EPTSimple Questions Simple Answers - Page 332
| Forum Index > Tech Support |
|
Count9
China10928 Posts
| ||
|
Myrmidon
United States9452 Posts
Wait, your monitor isn't actually 1280x720, or is it? If it's not, it could be a difference of scaling algorithms. | ||
|
Cyro
United Kingdom20275 Posts
Ingame AA implementation is awful (and out of game has some unwanted side effects) and for various reasons an encoded video @60fps looks smoother than really achievable when playing a game yourself; The streamer won't see it as smooth as the stream looks | ||
|
Count9
China10928 Posts
| ||
|
Craton
United States17233 Posts
There's a big difference between upscaling a 720 video to 1080 and trying to use a 720 resolution on a 1080 monitor. | ||
|
Count9
China10928 Posts
| ||
|
skyR
Canada13817 Posts
| ||
|
Craton
United States17233 Posts
| ||
|
waffling1
599 Posts
i'm getting BSOD at random times. the problem report and code is as follows: + Show Spoiler + Problem signature: Problem Event Name: BlueScreen OS Version: 6.1.7601.2.1.0.256.1 Locale ID: 1033 Additional information about the problem: BCCode: 3b BCP1: 00000000C0000005 BCP2: FFFFF80002C88B70 BCP3: FFFFF8800AFD8E70 BCP4: 0000000000000000 OS Version: 6_1_7601 Service Pack: 1_0 Product: 256_1 Files that help describe the problem: C:\Windows\Minidump\052013-5506-01.dmp C:\Users\gwho\AppData\Local\Temp\WER-12823-0.sysdata.xml Read our privacy statement online: http://go.microsoft.com/fwlink/?linkid=104288&clcid=0x0409 If the online privacy statement is not available, please read our privacy statement offline: C:\Windows\system32\en-US\erofflps.txt i'm using intelburntest, cpuz, and realtemp. i pass 20 runs of standard stress level at 1.272V at 4.0GHz at max temps of 90C (i don't require that much voltage, necessarily - i haven't optimized yet. But my 3570k is relatively pretty crappy in terms of overclocking compared to others most.) GFlops are around 100 when not multitasking. my Load line calibtration is at "high", which is the second highest out of (extreme, turbo, high, med, low, normal, standard). My computer was stable while mining bitcoins over a vegas weekend on a Radeon 7790 using GUI miner with no problems. Once I started messing with some OC settings (could be cpu, memory, or gpu settings) the blue screen started kicking in. I am using the 13-4 version of the driver and the 2.7 version of the app sdk setup package. I've ran a windows memory check and it was fine. I ran disk checks and i think they were fine too. I've checked my harddrive sata connections and those aren't loose either. The Blue screen of death BSOD got to the point of messing with my settings, corrupting firefox addons, freezing firefox whenever i try to access their options... I was thinking the problem might be load line calibration, that it's not getting enough voltage when needed, which i know is a direct cause of blue screens of death sometimes (yeah i know it depends on the error code, but idk too much about that, despite googling it). Anyway, but my thinking doesn't seem right because when i'm testing it at full load, that's when there should be the most voltage droop, so why wouldn't it blue screen when testing, and instead blue screen at random times? One note: I have separated my files between SSD and HDD in a way that my HDD tends to go to sleep. perhaps, maybe perhaps it tends to blue screen when the HDD wakes up? - this is purely conjecture. Any help with my blue screen problem is very much appreciated, thanks! | ||
|
Craton
United States17233 Posts
Have you passed a 24 hour Prime95 run? I have 3 SSDs and 4 HDDs and 2 of my HDDs are asleep 95% of the time. I've never had an issue when they're waking (and you definitely shouldn't). HDD power draw is pretty negligible. Describe what you're doing at the times the BSODs have occurred. | ||
|
waffling1
599 Posts
i haven't done prime 95. i'm still in the stages of ball-parking it, doing simple ibt runs at various multipliers and voltages. BSOD happens really rnadomly. sometimes i'm just idling, sometimes im surfing the web, sometimes im making an edit ona paint-like program called SMART board, soetimes CamStudio to record what i wrote on SMARTboard. perhaps it's mainly related to some sort of memory or virtual memory. that's what the BSOD code seems to say. i'm not sure how to take it any further from here. as mentioned, is uspected it may be graphics drivers, but that is unlikely as well. The BSODs were happenining more and more frequently. I set everything to optimized defaults, then set the multiplier to x40, and the voltage to something plentifully higher than needed. seems stable so far, but i have no idea when it will BSOD again, since i was doing essentially the same thing as before. I had done optimized defaults before this, even earlier, and was still getting BSOD at x34, auto voltage. perhaps the autovoltage is too little? but that seems extremely unlikely for something that was set at stock multipliers and voltages. it just doesn't add up with any of my lines of thinking. | ||
|
Craton
United States17233 Posts
| ||
|
waffling1
599 Posts
I don't think it's other stuff b/c once i changed the OC (either on cpu, gpu, and/or memory), that's when it started being unstable Seems like i would be doing what you recommend either way, though. | ||
|
Craton
United States17233 Posts
| ||
|
Meow-Meow
Germany451 Posts
I can't access the page via thepiratebay.sx nor .se and get the following message: Access Denied. Access control configuration prevents your request from being allowed at this time. Please contact your service provider if you feel this is incorrect. Yesterday things were fine, is this maybe my university blocking the page? | ||
|
Ropid
Germany3557 Posts
| ||
|
Ropid
Germany3557 Posts
On May 22 2013 09:13 waffling1 wrote: good to know from experienced people what things are unlikely causes. i haven't done prime 95. i'm still in the stages of ball-parking it, doing simple ibt runs at various multipliers and voltages. BSOD happens really rnadomly. sometimes i'm just idling, sometimes im surfing the web, sometimes im making an edit ona paint-like program called SMART board, soetimes CamStudio to record what i wrote on SMARTboard. perhaps it's mainly related to some sort of memory or virtual memory. that's what the BSOD code seems to say. i'm not sure how to take it any further from here. as mentioned, is uspected it may be graphics drivers, but that is unlikely as well. The BSODs were happenining more and more frequently. I set everything to optimized defaults, then set the multiplier to x40, and the voltage to something plentifully higher than needed. seems stable so far, but i have no idea when it will BSOD again, since i was doing essentially the same thing as before. I had done optimized defaults before this, even earlier, and was still getting BSOD at x34, auto voltage. perhaps the autovoltage is too little? but that seems extremely unlikely for something that was set at stock multipliers and voltages. it just doesn't add up with any of my lines of thinking. On May 22 2013 10:23 waffling1 wrote: rather, i think it IS related to my OC, but can't find evidence for it. I don't think it's other stuff b/c once i changed the OC (either on cpu, gpu, and/or memory), that's when it started being unstable Seems like i would be doing what you recommend either way, though. I'm personally a fan of offset overclocking. In practice it looks like this: Starting with optimized defaults: 1) change Vcore voltage from "Auto" to "Normal" 2) the row named "DVID" will light up and be accessible; that's the offset and you increase this until it's stable That's pretty much it. Explanation is this: The CPU messages a "VID" value to the board. That's the voltage the CPU thinks it needs. What voltage the CPU will actually get from the board is called "Vcore". VID changes depending on how much power saving is used at that particular point in time. VID is determined by Intel after testing the CPU in the factory. It is set so that your particular CPU will run stable for all normal multipliers and power saving. If you want to look at VID in real time, there's a program named "CoreTemp" that shows it. The offset is needed because VID will be off when overclocking enough. The offset you've set will simply get added to VID and result in Vcore. I used the turbo multipliers to overclock as that's what I knew coming from my ASRock board. If you use the turbo multipliers, you also have to increase the Watts and Ampere limits on the same screen. Just use some ridiculous numbers like 1200 W and 300 A, the CPU will not eat more than it wants. About LLC settings, you were worried the voltage will drop too much and the CPU get unstable, but a stronger drop could instead actually increase stability of your PC. This is because a stress test like IBT causes the voltage to drop the most, and you experiment with voltage that will run IBT stable. Normal programs will all run at a higher voltage than IBT in your experiments. Normal programs don't check the stuff they do as they aren't a test, so you want them to get more voltage to be safe. This is just speculation about LLC and stability and might just be a stupid idea I have. If your 3570k behaves like mine, you will have to severely increase voltage until it gets stable. IBT does not seem to be a good test for stability except if you look at the needed voltage over a range of multipliers. Some concrete numbers about my PC that might help you: 1) IBT ran fine at 1.215 V at multiplier 46 2) prime95 ran fine at 1.235 V at multiplier 46 3) WHEA warnings disappeared from the Windows event log at 1.255 V at multiplier 46 4) IBT ran fine at 1.265 V at multiplier 47 5) the needed voltage when increasing multiplier is about +0.06 V "IBT" was just the standard test that's done with 10 passes after two minutes or so. "prime95" was the blend test, and that can still throw errors after hours of running. In hindsight, I should have just used the middle point between those 46 and 47 multiplier IBT voltage numbers, then played some games and checked the Windows event log for WHEA warnings. | ||
|
Flaiker
Germany235 Posts
On May 08 2013 20:12 Rollin wrote: It is also completely safe if you follow a guide, and don't set voltages what more experienced people or guides tell you to. When you're overclocking come back to us to help you get a nice overclock without mucking up anything  So, I spend all day installing it and here I am now. How do I overclock it without mucking up anything? :D As a reminder my current system: Gigabyte Z77X-D3H Intel i5 3570k Thermalright HR-02 Macho Rev B XFX Radeon HD 5870 1 GB G Skill Ripjaws 2*4 GB 1333 mhz (F3-10666CL9-4GBRL) Samsung 840 SSD 128 GB; Samung Hard Drive 500 GB Enermax Pro 82+ 425W Planning to use this pocessor for 4-5+ years if possible with overclocking. | ||
|
Ropid
Germany3557 Posts
Programs you might want to install: CoreTemp, CPU-Z, IntelBurnTest (or LinX), HWMonitor If you have enough patience, best would be to wait a few days and see if everything runs perfect at default settings without any overclocking. After you are convinced of that, I mentioned what method I like best in that post just before yours. Starting from default BIOS settings, set "Load Line Calibration" (LLC) to "High". Set the memory to use the XMP profile. Set CPU voltage to "Normal" (or something fixed and modest like 1.2 V). For the overclocking, go at it like this: 1) if the PC runs stable, increase multiplier 2) if the PC runs unstable, increase voltage (or offset) 3) if the CPU is too hot, decrease multiplier and stop "IntelBurnTest" or "LinX" are the stress test programs that produce the highest possible CPU temperature. Use one of those and check if CPU core temperatures are below 85 C (that's just my opinion). Normal programs will then run at 60 C or something. IBT and LinX are also a solid test of stability if you increase their memory use so that they run for several minutes for each pass of the test they do. You could also go at it backwards. Instead of slowly working your way up, you could just declare your cooler will probably do 1.3 V just fine, set that voltage (that takes some guessing if using offset voltage), then find the highest multiplier that will run IBT or LinX at 4 GB memory for 20 passes without crash or error. Don't touch 1.4 V, better stay below 1.35 V (that's just my opinion). This is a guide about overclocking on Ivy Bridge: http://sinhardware.com/index.php/overclockingoc-guides/116-ivy-bridge-overclocking-guide/126-ivy-bridge-overclocking-guide-with-ln2-guide-at-the-end That guide uses a Gigabyte board and has screenshots of the relevant BIOS settings. This guide comes with a ZIP file of a bunch of programs and explains how to use "prime95" and stuff, but isn't for your motherboard: http://www.overclock.net/t/1198504/complete-overclocking-guide-sandy-bridge-ivy-bridge-asrock-edition About stability, you should look for "WHEA" warnings in the Windows event viewer in the Administrative Events log. The CPU can apparently catch errors and fix those without the PC crashing. Windows will log that WHEA event when this happens. Increase voltage by a +0.005 V step each time you see those warnings. Those events should disappear after you increase voltage enough. I needed about +0.02 V total. You might want to look at a program named "SpeedFan" which can control fan speeds depending on temperatures. It's pretty annoying to use, but it seems to be the only Windows program that can do what it does. | ||
|
z0rz
United States350 Posts
| ||
| ||
