Any truth to the "30 caps is better" claim?

[Rick Yerger]

I've got my Northwood search narrowed down to two chips, one
has 12 capacitors and the other 30. The guy selling the 30-cap
claims it runs much cooler and overclocks better than 12-cap
versions, even though both chips have the same S-spec and
thermal/electrical specs.

Google brought up a few other people claiming the same thing,
but I didn't find any hard numbers to back it up. Do the 30-cap
chips really run cooler or overclock better?

 

[Richard Hopkins]

I've got my Northwood search narrowed down to two chips,
one has 12 capacitors and the other 30. The guy selling the
30-cap claims it runs much cooler and overclocks better
than 12-cap versions, even though both chips have the same
S-spec and thermal/electrical specs.

Well that's a new one on me. At end of day there are many reasons why one
chip might run cooler and overclock better than another, and most of them
are on the CPU core itself.

Mind you, before you commit to either of these, make sure it's the right
sort - one hopes you're talking about two 3.0 @ 800FSB chips here. If not,
suggest you give my reply on your previous thread a scan.
--


Richard Hopkins
Cardiff, Wales, United Kingdom
(replace nospam with pipex in reply address)

The UK's leading technology reseller www.dabs.com

 

[Rick Yerger]

Well that's a new one on me. At end of day there are many reasons why one
chip might run cooler and overclock better than another, and most of them
are on the CPU core itself.

Mind you, before you commit to either of these, make sure it's the right
sort - one hopes you're talking about two 3.0 @ 800FSB chips here. If not,
suggest you give my reply on your previous thread a scan.

I am, and did. Here's the auction: http://tinyurl.com/a67yp

He's claiming it does 250FSB at stock Vcore. Hope it's true.
From what I've been reading elsewhere, the major limitation on
the IC7-G may be with memory voltage -- the current PC4000/
4400 2-2-2-yadda-yaddas such as OCZ, Geil and Corsair want
at least 3.1V, some up to 3.5V, while the IC7-G maxes out at
2.8V unless it's modified.

Do you have a preference among those 3 memory brands, or
a particular model you'd recommend?

Thanks again for all your help.

 

[NoNoBadDog!]

I am, and did. Here's the auction: http://tinyurl.com/a67yp

He's claiming it does 250FSB at stock Vcore. Hope it's true.
From what I've been reading elsewhere, the major limitation on
the IC7-G may be with memory voltage -- the current PC4000/
4400 2-2-2-yadda-yaddas such as OCZ, Geil and Corsair want
at least 3.1V, some up to 3.5V, while the IC7-G maxes out at
2.8V unless it's modified.

Do you have a preference among those 3 memory brands, or
a particular model you'd recommend?

Thanks again for all your help.

Corsair does just fine at 2.8.

Bobby

 

[Wayne Youngman]

"Rick Yerger"

He's claiming it does 250FSB at stock Vcore. Hope it's true.

Hi Rick,

I have two P4 Northwoods, a 2.6GHz (d1) and a 2.8GHz (mo/30 capper). Both
are over a year and a half old now and have been overclocked all the time.

The 2.6 is running at 3.0GHz using 1.625 vCore

The 2.8 is running at 3.5GHz using (stock) 1.525vCore

So in my experience a 30-capper is indeed a very good overclocking chip, but
could just be luck of the draw?

Wayne ][

 

[Richard Hopkins]

He's claiming it does 250FSB at stock Vcore. Hope it's true.

Yeah, so do I. It's not beyond the realms of possibility by any stretch, and
call me cynical, but I can't help but subject any claims like this,
especially when the CPU concerned is up for sale, to a small amount of
sensible skepticism. If it's that good, why's it being sold? Even if it
"does" 250 @ default, does it do it stably, 24/7?

From what I've been reading elsewhere, the major limitation
on the IC7-G may be with memory voltage -- the current
PC4000/ 4400 2-2-2-yadda-yaddas such as OCZ, Geil and
Corsair want at least 3.1V, some up to 3.5V, while the
IC7-G maxes out at 2.8V unless it's modified.

If you want more voltage, if you can do without serial and parallel ports,
and if you can find one, get an IC7 Max3 instead. Voltage for voltage it's
a slightly better overclocker than the -G, and you can push the DRAM voltage
up to 3.2.

However, I'm using Corsair XMS4400 atm at 275MHz FSB 2.5/3/3/6 at 2.7 volts,
thus it's arguable that you won't need to go beyond the 2.8 offered by the
IC7-G.
--


Richard Hopkins
Cardiff, Wales, United Kingdom
(replace nospam with pipex in reply address)

The UK's leading technology reseller www.dabs.com

 

[Rick Yerger]

Yeah, so do I. It's not beyond the realms of possibility by any stretch, and
call me cynical, but I can't help but subject any claims like this,
especially when the CPU concerned is up for sale, to a small amount of
sensible skepticism. If it's that good, why's it being sold? Even if it
"does" 250 @ default, does it do it stably, 24/7?

Feeling out eBay sellers is part experience, part intuition and
part just dumb luck. I always steer clear of sellers with bad
or no feedback, or who're trying to run a business via eBay,
or who offer no warranty etc. This seller in this case is a
private party and offered a complete satisfaction guarantee,
so there isn't much risk on my part. In fact I'm far more
skeptical about the IC7-G motherboard which is due to
be delivered today. It was the last one a local retailer had,
and I've rarely had good experiences with "last-one"
merchandise.

If you want more voltage, if you can do without serial and parallel ports,
and if you can find one, get an IC7 Max3 instead. Voltage for voltage it's
a slightly better overclocker than the -G, and you can push the DRAM voltage
up to 3.2.

The lack of serial ports was a deal-breaker for me. I need at
least one.

However, I'm using Corsair XMS4400 atm at 275MHz FSB 2.5/3/3/6 at 2.7 volts,
thus it's arguable that you won't need to go beyond the 2.8 offered by the
IC7-G.

How much of a real-world performace difference would there
be between your specs and 250FSB 2/2/2/5?