Now Vs. Later Vs. Never

When we consider a product; we don’t have a choice of two, yeah or nay, thumbs up or thumbs down.

Rather, we have three choices, now, never, or later.

A good example of when we say “Now,” is the IWill KK266. We like this board very much, and think it’s an appropriate purchase NOW.

A good example of when we say “Never” would be cooling products we find inadequate to the task. A fan and heatsink have no chance of getting better sitting in your computer.

This leaves “Later.” If we think a product’s main fault is that it is too expensive, we say “Later.” Not “later” as in “yeah, right.” “Later” as in later. Not never, later.

Let me give you some examples of “later.”


The last time we took a long look at RDRAM, it cost way too much for what it did in the pre-Willy era. We said, “Later,” and we said what needed to happen for it to become viable: the price had to go down a lot, and a platform had to take proper advantage of it.

We look around now, and the price has gone down a lot, still more expensive than the competitors, but worth the extra cost if it provided significant advantages. There is now a platform available that takes advantage of RDRAM, the Tehama chipset.

Unfortunately for Intel, the Willamette chip lets RDRAM down. If Willamette were 20% better rather than 20% worse than an Athlon, we would not hesitate to recommend the more expensive RDRAM in appropriate cases.

It is because the current Willamette solution is worse than cheaper alternatives that we continue to say “Later” to RDRAM, not because “Rambus sucks forever.”


We have and continue to like DDR, but later, not now. Right now, it’s too expensive, but sooner rather than later, it won’t be.

We haven’t been shortsighted in saying “Don’t buy now.” We would have only been shortsighted saying, “Don’t buy ever.” It is those who say “Buy SDRAM Now, What’s DDR” then will say two months from now, “Forget SDRAM, Buy DDR” who are being short-sighted.


We said for quite a while that it was worth waiting for Palomino, another case of not buying now, but buying later, though our stance on this is weakening as we see current chips approaching 1.5Ghz and Palomino refuses to come into the barn.


We have a new generation of video cards. They are very expensive. They do have capacities which no doubt will have major gaming benefits, later. They show dramatic improvement over its nearest of kin in one area, which you may or may not use.

The vast majority of you don’t want to pay anything close to that price anyway. The vast majority of you will be very happy to get 80% of the performance for 25-33% of the price, later.

So we don’t say now, we don’t say never, we say later.

Time Is Not A Single Box; It’s A Matryoshka Doll

Matryoshkas are Russian dolls. They are a series of dolls that fit one right into another. See here for examples of what I’m talking about.

We’ve never been focused on just now. We always look down the road to see if better opportunities are likely to become available. But that doesn’t mean we’re thinking “outside the box,” maybe it looks that way to a person in a tiny box, but all we do is think inside a bigger box.

You get a different and better perspective when you’re in a bigger timebox than the NOW box. For one thing, the world won’t end now (and if it did, not even overnight is going to help you.)

In the computer hardware field, there is no point to buying something NOW unless it provides something you can really use NOW. If there isn’t anything particularly valuable for you to use NOW, and its value will only come to you LATER, then you should buy LATER, when it’s likely to be better and/or cheaper than it is NOW.

The only mistake you can make is to determine that if it isn’t good enough for you NOW, that it will NEVER be worth it. Later keeps your options open.

“It’s now or never” applies to love or filing deadlines, not computer equipment.:)

Email Ed

Be the first to comment

Leave a Reply