VOGONS


First post, by foey

User metadata
Rank Member
Rank
Member

Vogons.org 3DMark Score Table

I hope I'm not treading on anyones toes here but theres been a number of times where I wanted to find if one of my retro rigs is performing adequately. The 2001 Megathread on here is fantastic so I thought I would take it a step further with a leaderboard. I've taken the last couple of pages from the megathread which is a good start. If people can post a screenshot of their scores along with CPU speed e.t.c.
For older cards, if anyone can post their 3DMark 99/2000 Scores, I will upload them.

As we get more and more scores from a range of cards & CPUs we will have a good reference table for everyone to compare.

+ Let me know if there is anything I can add to the spreadsheet which will be useful
+ Open to suggestions on the order/layout

Last updated :- 30/06/2014

3DMark 2001 SE Default Settings
Download :- http://d1heja5wzvsadj.cloudfront.net/3DMark2001SE.exe
34xfghu.png

Edit :- I can't upload XLSX, I'll try and upload it to a host. Link to follow.

Last edited by foey on 2014-06-30, 09:55. Edited 1 time in total.

Cyrix Instead Build, 6x86 166+ | 32mb SD | 4mb S3 Virge DX | Creative AWE64 | Win95
ATC-S PIII Tualatin Win9x Build :- ATC-S PIII Coppermine Win9x Build Log [WIP] **Photo Heavy**

Reply 2 of 4, by idspispopd

User metadata
Rank Oldbie
Rank
Oldbie

You should add a OS column. According to the 2001 Megathread the OS can make a big difference on the same hardware.

Reply 3 of 4, by obobskivich

User metadata
Rank l33t
Rank
l33t
idspispopd wrote:

You should add a OS column. According to the 2001 Megathread the OS can make a big difference on the same hardware.

Agreed. I haven't personally seen any marked differences between Windows 2000 and XP, but Windows XP and Windows 7 can have fairly substantial differences in terms of the resulting score. I'd also add that looking at the raw 3DMark composite score isn't always a good indicator of how well the machine is "performing" (3DMark seems to just arbitrarily pick a number - measured fill-rate, frame-rate, etc be damned); that is to say it isn't reasonable or safe to assume two systems with similar/identical scores will necessarily perform similarly.

On the CPUs thing:

If you're going to note core revisions it might be worth noting them for models other than the Pentium 4 (and Prescott has two Ts). 😀

Reply 4 of 4, by AlphaWing

User metadata
Rank Oldbie
Rank
Oldbie

Yea an OS column would be great, and if hyper-threading is enabled on cpu's that have it.