I must be going NUTS by now

Hi again everyone. This is my 3rd post regarding a performance issue
I'm having. Please read on!

Everyone who has helped me up to now: thank you. I'm a couple of steps
closer to the cause. Please help me solve this thing.

--- THE ISSUE - A RECAP ---
From a "cold" boot, the performance of my system is approx. 40% lower
than after a reboot (tested with 3DMark05, 3Dmark06 and F.E.A.R.). But
also a programs like Sony Vegas Video shows the huge performance
difference.

Rebooting right after POST, will also result in a 40% performance
increase. So, it does not seem to have to do with Windows or drivers.
But something with the BIOS or hardware. Windows does not report any
hardware conflicts or faulty devices.

--- THE MEASUREMENTS ---
I've measured a great many things, hoping that someone is able to
track the cause from there. I've used the following programs: CPU-Z,
RivaTuner, 3DMark05 and SiSoft Sandra. Note the huge differences in
the 3DMark measurements, exemplary for F.E.A.R., other games and Sony
Vegas Video 6.0.

Cold boot Re-boot
CPU speed (MHz) 2200 2200
FSB (MHz) 200 200
GPU (MHz) 300 300
VRAM (MHz) 700 700

3DMark05 (3DMarks) 2400 3700
CPU test 1 (FPS) 1,9 3,0
CPU test 2 (FPS) 2,2 4,4

MIPS 19637 19660
MFLOPS 8947 8880
Integer 41263 41286
Floating point 44804 45023

RAM Int (MB/s) 4314 4271
RAM Float (MB/s) 4305 4255
Cache indexed (MB/s) 5272 5238

Notice that 3DMark reports CPU performance issues after a cold boot
where Sandra reports that the CPU is performing normally in both the
cold boot as the re-boot case.

--- TRIED SO FAR, BUT DID NOT WORK ---
* Drivers, patches, hotfixes, about everything that was recommended to
me in earlier threads, but all Windows related.
* Various BIOS settings:
Primary graphics adapter [pci][agp]
Search for MDA resources [yes][no]
VLink 8x supported [enabled][disabled]
AGP mode [4x][8x]
AGP 3.0 calibration cycle [enabled][disabled]
DBI output for AGP trans [enabled][disabled]
AGP aperture [32][64][128][256]
Cool & quiet [enabled][disabled]
Busspeed [AUTO][200 MHz]
* Two BIOS versions 0219 (latest final version) and 0221 (latest beta
version).

--- MY SYSTEM ---
* Asus A8V, VIA K8T800Pro chipset, BIOS versions 0219 and 0221
* AMD Athlon64 X2 4200+
* 2 x 1GB dual channel Corsair Value RAM
* AOpen 6800LE AGP8x
* 150 GB WD Raptor & 200 GB WD Caviar
* everything at stock speed

--- PLEASE HELP ---
Anyone any suggestions?

Arend van Beek wrote:
Hi again everyone. This is my 3rd post regarding a performance issue
I'm having. Please read on!

Everyone who has helped me up to now: thank you. I'm a couple of steps
closer to the cause. Please help me solve this thing.

--- THE ISSUE - A RECAP ---
From a "cold" boot, the performance of my system is approx. 40% lower
than after a reboot (tested with 3DMark05, 3Dmark06 and F.E.A.R.). But
also a programs like Sony Vegas Video shows the huge performance
difference.

Rebooting right after POST, will also result in a 40% performance
increase. So, it does not seem to have to do with Windows or drivers.
But something with the BIOS or hardware. Windows does not report any
hardware conflicts or faulty devices.

--- THE MEASUREMENTS ---
I've measured a great many things, hoping that someone is able to
track the cause from there. I've used the following programs: CPU-Z,
RivaTuner, 3DMark05 and SiSoft Sandra. Note the huge differences in
the 3DMark measurements, exemplary for F.E.A.R., other games and Sony
Vegas Video 6.0.

Cold boot Re-boot
CPU speed (MHz) 2200 2200
FSB (MHz) 200 200
GPU (MHz) 300 300
VRAM (MHz) 700 700

3DMark05 (3DMarks) 2400 3700
CPU test 1 (FPS) 1,9 3,0
CPU test 2 (FPS) 2,2 4,4

MIPS 19637 19660
MFLOPS 8947 8880
Integer 41263 41286
Floating point 44804 45023

RAM Int (MB/s) 4314 4271
RAM Float (MB/s) 4305 4255
Cache indexed (MB/s) 5272 5238

Notice that 3DMark reports CPU performance issues after a cold boot
where Sandra reports that the CPU is performing normally in both the
cold boot as the re-boot case.

--- TRIED SO FAR, BUT DID NOT WORK ---
* Drivers, patches, hotfixes, about everything that was recommended to
me in earlier threads, but all Windows related.
* Various BIOS settings:
Primary graphics adapter [pci][agp]
Search for MDA resources [yes][no]
VLink 8x supported [enabled][disabled]
AGP mode [4x][8x]
AGP 3.0 calibration cycle [enabled][disabled]
DBI output for AGP trans [enabled][disabled]
AGP aperture [32][64][128][256]
Cool & quiet [enabled][disabled]
Busspeed [AUTO][200 MHz]
* Two BIOS versions 0219 (latest final version) and 0221 (latest beta
version).

--- MY SYSTEM ---
* Asus A8V, VIA K8T800Pro chipset, BIOS versions 0219 and 0221
* AMD Athlon64 X2 4200+
* 2 x 1GB dual channel Corsair Value RAM
* AOpen 6800LE AGP8x
* 150 GB WD Raptor & 200 GB WD Caviar
* everything at stock speed

--- PLEASE HELP ---
Anyone any suggestions?


A silly question perhaps, does Windows Task Manager show two cores
running at the same performance level?

On Sun, 07 May 2006 13:09:52 -0400, Ray
wrote:

A silly question perhaps, does Windows Task Manager show two cores
running at the same performance level?

As far as SiSoft Sandra goes: yes. Both cores are at 100%.

As far as 3DMark05 goes: no. One core is at 100%, the other core at
about 10-20%. This is the same for the cold-boot as the re-boot
situation.

Arend van Beek wrote:
Hi again everyone. This is my 3rd post regarding a performance issue
I'm having. Please read on!

Everyone who has helped me up to now: thank you. I'm a couple of steps
closer to the cause. Please help me solve this thing.

--- THE ISSUE - A RECAP ---
From a "cold" boot, the performance of my system is approx. 40% lower
than after a reboot (tested with 3DMark05, 3Dmark06 and F.E.A.R.). But
also a programs like Sony Vegas Video shows the huge performance
difference.

Rebooting right after POST, will also result in a 40% performance
increase. So, it does not seem to have to do with Windows or drivers.
But something with the BIOS or hardware. Windows does not report any
hardware conflicts or faulty devices.

--- THE MEASUREMENTS ---
I've measured a great many things, hoping that someone is able to
track the cause from there. I've used the following programs: CPU-Z,
RivaTuner, 3DMark05 and SiSoft Sandra. Note the huge differences in
the 3DMark measurements, exemplary for F.E.A.R., other games and Sony
Vegas Video 6.0.

Cold boot Re-boot
CPU speed (MHz) 2200 2200
FSB (MHz) 200 200
GPU (MHz) 300 300
VRAM (MHz) 700 700

3DMark05 (3DMarks) 2400 3700
CPU test 1 (FPS) 1,9 3,0
CPU test 2 (FPS) 2,2 4,4

MIPS 19637 19660
MFLOPS 8947 8880
Integer 41263 41286
Floating point 44804 45023

RAM Int (MB/s) 4314 4271
RAM Float (MB/s) 4305 4255
Cache indexed (MB/s) 5272 5238

Notice that 3DMark reports CPU performance issues after a cold boot
where Sandra reports that the CPU is performing normally in both the
cold boot as the re-boot case.

--- TRIED SO FAR, BUT DID NOT WORK ---
* Drivers, patches, hotfixes, about everything that was recommended to
me in earlier threads, but all Windows related.
* Various BIOS settings:
Primary graphics adapter [pci][agp]
Search for MDA resources [yes][no]
VLink 8x supported [enabled][disabled]
AGP mode [4x][8x]
AGP 3.0 calibration cycle [enabled][disabled]
DBI output for AGP trans [enabled][disabled]
AGP aperture [32][64][128][256]
Cool & quiet [enabled][disabled]
Busspeed [AUTO][200 MHz]
* Two BIOS versions 0219 (latest final version) and 0221 (latest beta
version).

--- MY SYSTEM ---
* Asus A8V, VIA K8T800Pro chipset, BIOS versions 0219 and 0221
* AMD Athlon64 X2 4200+
* 2 x 1GB dual channel Corsair Value RAM
* AOpen 6800LE AGP8x
* 150 GB WD Raptor & 200 GB WD Caviar
* everything at stock speed

--- PLEASE HELP ---
Anyone any suggestions?

On the A7N8X there has been similar problems.

http://www.nforcershq.com/forum/real...ht=pci+nforce2

On my computer from time to time after a fresh start, computer clock will
not keep time, it will speed up. At the same time, if I check CPU speed with
MBM5, it will show variable, lower speed in MHz. If I use other cpu-speed
utilities, some will show normal, some will show that there is something
wrong. In this state, ASIO sound output will skip, and game performance will
be degraded.
A reboot fixes it. It is more common at certain frequencies.

It is true that cpu clock is not affected per se, but there are stories of
some kind of interrupt storm being set up, hampering performance. MBM5
measures CPUspeed by giving the cpu a task to perform, other measuring tools
use other techniques. I think neither of those are wrong, they just measure
different things. All points towards an interrupt storm being set up from a
device not identifiable.

Could you try removing some hardware or maybe disabling some onboard stuff
to check?

On Sun, 7 May 2006 22:04:50 +0200, "Egil Solberg"
wrote:

Hi Egil

On the A7N8X there has been similar problems.

http://www.nforcershq.com/forum/real...ht=pci+nforce2

I doubt if this concerns the same issue?

It is true that cpu clock is not affected per se, but there are stories of
some kind of interrupt storm being set up, hampering performance. MBM5
measures CPUspeed by giving the cpu a task to perform, other measuring tools
use other techniques. I think neither of those are wrong, they just measure
different things. All points towards an interrupt storm being set up from a
device not identifiable.

Could you try removing some hardware or maybe disabling some onboard stuff
to check?

Before I take my PC apart: what is the cause? Would it mean the
motherboard is faulty?

For instance, what if the onboard LAN causes an interrupt storm, what
could I do about it?

On Sun, 07 May 2006 21:23:56 +0200, Arend van Beek
wrote:

On Sun, 07 May 2006 13:09:52 -0400, Ray
wrote:

A silly question perhaps, does Windows Task Manager show two cores
running at the same performance level?

As far as SiSoft Sandra goes: yes. Both cores are at 100%.

As far as 3DMark05 goes: no. One core is at 100%, the other core at
about 10-20%. This is the same for the cold-boot as the re-boot
situation.


Coco, Paul, John, Ray... please?

Arend van Beek wrote:

For instance, what if the onboard LAN causes an interrupt storm, what
could I do about it?

Disable it in BIOS, or if there is a jumper onboard, try that.

In article , Arend van Beek
wrote:

On Sun, 07 May 2006 21:23:56 +0200, Arend van Beek
wrote:

On Sun, 07 May 2006 13:09:52 -0400, Ray
wrote:

A silly question perhaps, does Windows Task Manager show two cores
running at the same performance level?

As far as SiSoft Sandra goes: yes. Both cores are at 100%.

As far as 3DMark05 goes: no. One core is at 100%, the other core at
about 10-20%. This is the same for the cold-boot as the re-boot
situation.


Coco, Paul, John, Ray... please?

I'm still waiting to see the test results from running
the following three programs.

http://www.xtremesystems.com/pi/super_pi_mod-1.5.zip
http://www.techpowerup.com/downloads...pi_mod-1.4.zip
http://www.overclock.net/attachments...p?d=1125204458

Run each program from the cold boot.

Run each program after a warm reboot.

Select the same precision, like 1 million digits.

The first two programs have the ability to detect time base
measurement problems. The programs check for a user attempt
at cheating, and the program will crash if cheating is
detected. The third program should always run, as it has
no cheat protection, it doesn't check the time base, and
the third program should always produce a result.

If all three programs, tested twice each, work without a
problem, then the problem is not a time base error. If the
programs fail to complete, then it suggests that the
computer cannot measure "real time" properly.

Running these tests does not mean I have a cure for your
symptoms. These tests are meant to further add to your
collected set of symptoms. At this time, I am unaware of
any application that does simple tests for time base accuracy,
by comparing the behavior of TSC with 8254, APIC timer, or
the RTC in the Southbridge.

So run these tests, and tell me the results. How many
seconds does each one take, as stated by the program ?
Does each program manage to complete 1 million PI digits ?
Or do the programs crash ?

Paul

On Wed, 10 May 2006 00:02:32 GMT, (Paul) wrote:

http://www.xtremesystems.com/pi/super_pi_mod-1.5.zip
http://www.techpowerup.com/downloads...pi_mod-1.4.zip
http://www.overclock.net/attachments...p?d=1125204458

Run each program from the cold boot.

Run each program after a warm reboot.

Select the same precision, like 1 million digits.

So run these tests, and tell me the results. How many
seconds does each one take, as stated by the program ?
Does each program manage to complete 1 million PI digits ?
Or do the programs crash ?

These are the results from the test.

All 12 runs completed. No errors or crashes. CPU usage on all tests:
50% (but with different distributions - which may or may not mean
anything).

--- Results COLD BOOT 1M digits ---

SuperPI 1.1 1st 42.000 2nd 42.000
SuperPI 1.4 1st 42.141 2nd 42.141
SuperPI 1.5 1st 42.063 2nd 42.063

CPU usage distribution:
core1 60%
core2 40%


--- Results WARM REBOOT 1M digits ---

SuperPI 1.1 1st 42.000 2nd 42.000
SuperPI 1.4 1st 42.297 2nd 42.296
SuperPI 1.5 1st 42.344 2nd 42.297

CPU usage distribution:
core1 80%
core2 20%


2ND REPLY FROM ASUS

Also, I received a second reply from ASUS.

"As I Know,WD Raptor is a SATA HDD,how do you install your system?
make a floopy driver disk and press"F6" when install WinXP? and try
0221beta BIOS.If this can not help you,could you please replace
another HDD to have a try?"

I've already answered them that the 0221 beta BIOS didn't solve the
problem. And that Windows was indeed installed with a drivers from a
disk (how else?). Could the issue be SATA or HDD related as ASUS
suggests?

Thanks in advance.

Arend van Beek wrote:
Hi again everyone. This is my 3rd post regarding a performance issue
I'm having. Please read on!

Everyone who has helped me up to now: thank you. I'm a couple of steps
closer to the cause. Please help me solve this thing.

--- THE ISSUE - A RECAP ---
From a "cold" boot, the performance of my system is approx. 40% lower
than after a reboot (tested with 3DMark05, 3Dmark06 and F.E.A.R.). But
also a programs like Sony Vegas Video shows the huge performance
difference.

Rebooting right after POST, will also result in a 40% performance
increase. So, it does not seem to have to do with Windows or drivers.
But something with the BIOS or hardware. Windows does not report any
hardware conflicts or faulty devices.

I read some of the other posts you have made on the topic, as well as the
A8V manual.
I see that you have gotten wise advice both in other threads here as well as
in the nvidia NG, with suggestions still to be tried.
It might seem banal, but for the information those questions below should be
answered.

You did reinstall Windows after you changed motherboards?
Are ACPI 2.0 and APIC mode both enabled in BIOS?