VOGONS


ATI Radeon 9700Pro vs 9800Pro

Topic actions

First post, by tpowell.ca

User metadata
Rank Member
Rank
Member

Hello everyone.

For the purpose of vintage computing, from DOS to Windows 98, which of the two cards would you keep ?
More to the point, is there anything at all the 9700 does better than the 9800 ?

Both are AGP, both have 128MB and both are in near-mint condition with thermal pad/paste replaced.

Thanks

  • Merlin: MS-4144, AMD5x86-160 32MB, 16GB CF, ZIP100, Orpheus, GUS, S3 VirgeGX 2MB
    Tesla: GA-6BXC, VIA C3 Ezra-T, 256MB, 120GB SATA, YMF744, GUSpnp, Quadro2
    Newton: K6XV3+/66, AMD K6-III+500, 256MB, 32GB SSD, AWE32, Voodoo3

Reply 3 of 20, by swaaye

User metadata
Rank l33t++
Rank
l33t++

The 9700 doesn't support trilinear filtering optimizations (aka brilinear) so its trilinear filtering is slightly higher quality (and more demanding for the chip). I'm not sure if it can be seen outside of synthetic tests though. 9600/9800 added this capability.

9800 also adds that F-Buffer stuff. Not sure that has practical value for any games though.

Reply 4 of 20, by darry

User metadata
Rank l33t++
Rank
l33t++

Assuming your motherboard is compatible with both, I would use the 9800 (might not be AGP 3.3v, depending on chip version) but would keep the 9700 (AGP 3.3v compatible) as a spare . These cards generate a fair amount of heat (for their time, though nothing like the FX5800) but are not always equipped with an adequate cooling solution. This will probably have on impact on longevity (I have a couple of dead 9800 Pros) .

See http://www.playtool.com/pages/agpcompat/agp.html for compaibility info .

Reply 5 of 20, by The Serpent Rider

User metadata
Rank l33t++
Rank
l33t++

the 9700PRO for me is a little more interesting

I would avoid any long use of 9700 Pro. They are painfully hard to find in working condition and die very randomly. 9800 Pro dies randomly too, but quite easy to find.

I must be some kind of standard: the anonymous gangbanger of the 21st century.

Reply 6 of 20, by appiah4

User metadata
Rank l33t++
Rank
l33t++

The only advantage to keeping the 9700PRO is that it is exponentially rarer than the 9800PRO, but it's also more prone to failure over time.

I have a 9800 (in all respects equivalent to the 9700PRO, actually even built on a 9700PRO board with an R360 and 9800PRO BIOS) and a 9800PRO, and the speed difference is around 15%.

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 7 of 20, by Baoran

User metadata
Rank l33t
Rank
l33t
appiah4 wrote:

The only advantage to keeping the 9700PRO is that it is exponentially rarer than the 9800PRO, but it's also more prone to failure over time.

I have a 9800 (in all respects equivalent to the 9700PRO, actually even built on a 9700PRO board with an R360 and 9800PRO BIOS) and a 9800PRO, and the speed difference is around 15%.

'
I thought the reason for 9700 pros failures were often the shim problem and if you fix that they don't really fail more often than 9800pro?

Reply 8 of 20, by Scali

User metadata
Rank l33t
Rank
l33t
tpowell.ca wrote:

More to the point, is there anything at all the 9700 does better than the 9800 ?

No, the 9800 was just a slight update to the GPU. It's better in every way.
See here for info: https://www.anandtech.com/show/1077/2

The extent of the changes were limited primarily to improving Hyper-Z III, more specifically in enhancing the Z Cache and optimizing it for the next-generation of games. There have also been improvements to ATI's Z and Color compression algorithms that improve performance in AA modes, since the algorithms aren't active in non-AA environments.

http://scalibq.wordpress.com/just-keeping-it- … ro-programming/

Reply 9 of 20, by appiah4

User metadata
Rank l33t++
Rank
l33t++
Baoran wrote:
appiah4 wrote:

The only advantage to keeping the 9700PRO is that it is exponentially rarer than the 9800PRO, but it's also more prone to failure over time.

I have a 9800 (in all respects equivalent to the 9700PRO, actually even built on a 9700PRO board with an R360 and 9800PRO BIOS) and a 9800PRO, and the speed difference is around 15%.

'
I thought the reason for 9700 pros failures were often the shim problem and if you fix that they don't really fail more often than 9800pro?

The shim actually exists on the 9800PRO as well.

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 10 of 20, by Baoran

User metadata
Rank l33t
Rank
l33t
appiah4 wrote:
Baoran wrote:
appiah4 wrote:

The only advantage to keeping the 9700PRO is that it is exponentially rarer than the 9800PRO, but it's also more prone to failure over time.

I have a 9800 (in all respects equivalent to the 9700PRO, actually even built on a 9700PRO board with an R360 and 9800PRO BIOS) and a 9800PRO, and the speed difference is around 15%.

'
I thought the reason for 9700 pros failures were often the shim problem and if you fix that they don't really fail more often than 9800pro?

The shim actually exists on the 9800PRO as well.

Yes but the shape of the stock heatsink has been changed for 9800pro that the shim doesn't cause problems.

Reply 11 of 20, by tpowell.ca

User metadata
Rank Member
Rank
Member

Ok. So other than rarity, there is no meaningful reason to use a 9700Pro over a 9800Pro.
Is the DVI output as reliable between them ? I am aware that there were many cards with DVI compatibility issues.

  • Merlin: MS-4144, AMD5x86-160 32MB, 16GB CF, ZIP100, Orpheus, GUS, S3 VirgeGX 2MB
    Tesla: GA-6BXC, VIA C3 Ezra-T, 256MB, 120GB SATA, YMF744, GUSpnp, Quadro2
    Newton: K6XV3+/66, AMD K6-III+500, 256MB, 32GB SSD, AWE32, Voodoo3

Reply 12 of 20, by SPBHM

User metadata
Rank Oldbie
Rank
Oldbie

I remember removing the shim on my first 9500NP and gaining significant OC headroom, which means it was probably running extremely hot before the removal.

The Serpent Rider wrote:

the 9700PRO for me is a little more interesting

I would avoid any long use of 9700 Pro. They are painfully hard to find in working condition and die very randomly. 9800 Pro dies randomly too, but quite easy to find.

it's difficult for me to comment, but I've had 3 R300 cards so far and they were reliable (but all 9500s)

also on forums 9800PROs dying were a very common theme, but it's true that the number of 9700s around was a lot lower, so that doesn't mean much.

but yes, most 9700/9800 had very poor cooling.

Reply 13 of 20, by swaaye

User metadata
Rank l33t++
Rank
l33t++

Yeah I think the design was intentional. Maybe to protect the exposed die from heatsink impacts? The shim causes a gap, separating the die from the heatsink, and they used phase change compound that still conducts some heat with a gap. It was their first exposed die GPU on the desktop.

If the chip isn't at >90C, I don't know if that would be a cause of failures though. On the other hand, the fans like to die, and then it would most certainly overheat.

Reply 14 of 20, by The Serpent Rider

User metadata
Rank l33t++
Rank
l33t++

Both Radeon 9700 and 9800 suffer from BGA RAM failures. That's just my speculation, but only Samsung-based cards are affected the most. Infineon/Hynix cards are relatively safe to use.

I must be some kind of standard: the anonymous gangbanger of the 21st century.

Reply 15 of 20, by candle_86

User metadata
Rank l33t
Rank
l33t

Not really, but you could get a 256mb FireGL X1 and change the bios device Id and have a 256mb 9700Pro, alot cheaper than a 9800Pro 256mb

Reply 16 of 20, by Baoran

User metadata
Rank l33t
Rank
l33t
swaaye wrote:

Yeah I think the design was intentional. Maybe to protect the exposed die from heatsink impacts? The shim causes a gap, separating the die from the heatsink, and they used phase change compound that still conducts some heat with a gap. It was their first exposed die GPU on the desktop.

If the chip isn't at >90C, I don't know if that would be a cause of failures though. On the other hand, the fans like to die, and then it would most certainly overheat.

I don't think radeon 9700 pro has any kind of temperature sensor so you wouldn't know that a gap between the chip and the heatsink had been formed before the card dies.
I still have my old radeon 9700pro with the original box too and it works, but I don't think I would use it in any build before taking it apart and either removing the shim or putting something there that would fill the possible gap between the chip and the heatsink.

Reply 17 of 20, by silikone

User metadata
Rank Member
Rank
Member
tpowell.ca wrote:

More to the point, is there anything at all the 9700 does better than the 9800 ?

Yes, it's better at leaving lasting impressions among gamers and serving as rhetorical fuel for modern debates.

Do not refrain from refusing to stop hindering yourself from the opposite of watching nothing other than that which is by no means porn.

Reply 18 of 20, by The Serpent Rider

User metadata
Rank l33t++
Rank
l33t++

By the way you can softmod 9700 into 9800, which alleviates the performance gap.

I must be some kind of standard: the anonymous gangbanger of the 21st century.

Reply 19 of 20, by appiah4

User metadata
Rank l33t++
Rank
l33t++
The Serpent Rider wrote:

By the way you can softmod 9700 into 9800, which alleviates the performance gap.

Is the inverse possible? I have this card:

ATI-Radeon-9800.jpg

A Radeon 9800 built using a 9700PRO PCB and an R360 GPU. I want to softmod it into a Radeon 9700PRO (Because I would rather have one of those than a Radeon 9800) - if I flash the BIOS, will it work?

Retronautics: A digital gallery of my retro computers, hardware and projects.