|
ZombieApostate posted:I read that that could be a problem, so I kept a watch on it and it never raised nearly as far from where it was set as people complained, but I'll take your word for it rather than fry my chip . Unless your voltage is hitting over 1.4 you dont need to worry about frying your chip. If your cpu hits 100-105c it will just throttle itself back, it wont kill itself.
|
# ? Jun 1, 2015 00:39 |
|
|
# ? Jun 5, 2024 04:08 |
|
I've been happily (stabily?) using the below setup that I OC'd from stock since around October 2013ish. I've had the CPU for 2-4 years prior to that. With the mobo I'm using I couldn't get the CPU to throttle with load with an OC from BIOS so I've been using AMD Overdrive to software OC the CPU thus far. It's all air cooled and the CPU never gets any hotter than 60C under full load. It's been fully tested with prime95 and others for 24hrs+. Stock clock is 3.2 at vcore of 1.35 OC to: 4.0 at vcore of 1.375 Lately I've been having the issue of the OC not being in place occasionally when I boot. I'm not sure why this is. I've also noticed that when I reset the OC and restart, in the last month or so it's been crashing (very very rarely and only when OC). I'm using the same vcore and clock as I always have been and never had any issues. Is my chip just getting old and ornery?
|
# ? Jun 3, 2015 04:14 |
|
Don Lapre posted:Unless your voltage is hitting over 1.4 you dont need to worry about frying your chip. If your cpu hits 100-105c it will just throttle itself back, it wont kill itself. So do people just let their CPU bash against the temperature limit for hours at a time while running prime95? Because I haven't seen the temps go higher than 65 for anything else, but prime95 still made it climb all the way up. The only other intensive thing I do is games, though, and I also haven't seen any of them really max the CPU at this speed. Also, thanks for answering all my dumb questions. I haven't done this since the Athlon 2500+ days and a lot has changed.
|
# ? Jun 3, 2015 18:09 |
|
What stronger solvent should I use on stubborn thermal compound? Not talking about enthusiast stuff but 10 year old Dell weird stuff in a laptop. Mineral spirits? Acetone?
|
# ? Jun 3, 2015 21:40 |
|
Shaocaholica posted:What stronger solvent should I use on stubborn thermal compound? Not talking about enthusiast stuff but 10 year old Dell weird stuff in a laptop. Mineral spirits? Acetone? Tried a thin piece of plastic like a credit card (if it's hard), or 99%-100% isopropyl alcohol?
|
# ? Jun 3, 2015 22:14 |
|
Sidesaddle Cavalry posted:Tried a thin piece of plastic like a credit card (if it's hard), or 99%-100% isopropyl alcohol? Yeah the alcohol isn't working well. It works well on the die but not on the heatsink side.
|
# ? Jun 3, 2015 22:16 |
|
Shaocaholica posted:Yeah the alcohol isn't working well. It works well on the die but not on the heatsink side. Yeuch, that must be an additional layer of something caked on then. Whatever you do, don't use acetone. Acetone can damage a lot more than you ask for. EDIT: Looking around, I'm seeing some benzene derivatives like Goof-Off? Supposed to make sure to use that stuff outside. Sidesaddle Cavalry fucked around with this message at 01:05 on Jun 4, 2015 |
# ? Jun 4, 2015 01:02 |
|
Some coolers come with something put on them initially that seems like a firmer compound than the thermal paste we put on the CPU die. It's probably that way because it doesn't have to spread out - comes at a set thickness in a perfect square. Try tossing the cooler in the oven (low temp obviously) and then taking it out and carefully trying to remove that poo poo?
|
# ? Jun 4, 2015 01:07 |
|
Sidesaddle Cavalry posted:Tried a thin piece of plastic like a credit card (if it's hard), or 99%-100% isopropyl alcohol? 90/91% will do in a pinch (you can't get 99 some places, and I remember Bude having to get 91 from an actual pharmacist with actual 'do not ingest' labels because Australia is even more afraid of showing up in the OSHA thread than California is). 70% is ... not a joke, it can get the job done eventually, but that's like three times as much cleaning as 91. Benzene is terrifying.
|
# ? Jun 4, 2015 05:42 |
|
Well got it off with acetone. Not the nail polish stuff but the hardware store tin kind. Didn't hurt any of the metal using q-tips. Not sure why it would, its metal. Sometimes 91% just isn't enough. Not for OEM never-meant-to-be-serviced stuff.
|
# ? Jun 4, 2015 06:13 |
|
I was fearful for any additional plastic or PCB pieces the sink would have been attached to but I guess it shows how much I know about laptop heat sinks
|
# ? Jun 4, 2015 07:36 |
|
I got a couple gallon jugs of 99% iso from Amazon I think for maybe $30 each, I run an IT shop though and beyond thermal paste I also use them if someone spills a soda into a solid-state laptop. Give the lower half of the laptop an iso bath then hit it with one of those high powered Metro dusters and it comes out fine.
|
# ? Jun 7, 2015 16:23 |
|
Just wanting to chime in to say that I don't know poo poo about overclocking, but with a little advice from the OP and google I just got a Pentium G3258 with an Asrock Z97 Pro4 mobo and this chip is special. On stock cooling I have this 3.2 GHz CPU at 4.2 GHz with absolutely no problems. ya, it's a cheapo with core-count limitations, but for $70 it's god drat good.
|
# ? Jun 10, 2015 06:38 |
|
Dza posted:Just wanting to chime in to say that I don't know poo poo about overclocking, but with a little advice from the OP and google I just got a Pentium G3258 with an Asrock Z97 Pro4 mobo and this chip is special. On stock cooling I have this 3.2 GHz CPU at 4.2 GHz with absolutely no problems. ya, it's a cheapo with core-count limitations, but for $70 it's god drat good. I've overclocked at least 5 of those G3258 now, and all of them will hit 4.4 or 4.5ghz stable at 1.3v with the stock fan, so you know. With a liquid cooler I have one at 4.7 on 1.35v.
|
# ? Jun 10, 2015 18:56 |
|
.
Don Lapre fucked around with this message at 19:17 on Jun 10, 2015 |
# ? Jun 10, 2015 19:07 |
|
Zero VGS posted:I've overclocked at least 5 of those G3258 now, and all of them will hit 4.4 or 4.5ghz stable at 1.3v with the stock fan, so you know. With a liquid cooler I have one at 4.7 on 1.35v.
|
# ? Jun 11, 2015 05:08 |
|
The cooler master 212 evo won't fit my my case. What's the next best thing around the same price point? I want to do basic over clocking with my i5 4690k.
Boris Galerkin fucked around with this message at 18:55 on Jun 12, 2015 |
# ? Jun 12, 2015 18:52 |
|
Boris Galerkin posted:The cooler master 212 evo won't fit my my case. What's the next best thing around the same price point? I want to do basic over clocking with my i5 4690k. With that case you might be alright with a corsair h50.
|
# ? Jun 12, 2015 19:31 |
|
My GTX970 G1 is limited by voltage, can't push the core past 1583 Mhz, I'm ready to reflash but what would be a decent Vcore limit ?
|
# ? Jun 12, 2015 21:11 |
|
Sh4 posted:My GTX970 G1 is limited by voltage, can't push the core past 1583 Mhz, I'm ready to reflash but what would be a decent Vcore limit ? What software are you using to OC? I think only going +0.12mv is what is recommended right now for those cards. On EVGA precision it will only allow that much overvolting for that chipset I think and unless you're watercooling it you'll have thermal throttling before you need more voltage anyways. What have you done to your memory clock?
|
# ? Jun 12, 2015 21:17 |
|
VelociBacon posted:What software are you using to OC? I think only going +0.12mv is what is recommended right now for those cards. On EVGA precision it will only allow that much overvolting for that chipset I think and unless you're watercooling it you'll have thermal throttling before you need more voltage anyways. What have you done to your memory clock? Have the following power curve, Vcore is hard capped at 1.2620 V, I'm a long way from the thermal limit as the card never go past 69°C. Memory is at 3800 Mhz, didn't try to max it yet, working on core for now. Not sure what +0.12mw means since the hundreds of models seems to have different voltage curves
|
# ? Jun 12, 2015 21:31 |
|
Anyone know the recommended max temps/vcore acceptable for Phenom ii x4 series chips? I've got a fantastic 955 chip which is running 4.0Ghz at 1.375v (stock is 3.2Ghz at 1.350v), around low 50's Celsius under full load. I'm reading that I can safely go up to 1.5v but that seems a bit high, or maybe I'm used to intel chips. Replacing everything soon and going to push for benchmarks.Sh4 posted:Have the following power curve, Vcore is hard capped at 1.2620 V, I'm a long way from the thermal limit as the card never go past 69°C. Memory is at 3800 Mhz, didn't try to max it yet, working on core for now. Not sure what +0.12mw means since the hundreds of models seems to have different voltage curves I think for that card you'll just have to change base clock and go from there, EVGA precision has a handy offset slider that would work well. At the end of the day these new nvidia cards seem to throttle themselves even when temps are fine. E: for those interested (in ancient processors I guess) I found the recommended max vcore for the deneb phenom ii chips is 1.5v. Able to get a stable 4.1ghz at 1.425, peak temps at 53C. Apparently around 59C people see a lot of instability. VelociBacon fucked around with this message at 10:10 on Jun 19, 2015 |
# ? Jun 19, 2015 07:13 |
|
To preface I have a i5-4690k and am using an asus z97-a motherboard, I also have some kind of 3x120mm thermaltake `*~xtreme~*' closed loop cooler. On stock settings under load I barely get to 55c with prime-95's small fft test so I figure my system is safe for overclocking. I've changed the core ratio to 45, but let the automatic voltage settings do their thing. This leads to a CPU core voltage of 1.296 on 3 cores and 1.319 on the other (core 2 for some reason is special?). I see just baaaarely 80c when I do prime's small FFT test using these settings, it's more like 76-78. This is a temperature I feel safe to have for a 24/7 overclock. Especially because I've seen it multiple places that the small FFT tests are particularly unfair to the haswell's (meaning i am unlikely to ever put the cpu under that kinda load just playing video games or even compiling software) If I manually do the voltage myself am I likely to see that I can substantially lower the voltage, meaning less heat and more overclocking headroom? I've read things start getting out of hand at 1.4v so i'm under the impression I have another tenth of a volt to mess around with. Am I likely to get much more out of this cpu and keep it 24/7 stable? I've been googling this for a couple days just seeing where others have gotten theirs, and there are a lot of 4.6 and 4.7 ghz overclocks from people on air. Is it unreasonable to expect to get close to 5ghz with a "slightly better than average" closed loop cooler? I'm aware this is a very subjective question (and also nobody can predict what my individual cpu can do), but I don't really trust reading someones post on tomshardware, or overclockers.co.uk because a lot of people probably lie about what they've got for the pissing contests going on at those forums. Am I likely to go much farther with this and keep it stable?
|
# ? Jun 24, 2015 20:19 |
|
Incomplete Fish posted:To preface I have a i5-4690k and am using an asus z97-a motherboard, I also have some kind of 3x120mm thermaltake `*~xtreme~*' closed loop cooler. On stock settings under load I barely get to 55c with prime-95's small fft test so I figure my system is safe for overclocking. I guess it will be similar to my 3390K, these cpus have a ceiling that is around 4.6-4.8 depending on the silicon quality, I have a pretty high end watercooling setup so heat is not a problem but can't get mine past 4.7Ghz it is very unstable and at 4.8+ it will not even go past windows boot even with 1.6V. From what I've seen the luckiest people are able to boot at 5Ghz with 1.5-1.6V but it's not something you'd use every day. Also for some reasons Windows 10 is a lot more sensible to overclocking than W7, I had a stable OC at 4.7Ghz that was stress tested but on W10 I have to go back to 4.5Ghz because of constant blue screens, no idea why.
|
# ? Jun 24, 2015 20:41 |
|
I'm updating my video card soon and I wanted to ask in here what people think the ideal cooling setup would be in my system? I was thinking of the evga 980ti hybrid, and figured this setup might work best. It's has a reference fan + closed loop. Currently I have the top radiator reversed, pushing air from inside the case out through the top, but I feel like this is hurting the effectiveness of the cooling. Also, if I ended up just getting a non reference 980ti, what setup do you think would work better? Edit: Also, Factory Factory, if you want the .psd of that image to mess around with and update the pictures in the OP, let me know. Cheap Shot fucked around with this message at 17:50 on Jun 25, 2015 |
# ? Jun 25, 2015 17:45 |
|
Cheap Shot posted:I'm updating my video card soon and I wanted to ask in here what people think the ideal cooling setup would be in my system? Put the video card water cooler on the back 120mm fan mount. The video card radiator will put out a metric poo poo ton of heat. If you had to pick gpu or cpu rad in the front id do the cpu as the water temp in that rad will barely change.
|
# ? Jun 25, 2015 17:51 |
|
Don Lapre posted:Put the video card water cooler on the back 120mm fan mount. The video card radiator will put out a metric poo poo ton of heat. If you had to pick gpu or cpu rad in the front id do the cpu as the water temp in that rad will barely change. wow fast, thanks! So then this would be more effective? (The double rad will only fit at the top to my knowledge) and should I configure the top rad as push out or pull in? Cheap Shot fucked around with this message at 18:10 on Jun 25, 2015 |
# ? Jun 25, 2015 18:06 |
|
Cheap Shot posted:wow fast, thanks! So then this would be more effective? (The double rad will only fit at the top to my knowledge) Try it both ways though id probably have it exhaust air out. Remove any filter material that may be on top of the case if you have it exhaust air out. have front intake and top and back exhaust.
|
# ? Jun 25, 2015 18:17 |
|
Don Lapre posted:Try it both ways though id probably have it exhaust air out. Remove any filter material that may be on top of the case if you have it exhaust air out. Right on. The top has a metal mesh that I can't remove but I'll see which way works best. Thank you!
|
# ? Jun 25, 2015 18:18 |
|
Cheap Shot posted:wow fast, thanks! So then this would be more effective? (The double rad will only fit at the top to my knowledge) Depending on how much airflow you have coming in the front I would either go for the H100i as intake or outtake. If you have enough airflow from the front to have positive pressure even with all three radiator fans exhausting (you'd have more fans exhausting but theyd all be on rads which hamper their airflow), then I would have the radiator as outtake. If you don't have enough airflow for there to be positive airflow then I would have the h100i as intake to make sure you have positive pressure. It'll only contribute ~90w or so to the air in the case, whereas the 980ti is more like 275w. Also your artwork is always so drat well done, I love it.
|
# ? Jun 26, 2015 08:56 |
|
BurritoJustice posted:Depending on how much airflow you have coming in the front I would either go for the H100i as intake or outtake. If you have enough airflow from the front to have positive pressure even with all three radiator fans exhausting (you'd have more fans exhausting but theyd all be on rads which hamper their airflow), then I would have the radiator as outtake. If you don't have enough airflow for there to be positive airflow then I would have the h100i as intake to make sure you have positive pressure. It'll only contribute ~90w or so to the air in the case, whereas the 980ti is more like 275w. Thank you :3 Well, the front of the case is sporting two large-ish fans. I'm feeling that in the event that a hybrid is outside my budget, I might drop down to a non reference card. In that case the back fan wouldn't be a rad. Im assuming i'd need to reverse the top radiator to be an intake in order to keep positive pressure. My main concern would be keeping GPUs temps down, so would that be the best solution outside dropping the extra $$$ on that hybrid?
|
# ? Jun 26, 2015 16:28 |
|
Cheap Shot posted:Thank you :3 If you are buying a non ref card make the top exhaust to get all the hot air out of the case.
|
# ? Jun 26, 2015 16:38 |
|
Cheap Shot posted:Thank you :3 What case are you using exactly? Getting a good aftermarket air 980ti (get the MSI Gaming, its the best) is an easy way to save a bit of money, they are typically actually quieter at idle due to the lack of a pump (the MSI goes fully passive until a certain load level) and are competitive under load (within 1-2dBA). The temps will of course be higher, but you'll likely be limited by the power/voltage limits before you are temp limited. Assuming there are good front fans I would just leave the back fans as exhaust and not worry about it too much. You can min max in the extreme, but at the end of the day the minor differences aren't going to kill you. E: Also what Don Said.
|
# ? Jun 26, 2015 16:41 |
|
BurritoJustice posted:What case are you using exactly? Corsair 540, so come to think of it, I didn't need to include the psu in that picture. Thanks for all the info guys!
|
# ? Jun 26, 2015 22:09 |
|
If you want positive air pressure at all costs, I have a pair of lightly used Scythe Ultra Kaze 120mm hurricanes up for grabs.
|
# ? Jun 26, 2015 23:01 |
|
Cheap Shot posted:Corsair 540, so come to think of it, I didn't need to include the psu in that picture. The air 540 ships with two fairly high airflow 140mm fans at the front (they actually have higher speeds than I'd really like, subjectively) so positive pressure shouldn't be a problem.
|
# ? Jun 27, 2015 02:57 |
|
hey if you wanna stare at your own anus in the mirror all day im okay with that
|
# ? Jun 27, 2015 08:30 |
|
So I recently picked up a 212 EVO to be used with my 3570k on a gigabyte Z77X-U3DH. I've been running it at the stock 3.4/3.8 with the stock cooler (no heat issues I've noted while checking occasionally). I decided recently to see about OCing to the 4.2-4.5ghz range, along with buying some newer, non-1.65v RAM so I could stick with it for another year or 2. I've looked over the OP and some decent tutorials, but I'm a little lost with the different BIOS settings and just being completely new to this. I'd like to maintain the clock scaling element as mentioned in the OP but it's not really well laid out anywhere, and it seems like just upping the mulitplier won't do that, unless I'm misunderstanding. I'm also a little confused on the voltage settings and how that's all handled. This is mostly concerning because I've seen that 1. the gigabyte AUTO setting can push it pretty high (1.3+), and 2. IB doesn't like going that high without getting really loving hot. I guess the better question is can anyone point me to or write me out babby's first guide for dealing with this UEFI? Here's a good youtube video of the Z77X UEFI for more specific information (although with a 3770 mounted). https://www.youtube.com/watch?v=pMpI7SMGRVU I can also try to get some other photos of it if needed. Sidenote, The RAM I bought was 2x8G Mushkin Stealth DD3L, currently running at 1.35v/1600mhz through the XMP profile. Mazz fucked around with this message at 23:17 on Jul 7, 2015 |
# ? Jul 7, 2015 22:54 |
|
Mazz posted:So I recently picked up a 212 EVO to be used with my 3570k on a gigabyte Z77X-U3DH. I've been running it at the stock 3.4/3.8 with the stock cooler (no heat issues I've noted while checking occasionally). I decided recently to see about OCing to the 4.2-4.5ghz range, along with buying some newer, non-1.65v RAM so I could stick with it for another year or 2. With your processor you want to aim for a Vcore of under 1.300v, and temps under load of <75C (up to 80C would likely be fine but under 75C is safe). Basically you'll want to go into your BIOS, find where your multiplier is, and where your CPU Vcore setting is. A good place to start might be to set your vcore to MANUAL so that you can find some stable settings without the mobo modifying your vcore to unsafe levels. Set that vcore to 1.2v to start. So: 1) Set Vcore to 1.2v 2) Set Multiplier to give you a final ghz of 3.8, which is the highest you have at present and is 'stock'. 3) Run your stability testing software, record the values. CHECK YOUR TEMPS AND VCORE 4) If temps are ok and the system is stable with your testing, go back to BIOS, increase your multiplier to give you +0.1ghz and do another run of software stability testing. 5) If it is stable at this increased clock, and the temps are still ok, go back to BIOS and increase it again until you hit instability. If you hit temp limits but it's not crashing, that's the overclock limit of your build. -> If your software tests are crashing the pc or software, it's not stable, go back to BIOS and increase your vcore by 0.010v increments until it is stable or your temps are too high, in which case back off the clock to the last stable clock with allowable temps. Sort of a simplified list but that's where I would start. I'm not super familiar with your chip or BIOS.
|
# ? Jul 8, 2015 01:41 |
|
|
# ? Jun 5, 2024 04:08 |
|
Double posting with a question a day or so later: is there really no way to set an adaptive vcore ceiling on ASUS UEFI motherboards? I've dialed in my stable overclock (4.7ghz @1.220v on air, 4790k on a Maximus vii Hero) but adaptive keeps raising my vcore to 1.240v regardless of what I set. I'm on the latest bios.
|
# ? Jul 9, 2015 14:40 |