Sign in with
Sign up | Sign in
Your question

overclockers, please read, help me decide if its worth it

Last response: in Overclocking
Share
February 23, 2006 12:41:30 AM

OKay for all you overclockers im trying to make a decision about if i want to get into overclocking and hopefully with your help you can help me decide if its the thing for me.

Okay heres the thing. NOw is overclocking something you should only do for yourself and not others? WHat i mean is this, i have a computer i built for my family. THey use it for games mostly. Im thinking about selling my processor to a friend and buying another one, thats the same type. Curently i have a atlon 60 3000. Now im thinking about using the money and buying either a athlon 64 3200 ro 3400, or buying the new 3000 from newegg that are getting great reviews on overclocking. my rig by the way is

gigabyte GA-Kans socket 754
1 gig corsair extreme
thermaltake 480 watt

then try and overclock it to 2.2 or maybe 2.3 if lucky. Now goign with the 3000 and overclocking it would save me about $25 , but i also know theres no garuntee that i will actually be able to overclock it to 2.2 or 2.3 to begin with. NOw lets say i do manage a overclock to 2.2, which would give probally about a 900+on my frontside bus, but honstly how much of a improvement can you honstly expect between a 2.0 to 2.2 , maybe 2.3. lets say even it does work out to be a good overclocker and works good, when i get ready to move out, i wont be here to monitor it. Im thinking that could spell disaster if something happens and im not here to stop serious meltdown. NOt to mention there goes the waranty if something happens and i, and my brohters are out a computer . My family here, are barely computer literate. Should i even bother overclocking it when something can go wrong.

another thing is when it comes time to upgrade i sell the older parts in the computer to help pay for new stuff. Now when i avertise something on ebay i like to put info like "THis processor has never been overclocked". I am always honest to what i did with the computer components. IM not into stuff like" The product is for sell as is". when i get ready to sell the cpu i know im going to get less money when i say this cpu has been overclocked. I think in the long run when it comes to upgrading it will provide better luck.

RIght now i could sell my gig of corsair extreme, take that money buy regular stick of crusial 1 gig of ram, and put it to better processor. By time you spend the money to get good ram, a good motherboard, and good heatsink to overclock you can buy the normal stuff and use that money to actually buy a cpu thats is rated at the speed you were hoping to get out of your overclock with out overclocking it. Not to mention you will get a higher bid when you go to sell it , because people like stuff that hasnt been overclocked. I dont realy think overclocking would probally be good unless you probally save at least $60 or get a great overclock. Like those moble athlon xp 2800 that were getting 3200 speed, but what kind of gear did you have to have to keep it stable not to mention how long would your cpu last. 1, or maybe 2 years. Well let me know what everyone thinks about it. Thanks guys.
February 23, 2006 1:07:31 AM

save your dough and oc your current proc. or if u must, get the venice. u will hit 2.4 GUARANTEED u do some volt mods
February 23, 2006 1:07:50 AM

Im in tha same situation, my pc shared by everybody @ home, they use it 4 email and browsing, I use it for gaming, right now I owe an ASROCK K8VM800, 2x512 Corsair ValueRam and a Sempron 3000+ palermo runing @
1.8 Ghz.

The point is that I do all my OCing experiments @ night, I try different RAM timings and FSB combinations (damn you locked multiplier!!!, somebody there knows how to pinmod it?) and test it 4 results.

A week ago I tried lotsa stuff and It was stable at this conditions:
Processor: 220*9=1980Mhz
HTT: 220*4= 880
Ram: 181.5x2=363, 2-2-2-5 @ 1T, low voltage
Vcore: 1.20V
Mobo Temp: 40ºC
CPU temp: 41ºC

That's only one degree above normal (at night, at day normal temp is 40/48)

Anyways....that wuz doing fine, Super PI=45s, Sciencemark = 867
so I thought "excellent........lets' go for the Gig barrier!!!!" so, FSB = 225*9, booting........bakoom! blue screen, registry fscked up xD in console mode when I wrote "c:D ir" it said error and didnt display the directories xD, I said...."damn you viruses" and fixed it xD.....you can try that :p 

BTW, that failure was caused by the IDE controller? I read something about component tolerance to high fsb's, so that's my guess
Related resources
February 23, 2006 2:00:32 AM

lax the value ram timings. if u want 2225 pay for it. cuz tahts wats causing u to crash. and jack vcore to 1.4 its safe.+ more stable when oc. and vdimm to 2.8 if u ocing ram
February 23, 2006 2:22:31 AM

Get one of the 3000 754's venice cores, they hit 2.8 easy. Also, if you can, upgrade to socket 939 and get a 3000, although this isn't necessary.
February 23, 2006 11:03:38 AM

they dont hit 2.8 easy....of the majority of ocers how many get them to 2.8...u dont even have a 3000. not everyone has a dfi board u kno..
February 23, 2006 11:15:27 AM

Quote:
they dont hit 2.8 easy....of the majority of ocers how many get them to 2.8...u dont even have a 3000. not everyone has a dfi board u kno..

It isn't hard. All the 3000's hit 2.7, so I am almost certain they can make the boost up 100 mhz. Also, all you need to get that high is about 1.55 vcore...
February 23, 2006 8:14:21 PM

ocing aint just bout the cpu u kno...the motherboard has to be able to do the fsb too.. and stop talking like everyone has a dfi board.
February 24, 2006 12:00:13 AM

Almost all boards today support over 250+ fsb, even most low end cheaper builds. The only drawback should be core voltage, which it is hard to find boards that go very high; however, you shouldn't need that much to get a venice far.
February 24, 2006 1:11:24 AM

go to any oc forum.. very rarely does a 3000 get past 2.7
February 24, 2006 1:12:31 AM

thats probably because people want to do 1:1 with pc4800.
February 24, 2006 1:21:31 AM

having a 2.6 - 2.7 cpu with higher bandwidth ram almost always outperforms a 2.8+ with lower bandwith ram
February 24, 2006 1:55:14 AM

possibly. what if they still have high bandwidth ram though? :o 
February 24, 2006 7:29:35 PM

First off your current 3000 will oc fairly well. It's on a 10x multi right? 2.0ghz?

The 939 3000's run on a 9x multi and 1.8ghz.

The problem will be that you're running the wrong mobo to be oc'ing. I don't know exactly what it'll do, but if i were you, i'd go do my research on that first.


To get a good oc, you have to have any cpu, a decent oc'ing mobo, good ram (if you want to keep as much of the bonus oc as you can...), a good stable psu, a decent cooling system, and most importantly be willing to do some research in to what other people have already done, or NOT been able to do, with the components you have by searching for the information (asking for help, is rarely useful, because most people just don't care what you're trying to do And 99.99% of the people out there responding don't have your components, so they aren't even qualified to be opening their big mouth [no bitterness there i swear lol]).
February 24, 2006 7:31:54 PM

Quote:
First off your current 3000 will oc fairly well. It's on a 10x multi right? 2.0ghz?

The 939 3000's run on a 9x multi and 1.8ghz.

The problem will be that you're running the wrong mobo to be oc'ing. I don't know exactly what it'll do, but if i were you, i'd go do my research on that first.


To get a good oc, you have to have any cpu, a decent oc'ing mobo, good ram (if you want to keep as much of the bonus oc as you can...), a good stable psu, a decent cooling system, and most importantly be willing to do some research in to what other people have already done, or NOT been able to do, with the components you have.


ANY CPU???
-.-
February 24, 2006 7:36:03 PM

Pretty much...

Can you think of a 754 or 939 cpu that can't oc?

How about a p4 on ddr or ddr2 (screw rambus tech...)?
February 24, 2006 7:43:05 PM

Quote:
Pretty much...

Can you think of a 754 or 939 cpu that can't oc?

How about a p4 on ddr or ddr2 (screw rambus tech...)?

Clawhammer 754's suck at oc'ing.
Prescott's can't without powerful cooling.
3500 Winchester's weren't too great, neither were the Newcastles.
The FX51/53 sucked at OC'ing too. Shall I continue?
February 26, 2006 12:59:50 AM

Personally, I think it's the mobo :S, but right now I cant afford anything better, of course I want a better one, with better BIOS options to change the ratios & stuff to get better results @ OCing, 'bout tha voltage....I'll try 1.4 to see how far It will get, and will set mem v @ Normal or High (how lame are those 3 options! xD), once I get some time to do that.......

What other kind of problems can I get from Ocing (other than..):
Meltdown (very difficult)
Overvolting
Registry corruption
data corruption
stff
February 26, 2006 2:52:12 AM

Quote:
Personally, I think it's the mobo :S, but right now I cant afford anything better, of course I want a better one, with better BIOS options to change the ratios & stuff to get better results @ OCing, 'bout tha voltage....I'll try 1.4 to see how far It will get, and will set mem v @ Normal or High (how lame are those 3 options! xD), once I get some time to do that.......

What other kind of problems can I get from Ocing (other than..):
Meltdown (very difficult)
Overvolting
Registry corruption
data corruption
stff


Instability...lots of that.
Data corruption is rare.
February 26, 2006 2:21:17 PM

Sorry i usually don't respond on weekends...

So in your technical opinion, the 754 'suck' in general at oc'ing. That's funny, i hadn't heard that, you didn't just make that up did you? Because if you'll do a search, there's alot of people oc'ing 754's, and they wouldn't be doing that unless it was capable of oc'ing. PresHott's can't without proper cooling...so they do oc and well i might add (as to getting xtra performance out of them, well....that's a different story). The fx 51/53 are your best example, but they still fall in the "do they oc" range. I didn't ask if they oc'd well, i just said they oc. Even the 51/53s have another 10-15% you can get out of them. So far you haven't listed a single cpu that won't oc. =P Oc'ing...your results may vary. You could get a fx 57 that can only oc 5% or you could get one that oc's 35%...It's all still an oc.


And data corruption is rare? Hmm...Have you used a computer before? What are you doing here, BILL GATES, don't try to brainwash us!!!! You haven't used your own product in so long, that you've grown falsely confident! It must be your secretary writing in, while you dictate to her/him. I wouldn't call data corruption rare. When i've been oc'ing, i have gotten it fairly often, and even have had to reinstall windows on occasion. I've had to reinstall programs on several occasions, while oc'ing (benching programs especially, however opening any program before you're completely stable is risky).
February 26, 2006 3:04:52 PM

I have a Sempron64 2600 and 3000. The former does 2.2 on stock air. The latter does 2.4 on air. Both are in low budget boards and both were under a 100 bucks. Yeah I know, they've only got 128kb L2 but who cares.
February 26, 2006 6:27:37 PM

Quote:
Sorry i usually don't respond on weekends...

So in your technical opinion, the 754 'suck' in general at oc'ing. That's funny, i hadn't heard that, you didn't just make that up did you? Because if you'll do a search, there's alot of people oc'ing 754's, and they wouldn't be doing that unless it was capable of oc'ing. PresHott's can't without proper cooling...so they do oc and well i might add (as to getting xtra performance out of them, well....that's a different story). The fx 51/53 are your best example, but they still fall in the "do they oc" range. I didn't ask if they oc'd well, i just said they oc. Even the 51/53s have another 10-15% you can get out of them. So far you haven't listed a single cpu that won't oc. =P Oc'ing...your results may vary. You could get a fx 57 that can only oc 5% or you could get one that oc's 35%...It's all still an oc.


And data corruption is rare? Hmm...Have you used a computer before? What are you doing here, BILL GATES, don't try to brainwash us!!!! You haven't used your own product in so long, that you've grown falsely confident! It must be your secretary writing in, while you dictate to her/him. I wouldn't call data corruption rare. When i've been oc'ing, i have gotten it fairly often, and even have had to reinstall windows on occasion. I've had to reinstall programs on several occasions, while oc'ing (benching programs especially, however opening any program before you're completely stable is risky).


A chip that can't get anything more out of itself will not be stable at its stock speeds. Any chip can "overclock", sure, but not that many chips will really see more than a 15% boost in clock speed. Also, Prescotts have the sky as the limit in terms of what they are capable of provided you have powerful cooling and high voltage.
February 27, 2006 12:51:39 PM

So what r we arguing about =P.
February 27, 2006 1:46:55 PM

Back to the original point of the thread. I just, and by just I mean 2 months ago, upgraded from a Northwood to a venice and passed down the northwood to my family. When I had the northwood I pushed is fairly hard (the board sucked at OC'ing but whatev), but when I left home I reset the CMOS and upped the vcore one notch just to make sure everything stayed stable.

Point is, I didn't want to drive 3 hours home to fix a computer just bc the OC crashed. Personaly, I woudl say OC the thing at night, but keep records of what is stable so you can reset it beforeyou go to bed so inteh morning its fully functional. Then when you find a good and stable OC bench the hell outta it to test it. Once you leave I woudl def reset it back to stock.

On a side note: I think i read somewhere around here that over volting the cpu reduces the life of a proc to 3yrs+ still, and I bet something else will give out before the proc does. Over volting only really kills it when "noobs" go too far or good OC'ers get really greedy heh :p 

Just do it in small increments and you should be fine
February 27, 2006 2:00:48 PM

Quote:
Back to the original point of the thread. I just, and by just I mean 2 months ago, upgraded from a Northwood to a venice and passed down the northwood to my family. When I had the northwood I pushed is fairly hard (the board sucked at OC'ing but whatev), but when I left home I reset the CMOS and upped the vcore one notch just to make sure everything stayed stable.

Point is, I didn't want to drive 3 hours home to fix a computer just bc the OC crashed. Personaly, I woudl say OC the thing at night, but keep records of what is stable so you can reset it beforeyou go to bed so inteh morning its fully functional. Then when you find a good and stable OC bench the hell outta it to test it. Once you leave I woudl def reset it back to stock.

On a side note: I think i read somewhere around here that over volting the cpu reduces the life of a proc to 3yrs+ still, and I bet something else will give out before the proc does. Over volting only really kills it when "noobs" go too far or good OC'ers get really greedy heh :p 

Just do it in small increments and you should be fine


By that time, nobody will want the cpu.
Also, I don't buy that "it will die in a few years" bullcrap provided you have good cooling.
February 27, 2006 2:21:26 PM

lol i'm begining to like you AK...you remind me of one of my friends from back in the day.


Sort of a, "believe it when i see it", kind of guy.

OverVolting is pretty much the only thing thats going to kill a cpu. Common sense should tell you that though. It's just a matter of degree. You run 3v through a 2v transistor and it'll blow up. Lol, same thing is true for cpu's. They're designed to run with a certain amount voltage, if you put more then that, you have the possiblity of burning it out, and/or increasing electrical migration (ie reducing the useful life of your cpu).
February 27, 2006 3:00:29 PM

I run 3.5v through my memory, and they work fine don't they?
Reminds me of my dumbass friend who bought an Intel and is totally against oc'ing anything, claiming that it will break/die/overheat.
February 27, 2006 4:27:52 PM

Yeah normal Over volting is fine, say 1.6v through a venice core, but running 1.9V through it will dman near fry it even with good cooling. Maybe not right away but definately shorten the life span.

Also AK, your redline is designed for 3.5v+ but mine is designed for around 3v, it just depends on the component. I try'd over volting my 7800GTX and it did NOT like it at all lol. Just upped it .1v from stock and it spazzed, ill try it again this weekend.

Im hoping for 575-600mhz on water.

Still chasing 3Ghz on my venice to AK, my system is starting to like 2860Mhz, so it seems like its just a matter of time.
February 27, 2006 4:30:13 PM

Quote:
Yeah normal Over volting is fine, say 1.6v through a venice core, but running 1.9V through it will dman near fry it even with good cooling. Maybe not right away but definately shorten the life span.

Also AK, your redline is designed for 3.5v+ but mine is designed for around 3v, it just depends on the component. I try'd over volting my 7800GTX and it did NOT like it at all lol. Just upped it .1v from stock and it spazzed, ill try it again this weekend.

Im hoping for 575-600mhz on water.

Still chasing 3Ghz on my venice to AK, my system is starting to like 2860Mhz, so it seems like its just a matter of time.


I almost guarantee you won't hit 3ghz, as I haven't already done it myself. I have an XP90 and a vantec tornado, the best air cooling money can buy ><
Nobody hits over any 2.8 unless with LN2 or phase changing cooling.
February 27, 2006 4:59:25 PM

Yea...but they're designed to run up to 3.5 still under warrenty...
February 27, 2006 5:03:44 PM

Quote:
Yea...but they're designed to run up to 3.5 still under warrenty...


That's the only reason I don't run it at 4v; it could be done considering I cool it with a Vantec Tornado.
February 27, 2006 5:28:24 PM

Sorry i didn't see superfly had already posted that.
!