made a few upgrades to the Win98 machine , replaced the 600Mhz Katmai with a 700Mhz/100Mhz FSB slot1 Coppermine (which isn't officially supported, but works fine) . overclocked slightly to 720Mhz (the Geforce2 had visual glitches in 3Dmark even at 112Mhz FSB , odd , at 133 it won't even boot into Windows) , also overclocked the memory in the GF2TI up to 230Mhz from 200, bringing it up to GF2Ultra levels (raising the core clock did nothing, and this Hercules card has good ramsinks)

Sad thing is, my 2009 Compaq Mini with Atom and GMA950 gets lower 3DMark 2001 scores than your 700 mhz Pentium 3 rig. Funny how a "hyper threaded" 1.6 Ghz CPU cannot beat a single threaded 700 Mhz CPU, and funny how it has hyper threading but cannot even do a basic Pentium feature like out of order execution.

its rather cheap when some of these companies use CPUs that were meant for tablets and embedded types for laptops (and I use the term to describe netbooks too).

makes you wonder how these machines run Win7 or what can one do on these machines besides the basics (which I guess is what they were intended for), its part of the reason I never bothered with netbooks.

tried running 3DMark01 on a Toshiba Laptop bought in 2010 I believe , Intel Dual-Core T4200 @ 2GHz with 3GB of RAM and a GMA4500 , gave me a score of 3898

I expected a bit more from a 2Ghz Dual core.

Pentium mmx233MHZ plus 1mb cache l2
128mb sdram
80gb ide maxtor 7200rpm
gpu: ati radeon 7200 agp (the drivers from that cool gpu eat so many power from my processor D:)
the result:

Darkman wrote:

tried running 3DMark01 on a Toshiba Laptop bought in 2010 I believe , Intel Dual-Core T4200 @ 2GHz with 3GB of RAM and a GMA4500 , gave me a score of 3898

I expected a bit more from a 2Ghz Dual core.

I think you will get a better score with 2gb (2*1gb) memory
Intel GMA4500* uses system ram as video memory and with 3gb your ram is in single channel mode.
Dual channel doubles the bandwidth.

tbh there is no point in trying anything on a system like this, its kind of old and not very useful for very much, its also running Vista so its kind of slow even when just going through Windows (this would have made a decent WinXP laptop , its just Vista looked good on the box for unsuspecting buyers)

in terms of games its in line with the benchmark , it will run Max Payne decently, but stuff like Max Payne 2, Halo , MGS2 , San Andreas lag big time on it even at low settings.

Just for shits n giggles, I decided to try this on my newest retro machine:

P4 2.8 HT Northwood
ASUS P4C800
2G RAM
Radeon HD 2600 Pro (AGP)

I've never actually run any kind of benchmarking software on any machine I've ever owned so this was a first for me.

ibm5155 wrote:

Pentium mmx233MHZ plus 1mb cache l2 128mb sdram 80gb ide maxtor 7200rpm gpu: ati radeon 7200 agp (the drivers from that cool gpu […]
Show full quote

Pentium mmx233MHZ plus 1mb cache l2
128mb sdram
80gb ide maxtor 7200rpm
gpu: ati radeon 7200 agp (the drivers from that cool gpu eat so many power from my processor D:)
the result:

Your PMMX-233 beat my PII-300.

PII-300, 192MB EDO, Radeon 7000 PCI:

EDO really, really sucks.

P4 Preshott @ 3.9 + ASUS P4P800 + crippled ATI 9800 Pro running @ 300/300:

senrew wrote:

Just for shits n giggles, I decided to try this on my newest retro machine: […]
Show full quote

Just for shits n giggles, I decided to try this on my newest retro machine:

P4 2.8 HT Northwood
ASUS P4C800
2G RAM
Radeon HD 2600 Pro (AGP)

I've never actually run any kind of benchmarking software on any machine I've ever owned so this was a first for me.

Test 1 and test 2 - are there big FPS drop in 2 places (laser shoots from flyer and from mech)?

Standard Def Steve wrote:

Your PMMX-233 beat my PII-300. […]
Show full quote

ibm5155 wrote:

Pentium mmx233MHZ plus 1mb cache l2 128mb sdram 80gb ide maxtor 7200rpm gpu: ati radeon 7200 agp (the drivers from that cool gpu […]
Show full quote

Pentium mmx233MHZ plus 1mb cache l2
128mb sdram
80gb ide maxtor 7200rpm
gpu: ati radeon 7200 agp (the drivers from that cool gpu eat so many power from my processor D:)
the result:

Your PMMX-233 beat my PII-300.

PII-300, 192MB EDO, Radeon 7000 PCI:

EDO really, really sucks.

its not just the edo ram's fault, the PII has large cache so it needs to access ram less, and edo isn't really a lot slower than pc66 sdram anyway.
radeon7000 is a much slower card than the 7200, and it lacks the hardware T&L which is important for 3dmark, which slows down the cpu a lot.
my k6-3-300 scores 960pts with geforce2 anyway. 🤣

Gamecollector wrote:

senrew wrote:

Just for shits n giggles, I decided to try this on my newest retro machine: […]
Show full quote

Just for shits n giggles, I decided to try this on my newest retro machine:

P4 2.8 HT Northwood
ASUS P4C800
2G RAM
Radeon HD 2600 Pro (AGP)

I've never actually run any kind of benchmarking software on any machine I've ever owned so this was a first for me.

Test 1 and test 2 - are there big FPS drop in 2 places (laser shoots from flyer and from mech)?

Yes there is. Is that significant or something?

Pentium 3 Katmai 450mhz oc 600 mhz,348mb ram,agp tnt2 pro oc ultra level.

K6-2+ at 550mhz,256mb ram,voodoo 3 3000 agp

Some scores from my PCs:

Asrock 4Coredual
Pentium 4 640 [3.2ghz iirc]
FX5950 Ultra
Windows 2000
15621

MSI K7D
Dual Athlon MP1600
Wildcat VP990
Windows 2000
6793

Asus A7M266
Athlon XP2000
Voodoo5
Windows 98 FE
2720

Chaintech 6VIA5T
Pentium III 1.4S
Rage Fury Maxx
V2 SLI
Windows 98SE
Maxx using default: 1959
Maxx using 16 bit: 2244
V2 SLI using 16 bit: 736

Then tried the last rig with 3dmark2000 and 3dmark99max:

3dmark2000:
Maxx - 4129
V2SLI - 2295

3dmark99:
Maxx - 6014
V2SLI - 5158

does 3Dmark take dual CPUs into consideration? very few games do (anyone is more than welcome to list any apart from Quake3) so Im wondering if the benchmark does.

if it does than it might be even less reliable as a real measure of how fast real world gaming performance is.

Pretty sure 3DMark01 is single-threaded. But you might notice a little boost if you have dual CPUs as you can assign background processes and other crap to run on the second CPU.

that's what I thought, what caused me to wonder was this series of videos

www.youtube.com/watch?v=-6hju0brNJ8
www.youtube.com/watch?v=FZL76Hc032M

in the first video the guy runs a dual PII 400Mhz setup with a GF3Ti200 , and the score he gets in 3DMark01 is pretty close to the score I got with a 700Mhz Coppermine/GF2Ti

in the second video , he runs the same setup on 3Dmark2000, but gets a score much lower than I did (can't remember the number but I got something in the mid 6000s)

made me wonder if 3DMark01 was using those dual CPUs, either that or the DX8 features of the GF3 are making up the difference.

m1919 wrote:

Pretty sure 3DMark01 is single-threaded. But you might notice a little boost if you have dual CPUs as you can assign background processes and other crap to run on the second CPU.

I assumed just the optimizations to the cores of the modern processors would make a difference.

Darkman wrote:

that's what I thought, what caused me to wonder was this series of videos […]
Show full quote

that's what I thought, what caused me to wonder was this series of videos

http://www.youtube.com/watch?v=-6hju0brNJ8
http://www.youtube.com/watch?v=FZL76Hc032M

in the first video the guy runs a dual PII 400Mhz setup with a GF3Ti200 , and the score he gets in 3DMark01 is pretty close to the score I got with a 700Mhz Coppermine/GF2Ti

in the second video , he runs the same setup on 3Dmark2000, but gets a score much lower than I did (can't remember the number but I got something in the mid 6000s)

made me wonder if 3DMark01 was using those dual CPUs, either that or the DX8 features of the GF3 are making up the difference.

yes, DX8 features are what making the difference. havn't you even noticed that the nature test would only run on dx8 cards?
my k6-3+550 and gf2ultra scores 1550pts. with a faster cpu the gf2ultra is capable of over 6000pts, so you can see that the k6-3+ is the bottleneck.
however, when combined with a ti4600, the k6-3+550 scores 3050pts!