|
LCD Deathpanel posted:That would do it, but lol at being literally required to throw out your warranty and risk CPU damage to actually use the specialty chip you paid for specifically to overclock. You aren't 'literally required' at all.
|
#
?
Jun 8, 2013 20:03
|
|
|
|
| # ? Apr 29, 2024 18:13 |
|
|
The rule is still just to not gamble more than you can afford to lose.
|
|
#
?
Jun 8, 2013 20:04
|
|
|
Shaocaholica posted:Haven't you come across any of the posts ITT about popping the IHS? I really can't afford to murder a $350 chip. It's a huge upgrade for me even running at stock; I'm only interested in a minor overclock, though it would be nice if my CPU fan did not have to go into jet engine mode when number crunching. It's a shame Intel is so loving cheap--the IHC goo was the one complaint from Ivy that could easily have been remedied, a token to otherwise disappointed nerds who tend to spend a lot on devices that could potentially contain Intel silicon. Intel products had a reputation for quality unlike the jalopys VIA, Cyrix or even AMD produced. I guess that doesn't matter anymore considering the lack of competition. With a lovely OEM case and the stock heatsink, Haswell desktops are going to hot boxes. There were professional modders that one could mail their XBox to have chipped; perhaps this is just a stimulus to the deliding business.
|
|
#
?
Jun 8, 2013 20:45
|
|
|
Yudo posted:I really can't afford to murder a $350 chip. It's a huge upgrade for me even running at stock; I'm only interested in a minor overclock, though it would be nice if my CPU fan did not have to go into jet engine mode when number crunching. You're talking as if you're bound to fail. Its not brain surgery. Would you pay someone $100 to delid your Haswell if it came with a no-DOA guarantee? Because I'll take your money 
|
|
#
?
Jun 8, 2013 22:49
|
|
|
Shaocaholica posted:You're talking as if you're bound to fail. Its not brain surgery. Would you pay someone $100 to delid your Haswell if it came with a no-DOA guarantee? Because I'll take your money Isn't it just a matter of cutting some adhesive on the edge of the lid?
|
|
#
?
Jun 8, 2013 22:53
|
|
|
cstine posted:Isn't it just a matter of cutting some adhesive on the edge of the lid? Pretty much but you also have to stay clear of the surface mount components and clean up all the adhesive from both the IHS and CPU package for the goodness to work. Its mostly physical labor on your part and a few dollars worth of razor blades, TIM and cleaning stuffs if you don't have those already.
|
|
#
?
Jun 8, 2013 23:04
|
|
|
Shaocaholica posted:Pretty much but you also have to stay clear of the surface mount components and clean up all the adhesive from both the IHS and CPU package for the goodness to work. Its mostly physical labor on your part and a few dollars worth of razor blades, TIM and cleaning stuffs if you don't have those already.  (since I haven't really heard much about it occurring), but it's pretty much how I cracked a x850pro die years ago so I'd wonder if the added weight of modern CPU coolers would increase the chances of it happening. I missed the AMD delidding era, but I don't remember cracked cores being really common then since they stopped using direct die cooling that you needed a flathead to install (socket 370/socket A). future ghost fucked around with this message at 23:35 on Jun 8, 2013 |
|
#
?
Jun 8, 2013 23:31
|
|
|
cstine posted:Isn't it just a matter of cutting some adhesive on the edge of the lid? https://www.youtube.com/watch?v=XXs0I5kuoX4
|
|
#
?
Jun 8, 2013 23:32
|
|
|
LCD Deathpanel posted:Do they reattach the IHS after reseating it with new TIM somehow? I seems like you'd want to secure it in place with some new adhesive since there wouldn't be anything holding the IHS on the die except for the TIM itself. Seems like if the heatsink was large & heavy enough and became unbalanced when mounting/unmounting the HSF, that there would be a small chance you could crack the CPU die. Since there's a small gap by default between the IHS and the die, there would be some room to give on the sides of the die of the IHS isn't secured somehow. Maybe that's You should read this: http://forums.anandtech.com/showpost.php?p=33789592&postcount=77 But in a nutshell, the CPU looks like this from the factory:  And when you delid it looks like the image on the right:  The die is pretty much holding all the weight but its also supported by the socket retention mech. You could shim the IHS by something ever so slightly less think than the IHS gap which is miniscule.
|
|
#
?
Jun 8, 2013 23:47
|
|
|
cliffy posted:Do you have a very aggressive overclock going? If not, maybe you should reseat your heatsink because that sounds way too high. Every Intel processor I've owned in the core 2 era and beyond has idled around 30 degrees Celsius. I've tried 4 different heatsinks before I decided it was a false/stuck reading, sent it back to OCUK (yeah I know) who "tested" it with nothing reported. It's using an H50 right now so I really don't know what it could be other than lovely sensors.
|
|
#
?
Jun 9, 2013 00:02
|
|
|
^^ Bad pump in the H50 maybe? H50's aren't really high-end by closed-loop standards though since you need dual fans at relatively high speeds to get better performance out of them due to the radiator design. Shaocaholica posted:The die is pretty much holding all the weight but its also supported by the socket retention mech. You could shim the IHS by something ever so slightly less think than the IHS gap which is miniscule. During the socket-A era they came out with a bunch of products like this to prevent damage to the CPUs, although those were exposed dies so the risk of cracking or chipping was much higher than with IHS-covered chips. future ghost fucked around with this message at 00:33 on Jun 9, 2013 |
|
#
?
Jun 9, 2013 00:27
|
|
|
Shaocaholica posted:You should read this: drat this looks cool and like something I want to try. I don't even overclock. And my i5-2500k is really just fine so getting a whole new mobo + cpu is unnecessary at best. Can someone assure me that this makes no difference at stock clocks and that I won't be able to run a giant fuckoff tower heatsink fanless by doing this?
|
|
#
?
Jun 9, 2013 00:28
|
|
|
Aquila posted:drat this looks cool and like something I want to try. I don't even overclock. And my i5-2500k is really just fine so getting a whole new mobo + cpu is unnecessary at best. Can someone assure me that this makes no difference at stock clocks and that I won't be able to run a giant fuckoff tower heatsink fanless by doing this? edit: If you want to run an overclocked 2500K fanless using case airflow or with a single low-speed fan your options are pretty much: get an HR-02, since that's what they were designed to do. If your case airflow is half-decent you can get by without putting a fan on it at stock, although a low-speed fan is a good idea for overclocking or for 3xxx/4xxx chips. Can't find an HR-02 for sale anymore but this is about the same thing minus the 60lbs retention nut: http://www.amazon.com/Thermalright-BW-Support-Socket-Driver/dp/B008YTUN38/ref=aag_m_pw_dp?ie=UTF8&m=A3D1M5ET5Z3YT6 This is the same thing with a 120mm fan for $10 less but you'll need a long philip's head screwdriver for the install: http://www.amazon.com/ThermalRight-Macho-size-reduced-HR-02/dp/B008SAOCHG/ref=aag_m_pw_dp?ie=UTF8&m=A3D1M5ET5Z3YT6 future ghost fucked around with this message at 00:51 on Jun 9, 2013 |
|
#
?
Jun 9, 2013 00:30
|
|
|
LCD Deathpanel posted:During the socket-A era they came out with a bunch of products like this to prevent damage to the CPUs, although those were exposed dies so the risk of cracking or chipping was much higher than with IHS-covered chips. Haha thats still 'in-stock'. Might even be repurposed for Intel IVB/HSW. Direct die is still in fashion BTW. Its simple, elegant and removes any variables/thermal-degredation caused by an imperfect IHS. You're just limited to lighter weight coolers which pretty much means water blocks.
|
|
#
?
Jun 9, 2013 00:35
|
|
|
Aquila posted:drat this looks cool and like something I want to try. I don't even overclock. And my i5-2500k is really just fine so getting a whole new mobo + cpu is unnecessary at best. Can someone assure me that this makes no difference at stock clocks and that I won't be able to run a giant fuckoff tower heatsink fanless by doing this? Really loose data but its something like 10C at stock speeds and 20C at OC speeds. Really roughly speaking. Thats only for IVB and Haswell BTW. I thought thats what you meant; delidding a new CPU and not your 2500. Shaocaholica fucked around with this message at 00:39 on Jun 9, 2013 |
|
#
?
Jun 9, 2013 00:36
|
|
|
LCD Deathpanel posted:^^ I don't see how it'd be a bad pump if the Noctua NHU12P resulted in the same temperatures, and yes the H50 is in a dual fan push/pull. Like I said, I've tested it with multiple heatsinks, even the stock one, and every single one returned the same temperatures.
|
|
#
?
Jun 9, 2013 00:53
|
|
|
This is a much easier way to delid and works exactly the same with Haswell: http://www.overclock.net/t/1376206/how-to-delid-your-ivy-bridge-cpu-with-out-a-razor-blade The only difference is that, in the picture I posted, you can see the line of small surface mount parts next to the die. Just mount the CPU in the vice in such a way that you're hitting the side of the die without the components on it. That way, if you smack it too hard, the IHS doesn't slam into the tiny little components and knock them off. Also, you know. Just be careful. You are swinging a hammer at a block of wood that's touching a $300 processor. Don't miss!
|
|
#
?
Jun 9, 2013 01:45
|
|
|
Yeah, I've always done the vice+block method to remove IHSes and by comparison the razor way looks tedious and nerve racking. I made a video a while back showing that it takes literally seconds once you have the CPU clamped.
|
|
#
?
Jun 9, 2013 02:14
|
|
|
The problem with the vise method, IMO, is: 1)already mentioned but the new surface mount components on Haswell pose an additional risk 2)not everyone has access to a -good- vise or a block of wood or a hammer. Yeah you can buy all those tools but now you're already way over-cost compared to a razor which IMO is safer even if more tedious. Its not like you're doing lots of these.
|
|
#
?
Jun 9, 2013 02:24
|
|
|
Shaocaholica posted:The problem with the vise method, IMO, is: 1) If you clamp the CPU in the vice correctly (aligned such that the IHS slides on the BARE side of the PCB when you hit it, not the side with the surface mount things) there's no chance of hitting the surface mount parts. That is, you smack the side of the CPU with the surface mount parts on it, so the IHS slides on the other side of the die. 2.) You don't need a good vice or hammer (I used the cheapest possible poo poo I could find at Harbor Freight), and the risk of cutting a trace or cutting too deep and knicking the die seems greater than the risk of screwing something up with the vice/hammer method. There's youtube videos where people smack it way too hard and the CPU goes flying across the room and they end up fine. There's also Youtube videos where people slip with the razor and take a corner off of their dies. Then again, there are also videos of people taking blowtorches to their Sandy Bridges to reheat the solder to remove the IHS so I dunno. I think I draw the line at fire, personally. Your jiggahertz is NOT worth this: https://www.youtube.com/watch?v=P4Hp0xQhJwg forbidden dialectics fucked around with this message at 06:29 on Jun 9, 2013 |
|
#
?
Jun 9, 2013 05:18
|
|
|
Has anyone tried freezing the CPU and then twisting the IHS off? I know that used to work for some chips that heatsinks/spreaders cemented on back in the day. e: Not much, maybe a degree C or 2 improvement in temp for delidding a soldered IHS. Freezer used to work just fine for me when I did it but dry ice was also an option if you didn't want to wait several hours for the chip/card to get cool enough. \/\/\/\/\/ PC LOAD LETTER fucked around with this message at 08:05 on Jun 9, 2013 |
|
#
?
Jun 9, 2013 05:38
|
|
|
PC LOAD LETTER posted:Has anyone tried freezing the CPU and then twisting the IHS off? I know that used to work for some chips that heatsinks/spreaders cemented on back in the day. Thats not a bad idea but not sure if normal fridges would get the glue cold enough and again, the new surface mount parts could get damaged. Might help with the hammer+vise technique though. Nostrum posted:Your jiggahertz is NOT worth this: Haha, what gains did people get with that? I'd be interested in knowing what gains you'd actually get compared to just lapping the IHS. Wouldn't the die also have lumps of solder you'd have to clean off too?
|
|
#
?
Jun 9, 2013 06:43
|
|
|
Shaocaholica posted:You're talking as if you're bound to fail. Its not brain surgery. Would you pay someone $100 to delid your Haswell if it came with a no-DOA guarantee? Because I'll take your money I can't afford to break it; anyways coming from a fist generation i5 stock Haswell is fast as hell to me. In a few months I may feel differently. Sure I would pay someone to delid an expensive processor, just not the one that is saving me several commutes to the office a week. Anyways this is my preferred method for deliding: https://www.youtube.com/watch?v=lMzzUuvKWPM. Notice the elegance, the finesse, and the hearty roast to top it off at the end.
|
|
#
?
Jun 9, 2013 08:14
|
|
|
Yudo posted:Anyways this is my preferred method for deliding: https://www.youtube.com/watch?v=lMzzUuvKWPM. Notice the elegance, the finesse, and the hearty roast to top it off at the end. This kid should have gone with the method of tying a string around the base of the IHS and the other end to a doorknob.
|
|
#
?
Jun 9, 2013 10:33
|
|
|
Anandtech has their Haswell Ultrabook review up. One major change is that Haswell ULT no longer supports PCI-Express from the CPU, meaning it cannot be paired with discrete GPUs. This isn't too significant as this would have been a somewhat stupid combination anyway, but maybe someone had a plan for those PCI-E 3.0 lanes that aren't there anymore. They also integrated the chipset (which was die-shrunk from 65nm to 32nm to save power) onto the CPU package, alongside the CPU die. Overall this will save platform space, reduce costs, and improve power usage.
Alereon fucked around with this message at 17:11 on Jun 9, 2013 |
|
#
?
Jun 9, 2013 16:55
|
|
|
Does Intel own the moniker 'Ultrabook'? Do laptop makers have to meet Intel's requirements in order to call a product an 'Ultrabook'? Do consumers even know/care?
|
|
#
?
Jun 9, 2013 17:10
|
|
|
Shaocaholica posted:Does Intel own the moniker 'Ultrabook'? Do laptop makers have to meet Intel's requirements in order to call a product an 'Ultrabook'? Do consumers even know/care? Yes, Yes and since ultrabooks are basically Intel's way of making every other pc manufacturer catch up with macbook air, I think so.
|
|
#
?
Jun 9, 2013 17:12
|
|
|
Shaocaholica posted:Does Intel own the moniker 'Ultrabook'? Do laptop makers have to meet Intel's requirements in order to call a product an 'Ultrabook'? Do consumers even know/care?
|
|
#
?
Jun 9, 2013 17:15
|
|
|
Thats kind of what I'm getting at. You could make something that doesn't qualify as an Ultrabook but then you're not bound to Ultrabook reqs like using the ULT procs. You could almost get away with dimensions that are so close that you'd need tools to measure the difference.
|
|
#
?
Jun 9, 2013 17:18
|
|
|
Shaocaholica posted:Thats kind of what I'm getting at. You could make something that doesn't qualify as an Ultrabook but then you're not bound to Ultrabook reqs like using the ULT procs. You could almost get away with dimensions that are so close that you'd need tools to measure the difference. You don't get Intel's marketing and development help though, which is why the manufacturers care.
|
|
#
?
Jun 9, 2013 17:22
|
|
|
And more importantly, the Ultrabook initiative exists because manufacturers refused to build good laptops to compete with Apple, so they needed some external force to set a higher bar that they are required to adhere to (if they want said marketing and dev resources and to call their devices Ultrabooks).
|
|
#
?
Jun 9, 2013 17:27
|
|
|
I guess I kind of missed that part of the story. Was not having a PC equivalent of a MBA all that big of a deal? I mean, I'm an ultrabook owner because I liked the specs and footprint of the Asus UX32 but I'm constantly wishing it were more like a MBP rather than a MBA. I'd gladly trade some of the thiness for a beefier dGPU and a real 2nd SODIMM slot. edit: For me personally, if I were to buy another 'Ultrabook', I'd want 14" or smaller, 1080p IPS or better, GT3e dual/quad, 2 DIMMS, whatever thickness that can fit into. Maybe Intel just needs another spec to match the MBPr13 in addition to the MBA. Unless I'm totally wrong and theres a ton of good MBPr13 PC clones out there. Shaocaholica fucked around with this message at 17:45 on Jun 9, 2013 |
|
#
?
Jun 9, 2013 17:35
|
|
|
It was a huge deal when Apple started being THE portable computer for a generation (i.e. majority share laptop on college campuses). There was nothing that could compete with an MBP in terms of feature balance and form-factor in the PC world, and then the Air comes out and does even BETTER and packs an SSD at a price where PC thin-and-lights struggled to exist at all.
|
|
#
?
Jun 9, 2013 17:44
|
|
|
And I suspect it was especially critical from Intel's point of view not to end up with Apple as their single important customer. Anyway, the stuff in the Anandtech article is seriously impressive (this is in minutes, I assume):  Although the battery capacity was increased here, that's still >50% improvement once that is adjusted for, pretty much as Intel promised. Considering that I already can get 7 or so hours out of my Sandy Bridge T520, a Haswell based T540 would be amazing. Edit: as long as they fix the keyboard 
|
|
#
?
Jun 9, 2013 18:18
|
|
|
Shaocaholica posted:edit: For me personally, if I were to buy another 'Ultrabook', I'd want 14" or smaller, 1080p IPS or better, GT3e dual/quad, 2 DIMMS, whatever thickness that can fit into. Maybe Intel just needs another spec to match the MBPr13 in addition to the MBA. Unless I'm totally wrong and theres a ton of good MBPr13 PC clones out there.
|
|
#
?
Jun 9, 2013 18:40
|
|
|
Alereon posted:but remember that the machine isn't even thick enough for Ethernet so you're asking a lot. Thats my biggest gripe I guess. They are thin enough to the point that its unnecessary for most people. I'd rather have 1-2mm more height if it means having 2 dimms and some extra stuff. So to me, GT3e seems like a great fit for MBPr13. No one else is expecting Apple to pull that off?
|
|
#
?
Jun 9, 2013 18:51
|
|
|
Again, the 13" rMBP uses a 25W TDP processor, the lowest GT3e SKU is the Core i7 4750HQ which is 47W. That CPU already has a 2.0Ghz base clock so there's not a lot of room to bin things down, at that point you're so starved for CPU performance it makes much more sense to use either a quad-core with GT2 (itself still 37W) or a dual-core with GT3 (28W) to keep CPU performance acceptable. If you want more, you need to be asking yourself why you're looking at such small laptops, or alternatively why dual-core+GT3 isn't enough.
|
|
#
?
Jun 9, 2013 19:06
|
|
|
Ok, I can see that the GT3e parts are way over TDP budgets but the current MBPr13 has a i7-3540M option which is a 35W part. I guess it would make more sense to package GT3e with a dual core CPU, no? Surely that would fit into 35W and lower.
|
|
#
?
Jun 9, 2013 19:17
|
|
|
Alereon posted:Anandtech has their Haswell Ultrabook review up. One major change is that Haswell ULT no longer supports PCI-Express from the CPU, meaning it cannot be paired with discrete GPUs. This isn't too significant as this would have been a somewhat stupid combination anyway, but maybe someone had a plan for those PCI-E 3.0 lanes that aren't there anymore. They also integrated the chipset (which was die-shrunk from 65nm to 32nm to save power) onto the CPU package, alongside the CPU die. Overall this will save platform space, reduce costs, and improve power usage. Yeah, I think the primary usage of that particular PCIe port would have of course been a GPU, which isn't the sanest thing to couple with an ultra-low power part. The PCIe 2.0 ports off the chipset are available to hang peripherals off and provide plenty of bandwidth; on the desktop, the PCH<->CPU link is still only DMI 2.0 (x4 PCIe 2.0 equivalent); I imagine while internally the electrical connectivity is obviously different, the same protocol is used to make software's life easier. IIRC though, as far as back as 5 series the PCIe SerDes can also be utilized to drive an eDP port; deleting the PCIe core to save die space but retaining some of the transmitters/physical pins to support eDP makes sense to me.
|
|
#
?
Jun 9, 2013 19:28
|
|
|
|
| # ? Apr 29, 2024 18:13 |
|
|
Shaocaholica posted:Ok, I can see that the GT3e parts are way over TDP budgets but the current MBPr13 has a i7-3540M option which is a 35W part. I guess it would make more sense to package GT3e with a dual core CPU, no? Surely that would fit into 35W and lower.
|
|
#
?
Jun 9, 2013 19:51
|
|