VOGONS


First post, by Aebtdom

User metadata
Rank Member
Rank
Member

I was Just wondering.
I am taking the opportunity to benchmark some graphics cards in my spare time, just to have an overview to how they compare in performance.
Now I am wondering If i put my voodoo 5 5500 into single chip mode. Are the results comparable to the voodoo 5 4500?

My list will be:
Voodoo 1 4mb
Voodoo banshee 12mb
Voodoo 2 12mb
Voodoo 2 SLI 2x12mb
Voodoo 3 3000
Voodoo 4 4500 if it adds up
Voodoo 5 5500

All on a AMD XP 2600+

Later on I will add a TNT 2 M64 and Ultra
I740, g200, gf 4280,gf 5500, gf 5200, rage 128, savage 4 on at least 3dm 99.

Attachments

  • 20210309_144305.jpg
    Filename
    20210309_144305.jpg
    File size
    1.64 MiB
    Views
    726 views
    File comment
    for so far.
    File license
    Public domain

Builds:

Xp2400+ gf3 ti200 + vd2 SLI 12MB + 768MB + SB live @ WinXP & 98 Dualboot.

MMX233 + S3trio 3D/X2 8MB + 3Dfx Voodoo 1 + AWE64 + 64MB SDR @ msdos / win98.

Reply 1 of 9, by framebuffer

User metadata
Rank Member
Rank
Member

Yes, since the 5500 is basically a 4500 SLI and since there is no unified memory access on the 5500 for the VSA100 chips, so each chip has access only to 32MB of RAM and disabling one of them disables also the associated RAM

Windows 98 and SAMBA | Quake CPU Benchmarks | GeForce2: GTS vs MX400

Reply 2 of 9, by Aebtdom

User metadata
Rank Member
Rank
Member

Thank you for confirming.

Builds:

Xp2400+ gf3 ti200 + vd2 SLI 12MB + 768MB + SB live @ WinXP & 98 Dualboot.

MMX233 + S3trio 3D/X2 8MB + 3Dfx Voodoo 1 + AWE64 + 64MB SDR @ msdos / win98.

Reply 4 of 9, by seanneko

User metadata
Rank Newbie
Rank
Newbie
framebuffer wrote on 2021-03-10, 20:35:

Yes, since the 5500 is basically a 4500 SLI and since there is no unified memory access on the 5500 for the VSA100 chips, so each chip has access only to 32MB of RAM and disabling one of them disables also the associated RAM

How do you disable one of the chips on a 5500? I have a 5500 with faulty RAM which I'm trying to troubleshoot.

Reply 5 of 9, by Aebtdom

User metadata
Rank Member
Rank
Member

If you install Amiga Merlin 2.9 driver with the 3dfx tools.. you can change the quality of your Direct3D and OpenGL/Glide from single, fastest performance, 2x Anti Aliasing and 4x Anti Aliasing.

Builds:

Xp2400+ gf3 ti200 + vd2 SLI 12MB + 768MB + SB live @ WinXP & 98 Dualboot.

MMX233 + S3trio 3D/X2 8MB + 3Dfx Voodoo 1 + AWE64 + 64MB SDR @ msdos / win98.

Reply 6 of 9, by framebuffer

User metadata
Rank Member
Rank
Member
Doornkaat wrote on 2021-03-11, 04:20:

Afaik the V5500 doesn't support sideband addressing, which the V4500 does.
Does it support sideband addressing when run in single chip mode?

that's a good point, in fact the Voodoo4 has also AGP4x and Voodoo5 AGP 2x, so for completeness, since I haven't done any test myself and can not find valid references, I guess it's fair to say that Voodoo4 4500 AGP and Voodoo5 5500 AGP, with one VSA100 disabled, should have very comparable performance but probably not at 100.00%

seanneko wrote on 2021-03-11, 04:41:

How do you disable one of the chips on a 5500? I have a 5500 with faulty RAM which I'm trying to troubleshoot.

It's been a while, don't remember exactly but I'd look in advanced options (in driver panel)

Windows 98 and SAMBA | Quake CPU Benchmarks | GeForce2: GTS vs MX400

Reply 8 of 9, by Aebtdom

User metadata
Rank Member
Rank
Member

That would be the same.
Difference is: the voodoo 5 5500 AGP card is limited to AGP 2x 3,3V and the voodoo 4 4500AGP supports 1,5V agp 4x.
The voodoo 5 5500 PCI in this case runs on newer systems, even on the ones which have pci-e, as long as they have PCI slots. Thus they can have a little higer performance

Builds:

Xp2400+ gf3 ti200 + vd2 SLI 12MB + 768MB + SB live @ WinXP & 98 Dualboot.

MMX233 + S3trio 3D/X2 8MB + 3Dfx Voodoo 1 + AWE64 + 64MB SDR @ msdos / win98.

Reply 9 of 9, by Meatball

User metadata
Rank Oldbie
Rank
Oldbie

The 4500 can edge out the 5500 in single-chip mode, but it's unlikely most of us would notice during actual gameplay.

According to Dodge Garage, the speculation is RAM timings are tighter. You can see results in the link below. Some early testing was done with a 5500 simulating a 4500:
http://www.thedodgegarage.com/3dfx/bench.htm

"Notice the 4500 taking the top 3dfx honors? I suspect the single VSA-100 chip card has more aggressive ram timings due to it's non multi-chip configuration."

"V5-4500 166 mhz- Typical 4500 card with one thing to note. When I first started doing the benchmarks I was running a 5500 in single chip mode to simulate a 4500. When I happened to install a 4500 in the test computer I noted that the benchmarks were slightly faster! I'm fairly sure that the 4500 has more aggressive ram timings since it doesn't have to sync to another processor."

***->WINNER, 1ST PLACE<-***
2022 #QUAKE3totheMAX -560.5fps-
Brain Drain Retro LAN https://discord.com/channels/799008837918261328
Windows ME
NForce2 A7N8X-E DLX
Athlon 848/154MHz
DDR@411MHz (2-3-3-3)
GeForce 256 DDR@144/344MHz
ESS Maestr0-1