Is my AMD system underperforming?

[P Collins]

Hello all. I've had to do lots of integer/FP arithmetic with work and
although my CPU is a bit old, it appears to be underperforming based on
the test results. My system is a:

AMD XP 2500+ Barton 333MHz FSB, 2x512MB DDR333 RAM, Abit NF7
motherboard. The RAM is downclocked from its stock DDR400 speed to
match my CPU FSB.

I am not overclocking it as I require my system to be stable + produce
no arithmetic errors.

In fact as it is, the system sometimes overheats >70 Celcius if I leave
it computing overnight with the side panel of the case closed, during
summer.

My test results are below:

------------------------------------------
CPUBench2003 report

Data reported on 18:31:49,Friday 8/7/2005

Tester informations
Tester Name: Anonymous
System Name: AMD XP 2500+ Barton 333MHz FSB, 2x512MB DDR333 RAM
System Details:





CPU Informations
CPU Name: AMD AthlonXP
CPU Codename: Barton (0.13 um)
CPU Manufacturer: Advanced Micro Devices, Inc.
Vendor ID: AuthenticAMD
CPU Default Frequency Range: 1.7 - 2.0+ GHz
Current Frequency: 1830.02MHz
Family, model, stepping: 6,a,0 Signature: 0x6a0
CPU Platform: SocketA(PGA462 Socket)
CPU Stepping Mask: Unknow
Bus type: Alpha EV6 (Double Data Rate )
L1 data cache: 64KByte
L1 Instruction cache: 64KByte
L2 cache: 512KByte
L2 cache location: External on SECC Cartdrige
CPU Default Voltage: 1.65v
CPU Current Voltage: Cannot detect (beta version)





Standard CPU Features
Built in FPU: Supported
Virtual Mode Extensions: Supported
Debugging Extensions: Supported
Page Size Extensions: Supported
TimeStamp Counter: Supported
Model Specific Regs: Supported
Page Address Extensions: Supported
Machine check Extensions: Supported
CMPXCHG8 instruction: Supported
Built in APIC: Supported
reserved Not Supported
Fast System Call: Supported
Memory type Range Regs: Supported
Page Global Enable: Supported
Machine check Architecture: Supported
CMOV Instruction: Supported
Page Attribute Table: Supported
Page size Extensions: Supported
96 bit Processor Number: Not Supported
CLFLUSH Instruction: Not Supported
reserved Not Supported
Debug trace store: Not Supported
ACPI Supported: Not Supported
MMX Instructions: Supported
Fast FPU Save/Restore: Supported
Streaming SIMD Extensions: Supported
Streaming SIMD Extensions 2: Not Supported
Self Snoop: Not Supported
Jackson Technology: Not Supported
Thermal Monitor On chip: Not Supported
IA64 Processor Capable: Not Supported
reserved Not Supported





Extended CPU Features
Built in FPU: Supported
Virtual Mode Extensions: Supported
Debugging Extensions: Supported
Page Size Extensions: Supported
TimeStamp Counter: Supported
Model Specific Regs: Supported
Page Address Extensions: Supported
Machine check Extensions: Supported
CMPXCHG8 instruction: Supported
Built in APIC: Supported
reserved Not Supported
Fast System Call: Supported
Memory type Range Regs: Supported
Page Global Enable: Supported
Machine check Architecture: Supported
CMOV Instruction: Supported
Page Attribute Table: Supported
Page size Extensions: Supported
reserved Not Supported
reserved Not Supported
reserved Not Supported
reserved Not Supported
MMX Instructions Extension: Supported
MMX Instructions: Supported
Fast FPU Save/Restore: Supported
reserved Not Supported
reserved Not Supported
reserved Not Supported
reserved Not Supported
x86 - 64 Technology: Not Supported
Enhanced 3DNow!: Supported
3DNow! Instructions: Supported





OS Informations
Operating System: Microsoft Windows XP
OS Revision: Microsoft Windows version5.1.2600
System Name: Computer
User name: Anonymous
Windows Directory: C:\WINDOW
System Directory: C:\WINDOW\System32
Service Pack Installed:
Session Boot Mode: Normal Boot
Stand-By Mode: Enabled
Stand-By Timeout: 20 Minute(s)
Network status: Detected
Video card Name: RADEON 9800 PRO
Video Card Driver: ati2dvag.dll
VendorID: 4098
DeviceID: 20040
SubSystemID: 135170
Actual Resolution: 800x600 Pixels
Number Of Monitors: 1Monitor(s)
Physical Memory: 1024 MByte
Available Memory: 732 MByte
Page File Size: 2462 MByte
Available Page File: 2127 MByte
User Size: 2047 MByte
Minimum Application Address: 0x65536
Maximum Application Address: 0x2147418111
Allocation Granularity: 0x65536 Byte





All names and symbols are trademarks of respective owners -
CPUBench2003, all rights reserved

 

[kony]

Hello all. I've had to do lots of integer/FP arithmetic with work and
although my CPU is a bit old, it appears to be underperforming based on
the test results. My system is a:

AMD XP 2500+ Barton 333MHz FSB, 2x512MB DDR333 RAM, Abit NF7
motherboard. The RAM is downclocked from its stock DDR400 speed to
match my CPU FSB.

I am not overclocking it as I require my system to be stable + produce
no arithmetic errors.

In fact as it is, the system sometimes overheats >70 Celcius if I leave
it computing overnight with the side panel of the case closed, during
summer.

Either your room ambient temps are too high, your case
ventilation is too low, or your heatsink is quite
inadequate. Relatively speaking it's very easy to keep a
Barton 2500 below 60C. See if system is dust-clogged and if
all else seems ok, try remounting the heatsink with a TINY
bit of thermal grease. 1/2 grain of rice sized blob is
plenty, if not overapplication, on a barton with a
good/smooth heatsink base.

My test results are below:

------------------------------------------
CPUBench2003 report

Data reported on 18:31:49,Friday 8/7/2005

Tester informations
Tester Name: Anonymous
System Name: AMD XP 2500+ Barton 333MHz FSB, 2x512MB DDR333 RAM
System Details:

<rest snipped>

You didn't provide specifics on what leads you to believe
it's underperforming. Is it because of someone else
reporting higher CPU bench scores? If so, is it possible
their Barton 2500 is overlocked?

 

[P Collins]

Hi Kony.

I'm using some Artic Silver putty on the CPU. I think my room temp can
be part of the problem, because in summer it gets up to mid-20's
Celcius. The only fans in my case are:

1 fan on the Abit NF7 motherboard.
1 fan on the CPU itself.
1 fan on my Radeon 9800 Pro.
1 fan sucking air from inside into the Antec Truepower 350W PSU.
1 fan blowing air out from the Antec PSU.

So really, the only fans blowing out from my case is my PSU fan. All
the other fans blow around inside.

As for why I think my system is underperforming:
Test system: AMD XP 2500+ Barton 333MHz FSB, 2x512MB DDR333 RAM, Abit
NF7 motherboard.

My overall CPU Bench 2003 score: 5675
Score of a Intel P4A 2.4GHz CPU: 5758 (note my CPU is supposed to be
equivalent to 2500+)
Score of an AMD Athlon XP 2400+: 5861 (my CPU should beat this!)
Score of an AMD Athlon XP 2600+ (266MHz FSB): 6146

The results are the stored results in CPU Bench 2003, I don't know if
the processors there are overclocked. Doubtful, or the XP 2600+ would
beat my 2500+ by a very long shot.

Thanks.

 

[General Schvantzkoph]

Hi Kony.

I'm using some Artic Silver putty on the CPU. I think my room temp can
be part of the problem, because in summer it gets up to mid-20's
Celcius. The only fans in my case are:

1 fan on the Abit NF7 motherboard.
1 fan on the CPU itself.
1 fan on my Radeon 9800 Pro.
1 fan sucking air from inside into the Antec Truepower 350W PSU.
1 fan blowing air out from the Antec PSU.

So really, the only fans blowing out from my case is my PSU fan. All
the other fans blow around inside.

As for why I think my system is underperforming:
Test system: AMD XP 2500+ Barton 333MHz FSB, 2x512MB DDR333 RAM, Abit
NF7 motherboard.

My overall CPU Bench 2003 score: 5675
Score of a Intel P4A 2.4GHz CPU: 5758 (note my CPU is supposed to be
equivalent to 2500+)
Score of an AMD Athlon XP 2400+: 5861 (my CPU should beat this!)
Score of an AMD Athlon XP 2600+ (266MHz FSB): 6146

The results are the stored results in CPU Bench 2003, I don't know if
the processors there are overclocked. Doubtful, or the XP 2600+ would
beat my 2500+ by a very long shot.

Thanks.

The difference between the 2400+ and your 2500+ is 3%, that's
insignificant. Lot's of things can account for that, it's so small that it
could be nothing more then a measurement artifact. Any difference in your
OS patch level, what anti-virus program you are running, any other daemons
that are running. I wouldn't about tiny differences like that. If your
machine was running at half speed then I'd look for a problem (which is
much more likely to be caused by software then hardware), but not when you
are running 5% less then you expect.

As for you cooling problem, put in a case fan, if that doesn't do it get a
better CPU cooler.

 

[kony]

Hi Kony.

I'm using some Artic Silver putty on the CPU. I think my room temp can
be part of the problem, because in summer it gets up to mid-20's
Celcius.

Not a problem, that would still mean CPU is at 50C over
ambient temp. The issue would then seem to be case flow or
heatsink adequacy itself.

The only fans in my case are:

1 fan on the Abit NF7 motherboard.
1 fan on the CPU itself.
1 fan on my Radeon 9800 Pro.
1 fan sucking air from inside into the Antec Truepower 350W PSU.
1 fan blowing air out from the Antec PSU.

So really, the only fans blowing out from my case is my PSU fan. All
the other fans blow around inside.

As per AMD recommendations, and further needed by use of a
higher powered video card, you should have a case rear
exhaust fan below the power supply, on rear case wall, and
plenty of unobstructed front-bottom case intake area.

As for why I think my system is underperforming:
Test system: AMD XP 2500+ Barton 333MHz FSB, 2x512MB DDR333 RAM, Abit
NF7 motherboard.

My overall CPU Bench 2003 score: 5675
Score of a Intel P4A 2.4GHz CPU: 5758 (note my CPU is supposed to be
equivalent to 2500+)
Score of an AMD Athlon XP 2400+: 5861 (my CPU should beat this!)
Score of an AMD Athlon XP 2600+ (266MHz FSB): 6146

The results are the stored results in CPU Bench 2003, I don't know if
the processors there are overclocked. Doubtful, or the XP 2600+ would
beat my 2500+ by a very long shot.

Your score is close enough.
Some users may've shut down all background processes & apps.
Did you?

They may have high-end 2,2,2,5 memory, do you?
How about dual-channel memory?
They may have tweaked and fiddled with several bios settings
or other parameters of their system- and we dont' even know
if their system is as stable as you express need for yours
to be.

Frankly, the easiest way to increase performance is to raise
the FSB about 5MHz. While I understand your desire for
stablity, overclockers take the chip up past 2.2GHz and
still can keep it stable. You're only 5% under the
expected average of the 2400 and 2600 reported above.

However, recognize that each type of performance test
stresses different things. Some emphasize raw clock speed
over cache size. That is what appears to be happening in
the CPU Bench test, since it is giving an Athlon 2600 with
lesser L2 cache, but higher overall MHz speed, a better
score. In a different benchmark, your Barton 2500 may
outperform the non-Barton 2600. This is one of the problems
with isolated synthetic tests. Your CPU should have
significantly better performance at anything that makes use
of the larger cache.