• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Benchmarking: 3850AGP and Q6600

Joined
Aug 30, 2006
Messages
7,247 (1.06/day)
System Name ICE-QUAD // ICE-CRUNCH
Processor Q6600 // 2x Xeon 5472
Memory 2GB DDR // 8GB FB-DIMM
Video Card(s) HD3850-AGP // FireGL 3400
Display(s) 2 x Samsung 204Ts = 3200x1200
Audio Device(s) Audigy 2
Software Windows Server 2003 R2 as a Workstation now migrated to W10 with regrets.
This is a basic benchmarking template to compare performance improvements when upgrading an AGP system.
  1. Direct comparison showing UPGRADES from Radeon 9700 to 3850 with same CPU and same DDR RAM, and
  2. Direct comparison showing UPGRADES from P4 to Q6600 with same GPU and same DDR RAM. Note that the Q6700 results are from revin. Also on Conroe865PE, but his system is based on "overclocked" and not stock GPU.
Blanks and ?'s will be filled in when I get time to run the benchmarks (and data from other users :) )
The "..." means no material change (since CPU and not GPU test).

System specs are as follows:
s478 tests using ASUS P4P800-VM
s775 tests using ASROCK Conroe865PE or 775i65G
CPU and speed as shown in table
GPU at stock speeds, except "HD 3850 OC" which is at 750/1000 and "7800GS+ OC" which is at 670/800
Mainboard RAM is dual channel DDR at stock (default)


P.S. These benchmarks are NOT memory-speed optimised or "overclocked" optimised. A few % here and there in not the point of this benchmark comparison. We are looking for order-of-magnitude change due to CPU and GPU swaps.

Capture023666.jpg
 
Last edited:
Thank you for the time/effort for the data!:toast:

If it is possable, could you break down the CPU/SM2/SM3 data?

I, myself like to see a look at what CPU's add to ie 3DM06, and gauge just what my SM2/3 video score's compare to other cards.
Especially what those 3850's are doing:rockout:

Again, a BIG Thank you
 
so are you using the 4coredual for the testing
 
for the 478 sys what are complete specs plase?


:toast:
 
s478 spec as follows:
STOCK CPUs. STOCK DDR. Stock GPU.

P4HT Northwood 2.8@2.8 on ASUS P4P800-VM
P4EE Galatin 3.2@3.2 on ASUS P4P800-VM
Q6600 on ASROCK 775i65G
DDR1 dual channel stock speeds

I will add more s478 benchmarks when I get time ;)

***

I used the tool 3dmark06 score calculator: http://www.xtremesystems.org/forums/showthread.php?t=102058 to test if the Q6700 is a good upgrade over Q6600 on my HD3850 system. But calculator suggests only extra 300 points in 3Dmark06 if I get the Q6700. So I'll sit still for the moment. A 3850x2 AGP would be a treat!

***

Lightmark 2007 added! LINK http://forums.techpowerup.com/showthread.php?t=59445
 
Ok yesterday I reached 12054 points at 3Dmarks06

FSB = 294Mhz
Multiplier = 11
RAM = 2 - 3 - 3 - 6

In Bios I've activated Spread Spectrum or 3Dmarks06 crash
Enhalt ...........
Virtualization ..............

All the rest in the same screen disabled

I also overclock my 3850 to 802/1002

I think I can get a little more, any suggestion?
 
Guys new record 12341!!! by overclocking a little more my HIS, WOW :rockout:
 

Attachments

  • 3Dmarks6 Record.jpg
    3Dmarks6 Record.jpg
    31.4 KB · Views: 613
I had to repeat the test to grab also the details and ...........

Same config.

x11 Multiplier

294Mhz FSB

His 3850Agp @824/1035

12349 points!!!! New Record
 

Attachments

  • 3Dmarks6 Record 2b.jpg
    3Dmarks6 Record 2b.jpg
    31.7 KB · Views: 554
Very interesting numbers .. especially the 3Dmark 03 , showing that scalability in comparison with the CPU type , is not that huge .

In reverse, the 3Dmark 06 score , its like a large Neon sign ... yelling Upgrade e e :D
In real life i prefer true fps deference .

But i do not afraid to use 3Dmark 06 ... i am almost good at scores :nutkick:
 

Attachments

  • 3dmark-06-3850_34.jpg
    3dmark-06-3850_34.jpg
    117.3 KB · Views: 602
Aqua mark from my system .
NX 6800GT
DisplayWidth: 1024
DisplayHeight: 768
DisplayDepth: 32
AntialiasingMode: 0
AntialiasingQuality: 0
AnisotropicFiltering: 4
DetailLevel: 4
AvgFPS: 59.349174
MinFPS: 25.072353
MaxFPS: 139.165009
AvgFPSRender: 85.708649
AvgFPSSimulation: 192.982971
AvgTrianglesPerSecond: 17864858
MinTrianglesPerSecond: 3123980
MaxTrianglesPerSecond: 39.153.842
AquamarkScoreRender: 8571
AquamarkScoreSimulation: 9649
AquamarkScore: 59349

HD3850

DisplayWidth: 1024
DisplayHeight: 768
DisplayDepth: 32
AntialiasingMode: 0
AntialiasingQuality: 0
AnisotropicFiltering: 4
DetailLevel: 4
AvgFPS: 75.911545
MinFPS: 47.290668
MaxFPS: 140.000000
AvgFPSRender: 119.096458
AvgFPSSimulation: 209.314163
AvgTrianglesPerSecond: 22850342
MinTrianglesPerSecond: 2959034
MaxTrianglesPerSecond: 102.509.923
AquamarkScoreRender: 11910
AquamarkScoreSimulation: 10465
AquamarkScore: 75911

Not bad ... :)

My new HD3850 AGP does boost things , even with a single piston engine. :p
 
Real world game performance is much better than Aquamark and Furmark suggest (they are OpenGL based and nVidia is the king of OpenGL)
 
About Aqua mark.

From my point of view , what i keep as hard evidence about the performance of my two cards, are the Max Triangles Per Second.

The Max Triangles Per Second its in my eyes ,
the evidence of true performance between the cards .

The HD3850 haves triple power here. ;)
 
Only test I've run as of now....
 

Attachments

  • Image2.jpg
    Image2.jpg
    108.3 KB · Views: 576

Attachments

  • lighstmar2008.jpg
    lighstmar2008.jpg
    47 KB · Views: 528
  • lighstmar2008-2.jpg
    lighstmar2008-2.jpg
    48.4 KB · Views: 566
  • lighstmar2008-3.jpg
    lighstmar2008-3.jpg
    47.8 KB · Views: 505
  • lighstmar2008-4.jpg
    lighstmar2008-4.jpg
    47.9 KB · Views: 539
Last edited:
our systems are basically the same? How do you double my score? :banghead:
 
.....
 

Attachments

  • atlas1.jpg
    atlas1.jpg
    56.9 KB · Views: 617
  • atlas2.jpg
    atlas2.jpg
    43.9 KB · Views: 682
  • atlas3.jpg
    atlas3.jpg
    40.1 KB · Views: 619
........ :banghead:
 

Attachments

  • CPUZZ.jpg
    CPUZZ.jpg
    93.8 KB · Views: 518
Well i was had your board , four years back , sold it and got the P4C800E , for two reasons.
1) Locks AGP PCI timings
2) Its a damn expensive I875 board.

The MSI cost as 70 EUR , and the ASUS 200 EUR .
 
Well, to upgrade to a new P4 socket 478 board right now would be useless considering I could use that money and get a MB, PCI-E Graphics card, Memory and CPU for just a bit more than it would cost to buy a new MB. Funny thing, games play fine but it benchmarks like crap. :confused:
 
Benchmarks are not games , thats true .

Keep the system as is , and be happy.

When i was building my system, i had in mind top of the cream hardware,
so to really last in time.

And it does.

But i had pay for it an extreme amount of money , the OCZ 2x512 cost me as 300 EUR . (27/12/04)

If it was not perform so extremely good , i would had set it on fire with petrol , and enjoy the view of the flames. ;)



.
 
Last edited:
OK, I've updated the scoreboard for Lightsmark 2008, and added ssome new CPUs for non-graphics test.
 
Fumark in the play . :)
 

Attachments

  • atlas4.jpg
    atlas4.jpg
    27.9 KB · Views: 578
just downloaded it and ran it. Also ran 3dmark 05' , everything stock except CPU @ 3.4GHz
 

Attachments

  • furMark.jpg
    furMark.jpg
    63.3 KB · Views: 541
  • futuremark05.jpg
    futuremark05.jpg
    172.2 KB · Views: 507
Last edited:
update: Furmark
 

Attachments

  • furmark2.jpg
    furmark2.jpg
    64.3 KB · Views: 523
Back
Top