darkl3ad3r oh man that sucks :/ my 7700k arrived a week ago but it's still sitting on my desk, waiting for the delid die mate to ship... it was supposed to ship today but they postponed it till March 10... oh well wish me luck :p I really need 5ghz in my life
Best of luck buddy. I do think you should try it before delidding. My chip actually runs cool at 4.8 and no need to delid. Who knows you might get lucky there too.
Have you tried using a normal amount of thermal paste and see what it does to temperatures? I mean, if you get more then a 10 degrees celsius aprovement then? I'm thinking if it does work as an isolator when you apply too much?
you're talking something under 1mm layer of paste at a guess, even if you used a whole tube. i dont think that is going to make much difference temperature transfer wise. for the sake of over heating the cpu by not covering the entire heat sink, i rather use more than not enough.
Likely won't happen. Last thing I want to do is chip a core or destroy a mobo if I don't have to. Besides, heatspreader on or off it is still just copper mass and LN2 gets cold enough to cold bug the chip. Yes, I know there are ways around that so you can run "full pot" and if my chip is that good and requires "more cold" then I will. but, not without the heatspreader.
You don't need to treat it like it's a bomb. I delidded my 3570K with a craft knife, scratched it several times, and got my fingers all over the contacts and it works just fine.
LOL ^ this ^ There is nothing wrong with touching the contact points and that scratched tabletop is actually a VYCO drawing board cover. There is no way it could damage anything.
So is this a recommend course of action if you're just using it regularly, or is it exclusively worth the risk if you're OC'ing? Other than that - Great video. Nice and comprehensible. Not a lot of technical jargon that kind of derails novices like myself. I rather enjoy how quiet and casual you are about the whole thing. Very comfortable to listen to.
Thanks for the feedback and I'm glad you liked the video. I/we do a bi-weekly podcast too that dives into some technical details about gaming and computers. you should check it out. That is, if you are into that sort of thing. www.hardwareasylum.com As for delidding in general. My recommendation is to "use your own judgment". The process is not exclusive to overclocking even if the process is in response to overclocking, eg, lower temps etc... the benefits apply to the entire spectrum. Thing is, and I mentioned this in a comment before, nobody "needs" to delid a chip and many adverse heat issues can often be solved by tweaking your voltage settings. As I have noticed some ASUS and Gigabyte boards favor voltage and feed the processor more than it needs. Likewise, doing a delid will make it impossible to warranty and might also be a detractor if you try to sell your hardware in the future. Hope this helps.
The difference between liquid metal and Gelid GC on top of the IHS (non-delidded) is already ~3-4 degrees and the difference gets bigger when used on the Cpu-die/chip, because thats where the crappy stuff was. Also one of the most important things in delidding is to reduce the gap between IHS and the die, this wont happen if you use glue under the IHS. it has to be used on the sides or not at all to achieve best results. I really recommend doing thorough testing now and maybe do the process again with CL Liquid Ultra and no glue under IHS. Superb 7700K btw ;)
People tell me it is a good chip and I would tend to agree. As for additional testing, not going to happen, my chip is already in the upper 5% and tweaking it any more isn't going to get me much. I've already run this chip on my Phase and will be doing LN2 next and neither respond well when using liquid metal so the TIM will stay. I'm also familiar with the whole "glue all sides" argument and can honestly say I have done both before and noticed no perceivable difference. For me I find that the Rockit 88 Relidding tool works really well at applying good pressure to the heatspreader to gives you a tight seal. As for "glue" RTV spreads quite well and leave very material between the heatspreader and the substrate. What is left is simply just filling an airgap that would have been there to being with. Overall It's much like mounting an LN2 container. if you do it correctly you'll get good results. While having uneven spring pressure and things go sideways fast. No amount of additional testing is going to change my thermal or delidding results however, it might make for a good follow-up video when my Z270 motherboard reviews are finished up.
Hardware Asylum Well that explains a lot :) But isnt Gelid TIM drying out really fast inside the cpu? You might have to reapply it, but seems like youre gonna do it anyway at some point :p
The whole "drying" of TIM has been an armchair argument for years. Honestly I'm not worried, the Intel TIM and the TIM used on graphics cards, heatsinks and anything from the OEM space has long since dried by the time we get it and the stuff still works. I can also cite a classic old question on "how often you need to replace thermal compound?". If you go back far enough it was simply a crafty marketing message to get people to use Arctic Silver. Sure there are some thermal benefits but, do so is cleaning the dust from your heatsink which is all part of the whole "Replace your thermal paste" process. :) However, with that being said, the nice thing about having a proper delidding tool is that it can be used more than once. :)
Supports almost every LGA 115x generation processor. By almost I mean ALL of them EXCEPT for Sandy Bridge, the 2000 series) those were soldered and don't need to be lidded.
The delidder supports LGA 115x CPUs but not all can be delidded. Sandy Bridge (LGA 1155) was the last soldered CPU generation while everything after that came with thermal paste (TIM). (repost of my comment from 5 months ago, it seems to have gotten lost.) and yes that Snap is epic!!
at the moment (2019) all the latest chips are soldered, they do not have any thermal paste to replace. you can delid and replace the solder with liquid metal but you will typically only get a 2 or 3 degree drop in temp for your trouble. it's mostly not worth doing any more because with soldered chips it is very easy to damage the chip
My CPU needs 1.315 volts to run Microsoft FSX SP2 with GEX North America, Rex World-Wide Airports, and Cessna C441 by Alabeo add-ons. And, that is at stock speed. [EDIT, not delidded yet - tool on order] I have been able to get into Windows 10 at 5.2GHz 1.445 volts, however, I have not benchmarked the CPU because of high temps. - - Your CPU volt requirement is fantastic, not just good. Thank you for the video.
Very welcome. I got lucky with this CPU as there is a good majority out there that won't run 5Ghz without a bunch of voltage and heat. Trick is to find where your processor likes to run and just keep it there. some are limited to stock speeds while others can handle a little more.
excellent stop motion animation! Very educational! Fuck all the haters in this comment section man like you literally didn't do anything wrong, this couldn't have gone any better and yet people still complain.
nice vid, nice results, i disagree with using avg thermal paste over LM if already going to the trouble of delidding, and definitely would use a lot less than shown for both the die and IHS... still very helpful video
I had addressed the question about LM over normal paste in other comments. I'll be using this processor under LN2 and my SS Phase cooler which required normal paste to actually work. For aircooling, you might as well use LM, or use normal paste they both work.
There were two parts to this video. The first was to show how easy it was to delid a processor using the tool from Rocketcool. The second was to show the performance advantage. Yes, LiquidMetal is the TechuTainment choice in thermal compound however, I an an extreme overclocker and LM does nothing for subZero overclocking. I attained both of my goals using the method in the video and the bonus goal of getting people to ask my why I didn't use Liquid Metal. More comments = moar betta :)
Actually your motherboard plays a huge factor in how much voltage the chip requires to reach a certain speed. I have run this chip on Asus, EVGA and Gigabyte motherboards and they all have different voltage requirements to run the processor at the same speed. Heck some of them over volt the processor at the default speed. It may seem odd but I checked each of them with a meter to confirm. 1.35v is a very standard setting and some chips require it while others don't. I just got lucky with this one.
Correct. This is a LGA 115x delidder and will work on all of them EXCEPT Sandy Bridge (2000 series) That processor generation was soldered and doesn't need to be delidded. There are different alignment tools so make sure to get the right one. I'm working on a video for the LGA 2066 delidder from Rockitcool and should be out next week.
You seem to be responding to comments so Il give it a shot. Do you really need to glue the lid of the chip back on? I would assume you could just rely on the pressure of that little arm to hold the lid onto the chip but i've never delided before so I have no idea.
You are above 1.35V and no one recommends this Volt for everyday use for this chip. Not only that anything above 1.35V I get big spike in temps above 85c :S Altho I'm using corsair liquid cooler "(
Fikret Alnasiri i dont care what the others say i use my one experience, i delid my CPU and put liquid metal, so i have also a corsair 120mm liquid cooler. And my temps are at Gaming 65C. I also can oc my chip to 5.0 ghz with 1.53v No Problem with a temp from 87C but this i only use for benchmark :). And 1.35 v or 1.38v for 24/7 Gaming doesnt make a BIG different even when you reach 85C all okay with that
Fikret Alnasiri My 6700k hit the good OC at 4.6 with 1.355 and get 71•c 100% aida64 test custom 360 water cooling EK was estabel in 4,7 too but i get 77•c and i dont like IT! Recomend 4.6
Thank you for an informative video. I have to wonder why bother to re-lid the chip at all? I remember my old AMD chips in the 90's where the heatsink went right on the chip. Seems like you'd get the best cooling if you ditched the lid.
Good question, The old K6 procs were pretty darn amazing and held in the socket by pin friction. LGA procs are held in from the top using the heatspreader to provide even pressure. Without it the chip would never make contact with the pins. On top of that the actual chip sits below the retainer so your hestsink would never make contact. Short answer: Lid is required.
In my semi-expert opinion. No but, kinda. The only thing that would allow you to push the chip further is with extreme cooling. Yes, it will lower temps and might get you an extra 100Mhz for the temperature drop. For instance my chip that I used in the video has been to 5.3Ghz and after the delid, it would still only go to 5.3Ghz using a standard watercooler. On the phase cooler it would go higher but, after the delid it still had the same upper max just with a lower overall load temp.
+Hardaware Asylum its not possible to run without the IHS? I think it will run very cooler...... Mine is running at 77ºC at 4,9Ghz 1,22V with a overdue cheap paste. Without delid.
You physically cannot run the chip without the IHS unless you modify the socket. In doing so you also cannot regulate the pressure on the pin pad which can damage the socket and cause the chip not to run. But it can be done if you are planning a suicide bench run.
Hardware Asylum a lot of people direct die cool, but that also adds risks as you said, and if you add even a tiny bit to much pressure you can crack the cpu
Thanks so much man! You really helped me with choosing the best DELID TOOL :D. I have 2 questions and I hope you answer me ^^, 1- Why you didn't use the LIQUID ULTRA THERMAL COMPOUND? and will it dry so fast if I use it?? 2- is it okay if I didn't remove the original Glue from the chip? Will it stick again? Or I need a glue like the one in this video? And thanks ^_^
1) Honestly, there are a ton of different compounds out there and everyone is claiming theirs is the best. I didn't use LUTC because I didn't have any. Never used it and prolly never will. I also was planning to put this chip under LN2 at some point and needed a compound that I knew would work well in those conditions. #bottomline, our mileage may vary and honestly use what you have, if it sucks try something else. 2) As for the glue, It "should" be removed, part of the reason people claim delidding is needed is because the glue Intel uses will space the chip away from the heatspreader. if you leave the glue and put your own it will can make the thermal problem worse.
Great video! I just wanted to ask, If I decided to delid my cpu and to not glue the ihs back, do I still remove all the glue/adhesive from the ihs and pcb so there will be no glue at all or do I not remove the glue and just place the ihs on top of the pcb once I have applied the LM. Thanks!
So, there are several questions here. The purpose of removing the glue is to ensure there is a good surface for re-attaching the top, if you don't remove the glue the lid will never fit down like it should. The reason you glue the top back on is to prevent it from moving around. As you may have noticed the CPU socket latch slides across the top of the CPU. If the lid isn't glued down it will move with the latch and will spread your thermal paste around. One thing that isn't discussed much is stress on the substrate. As you may have heard you can damage a CPU by bending the corner, this is due to the substrate being very thin and the glue + lid together from the factory help to prevent damage. If the lid isn't glued back on or is glued while in the CPU socket the substrate can bend preventing the CPU from making contact with the lid and/or pins in the socket. Basically gluing the lid back on is kinda required.
I'm assuming that "they" = "Intel" Yes "they" would be able to tell. It is a bit of a unknown if "they" would bother to do anything about it. Why don't you give it a shot #goodideas
Well it has been something like 4 years, links change and RockItCool stopped selling on Amazon shortly after this video went live. I suppose it is time to update the description. Thx for the heads up.
@Hardware Asylum Hey man, Great video. I was just wondering where I could buy one of these from.. I'm in the UK but your link doesn't work in the description. Also pretty sceptical doin' this myself but great job. :)
Thanks, I'm glad you liked the video. I'm pretty sure the Rockit guys sell worldwide, which is a limitation of Amazon in general. Try their website with the following URL. rockitcool.myshopify.com/ You can also contact them directly from the website.
i know its late but today, i tried bumping my oc from 4.8ghz at 1.370 volts to 5ghz and around 1.430 volts but it kept crashing. if i delid my cpu will help me reach a higher oc?
Never too late!! Overclocking is a balance between Temperature, Frequency and Voltage. I'm going to flat out say that 1.43v is way too much voltage, most of that is getting converted to heat and is why your system is crashing. Delidding isn't going to help that. However, I cannot stop you from trying and if anything it will help make your 4.8Ghz OC a little more stable.
Hello everyone. Is it more dangerous to delid a CPU that worked for a few months than a brand new? I would like to do it on mine but it ran for months at high temperatures and I don't know if the glue could have solidified or something.
Unfortunately stayed with Intel for Thunderbolt 3. Regretting it. ITX build with ID IS-50 cooler (says it is fine for 130W CPU). Maxed out fan, i7-7700K not overclocked, memory with XMP applied. Prime95 and it is hitting 100 Degrees C and throttling. Computer shop told me that i7-7700K is known to be dodgy.
Sorry to hear that. Thing is you are already at a disadvantage. 1) ITX builds are prone to have extremely poor venting, no matter how many fans you put in there or how open the case is you will always run hotter. 2) the cooler you have isn't very good, low profile fan, low profile fins, not much to remove excess heat. 3) Running Prime95 is great for testing but it is hardly "real world" I'm not surprised you are throttling. Of course you can monitor temps 24/7 but what fun is that. What I have found helpful with SFF builds is to make sure that you have a clear path of cool air directly to your CPU cooler. I have yet to see a builder create a cooling shroud and figure "oh there is a vent right here, or there is a fan blowing into the case so it is ok. Unless it is pulling directly from the outside the internal exhaust will get reused you will eventually overheat your system. Build a cooling shroud and you'll likely see a vast performance gain.
Hardware Asylum It occurred to me, but I have terrible 3D design skills and no 3D printer. Case is Lian Li PC-Q21B. I drilled another fan spot in the top for a 120mm fan. I have a 120x120x12mm fan in the bottom sucking air in under the R9 Nano. I tried with the top fan blowing in and out. It is a terrible cooler. I can see how the plate is only copper plated, and not absolutely smooth. Ordered the ID IS-60 now, with the proper copper plate. Will see how it goes. But considering delidding.
Like I said you are at a disadvantage, those cases are just small and never really designed to handle systems that put out more than 150w of heat (entire system). You can add fan after fan but unless you can get fresh air to your CPU and (maybe in the video card) your temps will likely never improve. You legit do not need 3D modeling or 3D printing skills to create a shroud. I know mainstream media claims that is the only way to do "anything" you can get by with way less. I built a ITX system in a Silverstone Raven and used some foam insulation to create a shroud between the CPU fan and a fan opening on the top panel. I had to use a wire fan grill to keep the foam from touching the fan blades but it worked awesome and allowed the CPU fan to pull air directly into the CPU cooler and lowered load temps by over 30c. Without the shroud it was never pulling fresh air into the case despite having three intake fans and having the CPU fan less than an inch away from the fan hole on the side of the case. It just sat there using warm case air until it overheated. Of course delidding will help but unless you can lower your internal case temperature you aren't addressing the real problem.
Hardware Asylum Normally I would agree, but before Ethereum crashed, I was 24/7 mining on R9 Nano, and Prime95 on CPU i7-6700K (not just stress testing, but for finding Mersenne Primes) and it was fine. If I had not modded the case then you would be 100% spot on.
Could be, I will be putting this chip under LN2 and my SS Phase so I needed to use some compound I knew would work subzero. Different strokes and all that.
Yeah. Most reviews I've seen have shown temp reductions of around 20degree! on LiquidMetal after de-lidding. That's just too much of a difference in my opinion to pass up a de-lid if one wishes to OC to over 4.8 Keeping temps down is always a good thing IMO. Can't believe the utter lack of quality in this regard from Intel. I mean, really....
I mentioned this in another reply but there is really no reason to delid even if your temps are way high because the paste they use has been vetted to work well enough that they can offer a warranty. Of course overclocking is a different story and it depends on what style of overclocking you are planning to do. On air a delid can lower fan speeds, on extreme you can maintain speed with less voltage. As for temp differences, I see that as chip quality more than anything.
Sure, if one doesn't plan to overclock, then I'd say probably don't bother. Lower temps though aren't a bad thing regardless IMO. Having one's system run cooler, helping the longevity of the CPU with lower temps, and not contributing such higher temps to the surrounding ambient temp of one's room are pluses IMO.
how the f you guys are able to archieve that much oc at that V ? i mean wtf ? my 7700k at stock was using like 1.25 - 1.3 and was performing 4,5 and i never touched any bios settings.. i have still problems with random instant temp spikes from 35 to 60 and back to 30 in 1-2 seconds.. any way to work around this ?
I just got lucky with my chip. It was already really good and didn't need to be delidded BUT, I wanted to test the Rockit 88 and see what kind of gains it would render. Getting to your temp spikes. those are 100% normal and happen when your system is running some background process. You can see them by pulling up the task manager and sorting by CPU activity. As for how to reduce them? Aside from killing the background processes the only thing you can do is to disable the BCLK voltage adjustment in the motherboard bios. That will reduce the natural overvolting and should reduce the temperature of the spikes but, won't get rid of them completely.
there is smth wrong with those spikes and ppl and lots of ppl are reporting the same problem, intel also admits that they are investigating this issue communities.intel.com/thread/110728.. now as far disabling the BLCK voltage adjustment option how will this affect the system ? and will it have any stability issues ? im not much of an oc guy i just want this shit to run at 4,5 which was at default but had to lower it at 4,2 cuz of those spikes driving that temp up to 68C.. atm at 4,2 wont go more than 58 but still...
Disabling the BCLK voltage adjustment just keeps the processor from getting too much voltage. Many boards feed the Kaby Lake with more voltage than it needs to run and causes excess heat. Bottom line: less voltage = less heat. As for your overclocking situation. The Core i7 7700K already runs at 4.5Ghz aka Turbo. www.hardwareasylum.com/reviews/motherboards/evga_z270_classified-k/page7.aspx
had to sync all cores on 42 value and lower it down to 4,2 so i could make those spikes not jumb above 60C cuz with default settings it was running at 4,5 (turbo) but the spikes were close to 80C.. so basicly i can go back to defaults and just disable BCLK voltage adj ?
Worth a shot. On a side note you claim to not be an OC Guy but based on the tweaks you have done I beg to differ. Overclocking isn't about top speeds and LN2, it is about tweaking systems to get better performance. I will say locking the cores to a certain speed will actually increase the thermal load, gotta remember the reported temp is typically just from the hottest core, not from the entire package.
The Gelid under the heatspreader should last forever, I used it on a 4770K that I delidded four years ago. It is still going strong. The Noctua paste is already gone considering I tore that system down already.
I have no answer for that, you can use whatever thermal paste you want. That wasn't really the point of the video. However since the question is on the table I didn't use LM because I plan to do LN2 overclocking and LM doesn't work subzero.
First of all, I didn't use crappy thermal paste/compound/TIM, Second, and this has been addressed already in the comments section, extreme overclocking. You can't use liquid metal with LN2, just doesn't work. Likewise, the temperature drop shown, even on an efficient CPU like mine is proof the process isn't pointless and can have lasting benefits.
I don't care what anyone says. Have you ever watched a video of someone deliding their processor using one of those deliding tools? These machined tools send such a tremendous shock through the entire processor at the moment the lid breaks free. This cannot be good for the millions of tiny circuits packed inside the chip. No way is that a good idea... You might as well throw it against a wall, it's about the same shock level! ... The razor method is the only way to go. It's gentle and non destructive as long as you make sure to keep the blade absolutely parallel with the circuit board of the processor, and not let the blade go in too much as to protect the raised chip in the center from getting scratched. Otherwise, the razor method is far superior to the shocking delding tools available today.
The whole reason for the tool is to hold the CPU in place while you remove the top. I have done the razor blade method before and it works great however with Skylake and Kaby Lake the substrate (the part the lid attaches to) is so thin that you can bend it just by putting the razor blade under the lid while you try to cut it away. So, in "this" case you need to use a tool.
why Intel is giving low quality liquid metal paste? don't we pay much money to buy a cpu? why to replace that with other??? did n't Intel had best paste to use or did not know it??
Four questions and none of them really ask anything. *thumb* However, here are your answers. 1) not liquid metal 2) just cause you buy something doesn't mean it has to conform to any one standard, 3) you don't have to replace the paste, as it it is not required 4) the paste is used such that they can warranty the chip and that is all that matters.
You could have used a lot less thermal paste the less the better, one small rice sized in the centre is the best, the pressure will spread it evenly. You would have had better temps, remember this for next time.
Hardware Asylum WAIT NOW THAT I HAVE YOUR ATTENTION how badly would you recommend relidding, and what's a good tool to get the old crusty glue off (I was gonna use a credit card lmao)
what, no bolding? wtf lol Relidding is rather important. without it the die won't make contact with your heatsink, burn up and likely cause quinnical depression of your wallet. So ya. 100% relid your processor. As for getting the glue off. ummm that's like in the video, legit I even recorded it. 4:43 and 5:08 Do I need to take away my "thanks for watching"?
Uh huh. Lets see... Newegg has the Ryzen 1700X listed for $369.99 whereas I can get an Core i7 7700K for $339.99. !700X is only fast in multi threaded applications (read: not games) and really isn't much faster than a Skylake Core i5 in most applications that matter. Of course this is based on my own personal testing for which I have published on my website. Of course AMD fans will be blind to all of that so... How is that any cheaper?
Oh yes, the classic misconception, I love how this myth keeps coming up. I will admit using too much paste is wasteful but will have NO impact on temperatures. The stuff spreads out to fill the gap between two surfaces and the extra just oozes off the sides. The only part we care about is where the two surfaces meet and is the main reason why people keep saying to use less.
Well, the reason people keep saying to use less is because thermal paste can work as an insulator. Otherwise you're fine, just make sure it's not overly excessive so it spills onto the board, which i actually think you applied enough on the die for it to do. The amount on the heatspreader, even tho too much, is not enough to make a difference.
"Some people think liquid metal works better than a regular thermal compound"...??? Liquid metal has thermals 20 times better than a thermal compound... There is a huge difference between them.. I mean look at any comparisons between liquid metal and a thermal compound. It is absolutely pointless to delid and then put a regular thermal compound back there... There will be a difference of 1-2 degrees if you are really lucky...
20x is a bit of an exaggeration, maybe 20% if you had to use 20 in a sentence. Thing is nobody is auguring that liquid metal is better however, it also has some major defects when it comes to extreme overclocking. and I'm not talking reddit/r/lookatmeIfiguredouthowtochangethemultiplier overclocking but, the kind of overclocking that uses LN2 and where liquid metal is not the correct choice.
After awhile you just roll your eyes and figure you can't police everything :) I'm excited for the next installment where I delid a LGA 2066, The comments for that one should be off the charts
Sounds good! By the way, I had a question about this vid too. The way you've shown it, it all looks deceptively easy using this RockitCool delidding tool. I'm guessing you've done this process lot of times. Have there been any failures? And if yes, what would you attribute them to? I've never done this before, but I'm thinking about it, and this looked so easy that I'm afraid its too good to be true (without any risks I mean).
Ok, so believe it or not, I have only delidded two (2) Intel chips and a couple of AMD K6-2's. One Intel chip was the 7700K in this video and the other was a 4770K Haswell that I did with a razor blade. Obviously there were some cuts in the video but what you see in the video is what it takes. As you can imagine the entire process is very mechanical and when you know the parts it goes quick and is why it looks easy, because it is. I have yet to run into any problems and that 7700K runs as cool as it did a year ago when I did this video. One thing that isn't in the video is what I did "after" I stopped recording. You see I didn't believe the numbers I got so I popped the lid off just to see how much compound really was between the die and the lid. Much to my surprise the connection was good. (good being: flat contact with the lid and very little compound between the die and lid). This gave me confidence that the numbers were correct. So, I reapplied the compound and glued the lid back on. Some things to keep in mind is that you really should glue all sides of the lid. The substrate is very thin and needs the glue for strength. And unless you are using liquid metal the amount of compound (TIM) you use doesn't matter. For Liquid Metal less is more, or so they say. Finally, delidding is optional, I did it for overclocking and performance under cold however others might do it for noise control or because they read about it on the internet. Almost all the comments in this video are the result of people trying to maximize their efforts to get the highest clocks or the lowest temps. Thing is unless you are running into issues it really isn't needed, just fun.
You would think that Intel would stop putting shitty ass paste on their Cpus by now. You should not have to do this at all for a 1000$ chip. I mean come on how much does 3 drops of good paste cost? An extra 1$ jeez man.
Core i7 7700K is only a 350 dollar chip. And, they have their reasons the largest of which is reliability, both during the mfg process and later in the hands/systems of consumers. Personally I think they should solder them on like they did with Sandy Bridge but.. I doubt that will ever happen again.
If the lid "plops" down on the edge the sealant will smear, thermal paste will not be where it should and generally you'll have a bad time. b-sides. this is a delidding video showing a tool designed for both parts of the job. There is already no shits given.
Damn, now that is a problem! I mean how do you counteract the loss of matter during the natural phase shift?!? And here I thought you just cared about what paste I used. Cause: LN2
Matter consists of small singularities all bound by the laws of covalent bonds. or is that mass I get my physics and chemistry confused over the years.
People risk destroying a $300 CPU because they what the best computing experience for a system they assembled themselves from a carefully prepared list of components that has been completely vetted and scrutinized by either a restrictive pocketbook or jury of their peers (or reddit *shudder*) with the added benefit that you can get a 10+ degree drop in temperatures which lowers fan speeds and overall system noise.
This was a pretty bad delid in my opinion I just think he had a godly 7700k from the silicon lottery. Singularity computer got his delidded 7700k to go down over 20°c he maxed his CPU at 58° load with a 4.8ghz OC (in a custom loop mind you). I wouldn't recommend delidding any processor besides kaby lake and up. Seems like these guys love running hit compared to Skylake or haswell. My friends 6700k won't go above 40° with Max load with a h110 aio and stock fans.
what the fuck?! 5.2 ghz at 1.25 volts... now thats an amazing chip!
THIS. I am barely stable at 4.8 1.25, pushing to 4.9 instantly takes me into the 1.35+ territory to get stable. 5.0? 1.45+. Fuck the silicon lottery.
darkl3ad3r oh man that sucks :/ my 7700k arrived a week ago but it's still sitting on my desk, waiting for the delid die mate to ship... it was supposed to ship today but they postponed it till March 10... oh well wish me luck :p I really need 5ghz in my life
Best of luck buddy. I do think you should try it before delidding. My chip actually runs cool at 4.8 and no need to delid. Who knows you might get lucky there too.
wow, 1.25 is amazing, i`m at 5ghz @1.35v without delid it gets above 90 when benchmarking and around 71 when gaming. will delid soon
Mihai Pop yea.. 1.25v @ 5.2 doesn't seem right to me.
dude that's alot of thermal paste over there !
Better to have too much then not enough.
@Hardware Asylum
Will this work on a i5 2500k?
No, 2000 series (Sandy Bridge) chips are soldered, like they should be, so attempting to delid them would actually destroy them in the process.
Have you tried using a normal amount of thermal paste and see what it does to temperatures? I mean, if you get more then a 10 degrees celsius aprovement then? I'm thinking if it does work as an isolator when you apply too much?
you're talking something under 1mm layer of paste at a guess, even if you used a whole tube. i dont think that is going to make much difference temperature transfer wise. for the sake of over heating the cpu by not covering the entire heat sink, i rather use more than not enough.
Intel's failures is Rocketcools gain. Ordered mine a few days ago for my i7-7700k.
5.2GHz @ 1.25V? Wow, your chip is amazing, OC the hell out of it
Plan to do some LN2 overclocking later this spring. Main reason for the choice in thermal compound.
Hardware Asylum why not direct dye ln2?
Likely won't happen. Last thing I want to do is chip a core or destroy a mobo if I don't have to. Besides, heatspreader on or off it is still just copper mass and LN2 gets cold enough to cold bug the chip. Yes, I know there are ways around that so you can run "full pot" and if my chip is that good and requires "more cold" then I will. but, not without the heatspreader.
But liquid metal does work better...
Clinton Fisher Yes it does but he doesn't not know about it. Just look at how carelessly he does everything.
You don't need to treat it like it's a bomb. I delidded my 3570K with a craft knife, scratched it several times, and got my fingers all over the contacts and it works just fine.
Clinton Fisher Poor 3570k.
he makes me cringe when he is touching the contact points underneath the CPU as well as placing it on that scratched tabletop
LOL ^ this ^ There is nothing wrong with touching the contact points and that scratched tabletop is actually a VYCO drawing board cover. There is no way it could damage anything.
Good clear sensible technique that was easy enough to understand.
So is this a recommend course of action if you're just using it regularly, or is it exclusively worth the risk if you're OC'ing?
Other than that - Great video. Nice and comprehensible. Not a lot of technical jargon that kind of derails novices like myself. I rather enjoy how quiet and casual you are about the whole thing. Very comfortable to listen to.
Thanks for the feedback and I'm glad you liked the video. I/we do a bi-weekly podcast too that dives into some technical details about gaming and computers. you should check it out. That is, if you are into that sort of thing. www.hardwareasylum.com
As for delidding in general. My recommendation is to "use your own judgment". The process is not exclusive to overclocking even if the process is in response to overclocking, eg, lower temps etc... the benefits apply to the entire spectrum.
Thing is, and I mentioned this in a comment before, nobody "needs" to delid a chip and many adverse heat issues can often be solved by tweaking your voltage settings. As I have noticed some ASUS and Gigabyte boards favor voltage and feed the processor more than it needs.
Likewise, doing a delid will make it impossible to warranty and might also be a detractor if you try to sell your hardware in the future. Hope this helps.
The difference between liquid metal and Gelid GC on top of the IHS (non-delidded) is already ~3-4 degrees and the difference gets bigger when used on the Cpu-die/chip, because thats where the crappy stuff was. Also one of the most important things in delidding is to reduce the gap between IHS and the die, this wont happen if you use glue under the IHS. it has to be used on the sides or not at all to achieve best results.
I really recommend doing thorough testing now and maybe do the process again with CL Liquid Ultra and no glue under IHS.
Superb 7700K btw ;)
People tell me it is a good chip and I would tend to agree.
As for additional testing, not going to happen, my chip is already in the upper 5% and tweaking it any more isn't going to get me much. I've already run this chip on my Phase and will be doing LN2 next and neither respond well when using liquid metal so the TIM will stay. I'm also familiar with the whole "glue all sides" argument and can honestly say I have done both before and noticed no perceivable difference.
For me I find that the Rockit 88 Relidding tool works really well at applying good pressure to the heatspreader to gives you a tight seal. As for "glue" RTV spreads quite well and leave very material between the heatspreader and the substrate. What is left is simply just filling an airgap that would have been there to being with.
Overall It's much like mounting an LN2 container. if you do it correctly you'll get good results. While having uneven spring pressure and things go sideways fast. No amount of additional testing is going to change my thermal or delidding results however, it might make for a good follow-up video when my Z270 motherboard reviews are finished up.
Hardware Asylum Well that explains a lot :) But isnt Gelid TIM drying out really fast inside the cpu? You might have to reapply it, but seems like youre gonna do it anyway at some point :p
The whole "drying" of TIM has been an armchair argument for years. Honestly I'm not worried, the Intel TIM and the TIM used on graphics cards, heatsinks and anything from the OEM space has long since dried by the time we get it and the stuff still works.
I can also cite a classic old question on "how often you need to replace thermal compound?". If you go back far enough it was simply a crafty marketing message to get people to use Arctic Silver. Sure there are some thermal benefits but, do so is cleaning the dust from your heatsink which is all part of the whole "Replace your thermal paste" process. :)
However, with that being said, the nice thing about having a proper delidding tool is that it can be used more than once. :)
If you get a good tool take your time and make sure everything is lined up properly this really does work well.
Holy fucking hell you go heavy with the tim xD
I have lots so, there is no need to skimp.
What cpu gens does this support? 2:49 That snap is epic!
Supports almost every LGA 115x generation processor. By almost I mean ALL of them EXCEPT for Sandy Bridge, the 2000 series) those were soldered and don't need to be lidded.
The delidder supports LGA 115x CPUs but not all can be delidded. Sandy Bridge (LGA 1155) was the last soldered CPU generation while everything after that came with thermal paste (TIM). (repost of my comment from 5 months ago, it seems to have gotten lost.) and yes that Snap is epic!!
My kit is in the mail atm. Can't wait to get it. My chip runs way too hot on an h115i.
Good times!
theripper121 where did u get it from?
If I remember correctly I just googled Rockit delidder and it will take you to a web store for purchase. It's a bit pricey but it works really well.
Also, holy SHIT your chip is fucking epic, 5.2 stock at 1.25 at only 80 degrees, thats fucking amazing, not to even mention 70 afterwards!!!
Thanks, I got lucky with the "go to newegg.com and buy a processor" lotto.
at the moment (2019) all the latest chips are soldered, they do not have any thermal paste to replace. you can delid and replace the solder with liquid metal but you will typically only get a 2 or 3 degree drop in temp for your trouble. it's mostly not worth doing any more because with soldered chips it is very easy to damage the chip
And that is why it says 7700K in the title :) You really should clarify that you are speaking of the 9000 series processors.
My CPU needs 1.315 volts to run Microsoft FSX SP2 with GEX North America, Rex World-Wide Airports, and Cessna C441 by Alabeo add-ons. And, that is at stock speed. [EDIT, not delidded yet - tool on order]
I have been able to get into Windows 10 at 5.2GHz 1.445 volts, however, I have not benchmarked the CPU because of high temps. - - Your CPU volt requirement is fantastic, not just good. Thank you for the video.
Very welcome. I got lucky with this CPU as there is a good majority out there that won't run 5Ghz without a bunch of voltage and heat. Trick is to find where your processor likes to run and just keep it there. some are limited to stock speeds while others can handle a little more.
excellent stop motion animation! Very educational!
Fuck all the haters in this comment section man like you literally didn't do anything wrong, this couldn't have gone any better and yet people still complain.
I know right. Thanks btw it took me awhile to get everything edited down and I feel it turned out pretty good.
I wonder what the temps would be if he puts cooler directly onto the crystal. I bet it'll run cool as a champ.
I'm sure an aura would consume me with a faint blue glow of healing light.
Lucky guy. My i7 7700k is running 97C at 5.2GHz and 1.45v non-delidded. Wouldn't even bother to delid if I had 80C at that overclock.
Now what would be the fun in that :) I'm fixing to do LN2 later this summer, it should be a blast.
I would seriously recommend you down clock if you have those temps, almost thermal throttling my guy.
had to lower the volts to 1.2 to maintain a below 90c lol
bruh I'm running at a mild 4.8 ghz at 1.275v and it hits 95°C my CPU is SHIIIEETT
@@AntaresSQ01 same problem here with the 7740x
I've just delidded my 7700k 5 mins ago, and the temp is smooth as f*** the temp is under 70 degree, I'm using cooler master evolution 212
EVO!! Reddit does love the CM Evo 212. It's cheap and works good.
nice vid, nice results, i disagree with using avg thermal paste over LM if already going to the trouble of delidding, and definitely would use a lot less than shown for both the die and IHS... still very helpful video
I had addressed the question about LM over normal paste in other comments. I'll be using this processor under LN2 and my SS Phase cooler which required normal paste to actually work. For aircooling, you might as well use LM, or use normal paste they both work.
I used a Delid Die Mate 2 but still a very nice Tutorial. THX :D
But why wouldn’t you use liquid metal? It works great for water cooling and you’ll never have to worry about it.
There were two parts to this video. The first was to show how easy it was to delid a processor using the tool from Rocketcool. The second was to show the performance advantage. Yes, LiquidMetal is the TechuTainment choice in thermal compound however, I an an extreme overclocker and LM does nothing for subZero overclocking. I attained both of my goals using the method in the video and the bonus goal of getting people to ask my why I didn't use Liquid Metal. More comments = moar betta :)
No way are you getting 5.2 Ghz @ 1.25V... That chip requires at least 1.35V to get 5Ghz! .... How is that possible?
Actually your motherboard plays a huge factor in how much voltage the chip requires to reach a certain speed. I have run this chip on Asus, EVGA and Gigabyte motherboards and they all have different voltage requirements to run the processor at the same speed. Heck some of them over volt the processor at the default speed. It may seem odd but I checked each of them with a meter to confirm. 1.35v is a very standard setting and some chips require it while others don't. I just got lucky with this one.
So this delid kit only work for lga 115x processors?
Correct. This is a LGA 115x delidder and will work on all of them EXCEPT Sandy Bridge (2000 series) That processor generation was soldered and doesn't need to be delidded. There are different alignment tools so make sure to get the right one. I'm working on a video for the LGA 2066 delidder from Rockitcool and should be out next week.
+Hardware Asylum im just curious but does lga 775 dellider kit exists?
Not that I am aware of. The LGA775 processors were soldered so there was no need to delid them.
You seem to be responding to comments so Il give it a shot. Do you really need to glue the lid of the chip back on? I would assume you could just rely on the pressure of that little arm to hold the lid onto the chip but i've never delided before so I have no idea.
You don't need to glue the lid back on, though it does help keep the lid from walking around when you are reinstalling the CPU.
fucking hell 5.2 my skylake 6700k wont hit 4.6 with 1.35 v 😠😠 I'm keeping it at 4.5
Fikret Alnasiri my skylake i7 hits 4.75 ghz at 1.38v
You are above 1.35V and no one recommends this Volt for everyday use for this chip. Not only that anything above 1.35V I get big spike in temps above 85c :S Altho I'm using corsair liquid cooler "(
Fikret Alnasiri i dont care what the others say i use my one experience, i delid my CPU and put liquid metal, so i have also a corsair 120mm liquid cooler. And my temps are at Gaming 65C. I also can oc my chip to 5.0 ghz with 1.53v No Problem with a temp from 87C but this i only use for benchmark :). And 1.35 v or 1.38v for 24/7 Gaming doesnt make a BIG different even when you reach 85C all okay with that
Fikret Alnasiri My 6700k hit the good OC at 4.6 with 1.355 and get 71•c 100% aida64 test custom 360 water cooling EK was estabel in 4,7 too but i get 77•c and i dont like IT! Recomend 4.6
lol wtf my 8600k hits 4.5 with 1.176v
Thank you for an informative video.
I have to wonder why bother to re-lid the chip at all? I remember my old AMD chips in the 90's where the heatsink went right on the chip. Seems like you'd get the best cooling if you ditched the lid.
Good question, The old K6 procs were pretty darn amazing and held in the socket by pin friction. LGA procs are held in from the top using the heatspreader to provide even pressure. Without it the chip would never make contact with the pins. On top of that the actual chip sits below the retainer so your hestsink would never make contact. Short answer: Lid is required.
I can get my 7700k to 5.0@ at 1.28v, but the temps are horrible. I'm hitting around 90c, now if I would delid maybe I could push it further?
In my semi-expert opinion. No but, kinda. The only thing that would allow you to push the chip further is with extreme cooling. Yes, it will lower temps and might get you an extra 100Mhz for the temperature drop.
For instance my chip that I used in the video has been to 5.3Ghz and after the delid, it would still only go to 5.3Ghz using a standard watercooler. On the phase cooler it would go higher but, after the delid it still had the same upper max just with a lower overall load temp.
+Hardaware Asylum its not possible to run without the IHS? I think it will run very cooler......
Mine is running at 77ºC at 4,9Ghz 1,22V with a overdue cheap paste. Without delid.
You physically cannot run the chip without the IHS unless you modify the socket. In doing so you also cannot regulate the pressure on the pin pad which can damage the socket and cause the chip not to run. But it can be done if you are planning a suicide bench run.
Hardware Asylum a lot of people direct die cool, but that also adds risks as you said, and if you add even a tiny bit to much pressure you can crack the cpu
Thanks so much man! You really helped me with choosing the best DELID TOOL :D. I have 2 questions and I hope you answer me ^^,
1- Why you didn't use the LIQUID ULTRA THERMAL COMPOUND? and will it dry so fast if I use it??
2- is it okay if I didn't remove the original Glue from the chip? Will it stick again? Or I need a glue like the one in this video?
And thanks ^_^
1) Honestly, there are a ton of different compounds out there and everyone is claiming theirs is the best. I didn't use LUTC because I didn't have any. Never used it and prolly never will. I also was planning to put this chip under LN2 at some point and needed a compound that I knew would work well in those conditions. #bottomline, our mileage may vary and honestly use what you have, if it sucks try something else.
2) As for the glue, It "should" be removed, part of the reason people claim delidding is needed is because the glue Intel uses will space the chip away from the heatspreader. if you leave the glue and put your own it will can make the thermal problem worse.
Hardware Asylum Thanks so much bro! ^_^
Great video! I just wanted to ask, If I decided to delid my cpu and to not glue the ihs back, do I still remove all the glue/adhesive from the ihs and pcb so there will be no glue at all or do I not remove the glue and just place the ihs on top of the pcb once I have applied the LM. Thanks!
So, there are several questions here. The purpose of removing the glue is to ensure there is a good surface for re-attaching the top, if you don't remove the glue the lid will never fit down like it should. The reason you glue the top back on is to prevent it from moving around. As you may have noticed the CPU socket latch slides across the top of the CPU. If the lid isn't glued down it will move with the latch and will spread your thermal paste around.
One thing that isn't discussed much is stress on the substrate. As you may have heard you can damage a CPU by bending the corner, this is due to the substrate being very thin and the glue + lid together from the factory help to prevent damage. If the lid isn't glued back on or is glued while in the CPU socket the substrate can bend preventing the CPU from making contact with the lid and/or pins in the socket.
Basically gluing the lid back on is kinda required.
Terrific! Well done!
Thanks, just a start for some new vids I/we have planned.
Strange, I just bought this tool and it looks different than the one in this video... did I get a bad unit???
also this is a dumb question but can they tell if its been delided if u clean it completely?
I'm assuming that "they" = "Intel" Yes "they" would be able to tell. It is a bit of a unknown if "they" would bother to do anything about it. Why don't you give it a shot #goodideas
your link takes me to a trailcam on amazon btw lol
Well it has been something like 4 years, links change and RockItCool stopped selling on Amazon shortly after this video went live. I suppose it is time to update the description. Thx for the heads up.
@Hardware Asylum Hey man, Great video. I was just wondering where I could buy one of these from.. I'm in the UK but your link doesn't work in the description. Also pretty sceptical doin' this myself but great job. :)
Thanks, I'm glad you liked the video. I'm pretty sure the Rockit guys sell worldwide, which is a limitation of Amazon in general. Try their website with the following URL. rockitcool.myshopify.com/ You can also contact them directly from the website.
Excellent video
Thanks I appreciate the feedback.
i know its late but today, i tried bumping my oc from 4.8ghz at 1.370 volts to 5ghz and around 1.430 volts but it kept crashing. if i delid my cpu will help me reach a higher oc?
Never too late!! Overclocking is a balance between Temperature, Frequency and Voltage. I'm going to flat out say that 1.43v is way too much voltage, most of that is getting converted to heat and is why your system is crashing. Delidding isn't going to help that. However, I cannot stop you from trying and if anything it will help make your 4.8Ghz OC a little more stable.
Hello everyone. Is it more dangerous to delid a CPU that worked for a few months than a brand new?
I would like to do it on mine but it ran for months at high temperatures and I don't know if the glue could have solidified or something.
Danger level is the same. The nice thing about the glue and TIM that Intel uses is that it doesn't change much over time.
What are the normal temperature of this cpu when not gaming???
Unfortunately stayed with Intel for Thunderbolt 3. Regretting it.
ITX build with ID IS-50 cooler (says it is fine for 130W CPU).
Maxed out fan, i7-7700K not overclocked, memory with XMP applied.
Prime95 and it is hitting 100 Degrees C and throttling.
Computer shop told me that i7-7700K is known to be dodgy.
Sorry to hear that. Thing is you are already at a disadvantage.
1) ITX builds are prone to have extremely poor venting, no matter how many fans you put in there or how open the case is you will always run hotter.
2) the cooler you have isn't very good, low profile fan, low profile fins, not much to remove excess heat.
3) Running Prime95 is great for testing but it is hardly "real world" I'm not surprised you are throttling. Of course you can monitor temps 24/7 but what fun is that.
What I have found helpful with SFF builds is to make sure that you have a clear path of cool air directly to your CPU cooler. I have yet to see a builder create a cooling shroud and figure "oh there is a vent right here, or there is a fan blowing into the case so it is ok. Unless it is pulling directly from the outside the internal exhaust will get reused you will eventually overheat your system.
Build a cooling shroud and you'll likely see a vast performance gain.
Hardware Asylum It occurred to me, but I have terrible 3D design skills and no 3D printer.
Case is Lian Li PC-Q21B. I drilled another fan spot in the top for a 120mm fan. I have a 120x120x12mm fan in the bottom sucking air in under the R9 Nano. I tried with the top fan blowing in and out.
It is a terrible cooler. I can see how the plate is only copper plated, and not absolutely smooth.
Ordered the ID IS-60 now, with the proper copper plate. Will see how it goes. But considering delidding.
Like I said you are at a disadvantage, those cases are just small and never really designed to handle systems that put out more than 150w of heat (entire system). You can add fan after fan but unless you can get fresh air to your CPU and (maybe in the video card) your temps will likely never improve.
You legit do not need 3D modeling or 3D printing skills to create a shroud. I know mainstream media claims that is the only way to do "anything" you can get by with way less. I built a ITX system in a Silverstone Raven and used some foam insulation to create a shroud between the CPU fan and a fan opening on the top panel. I had to use a wire fan grill to keep the foam from touching the fan blades but it worked awesome and allowed the CPU fan to pull air directly into the CPU cooler and lowered load temps by over 30c.
Without the shroud it was never pulling fresh air into the case despite having three intake fans and having the CPU fan less than an inch away from the fan hole on the side of the case. It just sat there using warm case air until it overheated.
Of course delidding will help but unless you can lower your internal case temperature you aren't addressing the real problem.
Hardware Asylum Normally I would agree, but before Ethereum crashed, I was 24/7 mining on R9 Nano, and Prime95 on CPU i7-6700K (not just stress testing, but for finding Mersenne Primes) and it was fine. If I had not modded the case then you would be 100% spot on.
Awesome. I'm glad I could help you waste my time.
Good luck with your delidding,
I'm kinda thinking that liquid metal has something to do with those lower temps most others are getting .
Could be, I will be putting this chip under LN2 and my SS Phase so I needed to use some compound I knew would work subzero. Different strokes and all that.
Yeah. Most reviews I've seen have shown temp reductions of around 20degree! on LiquidMetal after de-lidding.
That's just too much of a difference in my opinion to pass up a de-lid if one wishes to OC to over 4.8
Keeping temps down is always a good thing IMO.
Can't believe the utter lack of quality in this regard from Intel. I mean, really....
I mentioned this in another reply but there is really no reason to delid even if your temps are way high because the paste they use has been vetted to work well enough that they can offer a warranty. Of course overclocking is a different story and it depends on what style of overclocking you are planning to do. On air a delid can lower fan speeds, on extreme you can maintain speed with less voltage. As for temp differences, I see that as chip quality more than anything.
Sure, if one doesn't plan to overclock, then I'd say probably don't bother.
Lower temps though aren't a bad thing regardless IMO.
Having one's system run cooler, helping the longevity of the CPU with lower temps, and not contributing such higher temps to the surrounding ambient temp of one's room are pluses IMO.
my MoBo: asrock z170 extreme4
1.3v can hold 5ghz? what do you think?
There is only one way to find out and, it is not by asking me.
kinda cool to see you using instruments that i calibrate in a lab to solidify your results. :D
I figured someone needed to actually do things right for a change.
Don't see too many meters these days with dual K type thermocouple inputs.
It is true, Someone, several years ago, suggested this particular meter just for the dual inputs.
7:01 - Thank you for that explanation. I didn't know what to do with those white things.
What temperatures are you getting now??
No clue I haven't run that cpu in years. Though I still have it in the collection.
great... know if the product would be available.
great video though
It is available, though seems to be out of stock on Amazon. You can visit their website and buy direct.rockitcool.myshopify.com/
how the f you guys are able to archieve that much oc at that V ? i mean wtf ? my 7700k at stock was using like 1.25 - 1.3 and was performing 4,5 and i never touched any bios settings.. i have still problems with random instant temp spikes from 35 to 60 and back to 30 in 1-2 seconds.. any way to work around this ?
I just got lucky with my chip. It was already really good and didn't need to be delidded BUT, I wanted to test the Rockit 88 and see what kind of gains it would render.
Getting to your temp spikes. those are 100% normal and happen when your system is running some background process. You can see them by pulling up the task manager and sorting by CPU activity. As for how to reduce them? Aside from killing the background processes the only thing you can do is to disable the BCLK voltage adjustment in the motherboard bios. That will reduce the natural overvolting and should reduce the temperature of the spikes but, won't get rid of them completely.
there is smth wrong with those spikes and ppl and lots of ppl are reporting the same problem, intel also admits that they are investigating this issue communities.intel.com/thread/110728.. now as far disabling the BLCK voltage adjustment option how will this affect the system ? and will it have any stability issues ? im not much of an oc guy i just want this shit to run at 4,5 which was at default but had to lower it at 4,2 cuz of those spikes driving that temp up to 68C.. atm at 4,2 wont go more than 58 but still...
Disabling the BCLK voltage adjustment just keeps the processor from getting too much voltage. Many boards feed the Kaby Lake with more voltage than it needs to run and causes excess heat. Bottom line: less voltage = less heat.
As for your overclocking situation. The Core i7 7700K already runs at 4.5Ghz aka Turbo. www.hardwareasylum.com/reviews/motherboards/evga_z270_classified-k/page7.aspx
had to sync all cores on 42 value and lower it down to 4,2 so i could make those spikes not jumb above 60C cuz with default settings it was running at 4,5 (turbo) but the spikes were close to 80C.. so basicly i can go back to defaults and just disable BCLK voltage adj ?
Worth a shot.
On a side note you claim to not be an OC Guy but based on the tweaks you have done I beg to differ. Overclocking isn't about top speeds and LN2, it is about tweaking systems to get better performance. I will say locking the cores to a certain speed will actually increase the thermal load, gotta remember the reported temp is typically just from the hottest core, not from the entire package.
What thermal compound did you use?
I'm guessing you didn't watch the video? Noctua and Gelid
Yeah, how long that gonna last til you are gonna have to replace it?
The Gelid under the heatspreader should last forever, I used it on a 4770K that I delidded four years ago. It is still going strong. The Noctua paste is already gone considering I tore that system down already.
That amazon link is broken or does not work.
It hasn't worked for awhile. Hate to say it but check out the Rockitcool webstore. rockitcool.myshopify.com/
Do you buy the paste in bulk?
Nope
why shouldn't i use liquid metal ?
I have no answer for that, you can use whatever thermal paste you want. That wasn't really the point of the video. However since the question is on the table I didn't use LM because I plan to do LN2 overclocking and LM doesn't work subzero.
for me 4.8 1.27v. 5.0 1.46v for de i on Asus maximus hero
Not bad
The fact that op thinks liquid metal isnt the best tim solution says enough.
As was mentioned in the comments section this processor has been used under LN2 which is *NOT* the best TIM for that situation.
touche.
Mine can hit 5.1 havent tested 5.2 but its in a dan a4 sfx and CRAP cooling.
why not cover the whole PCB with a spoon full of termal paste? The more... the better =)
Maybe next time.
has anyone tried to delid the ryzen 1800x yet?
Yes, kinda not worth it though, Ryzen is soldered already
Why do you use crappy thermal paste .. isn't the point to put good thermal paste, like coolaboratory liquid or Kryonaut .. seems pointless otherwise
First of all, I didn't use crappy thermal paste/compound/TIM,
Second, and this has been addressed already in the comments section, extreme overclocking. You can't use liquid metal with LN2, just doesn't work.
Likewise, the temperature drop shown, even on an efficient CPU like mine is proof the process isn't pointless and can have lasting benefits.
To be fair liquid metal has been proven to lower temps even further.
Richard Dale he wants to ln2 cool later on, and liquid metal doesn't work well with sub 0
I don't care what anyone says. Have you ever watched a video of someone
deliding their processor using one of those deliding tools? These
machined tools send such a tremendous shock through the entire processor
at the moment the lid breaks free. This cannot be good for the millions
of tiny circuits packed inside the chip. No way is that a good idea...
You might as well throw it against a wall, it's about the same shock
level! ... The razor method is the only way to go. It's gentle and non
destructive as long as you make sure to keep the blade absolutely
parallel with the circuit board of the processor, and not let the blade
go in too much as to protect the raised chip in the center from getting
scratched. Otherwise, the razor method is far superior to the shocking
delding tools available today.
The whole reason for the tool is to hold the CPU in place while you remove the top. I have done the razor blade method before and it works great however with Skylake and Kaby Lake the substrate (the part the lid attaches to) is so thin that you can bend it just by putting the razor blade under the lid while you try to cut it away. So, in "this" case you need to use a tool.
why Intel is giving low quality liquid metal paste? don't we pay much money to buy a cpu? why to replace that with other??? did n't Intel had best paste to use or did not know it??
Four questions and none of them really ask anything. *thumb* However, here are your answers.
1) not liquid metal
2) just cause you buy something doesn't mean it has to conform to any one standard,
3) you don't have to replace the paste, as it it is not required
4) the paste is used such that they can warranty the chip and that is all that matters.
You could have used a lot less thermal paste the less the better, one small rice sized in the centre is the best, the pressure will spread it evenly. You would have had better temps, remember this for next time.
that's cool but will it run microsoft flightsimulator?
No clue. I'm not a terrorist.
lmfao
Out of all these comments I read while waiting on my delid tool this one is hands down the best. This comment alone gets my sub lol.
"Free performance"
*thermal compound, delid tool, glue for the IHS*
Holy crap. you got text to be bold on RUclips? You'll have to teach me that trick one of these days.
Hardware Asylum you put an *asterisk* on both sides of the text
Yes, very well aware. I was being a little sarcastic in contrast to ole Filthy Skipper helping with my lack of keywords.
Thanks for watching!
Hardware Asylum WAIT NOW THAT I HAVE YOUR ATTENTION how badly would you recommend relidding, and what's a good tool to get the old crusty glue off (I was gonna use a credit card lmao)
what, no bolding? wtf lol
Relidding is rather important. without it the die won't make contact with your heatsink, burn up and likely cause quinnical depression of your wallet. So ya. 100% relid your processor.
As for getting the glue off. ummm that's like in the video, legit I even recorded it. 4:43 and 5:08
Do I need to take away my "thanks for watching"?
why...why thermal paste....whyyyy!!!! after delidding use the liquid metaaallll!!!
Thanks for watching and commenting. I hear what you are saying and will reply with, Liquid metal doesn't do well under Liquid Nitrogen.
cheaper to sell the 7700k anf get r7 1700 for more performance then any deliding would give
Uh huh. Lets see... Newegg has the Ryzen 1700X listed for $369.99 whereas I can get an Core i7 7700K for $339.99. !700X is only fast in multi threaded applications (read: not games) and really isn't much faster than a Skylake Core i5 in most applications that matter. Of course this is based on my own personal testing for which I have published on my website. Of course AMD fans will be blind to all of that so... How is that any cheaper?
He could get better temps if he didn't overuse the thermal compound on both the die and heatspreader. He used way to much, half would've done it.
Oh yes, the classic misconception, I love how this myth keeps coming up. I will admit using too much paste is wasteful but will have NO impact on temperatures. The stuff spreads out to fill the gap between two surfaces and the extra just oozes off the sides. The only part we care about is where the two surfaces meet and is the main reason why people keep saying to use less.
Well, the reason people keep saying to use less is because thermal paste can work as an insulator. Otherwise you're fine, just make sure it's not overly excessive so it spills onto the board, which i actually think you applied enough on the die for it to do. The amount on the heatspreader, even tho too much, is not enough to make a difference.
"Some people think liquid metal works better than a regular thermal compound"...??? Liquid metal has thermals 20 times better than a thermal compound... There is a huge difference between them.. I mean look at any comparisons between liquid metal and a thermal compound. It is absolutely pointless to delid and then put a regular thermal compound back there... There will be a difference of 1-2 degrees if you are really lucky...
20x is a bit of an exaggeration, maybe 20% if you had to use 20 in a sentence. Thing is nobody is auguring that liquid metal is better however, it also has some major defects when it comes to extreme overclocking. and I'm not talking reddit/r/lookatmeIfiguredouthowtochangethemultiplier overclocking but, the kind of overclocking that uses LN2 and where liquid metal is not the correct choice.
Forget the video. I'm at awe at how you haven't gotten around to throttling people for asking the same question one hundred and thirty one times.
After awhile you just roll your eyes and figure you can't police everything :) I'm excited for the next installment where I delid a LGA 2066, The comments for that one should be off the charts
Sounds good!
By the way, I had a question about this vid too. The way you've shown it, it all looks deceptively easy using this RockitCool delidding tool. I'm guessing you've done this process lot of times. Have there been any failures? And if yes, what would you attribute them to? I've never done this before, but I'm thinking about it, and this looked so easy that I'm afraid its too good to be true (without any risks I mean).
Ok, so believe it or not, I have only delidded two (2) Intel chips and a couple of AMD K6-2's. One Intel chip was the 7700K in this video and the other was a 4770K Haswell that I did with a razor blade. Obviously there were some cuts in the video but what you see in the video is what it takes.
As you can imagine the entire process is very mechanical and when you know the parts it goes quick and is why it looks easy, because it is. I have yet to run into any problems and that 7700K runs as cool as it did a year ago when I did this video. One thing that isn't in the video is what I did "after" I stopped recording. You see I didn't believe the numbers I got so I popped the lid off just to see how much compound really was between the die and the lid. Much to my surprise the connection was good. (good being: flat contact with the lid and very little compound between the die and lid). This gave me confidence that the numbers were correct. So, I reapplied the compound and glued the lid back on.
Some things to keep in mind is that you really should glue all sides of the lid. The substrate is very thin and needs the glue for strength. And unless you are using liquid metal the amount of compound (TIM) you use doesn't matter. For Liquid Metal less is more, or so they say.
Finally, delidding is optional, I did it for overclocking and performance under cold however others might do it for noise control or because they read about it on the internet. Almost all the comments in this video are the result of people trying to maximize their efforts to get the highest clocks or the lowest temps. Thing is unless you are running into issues it really isn't needed, just fun.
You put not better thermal paste than intel
Put that new TiM on like toothpaste.... maybe a better TiM than intel, but gonna have a big mess under there now..
Seriously no worse than what was under there to begin with.
True!
You would think that Intel would stop putting shitty ass paste on their Cpus by now. You should not have to do this at all for a 1000$ chip. I mean come on how much does 3 drops of good paste cost? An extra 1$ jeez man.
Core i7 7700K is only a 350 dollar chip. And, they have their reasons the largest of which is reliability, both during the mfg process and later in the hands/systems of consumers. Personally I think they should solder them on like they did with Sandy Bridge but.. I doubt that will ever happen again.
solder you can unsolder the x99 chips then use liquid metal
triggered me when he plopped the ihs down "in the middle" no shits given
If the lid "plops" down on the edge the sealant will smear, thermal paste will not be where it should and generally you'll have a bad time. b-sides. this is a delidding video showing a tool designed for both parts of the job. There is already no shits given.
Rockit's are so expensive omg
Like a pick 🤔
Currently unavailable :D!!!
Seems they were redoing their shop last week and had to halt sales. You can still buy it from their website, just not on Amazon... yet.
"thumb nuts"
"numb nutz"
Thumbscrews not thumbnuts. 🤒
NO WAY THAT IS STABLE AT 5.2 AT THAT VOLTAGE. SHOW US PRIME95 28.10 TEST.
I bet it fails in first minute hehe.
Wow, that is specific. Next time I have the rig running I'll consider it.
gallium on the die!
I am scared of Terminators.
Damn, now that is a problem! I mean how do you counteract the loss of matter during the natural phase shift?!? And here I thought you just cared about what paste I used. Cause: LN2
Hardware Asylum @ absolute zero strange anomalies happen. you can have the same mass be there then not.
Matter consists of small singularities all bound by the laws of covalent bonds. or is that mass I get my physics and chemistry confused over the years.
Spider Mcgavenport you have no idea what you are talking about. But you're totally right.
im just mad watching this 1.25 its a hoax
You Dont know how to do it properly!!!
And yet, here you are watching my video AND leaving comment.
Soooooo, basically, people risk destroying a $300 CPU to reduce the temperature of it by 10 degrees? xD
People risk destroying a $300 CPU because they what the best computing experience for a system they assembled themselves from a carefully prepared list of components that has been completely vetted and scrutinized by either a restrictive pocketbook or jury of their peers (or reddit *shudder*) with the added benefit that you can get a 10+ degree drop in temperatures which lowers fan speeds and overall system noise.
This was a pretty bad delid in my opinion I just think he had a godly 7700k from the silicon lottery. Singularity computer got his delidded 7700k to go down over 20°c he maxed his CPU at 58° load with a 4.8ghz OC (in a custom loop mind you). I wouldn't recommend delidding any processor besides kaby lake and up. Seems like these guys love running hit compared to Skylake or haswell. My friends 6700k won't go above 40° with Max load with a h110 aio and stock fans.
I still have the Newegg receipt.
Thanatorz it's more like 25+ degrees for most people. And delidding isn't nearly as dangerous as people make it seem like...
20-30c lower is normal due to how bad the lid is fitted on the 7700k