jonathonsunshine
27th February 2006, 05:31
Howdy all. I had the stock standard heatsink on my athlon 3000xp and it was getting up to 67 degrees celcius under load on a hot day at which point the mobo would shut down the computer. I replaced it with a custom heatsink/fan and now it never goes above 51.
I also did a fresh install of windows just after that. Also at the same time i moved onto version 1.1 of the xvid codec. I started noticing a lot of errors in the encodings i was making, They look like digital tv when it has a error, there is a broken frame, and then the picture sort of shimmers with some funky colors, until the next key frame. This would happen frequently.
I do about 50/50 one pass and two pass encoding. I get the corruption in both.
The settings i use, are H.263, Quarter Pixel, Global Motion Comp, B-VOPs (2, 1.5, 1), Packed Bitstream, Motion Search Precision on 6 Ultra High, VHQ mode 4, VHQ for bframes too, Chroma Motion and Trellis quanitization.
Also, i have noticed (inaudible) crc errors when encoding ac3 streams to mp3 using lame, even though i know they are clean streams, (they came off dvd's).
I reinstalled windows and went back to xvid 1.0.3 (the one in AutoGK 2.20) and although not as frequent i still notice the same errors in the encoding where the source file was clean.
Is it possible i damaged the cpus mmx/sse functions somehow when i replaced the heatsink? I havn't noticed any other problems, everyonce and a while i will have applications lock up or freeze but i figure thats par for the course, afterall, it is a computer.
I originally had this or a very similar message posted on the AMD forums. http://forums.amd.com/index.php?showtopic=69296
Here is a brief cut n paste off anything i thought might be relevant.
zeppelinrox - to me it sounds like a ram issue.
test the ram first with memtest.
jonathonsunshine - ok, i ran memtest86 for about 6 hours, and nothing came up.
i also ran Prime95, on torture tests.
When i select "Small FFTs (maximum FPU stress, data fits in L@ cache, RAM not tested much)" it runs ok, whien i select "In-place large FFTs (maximum heat, power consumption, some RAM tested)", it runs ok.
But when i select "Blend (tests some of everything, lots of RAM tested)", the moment i start it prime gets a "Windows - Application Error saying the Application failed to inititialize properly (0xc0000017).
bogey - Please list the system hardware.
From PSU
DC OUTPUT: 400W
+3.3v---28.0A(ORG), +5V---30.0a(RED), +12V---15.0A(YEL)
+5Vsb---2.0A(PURP), -5V---0.3A(WHITE),-12v---0.5A(BLUE)
P.G. SIGNAL (GRAY), GROUND (BLACK)
(+3.3V & +5V=220W Max) (+3.3V & +5V & +12V=380W Max)
CPU-Z Report
CPU-Z version 1.32.
CPU(s)
Number of CPUs 1
Name AMD Athlon XP
Code Name Barton
Specification AMD Athlon™ XP 3000+
Family / Model / Stepping 6 A 0
Extended Family / Model 7 A
Package Socket A
Technology 0.13 µ
Supported Instructions Sets MMX, Extended MMX, 3DNow!, Extended 3DNow!, SSE
CPU Clock Speed 2162.8 MHz
Clock multiplier x 13.0
Front Side Bus Frequency 166.4 MHz
Bus Speed 332.7 MHz
L1 Data Cache 64 KBytes, 2-way set associative, 64 Bytes line size
L1 Instruction Cache 64 KBytes, 2-way set associative, 64 Bytes line size
L2 Cache 512 KBytes, 16-way set associative, 64 Bytes line size
L2 Speed 2162.8 MHz (Full)
L2 Location On Chip
L2 Data Prefetch Logic yes
L2 Bus Width 64 bits
Mainboard and chipset
Motherboard manufacturer http://www.abit.com.tw/
Motherboard model NF7-S/NF7,NF7-V (nVidia-nForce2), 2.X,1.0
BIOS vendor Phoenix Technologies, LTD
BIOS revision 6.00 PG
BIOS release date 11/22/2004
Chipset nVidia nForce2 Ultra 400 rev. C1
Southbridge nVidia nForce2 MCP-T rev. A4
Sensor chip Winbond W83627HF
Graphic Interface AGP
AGP Status enabled, rev. 3.0
AGP Data Transfer Rate 8x
AGP Max Rate 8x
AGP Side Band Addressing supported, enabled
Memory
DRAM Type DDR-SDRAM
DRAM Size 1024 MBytes
DRAM Frequency 166.4 MHz
FSB:DRAM 1:1
CAS# Latency 2.0 clocks
RAS# to CAS# 3 clocks
RAS# Precharge 3 clocks
Cycle Time (TRAS) 5 clocks
# of memory modules 2
Module 0 GEIL DDR-SDRAM PC3200 - 512 MBytes
Module 1 GEIL DDR-SDRAM PC3200 - 512 MBytes
Software
Windows version Microsoft Windows XP Professional Service Pack 2 (Build 2600)
DirectX version 9.0c
bogey - The power supply +12V rail is low for that system. So disconnect the two spare hard drives and the optical drives to ease the load.
The memory timings may need changing. Try 2.5 3 3 7 and see if it's stable in the blend torture test.
jonathonsunshine - I have another power supply on hand. An Antec 550watt.
The stats on it are in another format, they read like this.
AC Input: 115V/12A, 230V/6A ; 60hz / 50hz
DC OUTPUT +5v +12v +3.3v -5v -12v +5v SB
MAX. 40A 30A 32A 0.5A 1.0A 2.0A
MIN. 0A 0.8A 0A 0A 0A 0A
+5v, +12v and +3.3v max. load 530w.
Question 1
Do you actually think my power supply could be causing these problems or are you just saying that it is generally low.
Question 2
Is it actually possible for a cpu to be physically damaged and still function?
I also did a fresh install of windows just after that. Also at the same time i moved onto version 1.1 of the xvid codec. I started noticing a lot of errors in the encodings i was making, They look like digital tv when it has a error, there is a broken frame, and then the picture sort of shimmers with some funky colors, until the next key frame. This would happen frequently.
I do about 50/50 one pass and two pass encoding. I get the corruption in both.
The settings i use, are H.263, Quarter Pixel, Global Motion Comp, B-VOPs (2, 1.5, 1), Packed Bitstream, Motion Search Precision on 6 Ultra High, VHQ mode 4, VHQ for bframes too, Chroma Motion and Trellis quanitization.
Also, i have noticed (inaudible) crc errors when encoding ac3 streams to mp3 using lame, even though i know they are clean streams, (they came off dvd's).
I reinstalled windows and went back to xvid 1.0.3 (the one in AutoGK 2.20) and although not as frequent i still notice the same errors in the encoding where the source file was clean.
Is it possible i damaged the cpus mmx/sse functions somehow when i replaced the heatsink? I havn't noticed any other problems, everyonce and a while i will have applications lock up or freeze but i figure thats par for the course, afterall, it is a computer.
I originally had this or a very similar message posted on the AMD forums. http://forums.amd.com/index.php?showtopic=69296
Here is a brief cut n paste off anything i thought might be relevant.
zeppelinrox - to me it sounds like a ram issue.
test the ram first with memtest.
jonathonsunshine - ok, i ran memtest86 for about 6 hours, and nothing came up.
i also ran Prime95, on torture tests.
When i select "Small FFTs (maximum FPU stress, data fits in L@ cache, RAM not tested much)" it runs ok, whien i select "In-place large FFTs (maximum heat, power consumption, some RAM tested)", it runs ok.
But when i select "Blend (tests some of everything, lots of RAM tested)", the moment i start it prime gets a "Windows - Application Error saying the Application failed to inititialize properly (0xc0000017).
bogey - Please list the system hardware.
From PSU
DC OUTPUT: 400W
+3.3v---28.0A(ORG), +5V---30.0a(RED), +12V---15.0A(YEL)
+5Vsb---2.0A(PURP), -5V---0.3A(WHITE),-12v---0.5A(BLUE)
P.G. SIGNAL (GRAY), GROUND (BLACK)
(+3.3V & +5V=220W Max) (+3.3V & +5V & +12V=380W Max)
CPU-Z Report
CPU-Z version 1.32.
CPU(s)
Number of CPUs 1
Name AMD Athlon XP
Code Name Barton
Specification AMD Athlon™ XP 3000+
Family / Model / Stepping 6 A 0
Extended Family / Model 7 A
Package Socket A
Technology 0.13 µ
Supported Instructions Sets MMX, Extended MMX, 3DNow!, Extended 3DNow!, SSE
CPU Clock Speed 2162.8 MHz
Clock multiplier x 13.0
Front Side Bus Frequency 166.4 MHz
Bus Speed 332.7 MHz
L1 Data Cache 64 KBytes, 2-way set associative, 64 Bytes line size
L1 Instruction Cache 64 KBytes, 2-way set associative, 64 Bytes line size
L2 Cache 512 KBytes, 16-way set associative, 64 Bytes line size
L2 Speed 2162.8 MHz (Full)
L2 Location On Chip
L2 Data Prefetch Logic yes
L2 Bus Width 64 bits
Mainboard and chipset
Motherboard manufacturer http://www.abit.com.tw/
Motherboard model NF7-S/NF7,NF7-V (nVidia-nForce2), 2.X,1.0
BIOS vendor Phoenix Technologies, LTD
BIOS revision 6.00 PG
BIOS release date 11/22/2004
Chipset nVidia nForce2 Ultra 400 rev. C1
Southbridge nVidia nForce2 MCP-T rev. A4
Sensor chip Winbond W83627HF
Graphic Interface AGP
AGP Status enabled, rev. 3.0
AGP Data Transfer Rate 8x
AGP Max Rate 8x
AGP Side Band Addressing supported, enabled
Memory
DRAM Type DDR-SDRAM
DRAM Size 1024 MBytes
DRAM Frequency 166.4 MHz
FSB:DRAM 1:1
CAS# Latency 2.0 clocks
RAS# to CAS# 3 clocks
RAS# Precharge 3 clocks
Cycle Time (TRAS) 5 clocks
# of memory modules 2
Module 0 GEIL DDR-SDRAM PC3200 - 512 MBytes
Module 1 GEIL DDR-SDRAM PC3200 - 512 MBytes
Software
Windows version Microsoft Windows XP Professional Service Pack 2 (Build 2600)
DirectX version 9.0c
bogey - The power supply +12V rail is low for that system. So disconnect the two spare hard drives and the optical drives to ease the load.
The memory timings may need changing. Try 2.5 3 3 7 and see if it's stable in the blend torture test.
jonathonsunshine - I have another power supply on hand. An Antec 550watt.
The stats on it are in another format, they read like this.
AC Input: 115V/12A, 230V/6A ; 60hz / 50hz
DC OUTPUT +5v +12v +3.3v -5v -12v +5v SB
MAX. 40A 30A 32A 0.5A 1.0A 2.0A
MIN. 0A 0.8A 0A 0A 0A 0A
+5v, +12v and +3.3v max. load 530w.
Question 1
Do you actually think my power supply could be causing these problems or are you just saying that it is generally low.
Question 2
Is it actually possible for a cpu to be physically damaged and still function?