- Feb 4, 2006
- 57
- 0
- 0
Hi all, just putting together my new rig, and thought there are findings that I think is better shared with everybody. This thread will include pictorial logs, benchmarks on coolers, findings regarding to various software/hardware features. Some you would have seen before, but some I have yet to see a thread or topic that discusses my findings. This thread will cover various things that had been asked many times. My thread will cover:
1. Ninja VS XP-120
2. Opty overclocking, also seeking advice on RAM timing
3. Stack-cool (ASUS exclusive cooling feature)
4. P180 installation
5. Silencing
6. PSU comparison
Not in any particular order, some will come later and some will come in just a few minutes.
There will also be questions asked, either because I can not fix myself or find a solution online.
Setup
First of all, my system:
Opty 165 CCBEW 0551VPMW + XP120/Ninja
A8N32-SLI DLX 1009 BIOS
2X1GB Kingston Value RAM
Sliverstone Strider 400W PSU w Active PFC
Creative Audigy 2
ASUS 6600 256 TOPS
Maxtor Diamond 9 120GB
Maxtor Diamond 10 120 GB
I originally have a XP-120 mounted with Cooler Master Nanofusion paste, This is the system removed from my old case and sitting in a open case config.
Little messy but I can live with it, since it?s only temporary.
Picture
Here?s the new system
Ninja Plus!
MX-1!Hopefully lives up to its name (beating Ceramique by 4 degrees, and since many reveals that ceramique is on par with AS5, that means also beating AS5 by 4~3C)
Cleaned XP-120 and my opty, cleaning done with ArcticClean
Blob of MX-1, at size of roughly a grain of rice.
Ninja mounted
Fan mounted, no probs! Even DIMM slot 1 is not obstructed much.
Ninja test system.
OCing
I gotta say, this Opty is a monster, runs stock clock at 1V, running down clocked 1.28GHZ at .85V and 2.56GHZ at stock volt! Great stepping?.
Screenshot of 1.28GHZ @ .85V
Screenshot of 2.56GHZ @ 1.35V
Both 12 hr dual prime stable. The RAM setting is also pretty good?however, can someone please help me tightening the timing? Latency/RAM Voltage is quite tricky to tackle?
Heatsink Benchmark!
Like I showed in the first post, I had a XP-120 with Nanofusion prior to my Ninja and MX-1.
Picture of test platform with thermometer.
Burn #1: 1.28GHZ @ .85V
Ambient = 28.9C Idle = 33 Load = 36
Burn #2: 2.56GHZ @ 1.35V
Ambient = 29C Idle =44 Load = 55
During the test, chipset hovers around 46~44C
The CPU runs cooler than in closed case setup, however, the chipset temp is almost full 10C higher! Rising to 46C max!
OK, here comes the ninja?..the setup is exactly like what I posted before, but here?s the picture anyway.
Burn #1: 2.56GHZ @ 1.35V
Ambient = 29.4C Idle = 43 Load = 52
Ambient = 27.5C Idle = 41 Load = 50
Ambient = 25.5C Load = 49
Ambient = 24.5C Load = 48
Burn #2: 1.28GHZ @ .85V
Ambient = 27.8C Idle = 33C Load = 34C
Chipset temperature = 44~46C
Through the test, because the temperature is constantly changing (bloody Melbourne weather...) so i check ever 30min for changing temperature, and i noted the results above.
This burn serves 2 purpose. One is to harden my MX-1 because it is said to reach optimum performance once the paste hardens. This is just a hands-on burn result, in a day or 2, I?ll do another 2 burns, one also at 2.56GHZ and see if there?s actually any difference, and the other at 1.28GHZ to see what happens to ninja when the processor is being underclocked.
Observations:
Stack-cool is an absolutely AWESOME feature, it does exactly what it does, spread the heat from both the chipset and CPU across the entire motherboard. When I switched off the rig, the cushion underneath it is hot, not just directly beneath CPU and chipset, but all across where the board sits.
Although the XP-120's airflow SHOULD cool the chipset on A8N32 better, it is on par with the ninja. I could hardly believe this, but the result speaks for itself.
Undoubtedly, the relationship between the chipset/CPU temp is directly related to the ambient air temperature, however, I am not entirely sure if the relationship is linear, parabolic or hyperbolic. From 29.5C ambient to 27.5C ambient, i see 2C drop in CPU, but from 27C ambient to 25.5, there's only 1 C drop. The best mathematical model would probably be linear with some anomalies.
Conclusion
Ninja beats XP-120 by 3C
However, there are also other factors to consider, my first mount is not exactly perfect, i used probably too much thermal grease, and nanofusion is not exactly the best grease out there.
At idle, both HS cools just as well, but under heavy load, Ninja outperforms XP-120 by 3C. As i said, this could also be due to poor thermal grease and installation.
Considering the error that could involve, i would say Ninja only wins on a small margin, and in a case where a side intake is avalible, XP-120 could beat ninja, so, it's a matter of finding the best airflow design for your particular case/board.
Coming up:
Ninja in tunnel VS normal config Ninja
Day 2 + Day 3 burn - Does MX-1 really improve? Can 2.56 @ Stock live 48HR of Prime?
1. Ninja VS XP-120
2. Opty overclocking, also seeking advice on RAM timing
3. Stack-cool (ASUS exclusive cooling feature)
4. P180 installation
5. Silencing
6. PSU comparison
Not in any particular order, some will come later and some will come in just a few minutes.
There will also be questions asked, either because I can not fix myself or find a solution online.
Setup
First of all, my system:
Opty 165 CCBEW 0551VPMW + XP120/Ninja
A8N32-SLI DLX 1009 BIOS
2X1GB Kingston Value RAM
Sliverstone Strider 400W PSU w Active PFC
Creative Audigy 2
ASUS 6600 256 TOPS
Maxtor Diamond 9 120GB
Maxtor Diamond 10 120 GB
I originally have a XP-120 mounted with Cooler Master Nanofusion paste, This is the system removed from my old case and sitting in a open case config.
Little messy but I can live with it, since it?s only temporary.
Picture
Here?s the new system
Ninja Plus!
MX-1!Hopefully lives up to its name (beating Ceramique by 4 degrees, and since many reveals that ceramique is on par with AS5, that means also beating AS5 by 4~3C)
Cleaned XP-120 and my opty, cleaning done with ArcticClean
Blob of MX-1, at size of roughly a grain of rice.
Ninja mounted
Fan mounted, no probs! Even DIMM slot 1 is not obstructed much.
Ninja test system.
OCing
I gotta say, this Opty is a monster, runs stock clock at 1V, running down clocked 1.28GHZ at .85V and 2.56GHZ at stock volt! Great stepping?.
Screenshot of 1.28GHZ @ .85V
Screenshot of 2.56GHZ @ 1.35V
Both 12 hr dual prime stable. The RAM setting is also pretty good?however, can someone please help me tightening the timing? Latency/RAM Voltage is quite tricky to tackle?
Heatsink Benchmark!
Like I showed in the first post, I had a XP-120 with Nanofusion prior to my Ninja and MX-1.
Picture of test platform with thermometer.
Burn #1: 1.28GHZ @ .85V
Ambient = 28.9C Idle = 33 Load = 36
Burn #2: 2.56GHZ @ 1.35V
Ambient = 29C Idle =44 Load = 55
During the test, chipset hovers around 46~44C
The CPU runs cooler than in closed case setup, however, the chipset temp is almost full 10C higher! Rising to 46C max!
OK, here comes the ninja?..the setup is exactly like what I posted before, but here?s the picture anyway.
Burn #1: 2.56GHZ @ 1.35V
Ambient = 29.4C Idle = 43 Load = 52
Ambient = 27.5C Idle = 41 Load = 50
Ambient = 25.5C Load = 49
Ambient = 24.5C Load = 48
Burn #2: 1.28GHZ @ .85V
Ambient = 27.8C Idle = 33C Load = 34C
Chipset temperature = 44~46C
Through the test, because the temperature is constantly changing (bloody Melbourne weather...) so i check ever 30min for changing temperature, and i noted the results above.
This burn serves 2 purpose. One is to harden my MX-1 because it is said to reach optimum performance once the paste hardens. This is just a hands-on burn result, in a day or 2, I?ll do another 2 burns, one also at 2.56GHZ and see if there?s actually any difference, and the other at 1.28GHZ to see what happens to ninja when the processor is being underclocked.
Observations:
Stack-cool is an absolutely AWESOME feature, it does exactly what it does, spread the heat from both the chipset and CPU across the entire motherboard. When I switched off the rig, the cushion underneath it is hot, not just directly beneath CPU and chipset, but all across where the board sits.
Although the XP-120's airflow SHOULD cool the chipset on A8N32 better, it is on par with the ninja. I could hardly believe this, but the result speaks for itself.
Undoubtedly, the relationship between the chipset/CPU temp is directly related to the ambient air temperature, however, I am not entirely sure if the relationship is linear, parabolic or hyperbolic. From 29.5C ambient to 27.5C ambient, i see 2C drop in CPU, but from 27C ambient to 25.5, there's only 1 C drop. The best mathematical model would probably be linear with some anomalies.
Conclusion
Ninja beats XP-120 by 3C
However, there are also other factors to consider, my first mount is not exactly perfect, i used probably too much thermal grease, and nanofusion is not exactly the best grease out there.
At idle, both HS cools just as well, but under heavy load, Ninja outperforms XP-120 by 3C. As i said, this could also be due to poor thermal grease and installation.
Considering the error that could involve, i would say Ninja only wins on a small margin, and in a case where a side intake is avalible, XP-120 could beat ninja, so, it's a matter of finding the best airflow design for your particular case/board.
Coming up:
Ninja in tunnel VS normal config Ninja
Day 2 + Day 3 burn - Does MX-1 really improve? Can 2.56 @ Stock live 48HR of Prime?