nForce3 BIOS settings - what is 1T/2T?

I have an ECS nForce3 v. 1.0A mobo with BIOS 1.0F (Sept. 2005) and am
having trouble getting Kingston PC3200 ValueRAM to work reliably at the
default BIOS settings.

If the 1T/2T parameter (burst mode timing? The manual doesn't say) is
set to 1T, this memory will fail diagnostics, unless the memory bus
speed is set to 333 MHz or slower. Unfortunately the BIOS defaults to
1T if any BIOS parameter is edited, even the time and date. I don't
like to use memory that isn't reliable at the defaults.

The mobo works fine at 1T with PNY PC3200 memory containing chips
marked "BRAVO".

I'm running an NForce 3 chipset also, and I couldn't get Kingston's
performance line of RAM to run reliably in my system. I replaced it with
Mushkin Level 1 PC3500 (and upped to voltage to 2.7V as recommended by
Mushkin) and it's rock solid. The Kingston PC3200 went into a NForce 2
chipset and works great. This certainly doesn't solve your problem, but I
have my doubts about getting your ValueRAM to run stable at 400MHz
regardless of the settings.

Fitz

Fitz wrote:

> I'm running an NForce 3 chipset also, and I couldn't get Kingston's
> performance line of RAM to run reliably in my system. I replaced it with
> Mushkin Level 1 PC3500 (and upped to voltage to 2.7V as recommended by
> Mushkin) and it's rock solid. The Kingston PC3200 went into a NForce 2
> chipset and works great. This certainly doesn't solve your problem, but I
> have my doubts about getting your ValueRAM to run stable at 400MHz
> regardless of the settings.

I guess it's unreasonable for Kingston to make its PC3200 memory
actually work at PC3200 speeds.

I tried two 512MB modules of 3-3-3-8 Mushkin PC3200 Enhanced Memory,
and they were the worst PC3200 sticks I've ever tried and would not
work reliably at 400 MHz, no matter how much I slowed the other
parameters. Maybe that's why one manufacturer's list of approved
modules for its nForce3 mobos lists Mushkin PC3500 but not Mushkin
PC3200. OTOH the PNY modules I've tried (Bravo chips) all worked fine
at 1T and 3-3-3-8, even though the SPD specified 3-4-4-8.