Dripping Reluctance

Dripping Reluctance

In the past, surveys we’ve taken of this audience showed a rather stunning lack of enthusiasm for new generation platforms, whether from AMD or Intel.

This survey is essentially no different. If anybody thinks most of this audience is straining at the bit to buy a Hammer or Prescott system soon, they are sadly mistaken

Interest has perked up a bit since our last surveys with the imminent appearance of socket 939, but few are enthusiastic. The mood even among those who think they’ll end up buying one can best be described as grudging.

It’s also pretty clear that adoption of Hammer is going to be a piecemeal operation, again, even among those ready to buy, and the decision on exactly when to buy is more personal-circumstances than technology-driven. Indeed, you get the feeling from the comments that many of those willing to upgrade now are following more of a Joe Sixpack-like “X years, then out” pattern than “Gotta have it.”

There’s this general doubt that anything on the CPU side likely to come down the pike the next couple years isn’t going to be a “Gotta have it” and that when circumstances nudge someone into buying; the upgrade cost is going to hurt more than before, and people don’t like that.

A bit contrarily, people don’t seem to like the stutter-step, one new feature at a time: PCI Express: DDR2: dual-core, even though that will save people money, primarily because it leaves anyone buying X at Y time feeled obsolesced when Z comes along a few months later.

For those who have given this more than a bit of thought, it seems like there’s some resentment that there aren’t clear signals as to when would be a good time to buy without having to make tradeoffs and doing without the next shiny toy coming down the pike. .

How much of a shift in attitude has there been in the last six months? In the past, the predominant theme has been that any serious thinking was moot due to high prices. Now, price is much less of a factor, and people have informed themselves of their options over the next year, and they’re not too enthused about what they see.

The only areas where one can detect much enthusiasm are for those new components that will deliver a big jump in performance. That has been and continues to be video cards. It’s not that a lot of people want to spend $400 for a video card, but those who are do show some signs of life.

Of the course of the next few articles, we’ll post many comments you’ve given, but to start, we’ll post one long but insightful letter that captures the general mood of the audience pretty well.

A Letter To The Editor

Well, that is quite of can of worms you’re opening but here goes.

As far as buying anything right now — not a chance. I’m an equal opportunity purchaser and do not stick to one brand over another. I buy what makes sense at the time I need to purchase (which incidentally I really need to do).

I’m an independent contractor so when I buy I buy parts for several systems at a time and spend a couple of weeks building them all. Not really too much of a hardcore gamer but do enjoy a good ole romp through 1940’s Europe.

I wouldn’t touch AMD right now with a ten foot pole — they have a decision making disorder that won’t be resolved anytime soon. I would consider a 939/64 combo but not without PCIe and DDR2 — by the time STABLE equipment comes in that configuration we’ll be into next year and onto the multi core business.

Multicore is a whole different story, I mean we have multicores now and have had them for a while — its called dual processors. I run a dual PIII [email protected]. now as my main machine. There is no way Intel or AMD is gonna be able to do multicore/dual processor for the same price point as we have now.

I think the industry is probably going to split into home machines, enthusiast, and workstations. price points of 600-1000, 1000-2000, and 2000 ->. The enthusiast market will be the competitive one, but can you do Multicore for $1000? Not without some compromises, and what will those compromises be and will they be worth it?

Just look at Xeons’ and Opterons’ and you see where the price lays now for that stuff. It will be interesting to see how these companies introduce multicores and differentiate them between server platforms. (it doesn’t take that much to run a file/web server)

Not to mention the software changes needed to support 64 bit and multicores. How many companies are going to invest the money to redesign their software to multithread? How much of this software wouldn’t take advantage of it if it were anyway. AKA: software price increase.

I think you will see a few software companies force users to move to the latest 64 bit version in order to keep the software (no patch to allow current software to work on 64 bit longhorn)(something is always broken when we make the bit move)(I know it’s supposed to be backwards compatible) — great way to get rid of software pirates.

Anyway — AMD’s decision making disorder will probably keep me out of there stable for the time being. AMD’s “surprise” announcements only hurt the company and its sales.

Let’s look at Intel, since I’m finding this soapbox a bit comfortable.

I have to admit upfront that the last two times I upgraded it was with Intel — not married to them by any means but it was a matter of timing, price and convenience.

But I won’t consider Intel at this point either. The next platform I want needs to have 64 bit, DDR2, PCIe, etc…. What makes me nuts about Intel is how they constantly have features built in to the designs and just sort of enable them at the appropriate moment. I always feel ripped off when that happens. And I think the 1066 FSB and of course 64 bit( well sort of) is the case right now. when Intel goes 1066 and the 64 bit stuff gets sorted out I will probably go that route. I don’t really want to, but I trust that that mobo makers will continue to make 775 socket boards for a long time to come, and you can always find old Intel processors around.

I’m actually thinking of just going Northwood, AGP, etc., because its cheap does everything I need it too right now and stretch it for the next 2 years while this whole mess sorts itself out.
Ah hell, maybe I’ll just stretch my Dual PIII a little longer. It’s old but it works and manhandles Photoshop, Dreamweaver and Flash like a pro.

I would rather see the focus shift to the software makers to tighten up bloated, buggy, bastardized, software and in turn make it smaller and more efficient. I believe the future of computers is in features and not more speed. I think you’ll see high end gamers turning to to multicores and dual processors to get more performance but the working man would rather have features. I think this will lead to more specific computers for different applications — we have some of that now but I think it will only grow.

To give an example, I’d personally would like to see a computer that would do double duty has a audio pre-pro with all the inputs and outputs. Talk about upgradeable to the latest sound decoders and what a PIP display! Now that I would pay for. I think the HTPC realm isn’t being done correctly yet. I just haven’t seen anything that can’t be done by a 200 Tivo and not crash quite so often.

Much more tomorrwo.

Ed