facebook rss twitter

Review: EPoX EP-8RDA3+ nForce2 Ultra 400

by Tarinder Sandhu on 20 May 2003, 00:00 4.5

Tags: EPoX

Quick Link: HEXUS.net/qarn

Add to My Vault: x

BIOS

EPoX retain the Award BIOS we saw on the 8RDA+. Let's take a look at the interesting sections now.

The advanced chipset features section contains almost everything you'll need to tweak the board. Voltages are in a separate section. System performance is either set to one of the four preset options, namely 'Optimal', 'Aggressive', 'Turbo', and 'Expert'. Each option sets the board up with certain parameters. Expert is where you'll be heading if you like a bit of tweaking. The FSB ceiling has been lifted to 250MHz, all in 1MHz increments. We'd like it if you could manually input your desired FSB, though.

The 8RDA3+ unlocks Athlon XP's multipliers well. We had options from 3x to 24x (5x - 24x were usable). Similarly, the memory frequency subsection has numerous options open to change. It's best to run memory synchronously to the FSB for high performance. The clock buffering that occurs during asynchronous settings can seriously hamper performance.

Memory timings have the same 4 presets as system performance. nForce2 boards appear to allow you to set memory latency parameters to unbelievable levels. tRAS is the product of tRCD+ tRP. In view of this, it cannot be set to anything below 4 clocks if you're running decent memory. There's little that you cannot toggle or change. The AGP frequency can be se to either 50MHz (why?) or 66-100MHz in 1MHz increments.

Lots of voltage is the order of the day. I've inputted the maximum on offer. The 2.2v is only recommended for short-term benchmarking and phase-change cooling. We don't think that a 0.13u CPU will take kindly to a prolonged bout of 2.2v voltage. AGP and DIMM voltages are both good. Vdd voltage sees a nice boost from previous models. The 8RDA3+ is specified to run at 200FSB with the nominal 1.6v Vdd voltage. Like most electronics, adding a little voltage will allow you to push your components that little bit further. 2.0v is more than ample adjustment. We'd have thought a fan would be necessary on the SPP when pushing the Vdd up by 20% above stock.

I've included this lengthy shot to highlight a couple of points. Firstly, the CPU temperature seemed a little low. In-socket thermistors are always prone to inaccurate readings. Secondly, the voltages under BIOS load are excellent, with very little fluctuation on the critical lines. Pretty good.