|
Well, I hope it's not the CPU, or your Haswell dream is at an end. Didn't you direct die it, too? HalloKitty fucked around with this message at 10:31 on Jul 16, 2013 |
#
?
Jul 16, 2013 10:27
|
|
|
|
| # ? May 13, 2024 08:20 |
|
|
HalloKitty posted:Well, I hope it's not the CPU, or your Haswell dream is at an end. Yeah I did direct die but I have a feeling its the mobo. No funny smells when it hard shutdown. I also do have dGPUs. I'll try taking them out.
|
|
#
?
Jul 16, 2013 14:51
|
|
|
So as it turns out I think my mobo AND CPU are dead. The CPU is probably gone and not covered since I delidded which I was prepared for. The mobo should be RMA-able but I'm not sure who to do it through. I got it off newegg. Should I RMA with newegg or directly with asus? edit: oh nm. Its been just over 30days so probably asus. Shaocaholica fucked around with this message at 15:52 on Jul 17, 2013 |
|
#
?
Jul 17, 2013 15:48
|
|
|
If it's within 30 days, Newgg. Otherwise, Asus.
|
|
#
?
Jul 17, 2013 15:49
|
|
|
Well, after buying 'temporary' PSU, mobo and CPU, I've determined that the CPU is done for and so is the mobo. Didn't think I was pushing the CPU that hard. 1.32V@4.5Ghz. Was running in the mid 80s direct die on an H110. The last thing I was doing was a prime95 stress test after I had to reseat the water block due to core3 having much higher temps than the other cores just out of the blue. After reseating and new TIM core3's temp was back to normal and prime95 ran fine for maybe 60min then the whole system just hard powered down. Anyway just sharing the info. Also, with the dead CPU in the good mobo, the whole system wouldn't power up at all. Just 'clicked' the PSU. edit: well I'm wrong. Die is cracked(not chipped) after some further examination. Not sure right now if the block is still uneven/concave or if its from heat stress. The last intensive thing I did before reseating was processing a few thousand files which was causing all cores to ramp up and down. The downtime being attributed to disk IO so the CPU was heating up and cooling down a few thousand times. Shaocaholica fucked around with this message at 02:14 on Jul 18, 2013 |
|
#
?
Jul 18, 2013 02:09
|
|
|
Yeah, I think I'll stay on my 2600K for awhile yet. Not ready to delid and go through all that plus the risk of a cracked CPU die. Already went through that with enough videocards.
|
|
#
?
Jul 18, 2013 02:47
|
|
|
Yeah the replacement CPU will still be delidded but not direct die. I don't trust the H110 block.
|
|
#
?
Jul 18, 2013 02:53
|
|
|
Delidding is perfectly safe if you're handy and make sure to space your razor beforehand; direct-to-die is some daredevil poo poo that, well, doesn't work out for a lot of reasons  I've got two scenarios for my incoming 4770K based build: one is I luck out like one of our posters did and end up with a chip that'll go to 4.5GHz or higher (performance parity in games, superior performance in applications to my 4.7GHz Sandy Bridge i7). That's regular install, nothing fancy except about $120 worth of heat sink and another $125 in fans for the case  If it won't hit that and I'm temperature walled, I'll delid it (safely) and reseat the IHS properly to grab that extra 10-20ºC and go from there. So long as I can achieve performance parity with my 2600K at 4.7GHz in games and benefit from the additional features in applications and chipset improvements, I'm solid on the upgrade. A 2600K is not a necessary upgrade for most uses right now at all, but I am looking to expand my audio work and I need more PCI-e bandwidth in addition to more SATA lanes, so I'm doing an overhaul that will have the benefit of giving my wife a badass computer she can make lesson plans with etc. without having to fuss with a rather pokey laptop.Regardless - no way in hell am I going direct to die in this day and age, with all the crap you have to do to make that even work, let alone make it safe, no offense intended it's just drat that's way more extreme than necessary for standard "high clocks" stable overclocking - that's more into the realm of, y'know, at least dry ice, maybe liquid nitrogen cooling suicide runs. Bad way to kill a good chip if you're not immaculately careful and/or have an effectively bottomless budget (like the Gigabyte in-house OC team, with their busted Titan PCBs laying around makin' people sad). Agreed fucked around with this message at 05:32 on Jul 18, 2013 |
|
#
?
Jul 18, 2013 05:29
|
|
|
Jeeze, was it the i5 or i7? Either way, I'm definitely waiting for Haswell-E, which should still be soldered. My poor old 2600k is still fine.
|
|
#
?
Jul 18, 2013 08:04
|
|
|
Yeah I'm slightly bummed I didn't/couldn't wait for Haswell-E(if it is indeed soldered).Agreed posted:Regardless - no way in hell am I going direct to die in this day and age, with all the crap you have to do to make that even work, let alone make it safe... I'm 80% confident in saying that my die crack was due to the shittiness of the H110 block. I would totally direct die again with a premium waterblock. I'm not sure if that's on par with LN2 but for me, its a lot easier to watercool these days than in the 'good old days' that I personally remember. Shaocaholica fucked around with this message at 15:33 on Jul 18, 2013 |
|
#
?
Jul 18, 2013 15:29
|
|
|
Shaocaholica posted:Yeah I'm slightly bummed I didn't/couldn't wait for Haswell-E(if it is indeed soldered). Now that I'm thinking about it more, it seems like if you're going to delid the CPU, you could probably replace the TIM with something like Collab. liquid metal or Arctic Silver Alumina if you were careful to avoid any exposed resistors. The Collab. stuff tends to form a bond between heatsink and IHS, and Alumina's a thermal adhesive. Normally both of these would be problematic if you were using them under a CPU heatsink, however if you're trying to replace the TIM on the CPU die it could keep the IHS from moving around after you've delidded. The downside is that you can't really remove the IHS later without cracking the core (in the case of Alumina), so you couldn't reapply the TIM later if necessary. The liquid metal stuff would still allow TIM replacement while keeping the IHS in place during installation though. future ghost fucked around with this message at 18:47 on Jul 18, 2013 |
|
#
?
Jul 18, 2013 18:33
|
|
|
Shaocaholica posted:I'm 80% confident in saying that my die crack was due to the shittiness of the H110 block. I would totally direct die again with a premium waterblock. I'm not sure if that's on par with LN2 but for me, its a lot easier to watercool these days than in the 'good old days' that I personally remember. I say this with love in my heart, so please don't take it the wrong way, but while that may very well be true (it was quite a lot of effort to try to get it to mount correctly) - I am 100% sure that it's cracked because of your decision to go direct-die in the first place
|
|
#
?
Jul 18, 2013 18:44
|
|
|
Agreed posted:I say this with love in my heart, so please don't take it the wrong way, but while that may very well be true (it was quite a lot of effort to try to get it to mount correctly) - I am 100% sure that it's cracked because of your decision to go direct-die in the first place I'm not going to argue with you there but you could also say its cracked because I decided to build a new computer
|
|
#
?
Jul 18, 2013 19:41
|
|
|
It's cracked because of the boundary conditions of the universe. There's nothing anyone could have done. Literally, because free will is an illusion.
|
|
#
?
Jul 18, 2013 19:46
|
|
|
Schrodinger's CPU
|
|
#
?
Jul 19, 2013 14:24
|
|
|
go3 posted:Schrodinger's CPU We observed it and compressed the waveform, though, so not anymore.
|
|
#
?
Jul 19, 2013 17:03
|
|
|
go3 posted:Schrodinger's CPU This cat is definitely dead though GOD drat IT DOGEN
|
|
#
?
Jul 19, 2013 17:09
|
|
|
All is fixed now. Heres the lovely part. CPU AND mobo were dead. In addition, the dead mobo would eat working CPUs so I ended up burning a replacement 4770K testing in the bad mobo. Good thing Frys had it covered. Oh fun times. vvv Yes but it was the ONLY way to be SURE. Also the chip wasn't a great sample anyway vvv Shaocaholica fucked around with this message at 17:50 on Jul 19, 2013 |
|
#
?
Jul 19, 2013 17:34
|
|
|
Shaocaholica posted:All is fixed now. Heres the lovely part. CPU AND mobo were dead. In addition, the dead mobo would eat working CPUs so I ended up burning a replacement 4770K testing in the bad mobo. Good thing Frys had it covered. Oh fun times. You're KILLING them, man, my god, you're killing them 
|
|
#
?
Jul 19, 2013 17:42
|
|
|
I was not planning on overclocking when I put together my current PC, although I did spend $10 to get the i5 5670K over the 5670 "just in case" and because I figured it'd have better resale value someday. I bought an ASRock H87 board that seemed to be a good value and it turns out that it's one of the ones that supports the "non-Z OC" overclocking on the H87 chipset so I think I might actually give that a shot. I only have a stock Intel cooler at the moment but I have a Hyper 212 Evo on the way, and will attempt to do some H87 overclocking once I get that installed. Anyone else tried overclocking on an H87 board yet? I think ASRock was the first but apparently most vendors have followed suit with BIOS updates to allow overclocking on their H87 motherboards. Other than actually having the ability to change the CPU multiplier, was there anything that made Z87 boards more suitable for overclocking than H87 boards? Parker Lewis fucked around with this message at 20:22 on Jul 19, 2013 |
|
#
?
Jul 19, 2013 20:19
|
|
|
Generally Z87 boards will be built-out further in the portion of the VRMs that sits on the motherboard. For mild to moderate overclocks, this doesn't make much difference, but for moderate-to-beefy overclocks, that prevents board components from overheating or, y'know, starting dramatic electrical fires. ASRock made a few boards built for Non-Z OC that are a tad beefier, like the Fatal1ty-branded board that got the very first Non-Z OC update, but in general a top-end H87 board only handles as much electrical power and heat as a bottom-of-the-barrel Z87 board.
|
|
#
?
Jul 19, 2013 20:49
|
|
|
So I have a 4670K and an Asus Z87M Plus. I've been sitting at 4.2 and it's been solid but is 1.28 core voltage too high? Noctua DH-14 for cooling btw.
|
|
#
?
Jul 20, 2013 03:58
|
|
|
That's absolutely fine. It's near the tippity top of where you want to be voltage-wise, but it's safe to run that 24/7 until you get sick of the machine and junk it.
|
|
#
?
Jul 20, 2013 05:55
|
|
|
H87 Overclocking: it works! I received and installed my Hyper 212 Evo last night and tried out the "Non-Z OC" features of the new BIOS for my ASRock Fatal1ty H87 Performance with my 4670K. Turning Non-Z OC on prompts you to select your desired frequency from a menu of 4.0, 4.2, 4.4 and 4.6 GHz, with the latter two options being listed in red. I selected 4.2 because I figured that was good enough for me and so far it's been 100% stable after a few hours of gaming. I haven't done any actual CPU stress testing yet but so far things are looking really promising for this $105 motherboard.
|
|
#
?
Jul 24, 2013 14:13
|
|
|
Hopefully whatever Intel's trying to pull to stop this won't be forced down other boardmakers' throats so this trend can keep on going. When you can slap the letter Z on any board with 3 CPU phases and then say it's for overclocking, there's some serious voodoo being used. Weren't we promised strap clocking was going to be a feature for Haswell chips earlier this year?
|
|
#
?
Jul 24, 2013 14:45
|
|
|
Well it is. For K SKUs only.  AnandTech did a podcast with one of their writers, who is not great at podcasts because it's his first one, hashing out complaints about Haswell. It's a good listen regardless, if you have the time, but re: -K SKU overclocking, it basically comes down to "Yeah, delidding is only worth about 200 MHz on average, but the fact that you have to delid to get that 200 MHz feels like a slap-on-the-face piece of  by Intel."
|
|
#
?
Jul 24, 2013 14:55
|
|
|
Sidesaddle Cavalry posted:Hopefully whatever Intel's trying to pull to stop this won't be forced down other boardmakers' throats so this trend can keep on going. When you can slap the letter Z on any board with 3 CPU phases and then say it's for overclocking, there's some serious voodoo being used. Weren't we promised strap clocking was going to be a feature for Haswell chips earlier this year? I'm interested to see how Intel plans on trying to stop the non-Z87 overclocking if they in fact are going in that direction. I guess I could see them introducing something new in future parts which could introduce some interesting "pre-overclock" and "post-overclock" split but surely the cat's out of the bag for people like me who are already running it? It's not like they're going to enforce some sort of mandatory BIOS update.
|
|
#
?
Jul 24, 2013 15:34
|
|
|
If you want to run Haswell Refresh, they will, I bet. And when that happens, they'll very quickly stop producing first-gen Haswell.
|
|
#
?
Jul 24, 2013 15:47
|
|
|
drat Intel is evil.
|
|
#
?
Jul 25, 2013 05:37
|
|
|
Well, they're a business, preserving market segmentation is in their own interest. We're just mad as nerds because enthusiast/gaming PCs aren't getting any cheaper.
|
|
#
?
Jul 25, 2013 17:24
|
|
|
I followed this guide: http://www.overclockers.com/3step-guide-to-overclock-intel-haswell But I have only messed around with the clock speed. With a 4670K, Gryphon and a Swiftech H220 I was able to go to 4.7GHz at 1.25. But while running AIDA64 it spiked above 90C. Prime95 was about 70C and slightly louder than what is acceptable to me. So I dialed it back to 4.6 and +0.199 adaptive should be below 1.2 effective but because of how adaptive works it spikes up to 1.279 generating a Huge Civ5 map. Prime95 and AIDA64 max out at 1.352. It's stable so far and stays quiet under normal usage, Prime95 makes the fans spin up a little to keep it at about 65C. AIDA64 still gives it serious workout, with spikes in 80-90 range but averages in the mid 70s. Is this safe? Should I bother with any of the other settings?
|
|
#
?
Jul 27, 2013 10:54
|
|
|
Alright, so I'm here for one specific purpose: running my 4670K at 4.8GHz at full load results in an estimated draw of about 90W. Is that correct/normal? The voltage never exceeds 1.297. Temps under that sort of load settle in the 65-70C range (cooler is a Hyper 212 EVO). You guys have told me numerous times that I won the lottery, and I believe you, I just want to make absolutely certain that there's nothing horrifically wrong with these numbers that I may not understand. For reference: 4670K @4.8GHz, voltage under load is 1.297v ASUS Z87-A mobo Hyper 212 EVO 65-70C full load Gonkish fucked around with this message at 11:23 on Jul 27, 2013 |
|
#
?
Jul 27, 2013 11:20
|
|
|
If it benched okay (i.e. 24hr Prime95) and it remains stable then the only thing you have to worry about are goons breaking in and stealing your rig. Also we want to know how you applied the thermal goop - unless it would require you to remove your heatsink, because you should probably never remove your heatsink again with that kind of luck - and what other parts you used (exotic fans, cases, blood pact with Satan) because holy poo poo that shouldn't be possible.
|
|
#
?
Jul 27, 2013 11:50
|
|
|
So I've read the first and last few pages, and come here to see if I should give overclocking a try. I'm enticed as all hell for the EXTRA FREE GRAPHIXXX that I could be getting, but also equally scared by all the processes and jargon (and also burning my house down). Can someone who's never tried overclocking before (but has an idea of what's going on) overclock a brand new rig, or is it something that should be 'eased' into or some such? I'm hoping to have a new PC bought and built within a few weeks, and I'm totally willing to sit down a learn whatever I need to and spend as much time as is necessary to test stability or whatever, so time and effort aren't the problem, just total lack of experience. Also sorry if this post is dumb.
|
|
#
?
Jul 27, 2013 12:13
|
|
|
Sir Unimaginative posted:If it benched okay (i.e. 24hr Prime95) and it remains stable then the only thing you have to worry about are goons breaking in and stealing your rig. Yeah, it's benched perfectly fine. The only time I get some instability is if I increase the multiplier to x49 (and even then, it's only after about half an hour of load). It stays perfectly stable at 4.8, though. Been running that since the first day, which was at the end of last month. No problems to speak of. As for thermal goop, nothing special really, I just used whatever thermal paste came with the Hyper 212 EVO; just a small dollop of it on the center of the CPU. Fans are just the 120mm Cooler Master that came with the heatsink, and then a Bitfenix Spectre 120mm, with both setup in a push/pull config. Aside from that, the case is just a Corsair Carbide 500R, which has two 120mm intakes on the front, a huge 200mm intake on the side panel (blowing right onto the PCI-E slots), a 120mm exhaust at the rear, and a 140mm Bitfenix Spectre Pro that I added on the floor of the case as an intake. I could actually add two 120mm or 140mm fans up on the top, but I don't feel a pressing need to do so. (I'm not counting the PSU because it's bottom-mounted and exhausts outside of the case, so it's not really aiding/inhibiting airflow.) The case has loving excellent airflow, and cable routing is fantastic, so cables messing up airflow isn't an issue. I removed the top hard drive cage as I wasn't using it, so the 120mm front fan on the top is blowing directly towards the PCI-E slots. Aside from that, everything is stock. Specs, for reference: 4670K @4.8GHz ASUS Z87-A mobo 8gb RAM (DDR 1600, 1.35v) 180GB Intel SSD 1TB WD Black HDD 500GB WD Blue HDD GTX 560Ti 1GB (which is going to get replaced soon enough) So, basically, I chalk it up to good air conditioning, a good case, and sheer dumb luck.
|
|
#
?
Jul 27, 2013 15:11
|
|
|
Gonkish posted:Alright, so I'm here for one specific purpose: running my 4670K at 4.8GHz at full load results in an estimated draw of about 90W. Is that correct/normal? The voltage never exceeds 1.297. Temps under that sort of load settle in the 65-70C range (cooler is a Hyper 212 EVO). You guys have told me numerous times that I won the lottery, and I believe you, I just want to make absolutely certain that there's nothing horrifically wrong with these numbers that I may not understand. Where are you getting that estimated power draw? If it's from HWiNFO, that looks to be about right for what the tool says about such an overclock, but it's definitely not accurate to marginal increase in power draw at the wall for that voltage, which is, to a first approximation, a doubling of the chip's 84W TDP. SSGT Anime posted:So I've read the first and last few pages, and come here to see if I should give overclocking a try. I'm enticed as all hell for the EXTRA FREE GRAPHIXXX that I could be getting, but also equally scared by all the processes and jargon (and also burning my house down). Can someone who's never tried overclocking before (but has an idea of what's going on) overclock a brand new rig, or is it something that should be 'eased' into or some such? I'm hoping to have a new PC bought and built within a few weeks, and I'm totally willing to sit down a learn whatever I need to and spend as much time as is necessary to test stability or whatever, so time and effort aren't the problem, just total lack of experience. There have been success stories ITT about people who had no clue, read through the OP and locked in the idea of what should be done, and then clocked the poo poo out of their chips. Overclocking is primarily information-intensive and then procedure-intensive, and the information is so you know what procedure to use. The OP is "why you do things" followed by "okay, now here are some specifics to apply what you've learned." Other guides usually focus on "Here's a detailed listing of steps to take." Either works and will get you where you want to go. As a building thread poster, I only tend to recommend against overclocking (or avoid talking it up) when someone is legitimately confused when it comes to computers. The most important things to bring to the table are an open mind and a willingness to learn. E: Also, if you're using a 700-series GeForce, the GPU thread OP reposted Agreed's excellent guide to navigating GPU Boost 2.0. Factory Factory fucked around with this message at 16:09 on Jul 27, 2013 |
|
#
?
Jul 27, 2013 15:58
|
|
|
Factory Factory posted:Where are you getting that estimated power draw? If it's from HWiNFO, that looks to be about right for what the tool says about such an overclock, but it's definitely not accurate to marginal increase in power draw at the wall for that voltage, which is, to a first approximation, a doubling of the chip's 84W TDP. It's from HWiNFO. It fluctuates a bit, but that's about what it stays at.
|
|
#
?
Jul 27, 2013 16:11
|
|
|
Ok, I have a question about Vcore and adaptive mode. I've settled on an overclock of 4.4 GHz @ 1.2v on my 4670k. I set the multiplier to 44 and set my voltage to 1.2 with adaptive mode on. However, according to hwinfo64 the voltage never drops below 1.2v and the clock speed doesn't go below 4.4GHz. I was under the impression that the cpu would ramp down its voltage and clock speed according to its needs? I've made sure that the minimum processor state was set to 5% under power options and it still won't budge below 1.2v and 4.4GHz. I tested under stock settings and both the clock speed and voltage were ramping up and down as needed. I tried using CPU-Z to see my vcore and it lists it at .888 which obviously isn't right.
|
|
#
?
Jul 31, 2013 07:20
|
|
|
BIFF! posted:Ok, I have a question about Vcore and adaptive mode. I've settled on an overclock of 4.4 GHz @ 1.2v on my 4670k. I set the multiplier to 44 and set my voltage to 1.2 with adaptive mode on. However, according to hwinfo64 the voltage never drops below 1.2v and the clock speed doesn't go below 4.4GHz. I was under the impression that the cpu would ramp down its voltage and clock speed according to its needs? I've made sure that the minimum processor state was set to 5% under power options and it still won't budge below 1.2v and 4.4GHz. I tested under stock settings and both the clock speed and voltage were ramping up and down as needed. I tried using CPU-Z to see my vcore and it lists it at .888 which obviously isn't right. I haven't used HWInfo much these days and instead primarily use CPU-Z to check core voltage and core speed. For me, it's generally been the most, if not more than, accurate of a few other tools I keep on my USB. It's entirely possible that the 0.888V reading is true because my 3570K is currently under zero load, undervolted to 0.936V and underclocked to 2.1GHz automatically (both would probably be even less if I hadn't overclocked and added voltage). I would try your settings again and use CPU-Z constantly instead, and see if it's reporting the way you would expect it to. As for the power plan options I just keep the "high performance" plan always on and my min/max states are 100% so I don't think that would make a difference. An Unoriginal Name fucked around with this message at 07:31 on Jul 31, 2013 |
|
#
?
Jul 31, 2013 07:28
|
|
|
|
| # ? May 13, 2024 08:20 |
|
|
An Unoriginal Name posted:I haven't used HWInfo much these days and instead primarily use CPU-Z to check core voltage and core speed. For me, it's generally been the most, if not more than, accurate of a few other tools I keep on my USB. It's entirely possible that the 0.888V reading is true because my 3570K is currently under zero load, undervolted to 0.936V and underclocked to 2.1GHz automatically (both would probably be even less if I hadn't overclocked and added voltage). Just double checked - I'm set at 4.4GHz @ 1.2v with adaptive and CPU-Z says i'm constantly fluctuating between .888 and .896v (it does accurately report 4.4GHz). The reason why I mentioned the power plan settings was because I noticed the voltage and clock speed weren't fluctuating even on stock settings. I was on high performance (min-max 100% cpu) and when I changed it to 5-100% the voltage and clock speed started fluctuating as expected.
|
|
#
?
Jul 31, 2013 07:35
|
|