|
4.4ghz, it's on about 1.35v, mid level llc. It's one of those boring chips that totally maxes at 4.6Ghz but needs about 1.5v to do that. Do you guys normally goose vdimm 24/7? This is Samsung b die.
|
# ? Mar 8, 2018 02:01 |
|
|
# ? May 15, 2024 02:59 |
|
GRINDCORE MEGGIDO posted:Off topic, but I just want to rage in the OC thread about how crap my 6600K is at clocking. Could be lovely VRM on your mobo, that’s what I was getting with asrock z370 pro4
|
# ? Mar 8, 2018 02:09 |
|
mewse posted:Could be lovely VRM on your mobo, that’s what I was getting with asrock z370 pro4 Hah, this is an asrock z170 pro4. Sounds like it.
|
# ? Mar 8, 2018 02:15 |
|
VelociBacon posted:What cooling are you running? Generally if you have some decent cooler you can just set your vcore to 1.25 or so and your cpu to 4.5Ghz and see how it goes. Cooler Master Hyper 212 Plus I can't get to 4.5ghz with 1.25 vcore. I freeze in Windows a minute after booting. Revol fucked around with this message at 03:35 on Mar 8, 2018 |
# ? Mar 8, 2018 03:17 |
|
Revol posted:Cooler Master Hyper 212 Plus You can go up to around 1.3v or start dialing back your frequency.
|
# ? Mar 8, 2018 04:11 |
|
3770K and an Asus P8Z77 board. I've decided, if I want to try to OC, that I should target 4.5 @ 1.25v and test stability. Apparently Ivy really runs up against a thermal wall above 4.5, and I know my CPU can handle 1.25v because it was getting that on a constant basis due to some aggressive motherboard settings that I wasn't aware about until four days ago. Then test for stability, and move voltages up or down until the crashing either stops or starts up, and find the sweet spot. Fair enough. But then how do I go to offset voltage once I find out whether this works? I'm kind of put off from OCing right now just because I keep my computer on and idling and any additional efficiency would be appreciated.
|
# ? Mar 8, 2018 07:34 |
|
I ran 1.43v on a 7600k with my Asrock z270 pro4 for a full year no issues. It's a cheap board bit it didn't seem too bad. (Until I broke the socket )
|
# ? Mar 8, 2018 07:35 |
|
Craptacular! posted:3770K and an Asus P8Z77 board. I'd test the overclocking with offset right from the get go. It'll save testing it again after you've found its happy place. Shame it doesn't support adaptive. Then you have no voltage increase at non turbo clocks. Try seeing what it does at stock (offset 0).
|
# ? Mar 8, 2018 13:59 |
|
GRINDCORE MEGGIDO posted:I'd test the overclocking with offset right from the get go. It'll save testing it again after you've found its happy place. I'll give it a shot. Is Intel's XTU an okay stress test for a CPU this old? I kind of prefer that or Realbench over IBT or P95 when I don't know what I'm doing, though as an Ivy chip there is no AVX2 to melt down. Maybe I should leave vcore offset at stock and move LLC from Auto to Medium. The old guides from 2012-2013 when this board were new had everyone going to Ultra High (just one step below Extreme), but in the years that have passed I think we all know better than that. Craptacular! fucked around with this message at 19:55 on Mar 8, 2018 |
# ? Mar 8, 2018 19:52 |
|
Craptacular! posted:I'll give it a shot. Is Intel's XTU an okay stress test for a CPU this old? I kind of prefer that or Realbench over IBT or P95 when I don't know what I'm doing, though as an Ivy chip there is no AVX2 to melt down. OCCT stress test has been my go-to for CPU testing.
|
# ? Mar 8, 2018 20:16 |
|
VelociBacon posted:OCCT stress test has been my go-to for CPU testing. I like occt. It seems to catch errors quite quickly and be cake to run. The version I used didn't use avx at default, which was fine for me.
|
# ? Mar 9, 2018 00:10 |
|
I get that avx slams cpus. What I don't get is where the avx instruction set is used. If I'm just an average computer user (office apps, illustrator, steam games, etc.), would I run across avx and if so, in what sort of applications?
|
# ? Mar 9, 2018 18:21 |
|
Video encoding is a fairly common example, I think
|
# ? Mar 9, 2018 18:33 |
|
What's the IntelBurnTest equivalent for Ryzen stress-testing?
|
# ? Mar 9, 2018 18:42 |
|
Dadbod Apocalypse posted:I get that avx slams cpus. What I don't get is where the avx instruction set is used. If I'm just an average computer user (office apps, illustrator, steam games, etc.), would I run across avx and if so, in what sort of applications? Video/audio encoding, some games might use it, image editing might use it. There's a huge space of things that might reasonably benefit from using avx somewhere. Hell I wouldn't be surprised if Excel uses it when you operate on a large spreadsheet
|
# ? Mar 9, 2018 19:18 |
|
Web browsing uses it with any modern web browser. Compilers are optimizing to use it automatically when available so newer games and apps should use it too.
|
# ? Mar 9, 2018 19:26 |
|
AVX is really aimed at parallelism, and it's not very widespread. Note that in typical use, AVX instructions aren't going to sit there processing for minutes on end. They will only occasionally be used. It's not like "This game that uses AVX will stress your CPU the same way Prime95 will". Web Browsing will sometimes use AVX instructions, but very minimally today. This may increase in the future. OpenSSL and some games do, but there honestly isn't much need for it and I don't expect it will be heavily utilized for a while. If you are running Blender you can see significant AVX instruction use, but again it'll only be a small percentage of your CPU cycles there. If your concerned about your heat or a CPU dip when benchmarking AVX, don't be.
|
# ? Mar 9, 2018 19:54 |
|
Revol posted:What is the best guide for overclocking my 2500K? It seems like everything made back in the day is now outdated. I tried to follow this guide from the first post, but Windows wouldn't boot. I assume that I'm running a later BIOS version complicates my ability to follow these guides, which are almost all written when the CPU first came out. I don't know from your BIOS, but I reccomend simply incrementally increasing the turbo speed. Move 1 core/2core/3core/4core (this is core COUNT not number) all up to 4.0 which is all but guaranteed with these chips. If that works leave everything alone for a week. Watch your stability, use something like CPU-Z to monitor your multiplier as you use the system. It's also a very good idea to spot check your temperatures. I try to keep mine under 70c. As you get above 4ghz then you can start thinking about slowly increasing voltage if necessary for stability. As you gain confidence move the numbers up, leading with the 1 core/2 core multipliers and having the others catch up. If you were patient enough not to have overclocked this chip yet, then you should have patience with this method.
|
# ? Mar 9, 2018 19:59 |
|
Lockback posted:If your concerned about your heat or a CPU dip when benchmarking AVX, don't be. Right on. Thanks for all the insight!
|
# ? Mar 9, 2018 20:09 |
|
Since learning about AVX's existence and when it's used, I felt bad for accidentally causing a logic loop that left my Ryzen processing h264 overnight for like eight hours. Then I remembered that it's at stock and you sick freaks do this for 24 hours to call your machines stable, so now I'm good. No but really, I'm actually starting to consider whether my video rendering machine should have a more powerful CPU and the gaming rig on the crummier one. If I actually encoded videos constantly instead of a couple hours of TV recordings per day I'd do it.
|
# ? Mar 9, 2018 20:28 |
|
I was running 4.5ghz last night, but I came home today to my system unresponsive, so back to 4.0.Jago posted:Move 1 core/2core/3core/4core (this is core COUNT not number) all up to 4.0 which is all but guaranteed with these chips. That would be the Turbo Ratio here? I don't understand what you mean by "core count not number". Does it mean that, according to this image, the clock slowly goes down when more cores are being used? (Image not mine, found on overclock.net)
|
# ? Mar 10, 2018 03:31 |
|
Is it generally acceptable when going for higher overclocks on an 8700k to use a negative AVX offset, or is it considered something of a poor man's overclock? (since you'd "lose" the full overclock for AVX instructions). Putting together an 8700k system that will generally be used for gaming and hosting media to stream elsewhere in the house.
|
# ? Mar 10, 2018 06:25 |
|
Can someone explain what the best practices are for overclock get my i5 7600k ? My new motherboard will be in tomorrow. I was running 4.8ghz @ 1.4v stable all year but I'd like to go higher so I've delidded and delidded with a closer IHS tolerance. I'll be running an Asus Rog z270 mobo now, and water cooling. So before I just set the multiplier to 48 and the voltage to 1.4 and that was it. But I've read about instead setting the upper limit of the turbo function instead.
|
# ? Mar 10, 2018 07:26 |
|
Revol posted:I was running 4.5ghz last night, but I came home today to my system unresponsive, so back to 4.0. IIRC leave the clock ratio auto and put the turbo ratio to 40 if you want 4GHz. And drat that's one old school Award BIOS there.
|
# ? Mar 10, 2018 19:07 |
|
You are both correct. The single core setting is pretty useless as these old systems generally have some load on other cores and won't turbo the one, but generally your per core turbo speeds should look something like the above image.
|
# ? Mar 10, 2018 20:13 |
|
Revol posted:I was running 4.5ghz last night, but I came home today to my system unresponsive, so back to 4.0. FYI, my old super-star 2500k was rock solid at 4.4GHZ and needed TLC to get 4.5 and above. I think you can probably get to 4.2-4.4 stable, 4.5 is one of those barrier points on that CPU.
|
# ? Mar 11, 2018 00:30 |
|
Yeah, I have a 3570K and a 3770K and they both did 4.4 without a huge amount of hassle, but 4.5 never worked right without huge voltage increases and it wasn’t worth it. They’re both still kicking at 4.4, after all these years.
|
# ? Mar 11, 2018 00:39 |
|
Jago posted:I think I'm finally done fooling around with this OC. I'm stable and overspent! Still, I totally love this chip and I can't wait to see how the dust settles out with new AMD stuff. I have plenty of power to wait it out! Hey! I forgot, I have this relevant poo poo from a while back in this very same thread! I had to loosen the DRAM timing while trying to mine over the summer, since I've done that, totally zero crashes.
|
# ? Mar 11, 2018 04:36 |
|
Lockback posted:FYI, my old super-star 2500k was rock solid at 4.4GHZ and needed TLC to get 4.5 and above. I think you can probably get to 4.2-4.4 stable, 4.5 is one of those barrier points on that CPU. Yeah - mine has been rock solid at 4.4 and 1.38V for 6 years, but every time I've tried to go to 4.5 I've seen instability pretty quickly. I could probably get there fine with a bit more voltage but I haven't felt motivated to push it.
|
# ? Mar 12, 2018 18:22 |
|
Eletriarnation posted:Yeah - mine has been rock solid at 4.4 and 1.38V for 6 years, but every time I've tried to go to 4.5 I've seen instability pretty quickly. I could probably get there fine with a bit more voltage but I haven't felt motivated to push it. Same here. Im still running my 2500k at 4.4Ghz at 1.32V+ something since release, never could get it stable at 4.5ghz.
|
# ? Mar 12, 2018 18:31 |
|
My 2500K does 4.3GHz on the stock voltages and 4.5GHz at 1.35. Since it's not my main machine anymore it's going along at 4.3GHz. Spent 4 years or so doing 4.5GHz, I replaced it when Broadwell-E came out, I think that was early 2016, and I got the 2500K in late 2011.
|
# ? Mar 12, 2018 19:45 |
|
Question: Previously I talked about how I delidded my 7600k and when reinstalling the CPU, I bent 2 pins in the motherboard. It still worked except it rendered 2 dimm slots unusable and the CPU would instantly peg at 99c during any sort of CPU load test. Ok so I swapped out the motherboard for an Asus Rog z270 mini itx. The CPU is still behaving the same way, except now it instantly shoots to 80c, not 99c. This is at 4.8ghz. While I have no issue with 80c, it's odd that it instantly spikes. Should it not gradually ramp up ? I've re-delidded a few times, double checked the paste, and used a thin layer of bonding material to stick down the IHS so there shouldn't be any air gaps. The cooler is a DeepCool Captain 240mm AIO water cooler. Idle temps are around 27c.
|
# ? Mar 13, 2018 02:39 |
|
craig588 posted:My 2500K does 4.3GHz on the stock voltages and 4.5GHz at 1.35. Since it's not my main machine anymore it's going along at 4.3GHz. Spent 4 years or so doing 4.5GHz, I replaced it when Broadwell-E came out, I think that was early 2016, and I got the 2500K in late 2011. Mine couldn't break 4.4GHz and I ran it 4.3GHz for 4 years before I sold the CPU and mobo off.
|
# ? Mar 13, 2018 03:05 |
|
jonathan posted:Question: They spike up very quickly. You could direct die cool it for another 5deg c drop or so.
|
# ? Mar 13, 2018 04:00 |
|
GRINDCORE MEGGIDO posted:They spike up very quickly. You could direct die cool it for another 5deg c drop or so. Thanks. I have some cable management to do and then I'll play with settings and do some stability tests till I'm happy.
|
# ? Mar 13, 2018 04:11 |
|
That’s a normal spike! Things get hot fast.
|
# ? Mar 13, 2018 06:02 |
|
jonathan posted:Question: What paste are you using between die and IHS? Delidding doesn't seem worth it unless you are willing to apply some type of liquid metal. You say you used a thin layer of sealant to reapply the IHS, did you remove all the old crud? And when you say it's air tight, it's supposed to have an air gap - not be completely sealed.
|
# ? Mar 13, 2018 08:43 |
|
mewse posted:What paste are you using between die and IHS? Delidding doesn't seem worth it unless you are willing to apply some type of liquid metal. Between die and IHS I used arctic silver. I was reading that the biggest difference is closing up the tolerance between the die and IHS by removing the glue between the IHS and PCB. So I popped the IHS, cleaned all the sealer/glue off, already a thin layer of arctic silver over the die, then reglued the IHS using the weakest JB weld. Then used the motherboard CPU retainer to keep the IHS clamped. I only glued down the 2 side tabs on the IHS. Not the entire perimeter. The air gaps statement, what I meant was there is no gap between IHS or die, or IHS and cooler. I test fit everything and checked the paste marking and reapplied.
|
# ? Mar 13, 2018 11:51 |
|
Most of what I've read suggests that replacing the TIM with non-conductive paste has little to no benefit. This is for an AMD chip rather than Intel but it illustrates why people use liquid metal: On top of this Arctic Silver performs a few degrees worse than Kryonaut (and the stock paste) so you might have worsened thermal performance from stock, even after taking in account the reduction in distance between die and IHS. Llamadeus fucked around with this message at 12:21 on Mar 13, 2018 |
# ? Mar 13, 2018 12:16 |
|
|
# ? May 15, 2024 02:59 |
|
Llamadeus posted:Most of what I've read suggests that replacing the TIM with non-conductive paste has little to no benefit. This is for an AMD chip rather than Intel but it illustrates why people use liquid metal: This could totally be it. Maybe I'll order some liquid metal and try again.
|
# ? Mar 13, 2018 20:15 |