sub-topic to: Overclocking and You.. Why?

[spinejam]

in ~ 8+ years of overclocking various components (cpu, memory, video cards...), this is what i've come up with:

(3) 2x1GB DDR2 kits of 800mhz 2.2vdimm single-sided ballistix from last year ( but i believe it was more of a manufacturing defect than from my overclocking.)

 

[yh125d]

About 3 years, so far nothing. Kind of hard to

 

[VirtualLarry]

I burned out the northbridge of a 650i mobo, due to punching in some overclocking settings that I got off of the internet. Poor northbridge had inadequate cooling, and seemingly cooked itself and degraded.

Other than that, I burned out the VRMs of a BX6-2 mobo, due to a failed tualatin mod. Shorted out VCC and gnd. CPU survived just fine.

Other than that, don't think anything else has ever failed.

 

[Phantomaniac]

I've had an Asus socket 478 motherboard die on me but I don't suspect overclocking to be the cause of its death, as with all of my overclocks I keep all voltages at default.

 

[someone16]

Probably killed my 780G board when I tried to up the memory speed from DDR800 to DDR1066. Seemed like my northbridge couldn't take the extra traffic and refuses to boot up again

 

[Duvie]

I have not burned anything out that I am aware of....I have chiped a northbridge core on and 845 chipset mobo when I was trying to install an aftermarket heatsink....

I have had memory sticks from Gskill go bad...though if it was do to my ocing I dont know....They were touted as rated for OC speeds that I dont think I went over....


Generally I have always followed a few rules...

~10-15% max vcore boost
~10-15% max nbcore boost with 10% being the point I actively cool if it is not...
only oc ram with heatspreaders....Dont go over rated vdimm....just buy sticks rated to do close to my end speed
make sure I have pci, sata, and agp/pci-express lockable options in the bios
have ample case cooling...case must be only a few degrees above ambient room temp

 

[theAnimal]

Nothing.

 

[Denithor]

In 12+ years of overclocking - nothing that couldn't be attributed to something else.

Actually, the only time I had a problem was with an "OEM" PSU I stupidly used with a P3-733 build. When the PSU went out it took everything but the HDD & CPU with it. And I mean everything.

But that's the only time I've lost parts in an overclocked system.

 

[Xcobra]

been doing it for about 8 years started with a 2600+ T-bred....nothing so far...

 

[Painman]

I toasted a RAM chip on a TNT2 Ultra; I learned the value of RAMsinks from that experience.

I killed a Slot 1 P3-700 with a screwdriver while trying to bolt it to an Alpha P3125 and compressing a bunch of foam rubber insulation at the same time (Peltier cooling). Screwdriver slipped, scraped a few circuit traces and knocked off some SMDs.

I had 1 stick from a set of Ballistix DDR500 go bad, but that was par for the course for that product

I try to be much more careful with my gear these days (says the guy who just lapped his Q6600) :Q

 

[LOUISSSSS]

maybe some crucial ballistix ram, but they were SHIT. and killed $45 from my wallet for buying the Tuniq to OC. because the stock would've killed my cpu

 

[PCTC2]

Uhh...

(1) eVGA 680i-A1 (NB/Mem Controller degradation due to heat)
(3) Crucial Ballistix 1GB DDR2-1066 (due to 680i killing them @ DDR2-1111)
(4) Samsung RAMBUS-800 256MB @ 888
and probably some other things too. Probably a few caps.

I liked this thread: OC Junkies

 

[palladium]

Geforce FX5200 - I used PowerStrip to OC the core and memory speeds to nearly double its stock settings ( no volt mod). After a while display artifacts start to show up every so often even after I return everything to stock. ( guess you can't call that 'killed' can you?).

 

[deanx0r]

If you are not breaking anything, you are not going fast enough.

Over 8 years, I've burned several sets of RAM. What I learned from 8 years of overclocking is that it rarely yields back performance/productivity.
In order to overclock, you usually have to spend premium on heatsink/fan, RAM or motherboard where you could have just gone for the faster processor instead.

Also. overclocks always go bad in the long run. My last CPU purchase was a C2D 2.4GHz overclocked over 3.6GHz. It ran fine for over two years until I was getting instability issues, which led to productivity loss, and me scratching my head over what may have gone wrong. I was obligated to revert back to stock settings, but amazingly did not notice any severe performance drop.

It was fun while it lasted, and after amassing some overclocking experience (positive and negative), I am just going to stay stock the next round.

 

[F1shF4t]

Never killed anything which was directly due to overclocking but may have been over the long term.

1. Had a couple mem sticks go bad randomly, which was hard to reproduce. In the end I threw them away.
2. Northbridge degraded on my gigabyte board, I think it was a sis one, reason was due to overheating
3. Gigabyte p35 dq6 board. Was no longer fully stable and the network port stoped working.

 

[imported_Holly]

Never killed anything so far (OCing since Pentium 75Mhz times) and honestly killing something directly by OCing is something only a n00b should do. Or pple enlarging their e-peens by "the 10GHz Prescott & liquid Nitrogene"

 

[maluckey]

If you are not breaking anything, you are not going fast enough.

Well, I guess that I'm AOK then!

I've killed two supposed quality power supplies, a F5900 XT vid card, 4 sticks of RAM, 1 CPU and 2 motherboards.

In all cases, I trace it directly back to something that I personally did.

The video card gave up it's life after a cross flash gone good with a self modded BIOS. It went so good that after just twenty minutes of playing UT2003 with much better FPS, smoke came off of the card and that was that!

The PSU's died from just sheer abuse. back when everyone thought that a PSU was a PSU, I ran 4 80mm Delta High speed fans, 4 HDD, another modded vidd card (cross flashed and modded 5950), A Barton CPU at 2.0 volts, 1 GB RAM at almost 3.0 volts for gaming. The sound was amazing (I needed a headset to not go deaf).

The PSU's (each time) let go and took out the vid card, the motherboard and one of the HDD. Don't believe that any PSU will do............headroom is your friend with ragged edge voltage loads all over the map.

The RAM diead of abuse attributed to overvolting/overheating beyond reasonability.

Mark

 

[Markfw]

I killed a DS3 Gigabyte board. Or it was just plain defective, or wasn't designed to handle a quad (the old hot Q6600)

 

[Drsignguy]

"Knock on wood", nothing! Been nice so far. I like "Duvie's" rule of thumb. I seem to follow those rules just the same.

 

[aigomorla]

ROFL!

where do i start...

ummmm....

if i get bored of something does killing it mean losing it as well?

I dont think ive had anything die on me if it wasnt put on constant stress all the time. In fact even pushed hard unless under constant load stress, your eq will most definitely outlast your overclock break point.

Now if your talking about death from overclock + stress, you should ask the F@H people what there hardware rate is.

 

[Gillbot]

Nothing that i'm aware of. I've killed hardware but none of it I would directly blame on overclocking.

 

[apoppin]

GPUs



i don't bother OC'ing them any more except to test them ..
-CPUs are built to OC imo

 

[ajaidevsingh]

I have killed:-

A 3870 x2 "Voltage mod gone wrong and heat"
A Celeron 420 & 1Gb of the finest generic DDR1 ram "OCed it and it died after about a day due to crap ASrock Mb"
A amd Socket A ECS Mb "idk why it died it seemed to OC okk, also died after a month or so the ahlon xp survived and i still have it"

 

[PCTC2]

Originally posted by: Holly
Never killed anything so far (OCing since Pentium 75Mhz times) and honestly killing something directly by OCing is something only a n00b should do. Or pple enlarging their e-peens by "the 10GHz Prescott & liquid Nitrogene"

Way to just stereotype and generalize killing hardware. The 680i is notorious for being hot and killing RAM. And some parts just don't hold up the same as other parts and seem to last... until they die. And then there's Aigo and others that kill for fun. I tried to kill my E6400 @ 1.55v 3.733GHz 24/7, but it's still chugging along (at stock speeds now. No longer holds an OC above 3.2 and the board it's on doesn't like to OC). But not everyone killing parts is a "n00b". Please do not clump us all together. Plus, I think my watercooling has killed a few parts too. And that's indirectly related to my overclocking. Even the best have made mistakes. And don't get me started on your Zalman Reservator.

 

[cmdrdredd]

Memory, old D9 ICs on DDR2 memory, would require 2.1v to boot and at 2.3v could do DDR2-1000 4-4-4-12 (from DDR2-800 4-4-4-12) over time just went bad.