VRM implications in performance?
in [Hardware]
|
Prev: Win XP home network problem --- how to delete old one
Next: Replacement of motherboard VGA dsub 15 connector at Digikey or mouser???
From: Giovanni Azua on 22 Jan 2008 17:37 hi, I have been using the wrong VRM part with a system equipped with 2x Paxville Xeon processors and had experienced no problems. I wonder if there is any performance loss because of this? anyway in a few days I will find out when I get the right VRM ... TIA regards, Giovanni
From: kony on 23 Jan 2008 00:23 On Tue, 22 Jan 2008 23:37:52 +0100, "Giovanni Azua" <bravegag(a)hotmail.com> wrote: >hi, > >I have been using the wrong VRM part with a system equipped with >2x Paxville Xeon processors and had experienced no problems. >I wonder if there is any performance loss because of this? >anyway in a few days I will find out when I get the right VRM ... If it has a compatible range of (it's programmable) outputs that the processor(s) require, or at least close enough to not cause damage, and is capable of the current, it can work ok. We'd need to know the technical specifics of the two VRMs to know if they are different enough to matter in some other way (or at least speculate about it, for example if one barely has enough current capability it might not last as long but it's not like you could mark a failure date on a calendar). If by performance you meant processing performance, no, so long as the system runs stabily that difference will not speed up or slow down the processor. If the processor were getting too high a voltage and that caused overheating, or too low a voltage and it increased the required charge time beyond a level possible based on the clock rate, that could instead cause computational errors which may cause an effect seemingly like stalling code if not outright indications of errors like a crash... but you make no mention of this happening, though it might be more likely at full load than lesser load.
From: Giovanni Azua on 27 Jan 2008 08:29
hi kony, Many thanks for your explanation. regards, Giovanni "kony" <spam(a)spam.com> wrote in message news:jejdp35r1a35spv74dvgp7tf0276fjs7p7(a)4ax.com... > If by performance you meant processing performance, no, so > long as the system runs stabily that difference will not > speed up or slow down the processor. If the processor were > getting too high a voltage and that caused overheating, or > too low a voltage and it increased the required charge time > beyond a level possible based on the clock rate, that could > instead cause computational errors which may cause an effect > seemingly like stalling code if not outright indications of > errors like a crash... but you make no mention of this > happening, though it might be more likely at full load than > lesser load. > |