2T - 1T confused.

Associate
Joined
12 Oct 2005
Posts
517
Location
Scotland
There are a few threads kicking around mentioning 2T losing a lot of performance over 1T. Most of the advice given out is telling people to change from e.g. 4 x 512mb sticks to 2 x 1 GB sticks so that they can run at 1T instead of 2T.

I have 2 x 1 GB Mushkin sticks in my machine and according to the post screen they have been running at 2T Dual channel since day one. do i have something set up wrong? Am i really losing the amount of performance that people say you do?

A confused and in need of explanations,
Zeph.
 
Soldato
Joined
17 Jan 2005
Posts
8,544
Location
Liverpool
Yeh you're losing out on a bit of performance there. Go into the BIOS and there should be an option in the RAM settings page called something like Command Per Clock (CPC), or an option set to 2T. Just switch that to 1T and it should be sorted:)
 
Associate
OP
Joined
12 Oct 2005
Posts
517
Location
Scotland
Changed :)

Is there any way i could test performance 1T versus 2T ie some program i could run for each?

[Edit - Superpi 1MB is about 1.5 seconds faster]
 
Last edited:
Soldato
Joined
12 Sep 2003
Posts
10,002
Location
Newcastle, UK
Zephyrus said:
One more question :p Does 1T/2T affect overclocking the cpu?

The command rate would affect the fsb you could get on the ram which is linked I guess to the cpu in a way. You'd probably reach a higher fsb with 2T but that would probably just cancel out the fact you could have less fsb while running 1T. You'd have to bench to find which was best...
 
Soldato
Joined
12 Sep 2003
Posts
10,002
Location
Newcastle, UK
Zephyrus said:
Changed :)

Is there any way i could test performance 1T versus 2T ie some program i could run for each?

[Edit - Superpi 1MB is about 1.5 seconds faster]

SuperPi, even 3DMark might show different scores, along with timedemo's for games.
 
Associate
OP
Joined
12 Oct 2005
Posts
517
Location
Scotland
Cheers.

Currently upping the cpu a few Mhz at a time again and priming for 5 mins till i find where it will become unstable again.

[edit - Nvm made no difference :p ]
 
Last edited:
Don
Joined
7 Aug 2003
Posts
44,276
Location
Aberdeenshire
Running memory at 2T makes little difference:-

1T
fear1t.JPG


2T
fear2t.JPG


You won't be able to overclock your memory as far, though it makes little difference as 1GB sticks don't overclock/get as tight timings as 512MB sticks any way.

Jokester
 
Last edited:
Man of Honour
Joined
25 Oct 2002
Posts
31,707
Location
Hampshire
Jokester said:
Running memory at 2T makes little difference:-
Jokester

To be fair though, FEAR is a horrendously GPU-limited benchmark even for people running a couple of cards such as yourself (I'm assuming you were running your playing settings and not 640x480 or similar).

Games which are more cpu/chipset/memory limited such as flightsims etc might experience a slightly larger performance discrepancy.

That said, the importance of memory speed/timings etc is often overplayed on the A64 platform I feel. For example the difference in real world performance between a value kit of RAM and a 'posh' one costing maybe twice as much isn't really all that big. Spending an extra £100 (or even £50) on RAM gains you far less performance than the equivalent amount extra spent on graphics card.

Gone a bit offtopic here, but it's just a general musing of mine that many people (including myself) have often gone for a 'balanced' system build (i.e. decent cpu, decent memory, decent gfx) when in reality we would be better off overloading on the gfx card and saving the pennies on memory speed/timings - assuming we are gamers, of course.
 
Back
Top Bottom