EPTComputer Build Resource Thread - Page 1545
| Forum Index > Tech Support |
When using this resource, please read FragKrag's opening post. The Tech Support forum regulars have helped create countless of desktop systems without any compensation. The least you can do is provide all of the information required for them to help you properly. | ||
|
Gumbi
Ireland463 Posts
| ||
|
Ropid
Germany3557 Posts
| ||
|
Cyro
United Kingdom20323 Posts
| ||
|
Gumbi
Ireland463 Posts
 | ||
|
Cyro
United Kingdom20323 Posts
On June 17 2013 00:04 Ropid wrote: I feel it's not that strange. Both coolers have the same number of heat-pipes at similar diameter. The D14 has one 120mm and one 140mm fan, the U14S has a strange 150mm fan at higher RPM. This monster is the NH-U14S, which uses Noctua's special NF-A15 PWM fan. This is a 150mm fan designed to occupy the same space as a 140mm fan, but uses 120mm fan mounting. I remember this, i accused them of stealing time lord technology :D On June 17 2013 00:29 Gumbi wrote: How is it helping you with OCing that 4670k Mine? I got a 4770k (: It's doing nICEly  | ||
|
Gumbi
Ireland463 Posts
| ||
|
Cyro
United Kingdom20323 Posts
That caused a pretty massive amount of confusion, because i made assumptions about things, and about what i could do or should be able to do that were wrong, so i wasted a lot of time. Now that i have a much better idea of what i am doing (my understanding is so many levels ahead, every single day) i'm pretty confident in clocking up, and i should be able to get 4.7 np np. I'l set my temps around stress test absolute max's, and then add the few fans i am going to and enable HT, which should make my max's in regular use (and forcing full CPU load with x264) to be about the same. The u14s prevents you from using the top PCI-E slot, the silver arrow and nh-d14 (and probably that other one you mentioned) do not, something to consider. My board for example (the z87x-ud3h) will only do x8/x8 SLI in the first two slots. I must also point out, how awful most of these reviews are.. I removed every case fan but the front intake, which I ran at 5V to prevent it from affecting acoustics while still providing adequate airflow. For air coolers, I added a Noctua 140mm rear exhaust fan and used the ultra low noise adaptor to ensure it didn't affect acoustics in any meaningful way. This is in line with the usage cases air coolers are designed for, and should be representative of the kind of airflow most users will have from their exhaust fan. And the tests tuned, so the better performing coolers are at 40c above ambient, for absolute max load temps. That's just not realistic. A stock front intake at 5v, and a single rear exhaust with ultra low noise adapter is just NOT something i would consider using if i felt the need to buy an nh-d14 or a u14s. They show the u14s winning by 4c here - i think that's flat out wrong, and it's ridiculous to have absolute max load temps in a test at 60c. With actual case airflow and an actual overclock (90c after 15 minutes of small fft) i think a lot of the results they have, and a lot of results other people post, would be a lot different. I mean for the love of god, the h80i does not beat the nh-d14 by 6 degrees. Linus's tests, for example, show the silver arrow with a pair of NF-A15's beating the h80i in single-fan (nf-f12) by 21.5c - and he has solid benches. He actually knows what the hell he is doing to some extent, i feel, a lot more than some of the things anandtech posts. edit: They show the NH-D14 as 1.4c better than the 212 evo, and the h80i 6c better than the nh-d14. Ridiculous. | ||
|
Ropid
Germany3557 Posts
It could be still interesting for the future regarding that 150 mm wide cooler. mATX is interesting because there's a lot of neat compact cases to choose from so a future PC might go that route and the cooler perhaps won't fit anymore. | ||
|
{ToT}ColmA
Japan3260 Posts
when i had it on auto (the vcore) it went to 1.5vc and i had some wooping 90degree -_- kinda bogus how high my vc needs to be to get a "stable" 4,4ghz. didnt really change any other settings in bios as i am not sure about ptt etc, i am running this on a ga z77x d3h (which in hintside) seems pretty bad with all the issues it has (read that gigabyte tw is utter shit and doesnt care to fix a lot of bugs and there are bios out from someone who fixed a lot of stuff they dont feel the need to fix -_- temps should be fine i guess, did go to 1200rpm on the macho to get those so it should be fine i guess..still kinda sad that this oc is like that, would ve expected smoother vc, i cant get 4,5ghz to work no matter what lol (1.45vc still bluescreen) | ||
|
Cyro
United Kingdom20323 Posts
i just tried to oc again for the last hour or so, just running p95 with 4,4ghz on ....1,380vc and seems stable (so far) 73c on hottest core You're doing it WRONG somehow. I don't even know where to begin, sorry. Sounds like a pretty big mess. And please, don't go over 1.35-1.4vcore under any circumstances. I would, but i know what i'm doing. What Gflops does Intelburntest show when you run tests? On June 17 2013 01:05 Ropid wrote: The top slot is usually a tiny x1 slot on ATX boards, the first graphics card slot technically being the second slot. It should only be problematic on mATX boards where there's only four slots overall, and the main graphics card slot is put on the first position. It could be still interesting for the future regarding that 150 mm wide cooler. mATX is interesting because there's a lot of neat compact cases to choose from so a future PC might go that route and the cooler perhaps won't fit anymore. Good to hear, thanks | ||
|
{ToT}ColmA
Japan3260 Posts
 | ||
|
Cyro
United Kingdom20323 Posts
And what OC guide told you it was ok to go to 1.5vcore? That was very silly of you. If it was bloomfield, sandy or haswell you could have a seriously degraded, damaged or dead CPU on your hands very quickly. What Gflops does Intelburntest show when you run tests? Give this info, and also, are you using LLC? | ||
|
{ToT}ColmA
Japan3260 Posts
i dont test with intelburntest so dont know what it would say, i just use p95. brb, gonna reboot and ckeck the settings and write them down T_T + Show Spoiler + settings no memory oc cpu vcore 1.370 vtt auto cpu pll 1.650 (default is 1.800) recommended throu a guide imc auto vcore llc auto vcore voltage response auto | ||
|
Cyro
United Kingdom20323 Posts
| ||
|
Gumbi
Ireland463 Posts
| ||
|
Ropid
Germany3557 Posts
On June 17 2013 01:30 {ToT}ColmA wrote: u r reading it wrong, i never put it myself on 1.5, no guide says that. i had it on auto to look how much vc the auto setting would set via cpu-z, had it on 4,4ghz and it went to 1.5vc. manually i set it to 1.3661 (well it put it to 1.370 in bios) and cpu z shows 1.380vc. i put it that way cuz cpu temp shows vid 1.3661 under load so i just used that to try, if i go 0.05v lower p95 crashes constantly after test 2/3 and i get whea errors. i dont test with intelburntest so dont know what it would say, i just use p95. brb, gonna reboot and ckeck the settings and write them down T_T + Show Spoiler + settings no memory oc cpu vcore 1.370 vtt auto cpu pll 1.650 (default is 1.800) recommended throu a guide imc auto vcore llc auto vcore voltage response auto I'll send you a PM with screenshots of how I've set up my BIOS (same board and CPU). | ||
|
Cyro
United Kingdom20323 Posts
| ||
|
{ToT}ColmA
Japan3260 Posts
+ Show Spoiler + Time (s) Speed (GFlops) Result 8.224 108.6964 3.526496e-002 Time (s) Speed (GFlops) Result 8.151 109.6691 3.526496e-002 Time (s) Speed (GFlops) Result 8.138 109.8352 3.526496e-002 Time (s) Speed (GFlops) Result 8.150 109.6804 3.526496e-002 Time (s) Speed (GFlops) Result 8.158 109.5712 3.526496e-002 Time (s) Speed (GFlops) Result 8.151 109.6670 3.526496e-002 Time (s) Speed (GFlops) Result 8.122 110.0603 3.526496e-002 Time (s) Speed (GFlops) Result 8.158 109.5746 3.526496e-002 Time (s) Speed (GFlops) Result 8.153 109.6356 3.526496e-002 Time (s) Speed (GFlops) Result 8.151 109.6658 3.526496e-002 | ||
|
Cyro
United Kingdom20323 Posts
LLC on second highest level should let you get stable on a lot lower set vcore. When people say a vcore setting, they usually mean with some LLC (it's common to run top 1-3 most intense llc setting) and it changes the number you have to type into bios for vcore quite a bit | ||
|
{ToT}ColmA
Japan3260 Posts
this new pc sure makes me spam a lot in these forums...never posted this much on tl.. | ||
| ||
