They are supposedly matched pairs. Not sure if it makes alot if diff but
thats what I used.


"Strange Brewer" wrote in message
...
Thanks for the feedback jaeger - really appreciated.

I guess two followup questions:

1) There is a 433 MHz HyperX KHX3500/512 with the same statistics as
the Corsair -- any reason I shouldn't get the Kingston since it is a
little cheaper?

2) Some of the vendors sell "2 packs" marketed as 1 GB, but if I buy
the chips separate they are cheaper. Nothing special about the 2 packs
right? (Yes, I'm planning to use the same memory chips in all 4
locations.)

TIA-
SB

On Tue, 1 Jul 2003 22:51:23 -0700, jaeger wrote:

In article ,
says...

In what I have read, the lower the memory latency, the better the
performance at a given bus speed. The Kingston HyperX DDR400
(KHX3200/512) has a 2-2-2-6-1 rating (@ 400 MHz), which seems faster
than the "recommended" Corsair CMX512-3500C2, which has a 2-3-3-7-T1
(?what is T1?) rating (@ 433 MHz).

Any compnay that adevertises "1T" operation as being better or faster is
being misleading. Intel chipsets ONLY use 1T, so this applies to any
RAM you'll put in there aside from non-compliant generic crap.

(Short version of this question: 400 MHz 2-2-2-6-1 seems like it would
be faster than 433 MHz 2-3-3-7-1? Is this right? If not, why? Educate
me!)

No. Go for the speed. You'll never notice lower latency unless you use
a synthetic benchmark. Why is too long to explain in a ng post, but
assuming you have a basic idea of how memory works-applications that are
memory bandwidth intensive are generally the LEAST affected by lower
latency. Because once the data starts streaming, latency has no impact.

1) WILL this HyperX chip work?

Yes.

2) WOULD it be faster to go with the Corsair?

No.