- Joined
- Oct 21, 2006
- Messages
- 263 (0.04/day)
| Processor | E6400 2.14ghz @ 3.6ghz (24/7) |
|---|---|
| Motherboard | Asus P5B-Dlx WiFi |
| Cooling | Scythe Infinity (pressure-mod) w/ Dual Silverstone FM121's |
| Memory | Mushking Extreme DDR-2 PC6400 |
| Video Card(s) | EVGA 7900GT - 1.5 volt mod @ 700/852 |
| Storage | Generic Samsung SATA |
| Case | TT Armor w/ 250mm side panel fan |
| Audio Device(s) | Sound Blaster Live 5.1 ( what else? ) |
| Power Supply | TT 500 pure power silent |
| Software | GiMP |
Side by side review: Mod'd Zalman 9500 vs. Scythe Infinity
(using identical 110CFM Silverstone fans)
----------------------------------
Test procedure:
1. All test data obtained with ambient temps of 20C +/- .05C
2. System powered down for one hour prior to each test to ensure equality of temps
3. Idle temp obtained after 30 minutes idle time, immediately after boot.
4. Load temp obtained during 30 minute Orthos (large in-place FFT's)
5. Tests run on an E6400 overclocked to 3.6ghz (no stock speed tests!)
data will be gathered at the max fan speed of 2400rpm, as well as a reading at 1600rpm's (for the quiet folks out there).
----------------------------------
but first, a bit of background info ...
...
...
when i modd'd my zalman 9500 with a 120mm fan ... i was sure i had inherited the earth; the diffference in temps and gain in overclock was *that* significant. the stock zalman 9500 is simply *not* a good overclock cooler; yet, with a 110CFM fan strapped on ... it jockey's to the top end of the air cooling options, but even so ... just to the fringe of such top-end-cooler-domains. suffice it to say ... with the modd'd zalman ... i was able to hit 4.0ghz ... but *NOT* 4.1ghz and beyond ...
therefore ... wanting to bench 'higher than the next guy,' i decided to investigate other air coolers, if only on a test basis.
so, enter the scythe infinity.
...
xmas allowed me a gift option, and considering the scythe infinity was available for 40 bucks or so, and with no shipping charge ... i decided to give it a try.
the same day that i purchased it, ... i lapped it:
...
...
as i was lapping, i noticed that despite the very shiney stock surface finish, the heatsink was actually not very flat ... higher on one side than the other; but, after about an hour with 400grit sand paper ... it evened out. i followed the 400 with 1000, 1500, and 3000 to finish.
i did a lot of reading before purchasing the infinity cooler, and the one thing i found was that, unanimously, most serious overclockers found that the infinity's mounting mechanism was lame; the push-pins don't provide enough mounting pressure to ensure good heat transfer. and so, i decided to modify the mounting with the somewhat well known and popular bolt modification.
it was a SERIOUS pain in the ass ... including removal of motherboard ... bloody fingers and all ... but in the end, i got it mounted. i wish i had remembered to take photos ... was so funny. i over-torqued the mounting screws at first and my motherboard was bent into a total U shape! after noticing the warping, i backed the screws off a bit and it evened out. to be honest, i was amazed my system even posted after bending the board as much as i did!
anyway .. i got the thing installed and subsequently set out to bench it against the previously installed modified zalman.
...
like the scythe cooler, the zalman had also been previously lapped and polished:
...
...
furthermore, the E6400 CPU used for testing was also previously lapped:
...
...
And, after a thorough, scientific and truly unbiased test of each heat sink ... this is what I found:
...
Zalman 9500 @ 2400RPM:
-------------------------------
cpu idle: 40 | mb idle: 23
cpu load: 61 | mb load: 25
Infinity @ 2400RPM:
-------------------------------
cpu idle: 36 | mb idle: 25
cpu load: 54 | mb load: 26
...
Zalman 9500 @ 1600RPM:
-------------------------------
cpu idle: 44 | mb idle: 24
cpu load: 68 | mb load: 25
Scythe Infinity @ 1600RPM:
-------------------------------
cpu idle: 37 | mb idle: 24
cpu load: 55 | mb load: 25
...
good grief! i guess surface area really does matter. the scythe infinity smoked the zalman across the board; most notable is how little of a temp difference there was when using the fan at high speed versus low speed. with the infinity, only a 1 degree drop or so ... but with the zalman and it's reduced surface area, a somewhat huge drop of around 7 degrees when moving from 2400RPM to 1600RPM's.
...
anyway, since the scythe infinity so totally dethroned my mod'd zalman ... suffice it to say, it's going to remain mounted. lord knows, it was a pain in the arse getting properly installed!
i will say this, though ... if you have a zalman 9500 and want to hit serious overclocks ... you *must* mount a high output fan, like the silverstone FM121 in order to get there. it's been a great little cooler once modified, but alas ... today has become it's retirement day.
here's the new kid on the block:
...
...
also, as a closing note ... if your going to use the infinity with an armor case that has the side panel fan installed, you'll have to move the fan to the outside of the side panel as illustrated below.
i'm sure the same procedure would be required for any other Fat Ass Air Cooler(tm), like the Tuniq, etc.
...
...
...
here's to bigger overclocks on air!
(using identical 110CFM Silverstone fans)
----------------------------------
Test procedure:
1. All test data obtained with ambient temps of 20C +/- .05C
2. System powered down for one hour prior to each test to ensure equality of temps
3. Idle temp obtained after 30 minutes idle time, immediately after boot.
4. Load temp obtained during 30 minute Orthos (large in-place FFT's)
5. Tests run on an E6400 overclocked to 3.6ghz (no stock speed tests!)
data will be gathered at the max fan speed of 2400rpm, as well as a reading at 1600rpm's (for the quiet folks out there).
----------------------------------
but first, a bit of background info ...
...
...
when i modd'd my zalman 9500 with a 120mm fan ... i was sure i had inherited the earth; the diffference in temps and gain in overclock was *that* significant. the stock zalman 9500 is simply *not* a good overclock cooler; yet, with a 110CFM fan strapped on ... it jockey's to the top end of the air cooling options, but even so ... just to the fringe of such top-end-cooler-domains. suffice it to say ... with the modd'd zalman ... i was able to hit 4.0ghz ... but *NOT* 4.1ghz and beyond ...
therefore ... wanting to bench 'higher than the next guy,' i decided to investigate other air coolers, if only on a test basis.
so, enter the scythe infinity.
...
xmas allowed me a gift option, and considering the scythe infinity was available for 40 bucks or so, and with no shipping charge ... i decided to give it a try.
the same day that i purchased it, ... i lapped it:
...
...
as i was lapping, i noticed that despite the very shiney stock surface finish, the heatsink was actually not very flat ... higher on one side than the other; but, after about an hour with 400grit sand paper ... it evened out. i followed the 400 with 1000, 1500, and 3000 to finish.
i did a lot of reading before purchasing the infinity cooler, and the one thing i found was that, unanimously, most serious overclockers found that the infinity's mounting mechanism was lame; the push-pins don't provide enough mounting pressure to ensure good heat transfer. and so, i decided to modify the mounting with the somewhat well known and popular bolt modification.
it was a SERIOUS pain in the ass ... including removal of motherboard ... bloody fingers and all ... but in the end, i got it mounted. i wish i had remembered to take photos ... was so funny. i over-torqued the mounting screws at first and my motherboard was bent into a total U shape! after noticing the warping, i backed the screws off a bit and it evened out. to be honest, i was amazed my system even posted after bending the board as much as i did!
anyway .. i got the thing installed and subsequently set out to bench it against the previously installed modified zalman.
...
like the scythe cooler, the zalman had also been previously lapped and polished:
...
...
furthermore, the E6400 CPU used for testing was also previously lapped:
...
...
And, after a thorough, scientific and truly unbiased test of each heat sink ... this is what I found:
...
Zalman 9500 @ 2400RPM:
-------------------------------
cpu idle: 40 | mb idle: 23
cpu load: 61 | mb load: 25
Infinity @ 2400RPM:
-------------------------------
cpu idle: 36 | mb idle: 25
cpu load: 54 | mb load: 26
...
Zalman 9500 @ 1600RPM:
-------------------------------
cpu idle: 44 | mb idle: 24
cpu load: 68 | mb load: 25
Scythe Infinity @ 1600RPM:
-------------------------------
cpu idle: 37 | mb idle: 24
cpu load: 55 | mb load: 25
...
good grief! i guess surface area really does matter. the scythe infinity smoked the zalman across the board; most notable is how little of a temp difference there was when using the fan at high speed versus low speed. with the infinity, only a 1 degree drop or so ... but with the zalman and it's reduced surface area, a somewhat huge drop of around 7 degrees when moving from 2400RPM to 1600RPM's.
...
anyway, since the scythe infinity so totally dethroned my mod'd zalman ... suffice it to say, it's going to remain mounted. lord knows, it was a pain in the arse getting properly installed!
i will say this, though ... if you have a zalman 9500 and want to hit serious overclocks ... you *must* mount a high output fan, like the silverstone FM121 in order to get there. it's been a great little cooler once modified, but alas ... today has become it's retirement day.
here's the new kid on the block:
...
...
also, as a closing note ... if your going to use the infinity with an armor case that has the side panel fan installed, you'll have to move the fan to the outside of the side panel as illustrated below.
i'm sure the same procedure would be required for any other Fat Ass Air Cooler(tm), like the Tuniq, etc.
...
...
...
here's to bigger overclocks on air!