Is my 3DMark2001 score a bit low?

[Enverex]

RAM timings are nothing to do with your processor. The better your RAM is the higher speed you can get out of it, with better timings. Changing that to - 2-2-2-2 should give you quite a big increase in speed if your RAM can handle it....

NS

[quarryman]

2-2-2-2

?!! The 7 can be knocked down to 2? Anyone know a good FAQ on RAM timings and what they do? I'd like to reduce them if the performance increase is worth it. My RAM is OCZ 3200 which should be good for o/cing.

[bouncin]

might be worth knocking them down to 2,2,2,6 or maybe even try 2,2,2,11 but 2,2,2,2 would actually slow the memory as the it would not have enough time to read the data and would make a lot of mistakes and then keep repeating the cycle until it was done.

[Enverex]

My old machine was set to 2-2-2-7 by default, and I thought it was strange that it let me set it so low, but it did, and I benched it set to 2 and I got a huge increase in speed. Not sure if it is just good RAM or something, but it seemed to work for me...

NS

[bouncin]

the ram must have been awesome cos tbh they are the lowest timings i have ever seen in a desktop pc. But if it worked for ya fairplay. btw what ram was it?

[Enverex]

Some crappy 128MB stick (now defective) that I picked up ages ago for my Athlon 500mhz System (Slot A). I was supprised it worked, especially when my score went up so much.

Bouncin - Are you in Stourbridge?

NS

[bouncin]

pretty close to stourbridge about 2 miles away

[chrisf6969]

From the SPD timings that its on, you probably have this stuff...
http://www.ocztechnology.com/product...r+Dual+Channel

So I would probably leave it there.

You could try it at 2-3-3-7 as that would give you a nice boost.

You probably have an old version of 3dMark2k1.
Have you downloaded the latest patch for it?
http://www.futuremark.com/download/?...001patch.shtml

That will probably fix you up. If not check windows update for any DX9.0 patches/upgrades. Then uninstall 3dMark and download the latest version of the full install, and reinstall.

I want to see the performance boost!! (and links!)

[Austin]

3Dmark2001 is very quirky. Do ensure you have the latest build (you can get a 5MBish update file). If still problematic try reinstalling it (or uninstalling then reinstalling). I'm sure that will get it to run. As for RAM timings here's my usual spin ...

The CL rating (Cas Latency) is usually quoted with the designed speed of the RAM (eg 333mhz for PC2700) but this and the numbers which (should) follow can be daunting.

RAM is effectively arranged in rows and columns (much like a spreadsheet) so then CAS is Column Address Strobe, RAS is Row Address Strobe. The RAS, CAS and the RAS-to-CAS latency are basically the time spent waiting to locate, read or write a given area of memory but if a consecutive block of memory is needed the significance of any latency is pretty minimal, so latency is of greatest importance when using memory randomly.

The problem is all that's often quoted about latency or RAM timings is the CL number, it's the most important but less than 1/5th of the story. If you line the timings all up you can state a more complete picture. You should get quoted CL2.0 as being 2-2-2-5 (but could also be 2-3-3-7) and CL2.5 as being 2.5-3-3-7. The 1st number is Cas Latency, 2nd RAS Precharge, 3rd RAS-CAS, 4th Act Precharge (IIRC). BIOS' may use diff notations but that's basically the principle and lower is always faster as it means less waiting. Other things to look out for are Command: T1 (faster) or T2 (normal) and Bank Mode: off, 2 (fast) or 4 (faster). Anything to do with Memory holes, caches (not L1 or L2), shadowing and the like should be disabled. AGP Aperture is best at 256MB but there isn't (technically) really a perf hit in going lower, it's only assigned if the system has it free and your gfx card's own RAM 'runs out'. Even though some of these settings enhance perf they can actually hold back the attainable speed (in mhz) so it's worth experimenting.

As always there are many factors, CL2.0 is NOT 25% faster in real world performance, even when you isolate the RAM in a test. When you o/c, you can raise some of these 'latency' settings in order to run at a higher mhz. So you may prefer 2-3-3-7 400mhz over the default 2-2-2-5 333mhz for example.

These figures are from back in the KT333 era. Anyway at the same clockspeed 2-2-2-5 is rougly 1.5% faster in Sandra's RAM Benchmark than 2.5-3-3-7. It does depend on what you do with the RAM as to the exact perf diff. Sysmark2000 shows a 4% improvement. Quake3 & 3Dmark2001 show 3% improvement. 2-3-3-7 gives pretty much the same perf as running 2.5-2-2-5. It seems modern perf shows bigger gains esp for SktA systems, anyway the most vital number is the first number, CL.

If adding a new DIMM to an existing DIMM rem that the limit and timings are dictated by the slower DIMM. 256MB DDR333 CL2.0 coupled with 256MB DDR266 CL2.5 will both effectively be limited to DDR266 CL2.5. Cas Latency should be selectable in the BIOS, SPD is automatic, if you don't find many settings for latency then try the latest BIOS or CTRL+F1 in the BIOS on Gigabyte mobos. If you do raise the speed of your RAM past it's rated speed or lower the timings make 100% certain you are perfectly stable as the consequences can be dire.

I'm not saying any of the above is remotely definitive by any means, just what I've gathered from notes I've made coupled with experience.

[chrisf6969]

WAY TOO MUCH INFO!! I feel like Im in school!

One more note: timings do make a notably large difference on P4 dual channel systems with FSB running faster than the ram, b/c the FSB/CPU can soak up that extra bandwidth unlike some other systems. For example him.... he's maxed out on a shuttle at 250FSB (highest I believe BIOS lets you go) but his ram is running at 200Mhz DDR400 which should still allow for decent timings.

It has been shown like at PCstats.com's bandwidth vs. latency article. That its better to run at 5:4 than 1:1 or 3:2 at 250FSB.

[quarryman]

Ah yes. Its a little clearer now So even though I my be limited by the FSB in the shuttle (which I'm pretty sure I am) I can set the timings lower to get an extra boost. Interesting that you say it really gives a good increase.

Just to clarify this is the RAM that's in my rig. You might be able to tell what I can achieve by the specs there. Either way I'll be downloading an updated 3DMark2001 tonight and will retest the system.

Then I'll overclock as much as I can (including the RAM timings) and re-test. Fingers crossed.

When this is done what should I use to test if the system is stable? I understand that booting to desktop isn't necessarily evidence that it'll run solid from then on.

Also, should I be concerned about temps when it is overclocked? The shuttle is a little furnace when it gets going. I'm planning on buying a few fans to reduce the noise but also cool it better but I'd imagine I'll have to use Speedfan to keep an eye on things.

[Austin]

According to the specs your RAM is designed to run at 2.6v (2.5v is more normal) and 2x200=400mhz or DDR400-PC3200 with 2-2-3-5 timings. You should find if running at 333mhz the timings will tighten and you'll get perf closer to if you were running at 400mhz. As usual you want to keep RAM synced to FSB as much as is possible.

To check you're stable it's best to make a DOS boot floppy disk with memory testing sw on it. Most people use Memtest86, using a boot disk means that if your RAM is not stable it won't cause corruption inside Windows. However since Windows handles RAM differently to a DOS program you can still experience instability inside Windows. You may want to run some Windows stress testing sw like Prime95 (torture test). It can be a long and tedious process, just be 100% certain you can run at the speed and timings you choose.

[quarryman]

As usual you want to keep RAM synced to FSB as much as is possible.

Is this what I'm not doing though? I thought that by setting the RAM to 333MHz I'm acheiving 5:4 ratio, ie not syncing them

[Austin]

Not so important on P4 if you can get a good FSB o/c and still keep the RAM running fast (as in close to DDR400-PC3200 speed). The main problem with running async comes from SktA mobos where many people run the RAM faster than the FSB which can often yield lower perf, only real exception being KT333/400 using 266FSB with asynced DDR333-PC2700 which does yield a very nice 5%ish boost IIRC. Running async should definitely not be a stability issue so long as your RAM does not run out of spec either with its pure mhz (or timings).

[chrisf6969]

the order for BIOS of the IC7 is 2-3-2-5

you're at DDR400 so you should be able to run perfectly with in spec. as 2-2-2-5 is the fastest that you can probably run keep it at DDR400 (5:4) ratio. called DDR333 option in BIOS.

and you can try 2-2-2-5, if not just put it at 2-3-2-5

[quarryman]

Ok 3DMark2001 is giving me hassle. Everytime I go to run it now it tells me I need to install Directx 8.1 or higher and all the correct drivers.

I am bang up to date on all the drivers. Cat 3.9, Directx 9.0b. On the advice of some people I even downloaded the most recent version of 3DMark2001. I run it and the message about drivers just pops up.

It tells me it has stored the error in a log file which I'll paste below. It seems to think I don't have dx8.0 installed? I could dig it up on disk but what help would it be installing it over 9.0b?

I'm posting this because it might help solve a few other problems I'm having.

Morrowind seems to crash as soon as I try to leave the boat at the beginning of the game. This problem has been documented and a solution has been suggested has been given but it hasn't worked for me yet.

Call of Duty demo crashes to desktop at a random point during play. I am informed that the VPU has been reset as it wasn't responding to driver commands and could not be recovered. Sometimes I can alt-tab out and close the error and keep playing, othertimes not. This error is documented as well and I will try it but not sure if it will work for the demo.

Although it seems these errors might be able to be fixed it seems problems have been popping up very soon since I have barely got the system 2 weeks. Even my TNT2 was better than this. My old Dell could handle anything I threw at it. God I miss those days

Application info
----------------

Name 3DMark2001 SE
Build 300

Error log
---------

DirectX8 not installed.

System info available: yes
Project available: no



System info
-----------

System Info Version 2.2
Installation ID 0x00000000
OEM ID

CPU INFO
CENTRAL PROCESSING UNITS
Manufacturer Intel
Family Pentium® 4
Internal Clock 2.54 GHz
Internal Clock Maximum 2.99 GHz
External Clock 200 MHz
Socket Designation Socket 478
Type <unknown>
Upgrade ZIF Socket
Capabilities MMX, CMov, RDTSC, SSE, SSE2
Version Intel Pentium(R) 4
CPUID 0x00000f29

CACHES
Level 1
Capacity 20 KB
Type Internal
Type Details Synchronous
Error Correction Type <unknown>
Associativity <unknown>

Level 2
Capacity 512 KB
Type External
Type Details Synchronous
Error Correction Type <unknown>
Associativity <unknown>

DIRECTX INFO
DirectX Version 8.0

DIRECTDRAW INFO
DirectDraw Version 5.3.0000000.900

DISPLAY DEVICES
Description Primary Display Driver
Manufacturer ATI Technologies Inc.
Name RADEON 9600 SERIES
Total Local Video Memory 128 MB
Total Local Texture Memory 128 MB
Total AGP Memory 54 MB
Display Driver ati2dvag.dll
Display Driver Version 6.14.10.6396
Driver WHQL Certified Yes
Max Texture Width 2048
Max Texture Height 2048
Max User Clipping Planes 6
Max Active Hardware Lights 8
Max Texture Blending Stages 8
Textures In Single Pass 8
Vertex Shader Version 1.1
Pixel Shader Version 1.4
Max Vertex Blend Matrices 37
Max Texture Coordinates 8
Vendor ID 0x1002
Device ID 0x4150
Sub-System ID 0x000210f1
Revision 0x00

DISPLAY MODES
//leaving this out

TEXTURE FORMATS
//leaving pixel formats out.

3D Capabilities AGP Texturing, Hardware Transform and Lighting, Positional Lights, Subpixel Accurate Rasterizing, Stencil Buffers, Range Fog, Table Fog, Vertex Fog, W-Fog, Specular Gouraud Shading, Anisotropic Filtering, Bilinear Filtering, Point Sampling, Trilinear Filtering, Additive Texture Blending, Dot3 Texture Blending, Multiplicative Texture Blending, Subtractive Texture Blending, Environmental Bump Mapping, Environmental Bump Mapping With Luminance, Cube Mapping, Factor Alpha Blending, Vertex Alpha Blending, Texture Alpha Blending, Texture Clamping, Texture Mirroring, Texture Wrapping, Guard Band Support, Rendering to a Window, N Patches, Mipmap LOD Bias Adjustment, Projected Textures, Volume Textures, Point Primitive Support, Full-Screen Anti-Aliasing, DXT Compressed Textures

DIRECTSHOW INFO
DirectShow Version 6.05.01.0900

REGISTERED DIRECTSHOW FILTERS
........more here
Filter ASX v.2 file Parser
Filter AVI Decompressor
Filter AVI Draw
Filter AVI Splitter
Filter AVI/WAV File Source
Filter BDA MPEG2 Transport Information Filter
Filter Color Space Converter
Filter DV Muxer
Filter DV Splitter
Filter DV Video Decoder
Filter Default DirectSound Device
Filter Default MidiOut Device
Filter DivX Decoder Filter
Filter DivXAntiFreeze
Filter File Source (Async.)
Filter File Source (Netshow URL)
....here....

Filter MPEG Layer-3 Decoder
Filter MPEG Video Decoder
Filter MPEG-2 Demultiplexer
Filter MPEG-2 Sections and Tables
Filter MPEG-2 Splitter
Filter MPEG-I Stream Splitter
Filter Microsoft MPEG-4 Video Decompressor
Filter Microsoft Screen Video Decompressor
Filter Morgan Stream Switcher
Filter Multi-file Parser
Filter NSC file Parser
Filter Ogg Multiplexer
Filter Ogg Splitter
Filter Overlay Mixer2
Filter QT Decompressor
Filter QuickTime Movie Parser
Filter SAMI (CC) Parser
Filter VBI Surface Allocator
Filter VGA 16 Color Ditherer
Filter Video Port Manager
Filter Video Renderer
Filter Vorbis Decoder
Filter WM ASF Reader
Filter WM ASF Writer
Filter WMP CD Filter
Filter WST Decoder
Filter Wave Parser
Filter Windows Media Audio Decoder
Filter Windows Media Multiplexer
Filter Windows Media Pad VU Data Grabber
Filter Windows Media Update Filter
Filter Windows Media Video Decoder
Filter Windows Media source filter
Filter XML Playlist
Filter XviD MPEG-4 Video Decoder

MEMORY INFO
Total Physical Memory 512 MB
Free Physical Memory 371.55 MB
Total Pagefile Memory 1.22 GB
Free Pagefile Memory 1.13 GB

MEMORY ARRAYS
Max Module Capacity 2 GB
Location System Board
Use System
Supported Error D&C None
Supported Speeds <unknown>
Supported Types <unknown>
Supported Voltages <unknown>

MEMORY SLOTS
Installed & Enabled Size 256 MB
Form Factor DIMM
Frequency 0 Hz
Slot A0 [Bank0/1]
Manufacturer
Type SDRAM
Type Details Synchronous
Enabled Size 256 MB
Total Bit Width 64
Data Bit Width 64

Installed & Enabled Size 0 bytes
Form Factor DIMM
Frequency 0 Hz
Slot A1 [Bank2/3]
Manufacturer
Type <unknown>
Type Details <unknown>
Enabled Size 0 bytes
Total Bit Width 0
Data Bit Width 0

MOTHERBOARD INFO
Manufacturer Shuttle Inc
Model FB61R
Version N/A
BIOS Vendor Phoenix Technologies, LTD
BIOS Version 6.00 PG
BIOS Release Date 09/17/2003
BIOS Properties Plug and Play, Flash, AGP

CARD SLOTS
Number 1
Designation PCI0
Type PCI
Characteristics 5.0V, PME Signal
Data Bus Width 32
Details Available, Long
Device Class <unknown>
Device Name
Device Manufacturer
Device Driver Date
Device Vendor ID 0x0000
Device ID 0x0000
Device Sub-System ID 0x00000000
Device Revision 0x00
IRQ 0

Number 2
Designation PCI1
Type PCI
Characteristics 5.0V, PME Signal
Data Bus Width 32
Details In Use, Long
Device Class Network Adapter
Device Name Realtek RTL8139 Family PCI Fast Ethernet NIC
Device Manufacturer Realtek
Device Driver Date
Device Vendor ID 0x10ec
Device ID 0x8139
Device Sub-System ID 0x813910ec
Device Revision 0x10
IRQ 18

Number 3
Designation PCI2
Type PCI
Characteristics 5.0V, PME Signal
Data Bus Width 32
Details In Use, Long
Device Class 1394 BUS Controller
Device Name VIA OHCI Compliant IEEE 1394 Host Controller
Device Manufacturer VIA
Device Driver Date
Device Vendor ID 0x1106
Device ID 0x3044
Device Sub-System ID 0x30441106
Device Revision 0x80
IRQ 20

Number 4
Designation PCI3
Type PCI
Characteristics 5.0V, PME Signal
Data Bus Width 32
Details Available, Long
Device Class <unknown>
Device Name
Device Manufacturer
Device Driver Date
Device Vendor ID 0x0000
Device ID 0x0000
Device Sub-System ID 0x00000000
Device Revision 0x00
IRQ 0

Number 0
Designation AGP
Type AGP
Characteristics 5.0V
Data Bus Width 32
Details In Use, Long
Device Class Display Adapter
Device Name RADEON 9600 SERIES
Device Manufacturer ATI Technologies Inc.
Device Driver Date
Device Vendor ID 0x1002
Device ID 0x4150
Device Sub-System ID 0x000210f1
Device Revision 0x00
IRQ 16

SYSTEM DEVICES
Description Intel(R) 82801EB PCI Bridge - 244E
Manufacturer Intel
Vendor ID 0x8086
Device ID 0x244e
Subsystem ID 0x00000000
Revision 0xc2

Description Intel(R) 82801EB LPC Interface Controller - 24D0
Manufacturer Intel
Vendor ID 0x8086
Device ID 0x24d0
Subsystem ID 0x00000000
Revision 0x02

Description Intel(R) 82801EB SMBus Controller - 24D3
Manufacturer Intel
Vendor ID 0x8086
Device ID 0x24d3
Subsystem ID 0xfb611297
Revision 0x02

Description Intel(R) 82865G\PE\P Processor to I/O Controller - 2570
Manufacturer Intel
Vendor ID 0x8086
Device ID 0x2570
Subsystem ID 0x00000000
Revision 0x02

Description Intel(R) 82865G\PE\P Processor to AGP Controller - 2571
Manufacturer Intel
Vendor ID 0x8086
Device ID 0x2571
Subsystem ID 0x00000000
Revision 0x02

Revision 3.0
Rate 4x, 8x (2x enabled)
Aperture Size 3.81 GB
Sideband Addressing supported (enabled)
Fast Write supported (enabled)

MONITOR INFO
MONITORS
Name Dell M770
Manufacturer Dell Computer Corp.
Max Width 0
Max Height 0

POWER SUPPLY INFO
BATTERIES

OPERATING SYSTEM INFO
Operating System Microsoft Windows XP
Version 5.1.2600
Service Pack
Locale IE
Desktop Width 1024
Desktop Height 768
Desktop BPP 32

HARD DISK INFO
HARD DISKS
Name IBM-DTTA-351010
Manufacturer (Standard disk drives)
Capacity 9.44 GB
Model IBM-DTTA-351010
Interface Type IDE

Name ST360021A
Manufacturer (Standard disk drives)
Capacity 55.90 GB
Model ST360021A
Interface Type IDE