Overclock.net › Forums › Intel › Intel CPUs › Why OC in small increments?
New Posts  All Forums:Forum Nav:

Why OC in small increments?

post #1 of 22
Thread Starter 
I was just wondering: Does increasing the oc in small increments increase the chances of a high overclock? For example, I noticed that there's no stability if I jump right to 3.1GHz or higher from 2.8GHz no matter what I do with the voltage, memory timings, etc.

But if I slowly increase the FSB in 5MHz increments (and do a stress test afterward, of course), will my chances of successfully operating at 3.1 or 3.2 (maybe higher) increase?

I mean, is this basically why overclocking is done this way?

By the way, which stress testing program do you feel is the best? I'm basically comparing Orthos, Prime95, OCCT, and maybe even Super Pi, even though I know it's used for short tests.

Right now, I keep using Orthos, and I only use the Small FFTs, and I'm kinda stuck on running the test for 10 hours. But I guess I wanna know how I compare to everyone here so I can know if I'm doing things the "professional" way.
post #2 of 22
I'm using OCCT and when it says it's stable, then it really is stable (30mn test). I run no intensive tasks, only occasional gaming, so why would I even need it stable for 10 hours when it doesn't even stay on for 10 consecutive hours?

The idea with using small steps is that you get the chance to "feel" any impediments to your OC. Say you're @2.8, jump @3.1 and it doesn't work. You need to descend @3.0, doens't work either. Why? You can't always find out if it's the ram, mobo ir cpu that is holding you back, so you need to decrease some more. Eventually when fine tuning you'll end up taking sall steps anyway.

My 2 cents, cheers!
Centurion
(14 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 @3.3GHz 1.25V Gigabyte 790X-DS4 XFX 5850 8GB OCZ DDR2-800 
Hard DriveCoolingOSMonitor
Corsair Force 3 TRUE Win 7 Pro x64 ACER 23" TN  
PowerCase
400W Silverstone Strider Antec P182 
  hide details  
Reply
Centurion
(14 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 @3.3GHz 1.25V Gigabyte 790X-DS4 XFX 5850 8GB OCZ DDR2-800 
Hard DriveCoolingOSMonitor
Corsair Force 3 TRUE Win 7 Pro x64 ACER 23" TN  
PowerCase
400W Silverstone Strider Antec P182 
  hide details  
Reply
post #3 of 22
so this is how you racked up 1k posts? man i looked at it and almost fell off my chair, anyway i think it is more personal preference imo. I mean I do not have time to bumb 5 mhz them prime for 24 hrs or 8 for that matter.

I set a goal and went for it , when it failed i upped the cpu voltage . There may be a better explaination.
    
CPUMotherboardGraphicsRAM
E8500 GA-P35-DS3L ATi 3870 512MB 2x2GB PC6400 
Hard DriveHard DriveOptical DriveOS
Samsung 840 Pro 256Gb WD 640 GB Samsung Blu-ray R/W Win 7 
MonitorKeyboardPowerCase
Samsung 2253BW DELL CORSAIR VX550 CM690 
Mouse
Razer DeathAdder 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
E8500 GA-P35-DS3L ATi 3870 512MB 2x2GB PC6400 
Hard DriveHard DriveOptical DriveOS
Samsung 840 Pro 256Gb WD 640 GB Samsung Blu-ray R/W Win 7 
MonitorKeyboardPowerCase
Samsung 2253BW DELL CORSAIR VX550 CM690 
Mouse
Razer DeathAdder 
  hide details  
Reply
post #4 of 22
Thread Starter 
Quote:
Originally Posted by smokinbonz View Post
so this is how you racked up 1k posts? man i looked at it and almost fell off my chair,
<_< Do I look like I'm after a high post count or anything like that? I'm here to learn, and that's all I care about. I don't care about anything that gets me recognition.
Quote:
Originally Posted by smokinbonz
anyway i think it is more personal preference imo. I mean I do not have time to bumb 5 mhz them prime for 24 hrs or 8 for that matter.

I set a goal and went for it , when it failed i upped the cpu voltage . There may be a better explaination.
The explanation may be that you have a better mobo for this cpu, or a batch that allows a better oc. I had a goal and I went right for it too. But there was some instability issues at all oc's above 2.9GHz. But I'm at 3.0GHz now, and there haven't been any problems yet (but I haven't really done anything yet besides use Firefox), and I slowly crept up to 3.0 from 2.8 by going 2.8, 2.85, 2.9, 2.95, 3.0. Before, I'd go to 2.8, and go "screw this, I want 3.2", and went for it and ran into instability.

So, this is what prompted my question.
post #5 of 22
Thread Starter 
And there's always time to do a 10 hour stress test. I set it when I'm done using the computer for the day and I'm on my way to bed.

I sit there for the first 5-10 minutes of the test to monitor the temperatures. When I'm happy with where they've leveled off, I go brush my teeth and stuff, and then go to sleep. By the time I actually fall asleep for the night, an hour has already passed on the stress test. By the time I get up, shower, shave and eat and maybe relax, the test has already reached 10 hours. So to say that there's no time for it means nothing to me. There's always time. I can do a stress test any time I'm not using the computer. And I can think of many times when I'm not using the computer for 6+ hours, and sometimes 12+ hours.
post #6 of 22
Actually i have no clue what you look like nor do i care, I was not trying to aggrivate you or anything in the least however you seem to have taken offense.
I also do not believe having a high number of posts means you have recognition, It is obvious at least to me the people around here who know whats goin on. I dont look at how many posts they have .

With regard to your issues, idk about the chipset that much, if its true the memtiming divider doesnt matter even when in infancy stages of overclocking. I belive i offered that advice a while ago, but i dont think you ever tried it as the guides explained 1 step at a time. Although i didnt mention it in the previous post. I did start that way. originally with the ram at 1:1 until i got the cpu stable then i started with changing the ram divider. Thats all i think i can offer you . Maybe its worthless to you IDK

Seriously youll be overclocking forever if you went 5 mhz and stress tested for 10 hrs. using your schedule you could do 5 mhz a day thats if your tests were flawless. so i dunno what kinda life you lead that may be perfect for you. for me i was happy with jumping in and being done with my oc after a few days. got curious when i got my ram back from rma and fiddled some more,getting a new cooler and some fans my rig is cherry. Im done

This is the part where i say your kinda getting on my nerves, either its me or its you. its in your sig so im gonna take this time to mention it. All i try to do is help people here i ask for help in return. I wasnt breakin your balls .
    
CPUMotherboardGraphicsRAM
E8500 GA-P35-DS3L ATi 3870 512MB 2x2GB PC6400 
Hard DriveHard DriveOptical DriveOS
Samsung 840 Pro 256Gb WD 640 GB Samsung Blu-ray R/W Win 7 
MonitorKeyboardPowerCase
Samsung 2253BW DELL CORSAIR VX550 CM690 
Mouse
Razer DeathAdder 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
E8500 GA-P35-DS3L ATi 3870 512MB 2x2GB PC6400 
Hard DriveHard DriveOptical DriveOS
Samsung 840 Pro 256Gb WD 640 GB Samsung Blu-ray R/W Win 7 
MonitorKeyboardPowerCase
Samsung 2253BW DELL CORSAIR VX550 CM690 
Mouse
Razer DeathAdder 
  hide details  
Reply
post #7 of 22
1.) OC'ing in small increments, I always imagined, was done so it wouldn't "shock" or "scare" the components with the sudden major change in voltage/timings/frequency/etc.

2.) Lately I've seen you get antsy and on the offensive here around OCN; what gives, dude?
    
CPUMotherboardGraphicsRAM
Intel C2Q Q9550 @ 3.8 Asus P5K Deluxe MSI GTX460 1GB Cyclone G.Skill DDR2 - 4GB 
Hard DriveOptical DriveOSMonitor
2TB RAID0 F3, 2TB F4 (x2) DVD burner, Blu-ray reader Windows 7 Ult. (x64) Dell U3011 
KeyboardPowerCaseMouse
Filco Majestouch Linear R Corsair HX650W CM Sniper (Blk Ed.) Razer DeathAdder 
Mouse Pad
QcK cloth pad 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel C2Q Q9550 @ 3.8 Asus P5K Deluxe MSI GTX460 1GB Cyclone G.Skill DDR2 - 4GB 
Hard DriveOptical DriveOSMonitor
2TB RAID0 F3, 2TB F4 (x2) DVD burner, Blu-ray reader Windows 7 Ult. (x64) Dell U3011 
KeyboardPowerCaseMouse
Filco Majestouch Linear R Corsair HX650W CM Sniper (Blk Ed.) Razer DeathAdder 
Mouse Pad
QcK cloth pad 
  hide details  
Reply
post #8 of 22
I don't think it really matters how you get there. Some people go up 5mhz at a time then run a quick stability test before going up further. When you can't get through a quick super pi or something, then you know at least your getting close. Assuming you've found the approximate max of each component first.

Stability is all personal preference really. But it really sucks to be in the middle of a CSS match and get a BSOD, because you didn't test to make sure its 100% stable.

Just my 2 cents...
#22
(13 items)
 
  
CPUMotherboardGraphicsRAM
Q9650 (VID 1.1625) Gigabyte UD3P (Rev 1.6) Evga GTX 285 OCZ ReaperX HPC 4GB @ 1092 5-5-4-15 PL8 
Hard DriveOptical DriveOSMonitor
WD640AAKS x2 Lite-on Sata 20x Windows 7 | OSX LG 22" 2ms 
KeyboardPowerCaseMouse
G15 Corsair HX620 Lian Li PC-A05B MX 518 
Mouse Pad
QCK Heavy 
  hide details  
Reply
#22
(13 items)
 
  
CPUMotherboardGraphicsRAM
Q9650 (VID 1.1625) Gigabyte UD3P (Rev 1.6) Evga GTX 285 OCZ ReaperX HPC 4GB @ 1092 5-5-4-15 PL8 
Hard DriveOptical DriveOSMonitor
WD640AAKS x2 Lite-on Sata 20x Windows 7 | OSX LG 22" 2ms 
KeyboardPowerCaseMouse
G15 Corsair HX620 Lian Li PC-A05B MX 518 
Mouse Pad
QCK Heavy 
  hide details  
Reply
post #9 of 22
There is a logic behind going up by small increments on your FSB...on the other hand, setting your desired speed straight to say from 2.4 to 3.6 on a quad also makes sense but you really have to know what you are doing with your RAM, Northbridge etc., otherwise you may end up adding voltage on your cores when what is really needed is e.g. relaxing your timings on your RAM, or adding more voltage on either the RAM or Northbridge. When you go up by small increments you run short stability tests the go up again until you reach your desired speed...then you run a thorough 2, 4, 8, 24 hour test to check your 24/7 o/c.
Quad Machine
(13 items)
 
  
CPUMotherboardGraphicsRAM
Q6600 G0 @ 3.6Ghz @1.39v L733/ E8500 ASUS Blitz Formula SE 8800 GTS 512 G92 stable @818/1917/1107 4x1GB 800Mhz Transcend Axeram N 5-5-5-15 @1066Mhz 
Hard DriveOptical DriveOSMonitor
WD 320GB SATA II+250GB Hitachi 7200rpm IDE LG DVD-RW 3ple-Boot: SuSE 10.3 - XP x64 - Vista Ultimate x64 ASUS 22" VW222S 1680x1050 
KeyboardPowerCaseMouse
Logitech MX3000 Laser Thermaltake Toughpower 750W SLI Ready Thermaltake Armor V8003SWA with 25cm side fan Logitech G9 
Mouse Pad
Alienware 
  hide details  
Reply
Quad Machine
(13 items)
 
  
CPUMotherboardGraphicsRAM
Q6600 G0 @ 3.6Ghz @1.39v L733/ E8500 ASUS Blitz Formula SE 8800 GTS 512 G92 stable @818/1917/1107 4x1GB 800Mhz Transcend Axeram N 5-5-5-15 @1066Mhz 
Hard DriveOptical DriveOSMonitor
WD 320GB SATA II+250GB Hitachi 7200rpm IDE LG DVD-RW 3ple-Boot: SuSE 10.3 - XP x64 - Vista Ultimate x64 ASUS 22" VW222S 1680x1050 
KeyboardPowerCaseMouse
Logitech MX3000 Laser Thermaltake Toughpower 750W SLI Ready Thermaltake Armor V8003SWA with 25cm side fan Logitech G9 
Mouse Pad
Alienware 
  hide details  
Reply
post #10 of 22
Well you have to boot it anyway so why not run a 1 minute superPi at each increment. I think most people who've been overclocking awhile know where to start, but eventually everyone ends up going to small increments when they reach a max oc.
Black Box 3
(13 items)
 
  
CPUMotherboardGraphicsHard Drive
i7 920 4GHz 1.248 VCore Gig EX58-UD3R 5870 2x160GB Hitachi RAID 0, 500GB WD 
Optical DriveOSPowerCase
ASUS DRW-2014L1T Windows 7 Ultimate 64bit SilverStone Zeus 850w Antec 902 
  hide details  
Reply
Black Box 3
(13 items)
 
  
CPUMotherboardGraphicsHard Drive
i7 920 4GHz 1.248 VCore Gig EX58-UD3R 5870 2x160GB Hitachi RAID 0, 500GB WD 
Optical DriveOSPowerCase
ASUS DRW-2014L1T Windows 7 Ultimate 64bit SilverStone Zeus 850w Antec 902 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › Why OC in small increments?