewperf looked nothing like that.

Paul
 
Reply With Quote
 
RayL12
Guest
Posts: n/a
 
      02-11-2011
On 11/02/2011 01:06, Paul wrote:

>
> 3DMark2006 is more up to date with current technology, but it is a
> 584MB download. That might be a better match to your card
> (no "999 frames per second" runs).
>
> http://majorgeeks.com/3DMark06_d4935.html
>
> Those are not perfect benchmarks either. But for the purposes you're
> putting them to (ratio comparison, using identical hardware), most
> of their sins will not be an issue. That should make a good enough test.


OK, Paul. It is a 25min d/l from the fastest choice site of Australia.
The MajorGeeks site was 1.5 days. Bless them!

Here is an issue that puts doubt in my mind. when I put all the new
hardware in the PC, I ran the old system boot. I was warned that it
would never be accepted by the new hardware; or visa versa.

And in the first case Win XP64 did not quite get to the desktop. How
ever, after running in Safe mode (I did nothing in Safe mode: it
wouldn't let me ) and then restarting, I did get it to start in
Normal mode. From here I removed what previous drivers I could using
Driver Cleaner Pro and installed the ones from the new install discs.

And, everything appears fine? This saved me from days(weeks) of re
installing applications to the tune of 29Gbs.

Do you think there could be underlying problems? Old mobo M3N78-VM;
new M4A89GTD Pro. Both Asus.

Thanks again, Paul...

....Ray.
 
Reply With Quote
 
Paul
Guest
Posts: n/a
 
      02-11-2011
RayL12 wrote:
> On 11/02/2011 01:06, Paul wrote:
>
>>
>> 3DMark2006 is more up to date with current technology, but it is a
>> 584MB download. That might be a better match to your card
>> (no "999 frames per second" runs).
>>
>> http://majorgeeks.com/3DMark06_d4935.html
>>
>> Those are not perfect benchmarks either. But for the purposes you're
>> putting them to (ratio comparison, using identical hardware), most
>> of their sins will not be an issue. That should make a good enough test.

>
> OK, Paul. It is a 25min d/l from the fastest choice site of Australia.
> The MajorGeeks site was 1.5 days. Bless them!
>
> Here is an issue that puts doubt in my mind. when I put all the new
> hardware in the PC, I ran the old system boot. I was warned that it
> would never be accepted by the new hardware; or visa versa.
>
> And in the first case Win XP64 did not quite get to the desktop. How
> ever, after running in Safe mode (I did nothing in Safe mode: it
> wouldn't let me ) and then restarting, I did get it to start in
> Normal mode. From here I removed what previous drivers I could using
> Driver Cleaner Pro and installed the ones from the new install discs.
>
> And, everything appears fine? This saved me from days(weeks) of re
> installing applications to the tune of 29Gbs.
>
> Do you think there could be underlying problems? Old mobo M3N78-VM; new
> M4A89GTD Pro. Both Asus.
>
> Thanks again, Paul...
>
> ...Ray.


You can move an OS from one system to another, as long as there is a
driver for the disk interface.

Some boards, have classical emulation of IDE (even when using a SATA port),
and the driver loaded for the port is actually a Microsoft driver. If two
boards use that same, Microsoft driver, then you can move a system over.

The second option, is a "bounce" install. Plug a card into the old PC (I use
an Ultra133 TX2 from Promise for that). Install the driver, and move the disk
over to the removable card, and make sure it can boot from the add-in card. Then,
shut down for the last time, move the card over to the other PC, and try to boot.
Since the driver was already there, it should work. Then, if you want to set the
new motherboard to AHCI or whatever, it's easy to install the driver for the
new hardware, without upsetting the driver for the Ultra133 that is being used
to host the system at the moment. When the new driver is in place, move the
drive over to the motherboard disk interface and try to boot. Finally, pull the
Ultra133 card out of the system. By using an add-in card, you can transition
the system with better odds of having a working driver while you do it.

Also, before moving the system, you can remove any unnecessary drivers from
Add/Remove. That would include the video driver (as you likely have a new
video card). Don't reboot the system. Simply remove the drivers, and shut down
the system. Then, you're ready to move the disk over to the new system.

Somewhere in that mess, you should also have a backup of C:, in case your
transition attempt fails. By keeping a clone around, you have a way to go backwards
and try again. I made a mistake once, in one of my moves, and had to return to
the backup, so it's a useful thing to have.

On older OSes, you might have had tricks such as using regedit to delete
everything under the Enum key. Alternately, on some of the older OSes,
you could "create a profile" and empty it. On the newer OSes, all you seem
to be able to do, is copy a profile, and select one at bootup time, which
doesn't really help. The ability to create an empty profile on the other
hand, cleans out the drivers, and allows the first boot on the new system
to be a clean one. But the last one I did, I just made sure a compatible
driver was present, and moved the thing anyway.

I was using an OEM WinXP (the kind you can buy from places like Newegg),
and strictly speaking, I shouldn't have been able to activate it on the
new hardware. But I didn't even need to phone Microsoft to make the transition.
Activating online works (perhaps because a fair number of components were
the same, with regard to hardware). My change started out as a motherboard
and video card change - it still needed to be activated, but the online
activation worked.

I've read of cases, where activation is so screwed up when you move the
disk, that the machine freezes up and is useless. So that's the other
extreme. Yet another reason to have a clone copy of the disk around.
I think in my case, I was prompted to correct the activation issue,
in the next 72 hours or something.

Paul
 
Reply With Quote
 
 
 
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
OT: Small server room cooling Matt Burks MCSE 10 07-14-2005 05:41 AM
Can't setup MC PC and TV/Extender in other room! =?Utf-8?B?QmFnZWw=?= Microsoft Certification 2 03-04-2005 02:59 AM
Re: OMG Hilary Duff locker Room at school 4481 Matt Berther ASP .Net 0 04-16-2004 08:48 PM
Re: OMG Hilary Duff locker Room at school 2251 Jtyc MCSE 5 04-16-2004 08:38 PM
Chat room type textarea Ady ASP .Net 2 11-06-2003 10:01 AM



Advertisments