A Few More Pebbles

From Goliath:

“After the posting of our first review last week there was a minor VIA-friendly website that did not want to believe the truth, foolhardy attacking us as well as Anand, who had come to a similar conclusion.”

If this were so foolhardy, then why did you end up retesting?

We aren’t a Via-friendly website, by the way; we’ve had too many quirky boards over the last year to be blindly in love or even like with their boards, and we’ve said so.

We are a very truth-friendly website, though, and simply like to consider all the factors, and whether they are significant or not.

If this were a baseball game, we’re in the bottom of the first, AMD scored a run at the top of the first, and is now facing
what looks to be a weak-hitting batter. Even if he strikes out, there’s still the rest of the lineup, and AMD is saying they’re leaving after five innings. Nobody should be calling winners at this point, at most, they should say AMD’s a little ahead.

“A new BIOS was supposed to change everything and KT266 would all over a sudden perform really well.”

We said no such thing. This is what we said:

“That along with other BIOS improvements look like it brings the Via board up to rough equivalence with the AMD 760 boards.”

Nonetheless, Goliath did half the right thing. Goliath retested the board using BIOS 1.10, but not BIOS 1.09, which is supposed to yield better scores.

Even using BIOS version 1.10, though, on the whole, Goliath found out that new BIOS closes about half the (not too big to start off with) gap between the Via and AMD chipset. This brings the Via board to within what we would consider rough equivalence with the AMD board. Goliath appears to consider minor differences to be significant; David does not.

David strongly suggests Goliath at least attempt to run at least a sampling of his benchmarks using BIOS version 1.09, which seems to be the highest performing BIOS so far.

Update 4/17/01: BIOS 1.09 appears to have been disawowed by MSI as being too aggressive; version 1.10 is now considered the “right” version. However, if R127 is supposed to be the “right” place for the resistor, the motherboard used by Tom’s Hardware is in the “wrong” place. Should MSI confirm this, the testing should be redone. Not for the sake of the
benchmarks (they should drop a bit across the board and increase the AMD lead), but for comparing relative stability.

For some reason, Goliath doesn’t talk much about stability at all in these reviews. David considers that to be as important if not more important than a couple percentage points in a benchmark. People should know how well the two boards compare.

Our suspicion is the MSI board may well strike out, not because it may do slightly less well than an AMD 760, but because it’s not terribly stable when pushed with aggressive settings. This would strike us as being a far better reason to say “No” to that particular board than a couple percentage point difference in performance, and certainly something to extensively explore when testing future Via DDR boards.

What Most Bothers Us About Goliath’s Sayings

It would not shock us if Goliath, after exhaustive testing, proved to the satisfaction of all that the AMD boards did do a bit better than the MSI Via board, which appears at the moment to be a weak contender.

However, we don’t understand why Goliath feels the need to declare an overall winner when the heavy hitters haven’t shown up yet. The silence is deafening from Asus and Abit. Perhaps they just didn’t rush the job. Perhaps they’re having problems and want Via to tweak the silicon first. If you wish to call the MSI Via board a loser, you may well be right, but why call Via a loser so fast?

The AMD 760/761 boards, while not especially overclocking friendly, seem to be solid boards. What makes us additionally reticent about recommending them is just how long they’re going to be around, something Goliath glosses over. AMD has done everything except scream, “WE DON’T WANT TO DO THIS!!!” It is an unfortunate truth that BIOS support tends to go to the current boards, not last month’s or quarter’s model. This is a major concern for prospective buyers, and a legitimate one.

Email Ed

Be the first to comment

Leave a Reply