XP2500+ and PC2700

M

Muhammad Berki

I recently upgraded my cpu to a Barton XP2500+ and overclocked it
successfully to 2Ghz from 1.8Ghz by increasing the FSB to 182. I am thinking
about getting a new nForce2 Ultra based motherboard replacing this KT400
board. Will I be able to push the FSB to 200 using Samsung PC2700 DDR? Would
I be able to clock it to 2.2Ghz or will I have to get PC3200 memory first?
Is the FSB determined by the motherboard or the memory?
I bought my CPU recently it is week 43 manufactured and have heard these
are permanently locked by AMD. Anyone managed to unlock these yet? Is there
any point unlocking them as 11x200 seems a good overclock.

Thanks all.
 
K

kony

I recently upgraded my cpu to a Barton XP2500+ and overclocked it
successfully to 2Ghz from 1.8Ghz by increasing the FSB to 182. I am thinking
about getting a new nForce2 Ultra based motherboard replacing this KT400
board. Will I be able to push the FSB to 200 using Samsung PC2700 DDR? Would
I be able to clock it to 2.2Ghz or will I have to get PC3200 memory first?
Is the FSB determined by the motherboard or the memory?
I bought my CPU recently it is week 43 manufactured and have heard these
are permanently locked by AMD. Anyone managed to unlock these yet? Is there
any point unlocking them as 11x200 seems a good overclock.

Thanks all.

I cannot speak of your specific motherboard (especially since you've
failed to mention make/model/etc), but in general it would be (mostly)
a waste of time to upgrade from a KT400 to an nForce2 board... perhaps
3-5% performance increase, but a KT400 board should be able to hit 200
FSB.

The first thing to replace is the memory. You need memory capable of
DDR400, PC3200 memory. Trying to run 200 FSB with PC2700 memory is
asking for problems and if you ran an asynchronous "-33" memory bus
you'd have significantly worse performance than keeping it at 182MHz.

Perhaps I should back up a bit... Did you try running it at 200MHz, or
just one "notch" above 182MHz? You might start having trouble around
184MHz simply because of the resulting high PCI bus speed, but if you
did 200MHz FSB, that may put the PCI bus back in spec (might depend on
the board, it's PLLIC, whether it has a 1/6 divider support).

Frankly, I wouldn't bother buying a new motherboard just to get it
200MHz higher anyway... that $ could be better spent upgrading to a
new CPU AND board later, for which you'd need at least the PC3200
memory anyway, so you might as well buy some good memory first.


Dave
 
P

Peter

You overclock to get 10% extra? Really not worth risking all the trouble you
get yourself into.
But then again, I never understood overclockers, I mean, instead of
rendering a 3D scene that took 60 sec, it now takes 54 seconds... big deal!
 
K

kony

You overclock to get 10% extra? Really not worth risking all the trouble you
get yourself into.
But then again, I never understood overclockers, I mean, instead of
rendering a 3D scene that took 60 sec, it now takes 54 seconds... big deal!

The thing is, people DO buy CPUs faster than the slowest model... if
the performance isn't worthwhile then wouldn't everyone buy the
cheapest CPU?

The key is know-how. Once you know you can make a better informed
choice.

You ever speed in your car? That seems like a bigger risk to me yet
it doesn't stop many people from doing so.
 
P

Peter

I rather pay for an extra 50% than a 10% speed increase with risks involved
(just wait until you have one of these extreme hot summer days...)
Not only that, how many times do we have a problem with our computer that
seems hard to solve. Overclocking your processor only adds to the confusion
once you do have problem.

People do want faster computers of course, but without the risks over
overclocking.
Since when do have more brains that Intel and AMD's technicians? There is a
reason why a processor has a max. speed.
Of course, lots of overclockers won't have problems, but some will. I hope
you're just not one of them.
 
M

~misfit~

Peter said:
I rather pay for an extra 50% than a 10% speed increase with risks
involved (just wait until you have one of these extreme hot summer
days...)
Not only that, how many times do we have a problem with our computer
that seems hard to solve. Overclocking your processor only adds to
the confusion once you do have problem.

People do want faster computers of course, but without the risks over
overclocking.
Since when do have more brains that Intel and AMD's technicians?
There is a reason why a processor has a max. speed.
Of course, lots of overclockers won't have problems, but some will. I
hope you're just not one of them.

Intel or AMDs have no control over what the thermal carachteristics of the
case that the CPU is going to be used in are like. They err on the side of
caution. Anyone with a modicum of sense can fairly easilly get a few (or a
lot) more MHz out of a CPU by paying attention to details.
 
K

kony

I rather pay for an extra 50% than a 10% speed increase with risks involved
(just wait until you have one of these extreme hot summer days...)
Not only that, how many times do we have a problem with our computer that
seems hard to solve. Overclocking your processor only adds to the confusion
once you do have problem.

People do want faster computers of course, but without the risks over
overclocking.
Since when do have more brains that Intel and AMD's technicians? There is a
reason why a processor has a max. speed.
Of course, lots of overclockers won't have problems, but some will. I hope
you're just not one of them.

You are making assumptions without any testing or accumulation of
data. It IS possible to have problems, as in anything else in life
done without the right equipment, experience, etc.

Of all the risks in life, drinking public water, exposure to solar
radiation, driving in the car, etc, might be considered a lot more
significant than overclocking. It's not for everyone, only those who
know how, accept the risks.

It isn't true that you're "out-thinking" Intel or AMD technicians when
overclocking, they spec their parts based on yield and market
demand... another detail to consider when trying an overclock, what
the yields are like, what others have had success with.

It all boils down to testing. The system MUST be thoroughly tested,
but this is necessary for ANY system, non-overclocked included, for it
to be considered "safe".

Overclocking has it's failures, but is proven to work too. If
"safety" and stability are important I do hope you're not using
windows?


Dave
 
P

Peter

Overclocking is always a risk. You can indeed do a simple test to see if
your system still works after you overclock it, but it's not the test of a
professional!!!!!

It's is very hard to predict the temperature inside a computer case. It
depends on how much power (=heat) is used at any specific time
(processor(s), cards, drives, etc) and also depends on the temperature
outside the case and an airflow that always has to be constant. Trust me,
the older a computer gets, the more irregular this airflow is (dust, old
components). Even a case that is removed or isn't screwed properly can make
a difference! If you're overclocking close to the limits, then it only takes
a slight rise in temperature to make destroy one single transiator, one of
the billions on a CPU. Only one needs to be destroyed to make the processor
useless. Maybe not something for you to worry about as a gamer, but for some
situations this can have disastrous consequences, either because the system
stops working, or because of the occurance of sudden strange problems that
take quite a while to solve (and time is money in a company!!). After all,
who thinks that the CPU is the culprit?



Let me ask you a question... how many system administrators overclock CPUs
in the company's servers?
None. Why? Because they understand the risks. No management would also allow
their administrators to overclock.

If they want more speed, then they'll spend more money (better or more
processors).

Every company wants to save money, but ask yourself why they don't accept
overclocking to save this money.



It's always the same reason why people overclock; money and frame rates in
games.

As a regular user it's hardly noticeable if you add some 8-10% speed.
Photoshop, Cinema 4D, Illustrator, compressing a file, etc. You won't
experience any noticeable difference, unlike you use a stopwatch.



But in game it's slightly, but then again, SLIGHTLY more visible. Instead of
30 frames, you might have 33; big deal!

I DO have money and because of that I buy my speed, not 10%, no, 300-400%
more than my old machine. Can you understand that I laugh in such a moment
about your 8-10% extra speed?



And your examples of people being killed (radiation, driving, drinking
water) are totally ridiculous. There are more things in life that are
considered bad situations than people dying you know ha!ha! For an amateur
like you it might not be important if something does go wrong, but for
companies or professional users.
 
K

kony

Overclocking is always a risk. You can indeed do a simple test to see if
your system still works after you overclock it, but it's not the test of a
professional!!!!!

That's a subjective opinion. If _you_ don't have the prerequisite
knowledge to overclock, don't, it may not be safe for you.

You have been jumping to conclusions based on assumptions.
You assume Intel engineers are the ones who make final determination
of what speed is stamped on the CPU. Further you assume this is the
fastest the CPU will operate properly by their testing.

It's is very hard to predict the temperature inside a computer case.

Actually it's not hard to get in the ballpark, but even easier to take
temps. Funny thing is we actually HAVE temp sensors now, and they
aren't always completely accurate but serve their purpose.

It
depends on how much power (=heat) is used at any specific time
(processor(s), cards, drives, etc) and also depends on the temperature
outside the case and an airflow that always has to be constant.

Actually the airflow doesn't have to be constant, but pointing all
this out is pointless, since all these factors which you go on to
describe, apply to NON-overclocked systems too.

Trust me,
the older a computer gets, the more irregular this airflow is (dust, old
components).

Above you made vague reference to "professional". Professionals clean
their equipment on regular intervals, at least when it's
time-effective. Still this applies to non-overclocked as well.

Even a case that is removed or isn't screwed properly can make
a difference! If you're overclocking close to the limits, then it only takes
a slight rise in temperature to make destroy one single transiator, one of
the billions on a CPU. Only one needs to be destroyed to make the processor
useless. Maybe not something for you to worry about as a gamer, but for some
situations this can have disastrous consequences, either because the system
stops working, or because of the occurance of sudden strange problems that
take quite a while to solve (and time is money in a company!!). After all,
who thinks that the CPU is the culprit?

This speculation you're making, is contradicted by real-world testing
and use. Years of it.
Let me ask you a question... how many system administrators overclock CPUs
in the company's servers?
None. Why? Because they understand the risks. No management would also allow
their administrators to overclock.

It's also because:

A) It's not their money. At home they have an incentive to
overclock. At work they're not getting paid to overclock, not their
job description.
B) The bottleneck is the LAN, not the system.
C) They don't have the time, nor could allow the server downtime, to
properly test.
D) Apathy
E) You assume it. You have not surveyed them.

If they want more speed, then they'll spend more money (better or more
processors).

Usually better drives or LAN, but that's beside the point.
Every company wants to save money, but ask yourself why they don't accept
overclocking to save this money.

Because there is a difference in downtime cost. It downtime costs
1000 times more to a company than an individual, it makes sense to
reduce the risk by 1000X. That's risk of failure in general, not
specifically from overclocking. That includes many things not usually
seen in PCs, including RAID 0 and redundant power, backup power,
employee screening, etc, etc.

It's always the same reason why people overclock; money and frame rates in
games.
So?

As a regular user it's hardly noticeable if you add some 8-10% speed.
Photoshop, Cinema 4D, Illustrator, compressing a file, etc. You won't
experience any noticeable difference, unlike you use a stopwatch.

If you assume 10%, then you're right at the border of what's noticable
to the user. Stopwatches aren't necessary to see larger improvements
or on longer jobs or jobs with real-time requirements, like video
encoding or gaming.

But in game it's slightly, but then again, SLIGHTLY more visible. Instead of
30 frames, you might have 33; big deal!

Actually it IS a very big deal. Obviously you're not an avid gamer.
30FPS is unplayable in many games, assuming a rough average rather
than a minimal framerate. In such situations every frame counts until
the rate gets higher. The reasons to overclock are similar to those
for buying a faster CPU, which we've (most of us anyway) also done
over the years.

I DO have money and because of that I buy my speed, not 10%, no, 300-400%
more than my old machine. Can you understand that I laugh in such a moment
about your 8-10% extra speed?

Do you understand that we laugh when you think you can buy as much
performance as an overclocker sees? Assuming 8-1% is random,
apparently to support an invalid argument on your part. If you feel
wise in your decisions, fine, whatever makes you happy. Many people
live in ignorant bliss.
And your examples of people being killed (radiation, driving, drinking
water) are totally ridiculous. There are more things in life that are
considered bad situations than people dying you know ha!ha! For an amateur
like you it might not be important if something does go wrong, but for
companies or professional users.

Actually it's quite relevant. You claim "risk", but in most other
areas of life a larger relative risk is acceptable.

The bottom line is that YOU, personally, are incapable of safely
overclocking. Likely there are other things you can do safely that
others can't, and likewise you would argue that it isn't inherantly
unsafe to do. The difference is experience. You don't have enough
experience to make your argument, instead jumping to conclusions not
supported by evidence then jumping back and forth between that and
performance benefit when it's clear that we DO want more performance,
even 10% at a time. CPUs are sold in speed grades quite because of
this.

If you like to pause a few extra seconds waiting on your system, using
that time to think that you know better than those who have properly
working, overcloked systems, then good for you? It is a choice. If
you can't safely overclock, don't. A lot of us can and have proven it
time and time again, but then we didn't have a mental block
interfering with our learning.


Dave
 
P

Peter

The fact that you deny everything I wrote, makes it very clear that
absolutley NO discussion with you is possible.
That's only because you think that I consider you a 'fool' and now your
emotions are taking control over your common sense. It's a very common
response from a home user who can built his own computer and now thinks he
knows everything. For your information, the A+ is the beginners exam that
most beginners start with... I've already past that stage many and many
years ago.

I was only pointing out the risks of overclocking and I really don't care
that you ignore them, as long as there are others who do read my ideas about
this matter with an open mind.

Have a nice day Pacman
 
K

kony

The fact that you deny everything I wrote, makes it very clear that
absolutley NO discussion with you is possible.


Well if you'd only realized that a couple days ago... ;-)

Discussions are possible, but you're not looking to discuss. You
expected to make a declaritative statement but then came across
someone with enough experience that they can manage and reduce the
risk to a minimal level.
That's only because you think that I consider you a 'fool' and now your
emotions are taking control over your common sense.

Actually no, but congratulations on jumping to a false conclusion
again. Anything to avoid the subject, right? The subject WAS
overclocking, and there is plenty of evidence that it can be done
safely, that your conclusion of "risk" isn't relevant to the extent
that the risk is from doing it in ignorance, not inherantly from the
overclocking itself.

If you had written "if you dont' know what you're doing, don't do it",
I would've agreed. Overclocking is not just entering the BIOS and
picking a bigger number, you have to consider the IMPACT of that
choice, and know how to troubleshoot and test, skills needed outside
of overclocking too.

Dealing with an overclocked system is NO DIFFERENT than setting up a
non-overclocked system, in that a non-overclocked system must be
extensively tested... don't even mention "professionals" as you did in
the last post if you don't consider a professional to be someone who
does enough testing to eliminate the risks from overclocking. A
professional doesn't trust their equipment, they PROVE it appropriate,
unless there's liability attached to the equipment's warranty, and in
that case, the warranty and thus liability would be gone when the
system is overclocked, so of course it wouldn't be overclocked in
that specific scenario. Otherwise there is no less validation needed
for a non-overclocked system, no assumption of decreased risk if it
isn't overclocked. It must work 100% correctly either way.
It's a very common
response from a home user who can built his own computer and now thinks he
knows everything. For your information, the A+ is the beginners exam that
most beginners start with... I've already past that stage many and many
years ago.

Do you really want to make this personal? It appears so, that you're
desperately trying to claim you have experience, therefore you know,
and since I disagree, I must not have as much experience?

Keep on assuming, some day the odds will catch up and you'll actually,
accidentally, assume right.
I was only pointing out the risks of overclocking...

Actually, you weren't. You didn't list risks but one or two rarities,
otherwise just a vague reference to "risk", then a description of what
you must assume is news worthy, mentioning cooling and cleaning, then
dodging back and forth with nothing to support your argument.

This isn't a bullshitting contest, there are plenty of real-world
examples of people successfully overclocking. There are plenty of
people who have overclocked dozens of systems or more... this great
risk you assume would've been a deterrent to that, wouldn't you think?
Yet over and over again they had success.

... and I really don't care
that you ignore them, as long as there are others who do read my ideas about
this matter with an open mind.

Have a nice day Pacman

Quite an irony. You mention "open mind" but that's not at all what
you've had. Perhaps you're conservative when it comes to computers.
You repeat what you're taught without testing the boundaries... at one
point is was generally accepted that the world was flat, but someone
tested that.

Is there a risk? Sure. There's also a risk to connecting a hard
drive to a new power supply and turning it on. There's a risk to
trusting a motherboard manufacturer to make a board stable enough that
it'll run properly NON-overclocked. There's a risk that Microsoft
will be found guilty of monopolizing the industry and be broken into
a dozen little pieces. There's a risk a small child will eat your
motherboard battery, BUT, eventually they learn what to eat, and
eventually someone trying to overclock, learns how, too... life IS
risk, but it's RELATIVE risk.

There's a risk you'll go out on usenet and find someone who knows
their stuff, can overclock without anywhere near the degree of risk
you only imply. There's a risk you'll find more than one "someone" if
you visit alt.comp.overclocking et al.

I do not advocate overclocking, I advocate having control over your
OWN system and making an informed choice about what you can do with
it... not your company's systems, or your wife's, or anyone else's.


Dave
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top