VOGONS


3DBENCH CPU Benchmark Database

Topic actions

Reply 61 of 296, by Markk

User metadata
Rank Oldbie
Rank
Oldbie

Update. I had a strange problem. On every other pc the disk worked ok, but on this 386 it kept showing packed file error. And, as it also didn't recognize the HD I had in the case, I thought it may be some sort of incompatibility between the controller and the m/b. So I replaced it and it works fine. Now, running the 1.0 version of 3dbench, I got almost the same results. 15.8fps on normal setting, and 17.5 by raising the at bus speed.

Reply 62 of 296, by retro games 100

User metadata
Rank l33t
Rank
l33t
Mau1wurf1977 wrote:

We could need more scores for ........ Super Socket 7 stuff ...

Here's a SS7 3DBench2 score.

Intel-PentiumMMX-66-2.5-166-P5BV3+Rev:B3-VIA_Apollo_MVP3-1MB-AGP-Voodoo3_2000-16MB-Defaults:Optomised-retrogames100. Score:125.2

I then tweaked the BIOS options, to make the system faster, and reran 3DBench2. Curiously, I got the same score.

Reply 63 of 296, by retro games 100

User metadata
Rank l33t
Rank
l33t

Intel-80486SX-25-1-25-AsusPVI486SP3-Sis496/497-256KB-PCI-S3_Virge_GX_Nitro-4MB-Defaults:Optomised-retrogames100
= 20.0

External cache disabled = 19.2
Internal cache disabled, External cache enabled = 7.8

CPU change. DX4-100 write back model. Score = 62.5
External cache disabled = 55.5
Internal cache disabled, External cache enabled = 11.9

Conclusion: There's probably more flexibility using a Slot 1 system (lots of bus speed options, CPU speed choices, and cache disabling), when you want to emulate the system speed of a 386.

Last edited by retro games 100 on 2010-10-17, 15:36. Edited 1 time in total.

Reply 64 of 296, by elianda

User metadata
Rank l33t
Rank
l33t

Could you please bench the 486SX25 with an ISA card? Most of the earlier 486er were ISA only, so this would fit well to the SX25. Although the Intel Saturn chipset might be a bit faster anyway.

Retronn.de - Vintage Hardware Gallery, Drivers, Guides, Videos. Now with file search
Youtube Channel
FTP Server - Driver Archive and more
DVI2PCIe alignment and 2D image quality measurement tool

Reply 65 of 296, by 5u3

User metadata
Rank Oldbie
Rank
Oldbie

Here are some low-end 486 results:

Intel-486DX-25-1-25-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-ISA-Tseng ET4000AX-1MB-Tweaked-5u3 12.9

Intel-486DX-25-1-25-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 20.0

Intel-486DX-33-1-33-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-ISA-Tseng ET4000AX-1MB-Tweaked-5u3 15.6

Intel-486DX-33-1-33-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 27.0

Intel-486DX-40-1-40-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-ISA-Tseng ET4000AX-1MB-Tweaked-5u3 21.2

Intel-486DX-40-1-40-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 32.2

Reply 66 of 296, by retro games 100

User metadata
Rank l33t
Rank
l33t
elianda wrote:

Could you please bench the 486SX25 with an ISA card? Most of the earlier 486er were ISA only, so this would fit well to the SX25. Although the Intel Saturn chipset might be a bit faster anyway.

I'm sorry, I don't have a single ISA video card! However, 5u3 has just benched marked this exact mobo (including same revision number) with an ISA card, and his SX25 (5u3, did you mean SX?) score drops from 20 to ~13 when switching from PCI (same video card as mine) to ISA. That's a very interesting and valuable performance change.

Reply 68 of 296, by retro games 100

User metadata
Rank l33t
Rank
l33t
5u3 wrote:
retro games 100 wrote:

(5u3, did you mean SX?)

Nope, I tested with a 486DX, but it shouldn't matter for this benchmark anyways.

OK! Edit: Ah, I see how you've done the benchmarking. Take one CPU, and adjust the mobo's bus speed for the 25, 33, and 40 speed tests.

Reply 70 of 296, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

Database has been updated (post # 2)!

Thanks for the 386 / Pentium and Socket A results 🤣

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 71 of 296, by 5u3

User metadata
Rank Oldbie
Rank
Oldbie

Some more 486 results coming up...

AMD-486DX2-25-2-50-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-ISA-Cirrus Logic GD5426-1MB-Tweaked-5u3 20.4

AMD-486DX2-25-2-50-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 35.7

AMD-486DX2-33-2-66-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-ISA-Cirrus Logic GD5426-1MB-Tweaked-5u3 27.0

AMD-486DX2-33-2-66-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 47.6

AMD-486DX2-40-2-80-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-ISA-Cirrus Logic GD5426-1MB-Tweaked-5u3 33.3

AMD-486DX2-40-2-80-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 58.8

Reply 72 of 296, by aitotat

User metadata
Rank Member
Rank
Member

Here is 486DX4-100 results with the same VLB graphics card that I previously tested in a 386. The CPU is Intel with Write Back cache enabled.

Intel - 486DX4 - 33 - 3 - 100 - Asus - VL/I-486SV2GX4 - SIS471 - 256kiB - VLB - #9 GXE64 (S3 864) - 2MiB - aitotat - 66.6

I decided to play with BIOS settings and found out that setting Local Bus Ready from Sychronize to Transparent increased results to 71.4. So please write the 71.4 to the table since I'm using that BIOS setting from now on.

Edit: 27.0 with Turbo disabled.

Reply 73 of 296, by elianda

User metadata
Rank l33t
Rank
l33t

I just noticed that CPU Clock and Bus Clock is exchanged in the table for my result of the 486SX (the only one).
Seems we lack alot results for SX - Systems.

Retronn.de - Vintage Hardware Gallery, Drivers, Guides, Videos. Now with file search
Youtube Channel
FTP Server - Driver Archive and more
DVI2PCIe alignment and 2D image quality measurement tool

Reply 74 of 296, by 5u3

User metadata
Rank Oldbie
Rank
Oldbie

Here is the last batch of my 486 bench runs. 12.9 to 90.9 on one board! 😁

Intel-486DX4-25-3-75-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 47.6

AMD-5x86-25-4-100-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 52.6

Intel-486DX4-33-3-100-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 66.6

AMD-5x86-33-4-133-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 71.4

Intel-486DX4-40-3-120-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 76.9

AMD-5x86-40-4-160-Asus-PVI 486SP3 rev 1.8-SIS 496/497-256kB-PCI-S3 Virge GX-4MB-Tweaked-5u3 90.9

Reply 76 of 296, by Markk

User metadata
Rank Oldbie
Rank
Oldbie

I have a couple of different 3dbench results for you!
AMD-80286-16-1-16-DFI-286-12/16E-Headland-none-ISA-OAK_OTI067-512kB-Markk-5.9
and another :
Intel-PentiumIII-133-7.5-1000-QDI_Advance10F-VIA_ApolloPro133A-None-AGP-GeForce2MX-32MB-Markk-656.6

Reply 77 of 296, by BastlerMike

User metadata
Rank Member
Rank
Member
retro games 100 wrote:

That must be the fastest socket 3 board on planet earth! I notice it's a PCI board. Looking at your awesome score, there doesn't appear to be any sign of the infamous PCI "suckage" that is commonly associated with PCI and this type of board. What's your view on this? Generally speaking, is a socket 3 PCI board perfectly satisfactory, or would you consider a decent VLB/ISA board to be superior to all PCI boards?

Can you explain what you exactly mean with "suckage" ?
Regarding the PCI bus performance it can be stated that it depends on the chipset.
Are there any good 486 chipset guides existing on the internet ?

Reply 78 of 296, by retro games 100

User metadata
Rank l33t
Rank
l33t
BastlerMike wrote:
retro games 100 wrote:

That must be the fastest socket 3 board on planet earth! I notice it's a PCI board. Looking at your awesome score, there doesn't appear to be any sign of the infamous PCI "suckage" that is commonly associated with PCI and this type of board. What's your view on this? Generally speaking, is a socket 3 PCI board perfectly satisfactory, or would you consider a decent VLB/ISA board to be superior to all PCI boards?

Can you explain what you exactly mean with "suckage" ?

I assumed (incorrectly perhaps) that generally speaking PCI performance was poor. Suckage = poorness.

Reply 79 of 296, by BastlerMike

User metadata
Rank Member
Rank
Member

Ok, then I understood it right and as I said, it depends on the chipset.
I have motherboards with all the chipsets that were common back then.
From my experience the following conclusions are proposed (these results accord with some pc magazine articles from that time):

First group (PCI bridged off the VL-Bus, mainly produced until mid-1994):

VLB Chipset / Bus Bridge device
VIA 82C486A / VIA 82C505
Opti 82C802G / Opti 82C822
ALI M1429 / ALI M1435
Contaq 82C596 / 82C599

VIA and Opti -> slowest PCI solutions (~15 MB/s)
ALI -> only bridge device with good performance (30-35 MB/s)
Contaq -> mediocre (20-25 MB/s)

Second group (true PCI designs):

Intel Saturn/Aries
VIA 82C496/82C505
SiS 85C496/85C497
ALI M1487/M1489
UMC UM8881F/UM8886F

No exact data at the moment, but I would suggest this performance order: Intel > UMC > ALI > SIS > VIA