ผมไปถามมาละแกบอกเทียบกันระหว่าง direct die vs บัดกรีเดิมๆ ไม่ใช่ direct die vs ผ่าใส่ปรอท
worst case คือต่างกัน 14C
ถ้ากรณีคนที่ผ่า P0 อยู่แล้วไป direct die อาจจะไม่เห็นผลเท่าไร เพราะผ่าแล้วมันหายไปประมาณ 10C เหลืออีกราวๆ 4C ที่จะเห็นผล
'Vanity' Direct die results before and after 9900k
HWL 420 GTX 360 GTX 360 GTS
Fan curves remained the same for testing for each test and before and after. (Corsair ml120 and 140 @ ~< 30% RPM)
Ambient of 24-24.5C
Avx off
5.2 @ ~ 1.337 vcore load *stability confirmed (8H Small FFT)
Prime95 Small FFT 1hour. CPU Package
Before: 84C
After: 77C
5.3 @ ~ 1.418 vcore load *stability not confirmed
Prime95 Small FFT 1hour. CPU Package
Before: 98C
After: 84C
Looped 1 hour of Final Fantasy XV benchmark using the 5.2 OC for a real use scenario... 5.3 was unstable and dismissed for this test
GPU Temp Caps at 50C (confirmed with furmark) with a water temp Cap of 38C (water up 2C from the cpu testing)
CPU Package
Before: 74C
After 71C
Caveats*
-Narrower diameter tubing used with the delid (switched to glass tubing)
-Mounting pressure, No fear when cranking down the stock IHS, as for the direct die, I held back a bit... Maybe i'll try tightening down 1/4 - 1/2 turn more with the delid. Or YOLO full turn?
-Conductonaught application margin of error.
-After testing was done right after filling the loop, some air pockets may exist somewhere within the 3 rads I got. (although likely negligible to make an impact)
***after cranking down the coldplate till i couldn't crank no more, the 5.2 OC after temp dropped another 2C, (i'd guess the 5.3 may be 4C if i was to test it again)
Comment