Zzzzzzzz . . . .
2003 is shaping up to be a snoozer. Just not a whole lot dramatic and exciting is going to happen for most of the next year. There will be flurries of activity here and there, but on the whole, it’s going to be dull.
It’s going to be dull because we’re on the backend of the .13 micron cycle. We’re not going to see dramatic jumps in CPU performance until we get to .09 micron processors, and that won’t take place for another year.
You know it’s going to be a slow year when most of the two biggest computer stories aren’t really computer stories.
AMD: Will They Make It?
This story could be subtitled, “Will the math match the mouth?” That’s the problem with AMD right now: it doesn’t.
This story is also an all-or-nothing proposition. It will either suddenly fizzle out if AMD posts decent results, or it will just get bigger and bigger as the year progresses.
AMD doesn’t have to do great in 2003; it just has to stop doing terribly. If they can just get back to lousy (say, lose just $100 million a quarter from operating businesses), they should be OK.
The only way they can realistically do that is to sell at least six million CPUs a quarter at a reasonably good price (say about $90 ASP). If they do that (or get close), I stop writing these “will they die” stories.
These stories continue to be written not because I don’t think they can do it but because they don’t think they can do it based on their sales projections.
Digital Rights Management Comes To Congress
2003 will be the first real year of the DVD-recorder, sales will skyrocket during the year and next Christmas. This will get Hollywood to join in the digital protection chorus big time.
Again, this story has an uncertain fate. The upcoming war with Iraq could hurt, or greatly help the passage of legislation. Remember, this is special interest legislation, and a lot of things slip through the cracks when the average Congressional mind is focused on something else.
Any digital rights legislation will almost certainly include an endorsement of the Palladium/LaGrande approach, but there are other possibilities that bear watching.
For just one possible example, Hollywood may well play cat-and-mouse with DVDs. They might start burning two-layer DVDs and try to legislatively prevent two-layer burners from being sold in the U.S..
What will be far more interesting in this neck of the woods is not the technical but rather the psychological reaction to the likely results on this community.
We have long warned that this community needs to get reasonable and get organized politically. Others have warned far longer (at least for the latter). For all intents and purposes, this has been completely ignored.
What happens if the expected happens, and the cause for free content crashes and burns in Washington? Will people wake up after pissing on the electric fence? Will they realize that they live in a real world, understand that real-world laws apply to them, too, realize that everything can’t be free, and get politically involved to defend reasonable interests.
Or will they not learn from this, conclude that society is hopeless, and veer off more and more away from society and towards criminality, getting smaller and smaller and eventually marginalizing themselves?
I would bet the latter.
Prices, Prices
A big reason why matters are so slack right now is not because interesting technology isn’t out there, but because it costs too much.
AMD’s apparent inability to make CPUs, and financial need to get as much money as they can for them, will probably keep CPU prices relatively high for some time to come.
Obviously if AMD goes down the tubes, this will remove the major impediment to higher pricing, and while the impact should not be exaggerated, Intel would be in for a few bumper years of profits until Via could present a, well via-ble challenge. 🙂
If AMD recovers, though, one can at least hope for price slippage in the latter half of 2003 and expect it in 2004.
Keep in mind that the rest of the world isn’t in a computer slump. It is only in the United States where computer sales have dropped a lot. Some of it is due to recession, and there will likely be some uptake on the business end, but it’s not going to be a rip-roaring recovery. The U.S. has reached the “replace-the-old-one” state.
However, both AMD and Intel are building production like happy days will soon be here again. If AMD finds itself in 2004 capable of pumping out eight, ten or even more million processors a quarter, we are likely to have a glut. If they can’t, unless they can get skyhigh pricing, we are unlikely to have AMD.
Flurries of the Year
Northwood/Dual DDR For the next few months, this is where the action is likely to be. However, this is more likely to be a continuous dribble than a flood.
How long the dribble lasts will depend largely on how much Intel can improve Northwood over the next few months. The more they can improve it, the longer the dribble.
Latecomers may well give a boost to the stream when Canterwood/low-speed HT chips shows up, partly because of their own merits. This may get boosted a bit if Clawhammer systems don’t present an attractive price/performance package.
TBred/Barton The 2400+ hasn’t had much traction at all in the overclocking forums so far. This is mostly due to the price. There’s plenty of interest in TBredBs; the real question is at what price?
It seems like the average prospective TBredB buyer would be delighted to pay $70 for an 1800+ TBredB upgrade, but not so interested in paying $150 for one.
Those people may have to wait a good long while for one. AMD is still reporting pretty high inventory levels (they only dropped from about $225 to $170 million during the third quarter). Those who have been buying the 1700+ TBredAs from Newegg have been finding that they’ve been getting processors made in August.
Put those two facts together, and it’s likely AMD has replaced a big backlog of Palominos with a smaller but still very sizable backlog of TBredAs.
Eventually, there will be a TBredB surge, but it’s likely to come later rather than sooner.
Per Barton, again we have the price barrier to contend with. People want Bartons, but they’re not going to pay $400 for them. {mospagebreak}
nVidia vs. ATI It will take a while to get started, but spring looks to the start of a truly competitive video card race.
The top of the line nv30 looks better than 9700 Pro, but not so much better that the medium-priced nv30 will blow away the 9700 Pro.
ATI will no doubt have an improved top-end video card, and that will also look pretty competitive against nVidia’s best.
Next spring/summer will probably be a good to upgrade an old video card.
Serial ATA Sneaks In Right now, it’s a curiosity and a bit of a kludge, but serial ATA will become an integrated part of motherboards during the year by the middle of the year.
Serial ATA drives will probably need some help to get going, and my guess is that we’ll get to see “designed for serial ATA drives from the ground up” drives next fall. This would also be an opportune marketing time to also introduce 10,000 RPM IDE drives.
DVD-Burners Don’t Sneak In This will be the cutting rather than bleeding-edge product of 2003. It won’t become standard until 2004, but it will get close by the end of the year.
The same will apply to HDTVs. That will probably become the big Christmas sales item, period, in the U.S. next year, with prices within shouting range of “normal” TVs.
The combo of the two will no doubt cause some very interesting nightmares in Hollywood heads.
LCDs Become Mainstream By this time next year, buying a CRT rather than an LCD will just be a matter of being cheap. Up to 19-inches, prices should decline to pretty reasonable levels. This could well drop the price of CRTs to more than reasonable levels during the course of the year.
DDR-II Cometh Motherboard generations usually last a few years. There was the Pentium board generation. Then there was the PII board generation, which we are now just slipping out of. The next generation boards will be son-of-PCI, serial ATA, dual-channel DDR-II motherboards, and the last to arrive will be DDR-II.
We see it a year from now, but it’s really meant to power the .09 and .065 micron generations of CPUs. Its future is 2004 and beyond.
So does .09 Intel has now pushed Prescott back to the end of the year. I suspect Intel will push it up a bit to more like Labor Day, but Prescott will be too pricey for most until 2004.
Prescott will be the next “fun” time for overclockers, “fun” meaning easy overclocks into virgin territory.
You Haven’t Mentioned Clawhammer!!!
You’re right. That’s because trying to figure out what AMD is going to do with it would leave anyone dazed and confused.
Intramural Fratricide
Next spring, it’s likely that something is going to come out. Just what and for how much is anybody’s guess.
It’s probably safe to say we’ll see an expensive Clawhammer with 1Mb cache on sale for the workstation bunch. It’s probably safe to say we’ll see a bit less expensive dual-processor Clawhammer come out for the same group.
Beyond that, who the hell knows?
Right now, it looks like we’ll see the spectacle of AMD putting out two lines of top-end processors. In one corner, Barton. In the other, Athlon 64. At this point, it doesn’t look like one will be much better than the other.
How is AMD going to manage this two-headed beast?
There’s a number of things AMD could do, and odds are, AMD will try all of them. 🙂 You don’t try to predict what AMD is going to do when the one thing you feel surest about is that they don’t know either and plan to play this by ear.
It’s probably safe to say that sticking with socket A now, and waiting for a .09 Clawhammer II in 2004 to shift platforms would be the prudent thing to do, and that’s about it.
Change of Audience
Most of the computer hardware sites can trace their beginnings back to the Celeron surge of 1998 and its aftermath. Those relative few that were around at that time got popular, and many more got started (including this one) as a result of it.
Four years is a long time to be devoted to a hobby, and it seems like a lot of those who were around back then are having, if not burn-out, then burn-down.
People find themselves satisfied with what they have; or find that speed isn’t everything, or find that as they get older, they find more interesting things to do, or find there are more important things they have to do.
Some no doubt have been alienated by the increasing commercialism of the hardware websites.
They don’t go back to calculators or iMacs; they just are less intense about it. So they withdraw somewhat.
Combine that with much lower pricing, and the demographics of the active audience (which was never old to begin with) becomes even younger. It sure at least seems that way. That doesn’t necessarily mean the audience has actually gotten younger, it just seems like it.
What does this mean? What ramifications does it have? You can be sure that it will have some, and given the relative lack of advancement in the field, that the proportion of the audience for whom everything is new will grow during 2003, and websites will have to adapt to that.
Be the first to comment