|
well after really pumping the volts i can push a little more OC out of my 6600k it seems 1.395 (in the bios) and LLC level 1 seems to let it be stable at 4.4, but last time i thought i was stable i eventually had dolphin start erroring out in a weird way that suggested an instability that's a sustainable voltage right? i dont really know where the line is, temps def start to push up with that much voltage though
|
#
?
Nov 7, 2016 10:58
|
|
|
|
| # ? May 16, 2024 15:00 |
|
|
That's a shitload of voltage for a 6700K - I wouldn't do it. 14nm lithography is not as forgiving as the 32nm used on the old 2500K.
|
|
#
?
Nov 7, 2016 13:19
|
|
|
Just put together a media editing computer for a friend with a 6800k and x99-a ii board with 4x16 ddr4 3000 corsair lpx ram. When I enable the xmp profile it says BCLK to 125 and when I try to boot up it says overclocked failed and kicks me back to bios. Any suggestions? Also did a prime 95 with an oc at 4.0 for a half hour, average temp was ~65 with highest spike at 81. Pretty happy about that.
|
|
#
?
Nov 17, 2016 14:19
|
|
|
JnnyThndrs posted:That's a shitload of voltage for a 6700K - I wouldn't do it. 14nm lithography is not as forgiving as the 32nm used on the old 2500K. 1.45 is still in factory spec though. I think calling 1.395 a "shitload of voltage" is being overly cautious. 1.395 is perfectly acceptable.
|
|
#
?
Nov 17, 2016 14:32
|
|
|
The Slack Lagoon posted:Just put together a media editing computer for a friend with a 6800k and x99-a ii board with 4x16 ddr4 3000 corsair lpx ram. When I enable the xmp profile it says BCLK to 125 and when I try to boot up it says overclocked failed and kicks me back to bios. Any suggestions? Make sure you have the latest bios first of all. If so you can try upping the memory voltage a bit. XMP doesn't always work 100%. Also make sure that memory is on their list. http://dlcdnet.asus.com/pub/ASUS/mb/LGA2011/X99-A/X99_Series_DRAM_QVL_for_i7-69xx_68xx_processors.pdf?_ga=1.36443872.1515668729.1474906914
|
|
#
?
Nov 17, 2016 15:50
|
|
|
Don Lapre posted:Make sure you have the latest bios first of all. If so you can try upping the memory voltage a bit. XMP doesn't always work 100%. Also make sure that memory is on their list. Bios is updated to a release from 11/08 and the ram is listed (CMK32GX4M2B000C15), though I do see that it says it works with 2 dimms but not 4 on that sheet. Could that be an issue? Could setting ram timings and speed manually work?
|
|
#
?
Nov 17, 2016 16:24
|
|
|
The Slack Lagoon posted:Bios is updated to a release from 11/08 and the ram is listed (CMK32GX4M2B000C15), though I do see that it says it works with 2 dimms but not 4 on that sheet. Could that be an issue? Yes, setting manual timings and speeds may work, or relax them a bit or up some voltage. Did you buy 2 2 dimm kits or a 4 dimm kit?
|
|
#
?
Nov 17, 2016 16:27
|
|
|
2 dimm kits, and RAM are slotted into recommended slots a1 b1 c1 d1
The Slack Lagoon fucked around with this message at 16:30 on Nov 17, 2016 |
|
#
?
Nov 17, 2016 16:28
|
|
|
Try just manually setting it all. Also try 100mhz bus speed. If you are going to run 4 dimms you SHOULD buy a 4 dimm kit as they are pretested together at XMP speeds.
|
|
#
?
Nov 17, 2016 16:33
|
|
|
Don Lapre posted:Try just manually setting it all. Also try 100mhz bus speed. If you are going to run 4 dimms you SHOULD buy a 4 dimm kit as they are pretested together at XMP speeds. I'll keep this in mind for future ram buying. Didn't realize that mattered if the ram was the same sku. Worst comes to worst, does running the ram at the 2133 it defaults to lose much performance for video editing?
|
|
#
?
Nov 17, 2016 16:40
|
|
|
Okay manually set ram timings and speed to 3000 and it looks like it's overclocked. Max voltage its hit for DRAM is 1.206 VCCIO PCH is 1.322 max VCCIO CPU is 1.306 max Looking at mobo (x99-a ii) specs it looks like these are within spec. Anything else I should check to make sure nothing will get damaged? What can I use to test stability of the ram?
|
|
#
?
Nov 18, 2016 00:21
|
|
|
It's hard to beat memtest86.
|
|
#
?
Nov 18, 2016 01:34
|
|
|
Pretty much all higher speed DDR4 is designed to work with 1.35V (not the standard 1.2V you'd see with DDR4 2133) so it may be worthwhile to tweak that. Also, if you're using auto settings you should also check what voltage you're pulling for System Agent / VCCSA, as Auto settings often apply a much higher voltage than what would work with some manual tweaking. I'm also on an Asus X99-A II - I've gotten away with a SA voltage of ~1.0->1.2V, VCCIO (CPU) of 1.05-1.15V, and VCCIO PCH of 1.05V (i.e., stock) when running DDR4 3200. SA and VCCIO seem to be the most important voltages for RAM stability. To test stability you could always get a linux thumb drive ISO, load it up, and then download/run google stress app test from apt get or whatever for an hour or two.
|
|
#
?
Nov 18, 2016 02:18
|
|
|
X99-A II master race
|
|
#
?
Nov 18, 2016 04:02
|
|
|
So here's an overclocking adventure. I've been toying around with my 6950x, this thing is a space heater. I got it to 4.3 ghz at 1.325 stable, or so I thought. I made the mistake of running the latest prime95 which is apperantly a no no on broadwell and haswell chips due to AVX instructions causing massive temperature spikes or some poo poo. True enough, that's what happened. My chip shot up to it's tjmax of 96 degrees and began to throttle. I learned of this issue and switched to the intel XTU stress test and Asus ROG realbench instead. Using realbench, the throttling stopped, but the temperatures were still too high and it would crash on top of that. I was using a corsair H110i GTX for cooling and it would spike to 89-94c still. So I researched, and found out the stock fans on that cooler are poo poo. I bought some Noctua NF-A14 iPPC 3000rpm suckers to slap on there. Dropped the temps a good 10 degrees. I managed to get 4.3 stable at 1.360 volts but every little increase with this chip shoots the temps up, I was hitting around 90c again. So I started looking into better cooling (oh god what am I doing). Researched different aio's, tried to buy a ekwb Predator 360 but the order kept messing up and I couldn't get it. I figured my corsair unit is a few years old at this point, I've maintained it by dusting out the radiator regularly but maybe it's just slowing down. I figured what the hell, I'll just buy a new h115i as a temporary solution to see if it helps the temps any. The new cooler hooked up with the noctua's brought the load temps down 11c, to 79c. 4.3ghz stable at 1.36 vcore with reasonable temps, finally. For shits I upped the vcore to 1.39 and clocked it at 4.4. It was actually stable and the temps stayed around 89c. Too high obviously but it was enough for a few benchmark runs, and to beat my timespy score. Now I'm looking into custom loops. I have a bunch of questions but that's probably better suited in the watercooling thread...
|
|
#
?
Nov 24, 2016 22:01
|
|
|
Custom loops are easy peasy lemon squeezy
|
|
#
?
Nov 24, 2016 22:18
|
|
|
I've ordered all of my parts now and I'm planning ahead - is there a coherent/clear guide for overclocking the i5-6600k? The OP has stuff for LGA 1155 and 2011 but nothing about 1151. Is the process more or less identical or should I be looking somewhere specific for guidance?
|
|
#
?
Nov 27, 2016 00:14
|
|
|
It's similar. For non pushing the envelope gains just increase your turbo multiplier until you hit a temperature limit or crash IBT and be done with it.
|
|
#
?
Nov 27, 2016 01:58
|
|
|
DarkEnigma posted:So here's an overclocking adventure. I've been toying around with my 6950x, this thing is a space heater. I got it to 4.3 ghz at 1.325 stable, or so I thought. I made the mistake of running the latest prime95 which is apperantly a no no on broadwell and haswell chips due to AVX instructions causing massive temperature spikes or some poo poo. True enough, that's what happened. My chip shot up to it's tjmax of 96 degrees and began to throttle. I learned of this issue and switched to the intel XTU stress test and Asus ROG realbench instead. Join us on the dark side.
|
|
#
?
Nov 27, 2016 22:36
|
|
|
I'm trying to squeeze the last bit of life out of my old system and failing to achieve what reports suggest should be possible. I have a q6600 on msi p965 platinum, 8gb crucial pc5300 ddr2 ram, with almost new corsair rmx650 psu. I've clocked it from 2.4 to 2.8ghz with no problem, but it won't boot if I try for 3ghz when 3.6 is apparently achievable. My bios seems very limited in the options available to me, no memory ratio etc. Do I just start upping voltages from auto until something works? e: cpu-z is showing my multiplier jumping from x6 to x9 and back, with my cpu clock going from 1.9 to 2.8ghz. Is that normal? Core vltage is generally 1.136 but occasionally spikes up to 1.224v. Again, is this normal? Lungboy fucked around with this message at 14:40 on Nov 28, 2016 |
|
#
?
Nov 28, 2016 14:21
|
|
|
Lungboy posted:I'm trying to squeeze the last bit of life out of my old system and failing to achieve what reports suggest should be possible. I have a q6600 on msi p965 platinum, 8gb crucial pc5300 ddr2 ram, with almost new corsair rmx650 psu. I've clocked it from 2.4 to 2.8ghz with no problem, but it won't boot if I try for 3ghz when 3.6 is apparently achievable. My bios seems very limited in the options available to me, no memory ratio etc. Do I just start upping voltages from auto until something works? Ah buddy you're going to need a memory ratio if you plan on using 667mhz RAM. C2Q is a whole different ballgame compared to the i5s/i7s. What FSB/Multiplier are you running? What's your motherboard? There's just a whole slew of questions to ask.. don't use auto voltage. use load-line calibration. 1.3 voltage is fine, some people say it's okay to go to 1.4. always monitor your temps just to be safe Here's a thing be careful http://www.bit-tech.net/hardware/cpus/2007/07/25/overclocking_intel_core_2_quad_q6600/2
|
|
#
?
Nov 28, 2016 23:57
|
|
|
Lube banjo posted:Ah buddy you're going to need a memory ratio if you plan on using 667mhz RAM. C2Q is a whole different ballgame compared to the i5s/i7s. What FSB/Multiplier are you running? What's your motherboard? There's just a whole slew of questions to ask.. don't use auto voltage. use load-line calibration. 1.3 voltage is fine, some people say it's okay to go to 1.4. always monitor your temps just to be safe Multiplier is locked at x9 as far as I'm aware, so FSB of 333 to get 3.0ghz. Motherboard is an MSI P965 Platinum, which is apparently a terrible overclocker. I can't find a memory ratio anywhere in the bios.
|
|
#
?
Nov 29, 2016 12:00
|
|
|
Lungboy posted:Multiplier is locked at x9 as far as I'm aware, so FSB of 333 to get 3.0ghz. Motherboard is an MSI P965 Platinum, which is apparently a terrible overclocker. I can't find a memory ratio anywhere in the bios. It might look something like this. Just make sure it's on the lowest setting. Some mobos do the mhz math for you.. From what I skimmed it might be tough getting that mobo to 400fsb. 
|
|
#
?
Nov 29, 2016 14:18
|
|
|
Lungboy posted:I'm trying to squeeze the last bit of life out of my old system and failing to achieve what reports suggest should be possible. I have a q6600 on msi p965 platinum, 8gb crucial pc5300 ddr2 ram, with almost new corsair rmx650 psu. I've clocked it from 2.4 to 2.8ghz with no problem, but it won't boot if I try for 3ghz when 3.6 is apparently achievable. My bios seems very limited in the options available to me, no memory ratio etc. Do I just start upping voltages from auto until something works? The Q6600 supports SpeedStep so unless you disable that your multiplier and voltage will jump around some when the CPU isn't working hard. Without RAM multiplier settings (really? If so, that sucks - any overclocking board should have them) you will be limited in what you can do since when you bump up BCLK your RAM will go faster in lockstep with the CPU and will probably reach its limit first. If you find out that you can change the RAM multiplier, just bump it down as you reach the rated speed of the RAM so that the RAM is always running within spec while you find the CPU max. Once you have the CPU at its max, you can try tuning the RAM multiplier higher to see if it's possible but you don't want to have multiple potential sources of instability to rule out at once if you can help it. Note that if you have a tiny bit to spend, Q9400s (or Xeon X3350s if your board supports them, they're basically Q9450s) are cheap and will probably go a bit higher than your 65nm chip. Oh, also on older systems some people found that SpeedStep was a liability for stable overclocks from what I recall. If you can't get the CPU to do what you think it should and your memory is running within limits, you can try disabling it and seeing if that makes a difference. I'd recommend leaving it on if it doesn't make a stability difference to help keep your power bill down. Eletriarnation fucked around with this message at 16:28 on Nov 29, 2016 |
|
#
?
Nov 29, 2016 16:21
|
|
|
The Slack Lagoon posted:I'll keep this in mind for future ram buying. Didn't realize that mattered if the ram was the same sku. It doesn't its just that the A-II is a poo poo board for 64gb+ of ram. I've been trying to make 128Gb over 8 slots forever with mutiple A-II boards and sometimes even 64gb would not work properly, switched to a different board and everything was much better. OC was shithouse with the A-II anyway with 5930K's Sh4 fucked around with this message at 00:13 on Dec 8, 2016 |
|
#
?
Dec 8, 2016 00:10
|
|
|
Ignoranus posted:I've ordered all of my parts now and I'm planning ahead - is there a coherent/clear guide for overclocking the i5-6600k? The OP has stuff for LGA 1155 and 2011 but nothing about 1151. Is the process more or less identical or should I be looking somewhere specific for guidance? Don't use auto voltages, they can over volt way higher then you need. Choose a reasonable vcore depending on the cooling you have and apply it as offset or adaptive. Also check it isn't going full retard with vccio or user agent voltages, sometimes they do if you set fast RAM speeds. Auto voltages can get off my lawn.
|
|
#
?
Dec 8, 2016 15:12
|
|
|
Welp, my 6600k won't boot at 5GHz. I pushed it up to 1.5V. I heard that 1.52 might be the upper limit, but I think I'm done. I don't know if there's some tricks, but 4.7-4.8GHz might be my limit, maybe a little more if I hosed with the BCLK. Edit: I got it to pass a 10 minute stress test with the vcore set to 1.45, and the multiplier set to 48 (standard 100MHz base). Temperatures hit the mid 70s with a 212 evo. Since 10 minutes probably isn't enough to be sure of anything, how long should I run my next test? 2 hours? 6? 12? 24? And why is my CPU getting 1.44V when I set it to 1.45? I don't have any dynamic Vcore poo poo active. zergstain fucked around with this message at 10:03 on Dec 10, 2016 |
|
#
?
Dec 10, 2016 09:32
|
|
|
zergstain posted:Welp, my 6600k won't boot at 5GHz. I pushed it up to 1.5V. I heard that 1.52 might be the upper limit, but I think I'm done. I don't know if there's some tricks, but 4.7-4.8GHz might be my limit, maybe a little more if I hosed with the BCLK. Vdroop is what is causing the loss of voltage. Set load line calibration to the first level if you want it to line up properly under load.
|
|
#
?
Dec 11, 2016 08:26
|
|
|
BurritoJustice posted:Vdroop is what is causing the loss of voltage. Set load line calibration to the first level if you want it to line up properly under load. My BIOS gives me the options Auto, High, and Standard for load line calibration. What do I choose? It's a Gigabyte motherboard. I'm also thinking I should set an offset with DVID so that it won't get the full voltage when it's just idling. I haven't seen much talk about setting individual per core turbo boost multipliers, and I was wondering if it's not generally done, and if so, why? I figure while I can't get it to go to 5GHz on all cores, maybe I could for single-threaded loads. Sure I'd have to test for 1..4 cores, so maybe that's why? Speaking of testing, I passed 6 hours of XTU, only to fail Prime95 in seconds. I think I have some tweaking to do. Finally, does the required voltage tend to grow exponentially with the CPU frequency? I haven't test extensively, but it seems like it requires significantly more additional volts per MHz when I hit 4.7GHz or so.
|
|
#
?
Dec 12, 2016 08:33
|
|
|
Try them all. Also yes, Most cpus hit a wall where voltage levels have to increase greatly for more performance.
|
|
#
?
Dec 12, 2016 16:41
|
|
|
zergstain posted:I haven't seen much talk about setting individual per core turbo boost multipliers, and I was wondering if it's not generally done, and if so, why? I figure while I can't get it to go to 5GHz on all cores, maybe I could for single-threaded loads. Sure I'd have to test for 1..4 cores, so maybe that's why? I wouldn't think it would make much sense having multiple cores running at different speeds - plus overclocking individual cores means you might have to set your apps manually to utilize the faster core in Windows with task manager (someone can correct me if that's wrong though, I remember when dual/quad cores were new, some old single thread apps used to need tweaks to run on dual core CPUs properly). Probably wouldn't make much difference and I'm not sure how apps would behave if you have something using multiple cores that are all clocked differently. Also with the OC settings, usually a good ballpark is 8-12 hours running stable with a stress test tool. If you're running your PC 24/7, let a stress test run for a full 24 hours to check for issues. BOOTY-ADE fucked around with this message at 20:33 on Dec 12, 2016 |
|
#
?
Dec 12, 2016 20:28
|
|
|
zergstain posted:My BIOS gives me the options Auto, High, and Standard for load line calibration. What do I choose? It's a Gigabyte motherboard. I'm also thinking I should set an offset with DVID so that it won't get the full voltage when it's just idling. 1. You want High LLC and setting it up as an offset is a good idea. 2. I dunno it just seemed like extra hassle for no real gains to me. 3. Yes, the voltage will need to go up by larger and larger amounts per mhz the higher you want to push the frequency.
|
|
#
?
Dec 13, 2016 07:16
|
|
|
Regrettable posted:1. You want High LLC and setting it up as an offset is a good idea. On my board, offset voltage wasn't applied correctly (it did nothing) on the bios the board came with. It's of course an Asus board. Just pointlessly putting that out there. What's the thread consensus on adaptive vs offset voltage? I found it pretty neat to use, as my chip ticks over at fairly low voltages but on load it gooses the voltage for the turbo clock.
|
|
#
?
Dec 13, 2016 20:34
|
|
|
BOOTY-ADE posted:I wouldn't think it would make much sense having multiple cores running at different speeds - plus overclocking individual cores means you might have to set your apps manually to utilize the faster core in Windows with task manager (someone can correct me if that's wrong though, I remember when dual/quad cores were new, some old single thread apps used to need tweaks to run on dual core CPUs properly). Probably wouldn't make much difference and I'm not sure how apps would behave if you have something using multiple cores that are all clocked differently. What I was talking about was the Turbo Boost settings, i.e. the maximum clock per number of active cores. Like maybe 4.7 for 4 cores, 4.8 for 3, ... As for loadline compensation, I left it on Auto, since it looks like it was picking High anyway. I ended up using the x264 stress test from the Skylake thread on overclock.net. The actual measured Vcore in order to survive 5 hours was like 1.344. I tried DVID, the minimum to even boot Windows pushed it up to like 1.38 when running the same test. By the way, is it likely I'll need to increase the voltage for the XMP profile? Next up, my GPU. I've heard conflicting information on whether Heaven and Valley are good ways to test the stability of my OC.
|
|
#
?
Dec 14, 2016 06:28
|
|
|
I lightly played with higher multipliers for lower numbers of cores and it was very difficult to stability test in addition to not being too meaningful of performance increase in benchmarks. By the time stuff needs the performance it's hitting all the cores and dropping you down to the all core multiplier anyways. I gave up on messing with it after a half hour or so, seemed like a waste of time. That was with a 6 core processor too which you'd think would have more room for idle cores, but the way Windows bounces stuff around between cores they all kind of got hit enough to drop to the all cores multiplier with any stress test I tried to run hitting only 5 or fewer cores. I think I had to drop down to 2, maybe 3, threads before it'd use the 5 core multiplier and even then it would bounce around.zergstain posted:By the way, is it likely I'll need to increase the voltage for the XMP profile? zergstain posted:Next up, my GPU. I've heard conflicting information on whether Heaven and Valley are good ways to test the stability of my OC. What videocard do you have? We'll be able to give you a good baseline so you don't waste hours testing speeds that are stable for 99% of people. I like Heaven because it hits most DX11 features, it's able to scale up enough to fully load any videocard out today, and I've been running it so long I know how it's supposed to look so I can tell artifacts right away. The best videocard stability test is something you can stand to live with for a while (as long as it is capable of driving GPU usage to 100% and hits a relatively wide feature set) because you'll probably use it across many GPU generations as you get used to it and know what to look for regarding artifacts. craig588 fucked around with this message at 06:51 on Dec 14, 2016 |
|
#
?
Dec 14, 2016 06:43
|
|
|
craig588 posted:I lightly played with higher multipliers for lower numbers of cores and it was very difficult to stability test in addition to not being too meaningful of performance increase in benchmarks. By the time stuff needs the performance it's hitting all the cores and dropping you down to the all core multiplier anyways. I gave up on messing with it after a half hour or so, seemed like a waste of time. That was with a 6 core processor too which you'd think would have more room for idle cores, but the way Windows bounces stuff around between cores they all kind of got hit enough to drop to the all cores multiplier with any stress test I tried to run hitting only 5 or fewer cores. I think I had to drop down to 2, maybe 3, threads before it'd use the 5 core multiplier and even then it would bounce around. craig588 posted:What videocard do you have? We'll be able to give you a good baseline so you don't waste hours testing speeds that are stable for 99% of people.
|
|
#
?
Dec 14, 2016 07:16
|
|
|
Dumb question: I've got a X99 choose system I put together a few months ago without overclocking intended to be used. The RAM in the system is DDR4-3000 RAM, and I've been leaving it at the default 2133 clockspeed for the chipset. When at that speed, it runs at 1.2V, but when set to full speed (briefly) it goes to 1.51V, which reading online seems to indicate is a rather high and on the boarders of acceptable voltage for DDR4. I've got no stomach for this, so I set it back for now. Is this a problematic voltage for that speed? How important is the "speeds over 2133 MHz are technically overclocking" thing anyway? (I'm not overclocking the CPU both because I'm too nervous to try and I tend to prioritize stability anyway.)
|
|
#
?
Dec 14, 2016 10:25
|
|
|
What ram do you have? I would expect 1.35v for ddr4 3000
|
|
#
?
Dec 14, 2016 14:25
|
|
|
zergstain posted:It's an MSI GTX 1070 with a 1557 MHz base clock. I have MSI Afterburner. Max out the power target and check what clock speed you end up with when 3d rendering first starts up, adjust the offset clock so you start at 2000MHz (You probably could start with 2050 but that might mean you have to move down instead of up and it''s more fun to move up as stability tests pass instead of down as they fail) and move up in 13MHz increments from there. You shouldn't have to stability test long, long enough to get the temperature to level out should be long enough to tell for the initial testing, probably 5 minutes of load or less. The clock speed will drop as temperatures increase, but it's designed to do that, but once you find a speed that crashes in that 5 minute heating up window back off 13MHz and that probably will be stable. If you have good cooling and power delivery you might see artifacts before crashes, Pascal has gotten better about running out of GPU potential around the same time as power delivery capability runs out, if you see artifacts, same deal back off 13MHz and that'll probably be your stable speed. One trap is after almost every in game crash and recovery your computer will still be running, but it'll ignore overclocking settings until you reboot, so I'd just reboot after any driver reset. You'll know it crashed because the screen will momentarily freeze and flicker black and then recover back to running whatever it was doing before. Once you get it stable like that there might be a bit more you can eek out from adjusting the voltage/clock curve. Hit CTRL+F in the latest version of Afterburner and it'll bring up that table, same 13MHz steps there and you can try raising some of the points the GPU drops down to as it heats up. Pascal doesn't seem to get great results out of more voltage, it seems to get a lot more heat for maybe a single 13MHz step increase which doesn't feel worth it to me so I've left the voltage slider alone in all of the Pascals I've overclocked. As for GPU Memory? I'd add 400Mhz and (hopefully) ignore it forever. If 400MHz has artifacts the highest factory memory overclock I was able to find is 250Mhz so that'll almost certainly be a stable setting. It's not that memory performance doesn't matter, but memory overclocking usually doesn't have as much headroom for performance gains as GPU overclocking does. The memory speed isn't stepped like the GPU speed so if you want to get really crazy with it you could get all the way down to the exact MHz that it gets artifacts. Once you get everything dialed in like that try running something for a few hours and see if it's still stable, crashes will almost always be indication to drop the GPU by 13Mhz and artifacts are more likely to be memory, but either could cause either fault.
|
|
#
?
Dec 14, 2016 14:53
|
|
|
|
| # ? May 16, 2024 15:00 |
|
|
craig588 posted:I lightly played with higher multipliers for lower numbers of cores and it was very difficult to stability test in addition to not being too meaningful of performance increase in benchmarks. By the time stuff needs the performance it's hitting all the cores and dropping you down to the all core multiplier anyways. I gave up on messing with it after a half hour or so, seemed like a waste of time. That was with a 6 core processor too which you'd think would have more room for idle cores, but the way Windows bounces stuff around between cores they all kind of got hit enough to drop to the all cores multiplier with any stress test I tried to run hitting only 5 or fewer cores. I think I had to drop down to 2, maybe 3, threads before it'd use the 5 core multiplier and even then it would bounce around. The tuning software that came with my Asus motherboard seems to have something that lets it try to test just a few cores instead of all, and will try to clock the chip up higher for these fewer-core scenarios. I suspect this is only really useful if your cooling is marginal and the only reason you can't hit all four cores at X GHz is because of thermal issues - so you set the all-core clock to one you can hit, and let it try to ramp up further on lower loads. Any time I've messed with it, I don't actually get anything useful, probably since I'm not hitting thermal issues on all cores.
|
|
#
?
Dec 14, 2016 16:08
|
|
