VOGONS


My 3DMark01 Mega Thread

Topic actions

Reply 600 of 804, by Aideka

User metadata
Rank Member
Rank
Member

Something a bit different. Apple Mac mini 2012 base model, with Core i5 3210M, Intel HD 4000 Graphics, 4 gigs of RAM, Windows 8.1 via Boot Camp: 21685 3dMarks.

8zszli-6.png

Reply 601 of 804, by Carlos S. M.

User metadata
Rank Oldbie
Rank
Oldbie

Running some 3DMark2001 tests on a Socket 423 P4 system (1.5 Ghz). I'll post results soon

What is your biggest Pentium 4 Collection?
Socket 423/478 Motherboards with Universal AGP Slot
Socket 478 Motherboards with PCI-E Slots
LGA 775 Motherboards with AGP Slots
Experiences and thoughts with Socket 423 systems

Reply 602 of 804, by Oldskoolmaniac

User metadata
Rank Oldbie
Rank
Oldbie

My Dell Precision 470, the video card is defiantly a bottleneck.

Motherboard Reviews The Motherboard Thread
Plastic parts looking nasty and yellow try this Deyellowing Plastic

Reply 603 of 804, by Standard Def Steve

User metadata
Rank Oldbie
Rank
Oldbie

The Phenom II X6 is a fast little bastard. It's slightly faster than my Q6700 clock-for-clock in single threaded mode. In multi-threaded apps it demolishes the C2Q.

Here's the 3DMark2005 result at 4.07GHz w/ 16GB of DDR3-1760.
3DMark2005 becomes just as CPU limited as 2001SE when the video card is faster than a GTX280.
One of these days I'll install XP on it and try 3DMark01.
By57Z4Z.png

94 MHz NEC VR4300 | SGI Reality CoPro | 8MB RDRAM | Each game gets its own SSD - nooice!

Reply 604 of 804, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie

To have a good starting point with how powerfull modern GPU's are :
GTX 580 (w/ +10% OC) + E5700 4GHz (basicly, a "x15" multiplier E5800) vs. 3DMark 01 SE @ 3840x2160 (4k) with MSAA 8x 😉
Score : 26986 pkt.

3DMark 01 SE GeForce GTX 580 (850-4400) [3840-2160] mini.png

To get actual FPS numbers into perspective :
Avg. 400+ FPS in first test (and at this settings) means, that GPU (at this clocks) had to produce over 3 300 000 000 Pixels per second of actual Pixel Fillrate (or ~3,3 GPix/s+).
That's around the maximum Theoretical Pixe Fillrate of Radeon 9800 XT and I don't even take into account MSAA 8x I added on top 😁

PS. Bandwidth requirements for this feat (NOT real value, more like a city/street accuracy... of a house 😉)
Theory :
(Res[x]*Res[y] * (bytes/color value+Bytes/Z Value) * AA Level) * Number of times the screen is overdrawn * FPS,
Practice : (3840*2160*(32/8 + 24/8) * 8 )*400 = 173GB/s * Overdraw
^that's without color compresion, texture fetch bandwidth, blending stuff bandwidth [if blending is used], and few other things.

Source : LINK

157143230295.png

Reply 605 of 804, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie
Carlos S. M. wrote:

Running some 3DMark2001 tests on a Socket 423 P4 system (1.5 Ghz). I'll post results soon

Wanna race ?

3DMark 01 SE HD 3850 (AGP) [1,76GHz] mini.png

;D

157143230295.png

Reply 606 of 804, by Carlos S. M.

User metadata
Rank Oldbie
Rank
Oldbie
agent_x007 wrote:
Wanna race ? […]
Show full quote
Carlos S. M. wrote:

Running some 3DMark2001 tests on a Socket 423 P4 system (1.5 Ghz). I'll post results soon

Wanna race ?

3DMark 01 SE HD 3850 (AGP) [1,76GHz] mini.png

;D

Well, you used a socket 478 willamette with an i865PE mobo and dual channel DDR400

I used a pure Socekt 423 system with an i850 and RAMBUS RAM (CPU-Z doesn't properly recorgnize the Socket 423)

Specs are:
Pentium 4 1.5 Ghz
ECS P4ITA
512 MB RAMBUS
60 GB IDE HDD
Geforce 7600 GT (just for the benchmark)
Windows XP Pro SP3

What is your biggest Pentium 4 Collection?
Socket 423/478 Motherboards with Universal AGP Slot
Socket 478 Motherboards with PCI-E Slots
LGA 775 Motherboards with AGP Slots
Experiences and thoughts with Socket 423 systems

Reply 607 of 804, by Carlos S. M.

User metadata
Rank Oldbie
Rank
Oldbie

Also, some benchmarks done on the Socket 423 build as well

What is your biggest Pentium 4 Collection?
Socket 423/478 Motherboards with Universal AGP Slot
Socket 478 Motherboards with PCI-E Slots
LGA 775 Motherboards with AGP Slots
Experiences and thoughts with Socket 423 systems

Reply 608 of 804, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie
Carlos S. M. wrote:
Well, you used a socket 478 willamette with an i865PE mobo and dual channel DDR400 […]
Show full quote

Well, you used a socket 478 willamette with an i865PE mobo and dual channel DDR400

I used a pure Socekt 423 system with an i850 and RAMBUS RAM (CPU-Z doesn't properly recorgnize the Socket 423)

Specs are:
Pentium 4 1.5 Ghz
ECS P4ITA
512 MB RAMBUS
60 GB IDE HDD
Geforce 7600 GT (just for the benchmark)
Windows XP Pro SP3

It's only DDR300 not DDR400 (locked RAM multiplier) 😉
Two more results :
1) Stock (DDR 266MHz) with my GF 7600 GS @ GT

3DMark 01 SE GF 7600 GS @ GT (AGP) [1,6GHz] mini.png

2) OC 2,17GHz (2,0V) HD 3850 (AGP)

3DMark 01 SE HD 3850 (AGP) [2,176GHz] mini.png

Is it possible to OC that 1,5GHz P4 to 1,6GHz (it's only +7%) ?

157143230295.png

Reply 609 of 804, by Standard Def Steve

User metadata
Rank Oldbie
Rank
Oldbie
agent_x007 wrote:

2) OC 2,17GHz (2,0V) HD 3850 (AGP)

You're basically tied with my best DDR/PIII-S/6800GT score.
Edit: And it's only because your Nature score is much higher than mine. 😜

PIII-S DDR 6800GT 3D01.png

94 MHz NEC VR4300 | SGI Reality CoPro | 8MB RDRAM | Each game gets its own SSD - nooice!

Reply 610 of 804, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie
Standard Def Steve wrote:

You're basically tied with my best DDR/PIII-S/6800GT score.
Edit: And it's only because your Nature score is much higher than mine. 😜

Here's more direct comparison 😀

3DMark 01 SE GF 6800 Ultra (AGP) [2,176GHz] mini.png

157143230295.png

Reply 611 of 804, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie

Celeron 333A + 7600 GS @ GT (w/glitches, but it completes the test) :

3DMark 01 SE GF 7600 GS @ GT (Celeron 333MHz) mini.png

157143230295.png

Reply 612 of 804, by Jade Falcon

User metadata
Rank BANNED
Rank
BANNED

Agent, do you have an hwbot account?

Reply 613 of 804, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie

No, I do not.
But putting ALL of my scores on hwbot... is simply too much effort.

157143230295.png

Reply 614 of 804, by Jade Falcon

User metadata
Rank BANNED
Rank
BANNED
agent_x007 wrote:

No, I do not.
But putting ALL of my scores on hwbot... is simply too much effort.

I or mr.horse can add them to the team account or I or mr.horse can setup a user with your name and add them all for you. Then pass the login info to you.

Pretty much every bench I seen you post here fallows there rules so I could just pull them off Vogons and post them on your behalf. I just don't want to post them without your permission.

Reply 615 of 804, by kithylin

User metadata
Rank l33t
Rank
l33t
Jade Falcon wrote:
agent_x007 wrote:

No, I do not.
But putting ALL of my scores on hwbot... is simply too much effort.

I or mr.horse can add them to the team account or I or mr.horse can setup a user with your name and add them all for you. Then pass the login info to you.

Pretty much every bench I seen you post here fallows there rules so I could just pull them off Vogons and post them on your behalf. I just don't want to post them without your permission.

Although.. unless you're near the top spot for a specific video card/platform there's literally no point in ever submitting there if you're just going to fall down in to the rest of the rabble for that platform. hwbot is all about records and being the very best globally, that's kind.. the entire point of the website. Overclock records. If you aren't setting one there's no point in submitting anything. I'd have to look, but not sure if any of his submissions are record-breaking or even close or not. I have a lot of computers of different era's around here and I've benchmarked em lots of times, and usually overclock the hell out of em as far as I can and still be stable. Even with all of that.. I rarely ever get to submit anything to hwbot though, most of my systems just aren't good enough vs what's already on there.

Reply 616 of 804, by Jade Falcon

User metadata
Rank BANNED
Rank
BANNED
kithylin wrote:
Jade Falcon wrote:
agent_x007 wrote:

No, I do not.
But putting ALL of my scores on hwbot... is simply too much effort.

I or mr.horse can add them to the team account or I or mr.horse can setup a user with your name and add them all for you. Then pass the login info to you.

Pretty much every bench I seen you post here fallows there rules so I could just pull them off Vogons and post them on your behalf. I just don't want to post them without your permission.

Although.. unless you're near the top spot for a specific video card/platform there's literally no point in ever submitting there if you're just going to fall down in to the rest of the rabble for that platform. hwbot is all about records and being the very best globally, that's kind.. the entire point of the website. Overclock records. If you aren't setting one there's no point in submitting anything. I'd have to look, but not sure if any of his submissions are record-breaking or even close or not. I have a lot of computers of different era's around here and I've benchmarked em lots of times, and usually overclock the hell out of em as far as I can and still be stable. Even with all of that.. I rarely ever get to submit anything to hwbot though, most of my systems just aren't good enough vs what's already on there.

Nope. Well that's how it works for personal hw points. But only for none popular hardware. They changed the points system a good bit. MR. Horse got 6.x points for a OC with a q6600 that came in 600 something place. But I came in fist with a 750mhz PIII and only gotten 2 points.

Team points or OC events are done differently. True the best scores get the most points, but any submitted bench helps. The more your team benchmarks the more points your team ranks up.

And still there are not that many benchmarkers that submit scores with older hardware. It's not hard to brake records with anything from the late 90's. It's around 2005 that people start to really benchmark hardware.

Reply 617 of 804, by kithylin

User metadata
Rank l33t
Rank
l33t
Jade Falcon wrote:

Nope. Well that's how it works for personal hw points. But only for none popular hardware. They changed the points system a good bit. MR. Horse got 6.x points for a OC with a q6600 that came in 600 something place. But I came in fist with a 750mhz PIII and only gotten 2 points.

Team points or OC events are done differently. True the best scores get the most points, but any submitted bench helps. The more your team benchmarks the more points your team ranks up.

And still there are not that many benchmarkers that submit scores with older hardware. It's not hard to brake records with anything from the late 90's. It's around 2005 that people start to really benchmark hardware.

Was the change recent? I may have to go make the site re-calculate all my scores... time to go browse their site and try to find info on changes then. Thanks for the heads up.

Reply 618 of 804, by Jade Falcon

User metadata
Rank BANNED
Rank
BANNED

It's been changed not long ago if I recall. But it's not all that big of a difference from the last change for personal HW points if I recall. I don't fallow the points system all that close. I only care about cups and team points.
I
They removed the score recalculate button in subs, you have to go to your profile setting page to force recalculate everything.

The HW points system gives more points the more submitted score with the hardware. If your after HW points bench popular hardware. You don't breed to even be in the top 100 to rank in the points with some hardware.

Reply 619 of 804, by Carlos S. M.

User metadata
Rank Oldbie
Rank
Oldbie

I ran 3DMark2001 SE on a custom P3-S 1.4 GHz build with a Gigabyte GA-6OXT and a Geforce 4 ti 4600, all hardware is runing at stock clocks

What is your biggest Pentium 4 Collection?
Socket 423/478 Motherboards with Universal AGP Slot
Socket 478 Motherboards with PCI-E Slots
LGA 775 Motherboards with AGP Slots
Experiences and thoughts with Socket 423 systems