THG Column: The End Is Near

<b><A HREF="" target="_new">The End Is Near - A THG Discussion Forum</A></b>

As the Microprocessor Forum kicks off this week, you have to wonder what the hell is going to happen to the chip business in 2003. Does AMD have the cash to make it through the ever-delayed Hammer? Is Intel going to have to adopt a scorch earth policy and leave everyone for dead just to keep it's 800 pound Gorilla status?

So, why don't you give your ideas about what is going to set the tech world on fire in the coming years. What will keep you enthusiastic about the PC? We'll keep an eye on it, and discuss some of the comments in the next couple of days.
16 answers Last reply
More about column near
  1. Well, I guess what's happening with AMD these days is also happening with some of the Giants in different industries all over the world. Take FIAT for instance, experts expect it to sell out to GM to keep the company afloat. Reason: The ABSENSE of TOP selling models or NEW innovative models in the past few years.
    AMD, on the other hand, has been sleeping on the ATHLON for quiet a while now. It should have charged quiet a while ago with its new line. Intel, in the past couple of months has made quiet a few leaps, DDR333,533FSB,PC1066,512Kcache,Hyperthreading and so on. YOU see, even though most of these steps INTEL took may not be exactly NEW, or ground breaking and dont reflect real increase in performance over comparative AMD processors...BUT they make news..and provide a refreshment in the companies yield of processors.
    Globaly, day after day, the industry is getting less and less forgiving, and making it harder to survive, especially for those sitting around, waiting for a miracle to happen.
    AMD has been sitting around like a DUCK...for quiet a while now.
  2. I honestly am somewhat sick and tired of these rants that Omid makes. Realistically, there's no real conclusion anyone can come to before the products are out. Save for the great hearsay about technical data sheets, but that's never good enough. Look at the example of the first Pentium 4, the Wilamette. It ran hot, was slow compared to the Pentium 3 and Athlon, and was expensive. It took some time, but Intel bounced back with the Northwood. Who knows? Maybe smaller companies like AMD will need another break as they first recieved with the Athlon.

    Is Intel going to have to adopt a scorch earth policy and leave everyone for dead just to keep it's 800 pound Gorilla status?

    That is <i>highly</i> unlikely for several reasons. Not only would killing all competition hurt the tech industry in general, but Intel would most likely be forced into a government seperation to prevent monopolization of the microprocessor market. Intel can't afford something of that caliber, and I think both companies know a product of that level would not be released.

    I myself am more interested in the most immediate products, that will ultimately affect the end users and enthuthusiasts. My biggest concerns now are the production and release of the Prescott Pentium 4, and the AMD Hammer. With the introduction of more cost efficient memory interfaces and the eventual dropping of the erratic company Rambus, I believe there is a significant move towards a faster and greater level of technology development in microprocessors.

    Soon enough, Intel will make the i845s...imagine dual channel Sdram...*shudder*
  3. CPU manufacturers are not listening to what the general public want. They don’t need the fastest processors for surfing the web and simple office tasks. The upgrade cycle for corporate PCs has increased from 2 to 4 years showing that they do not need the latest and greatest CPUs. What is needed is a new killer-app for all this unharnessed power but what we will probably see is more marketing along the line of Intel’s Internet Acceleration adverts that obviously are meant to mislead.

    What will make me upgrade after almost 3 years of not having touched my PCs core configuration is size and noise. Sure I’ll be getting a relatively fast CPU but I will not be shelling out for the most expensive as I have often done in the past. I want to be able to use my PC for listening to music and watching movies without a hum from the PC in the background. The amount of wasted space in a conventional case is also crazy but there are not currently enough half-height PCI and AGP cards around to make some of the smaller form factors a viable alternative to a conventional PC.

    "Simplicity is the ultimate sophistication."
    — Leonardo DaVinci
  4. I most forcefully agree. I personally find it cool to have a processor under the hood, which beats the crap out of the competition, but I also find the trend disturbing.

    In some places we do not need more power, but less price, less heat production and less power consumption. Laptops, and home computers.

    In laptops, the thing is pretty clear. Who wants to sit in the airport, with a thing in your lap, that produces enough heat to make your creases disappear.

    I also would like it if computers could again be what they were only a few years ago. Silent.
    Silence can only happen by 1# diminishing the heat generated by computer components and 2# diminishing the amount of electricity needed to run a computer thus eliminating the necessity to cool your powersource.

    I am currently working on the project of making a 1Ghz+ system work without a single fan (without water cooling), and although my experiments show promise, It should not have to be so darn hard.


    "Anyone who believes in psychokinetics, please raise my hand."
  5. Is AMD going to stay afloat another year? Is Intel's predominancy going to be finally broken? Who's going to break the 4 GHz barrier first?

    Answer: Who cares? Why care? The real issue here isn't speed or a couple of instruction sets more or less, the eternally unaddressed issue is system integration. Why have a 3 GHz CPU with a 133 MHz memory bus? Why have ATA-133 and serial ATA, when the PCI bus can't support it? What do we need 2 GHz internet PCs for, when we use 56k modems? Why buy the new 64-bit Hammer, when the OS won't support it ?(and don't anybody dare mouth the word "64bit WinXP"). And while we're here, let's take it a level further! In the final analysis, what's with this x86 architecture?

    What we're now using, is a pumped up, steroid stuffed version of the 8086, the 286, the 386 (after which, absolutely no progress architecture-wise has been made). Cool. All this blabber is about this: The Gods have failed us. Intel has failed to deliver. The x86 design was outdated, flaky and feature-poor (quite frankly it was the laughing stock of the market back then) when it was first introduced and since then there's been hardly any improvement worth mentioning. Microsoft... well, if I was Bill Gates I would be so embarassed being me, I'd change my profession to organic farming or something. The whole world (including yours truly) is using the worst operating system: monolithic, no breakthroughs since introduced, still can't shake the ghosts of 16 bits (and they have the nerve to talk about 64 bit -what are they going to do, give backwards compatibility for 32 and 16 bit?), and in the middle of this palaiolithic soup.... the registry. 30-something gigabytes of text, ripe for the taking.

    The Gods have failed us indeed. And from what little I've read, when Gods fail people, new Gods are created. The buying public would have the power to forcively request such a change, but poor information distorted by marketing strategies ensure they don't. People are spending their dosh for 'new' hardware anyway. Innovation is sacrifised on the altar of 'backwards compatibility'. If only everybody could understand that this 'backwards compatibility' serves no other purpose than to help clear storehouses of excess old stock. I mean, really now, what good does it do to you if you can, say, carry a memory module a hard drive and a floppy drive over to your brand new PC? Save you $100? When you 've just spend $1500?

    The marketing hype of the supergiants has got the buying public in chains. It's not just that they sell us coal for gold. It's not even that we're actually buying it. What's really disturbing is that, ultimately, for most people the scam passes unnoticed.

    The responsibility of the wizz-kids and the geeks and THG community is not to boo and jeer Intel in favour of AMD (remember that AMD is also producing x86 and it entered Intel's arena, not the other way around). Our responsibility is not to ask for less bad, our responsibility is to ask for outright good. Let people know what they could have and who's keeping them from having it.
  6. hmmm .... no ... the international system is at a recession.
    the system goes up then it goes down then it goes up then down and repeats process.

    However during such a down time in the world i do beleive thats when war most likely to occure. Especially with these terrorist. sooner or later WW3 will start as frustrations linger.

    Life is irrelivent and irrational.

    <A HREF="" target="_new"> My Rig </A><P ID="edit"><FONT SIZE=-1><EM>Edited by xxsk8er101xx on 10/14/02 12:28 PM.</EM></FONT></P>
  7. I hear every geek I know saying the same thing, myself included: smaller, cooler, cheaper, we have enough CPU speed. But yet, I don't see any of us running VIA C3 CPUs at home. When I went to build a small, quiet machine, I used an Intel CPU for their lower heat dissipation because, although the C3 900 *should* be powerful enough for anything I want to do at home (watch DVDs, read email, play MP3s, the occasional game) without a fan, the simple fact is that it is not.

    The folks at Microsoft always expect you to have a fast CPU to run their new stuff, and XP is simply unusable on older or slower hardware. Ditto Office. Ditto Visual Studio (.NET or 6). Do I blame Microsoft? Hardly. They are responding to market pressure and the fact that the CPUs are there. Face it, guys. We are NOT the target market. The target is the IT manager who will buy 100+ machines, or Dell and Gateway who will sell to hundreds of thousands. Those of us who build our own machines and actually understand what we are doing are a niche market, and probably always will be.

    The price of going Wintel is hardware efficency, but the benefit is productivity. This is only possible with a glut of horsepower. If I really cared about getting the most from my hardware, I'd run Linux or BSD. If our community of users were big enough to matter, there would be applications (games, _good_ office suites, DRIVERS!!, etc.) written for those OSes, but there aren't, and the trend is getting stronger, not weaker. Look at the future plans for Longhorn.

    If we want to effect the marketplace, we need to think about what our identity as a user group is and define ourselves as a target. The graphics guys (ATI and NVidia) have figured us out. The CPU guys have not.


    Oh, and as a side note, if you want to rant at Intel for something, rant at them for the loss of the next generation of the Alpha. That is a sin.
  8. <QUOTE>I hear every geek I know saying the same thing, myself included: smaller, cooler, cheaper, we have enough CPU speed. But yet, I don't see any of us running VIA C3 CPUs at home. When I went to build a small, quiet machine, I used an Intel CPU for their lower heat dissipation because, although the C3 900 *should* be powerful enough for anything I want to do at home (watch DVDs, read email, play MP3s, the occasional game) without a fan, the simple fact is that it is not. </QUOTE>

    Yea, and nay. I'd like to take a swipe at a VIA C3, because I know some applications, where it would fit. Problem is, I'd have to buy it online... Which, call me a skeptic, is something I would not like to do. A quick run-through some 25 hardware retailers in my area (northern europe), reveals the slowest (and cheapest) CPU to be a Duron 900 (200FSB) for some 40$... I can not find a single retailer in the whole country who carries the VIA C3-CPU !

    What you say is true, about the market going after those IT-exec's who buy 100+ machines at a time. But there is another market as well. Last week I did some consulting for the local public libraries. Their current systems for user terminals is about as old as the average two-handed sword. They'd like to update their systems, some 900 user terminals (used mainly for DB access), some 300 multifunction machines (used for internet etc.).

    They did some test with current systems to compare them to the old ones, and came up with a few problems:
    1# The old machines have worked nearly without interruption (24/7) for some 9 years now. Show me the Athlon/P4 system, which can guarantee it's fans to have a lifespan of above 70 000 hours.
    2# Sound radiation increased too much, a library is supposed to be quiet.
    3# Power consumption increased 4-fold.

    So, they decided to continue for another few years with their old systems...

    Mr. Jeeves

    "Anyone who believes in psychokinetics, please raise my hand."
  9. Here's something else to talk about. What's up with Duron? It's been stuck at 1.3 ghz for a while now. I think it's pretty lame not to move the duron to the new core. Wouldn't it cut costs to eliminate that core? theres a big lazy boy recliner sitting on top of the celeron 1.8 why? no competition. AMD is procrastinating! They need to move it or loose it. If they upgrade the duron and call it the duron xp or something intel will have to concentrate on that. I'm a performance guy weather it's my pc or my turbocharged civic with a 2.0 crv block mated to a vtec head i wan't the best. Just like when 3dfx went out of buisiness i went out and bought a gforce2 card. And as far as the video in intel vs. amd cooling, there is just no way that the pentuim 4 can be under 100 degrees ferenheit, because I just built a 2.0 ghz p4(boxed set)and with the cpu fan installed it was hot to the touch. not to mention that it was 141 degees F.
    By the way my cpu fan went out on my xp 1700+ yesterday and the computer just rebooted. i hapend to have the case cover off. i checked the temp in the bios aqnd it was 229 degrees F. Water boils at 212. i replaced the fan and its back at normal temp. it survived just like my old k63 450 were the same thing happend. they probably survived because i don't use cheap motherboards. i use epox. but have had good success wit chaintec, msi, and gigabyte. Does anyone have anything to say to this? also with my experience with amd for example i bought my cpu online seperate and it runs at 141 degrees F. my buddy bought the boxed set same cpu and his runs at 120 F max. i use the antec fan which is exactly the same as the amd. i set them side by side and they even spin at the same rpm. so by a boxed set beause it seems you get the pick of the litter as far as cpu. if you dont buy a boxed set make sure you get a copper heatsink.
  10. If AMD goes down its. Government wont break intel up. Example "Microsoft".

    Instead we'll have this:
    Intel Vs. VIA Vs. IBM (PowerPC)

    i really want all the chips on the same platform like Socket 7 cuz it was all good. Plug my Pentium for a cyrix or K6.

    <A HREF="" target="_new"></A>
  11. For me it is and shall be the want to play games which makes me upgrade periodically to maintain acceptable performance on the newer ones.

    ID has an amazing power over me. If the latest wolfenstein/doom/quake doesn't work with my current setup, then I must change my current setup.
  12. AMD has suffered just as all tech stocks have suffered from the recession, but I personally believe that holding off on the Hammer won't hurt them. How many times has Intel rushed a product to the shelves to find out they made a huge mistake. Don't forget that AMD has always been under the shadow of Intel. They are used to fighting for what they have. Intel hasn't faired much better during the recession. Percentage wise, Intel's stock has probably lost as much as AMD.
    Intel has been "the name" in computers since Comodore, Tandy and Amiga. The older generation is going to trust a name they know, even if a lower priced AMD is a better deal. Intel is living off of their name, and Intel is also more diversified.
    AMD has to survive a few more years before it will develop a name outside of the gamer/computer enthusiast realm. Another key to the success of AMD will depend on better advertising. Last football season I was so sick of seeing Intel's Blue Men on the TV that I turned the channel. They are also trying to brain wash kids with Aliens using the Pentium 4 processor (Intel's next generation of buyers).
    AMD must advertise their products better. I can only remember one AMD commercial, and that was about two years ago. In addition, the commercial was extremely vague on what AMD was all about. I personally believe AMD needs to advertise themselves as less expensive option to Intel's P4. It may appeal to the crowd seeking computers this Christmas on a budget.
    I've rambled on enough. With all of this being said, one thing is a must... the next AMD processor must run cooler! I'm sick of sitting next to a box that sounds like a Nascar race!
  13. I totally agree, x86 hardware platform and MS OSes belong to the stone age. But ... the same things happen in every field of economy. For instance, every enviromentalist will tel how obsolete is car engine concept you can find in 99.999% of cars made these days. But people are used to it, industry has developed many other things related, and so many people earn more if cars burn more oil.
    It's the same in IT industry. For normal gamer or IT enthusiast it's no big deal switching from Amiga to PC and then to ... PPC with Linux maybe. If there only are enough of new software titles he uses. For big corporates things are much different. In a giant information sistems there may be tens of thousands years of work (or even more) needed in their creation and even if new technologies are 100 times beter and easier to program, that still means, upgrade means hundreds years of work to get to the same point. So managers point is clear: it's much easier and cheaper to buy an even more steroid pumped but still obsolete CPUs or other HW, than invest in a group of few hundred people to make transition in a year or two.
  14. My semi-random thoughts:

    * AMD is struggling. This is nothing new for them though. Hopefully they will either survive by holding on, or drop down for a while and then pick up again with a new killer processor. (Much like they did with the Athlon.)

    * Intel is trying so very hard not to crush AMD that it is laughable. Intel really doesn't want to be the only kid on the block still playing, but AMD is making it hard on them. I think Intel will do their best to keep AMD from dying because they need the competition to avoid looking like a monopoly.

    * The Hammers are looking less and less impressive every time I see new benchmarks leaked. Hopefully AMD manages to get them out before they screw things up entirely.

    * I still don't trust the future scalability of DDR SDRAM. It's just too noisy and has too many data paths.

    * I still am not very fond of RDRAM either. It has a bad latency and was designed for high bandwidth, which is something that processors just don't need to be <i>that</i> scalable. No P4 is going to need the bandwidth of dual-channel 64-bit PC1200 RDRAM.

    * Home Internet Appliances and other such niche markets are never going to 'save' the PC market. Hardly anyone uses them even though processors are already more than fast enough to make them. Heck, my 133MHz Pentium surfed the web beautifully and was whisper-silent.

    * What <b>will</b> save the PC market is DX9 games. No matter how many people hate to admit it, it is games that drive people to upgrade their PCs. After all, I can run Word, Excel, Internet Explorer, WinAmp, and even Visual Studio quite nicely on my Celeron 500MHz. There hasn't been a reason to ugprade to a new PC for apps like that in <b>years</b>.

    However, if people start releasing games that look oh so sweet but need an uber-processor and ultra-graphics card just to run, people <i>will</i> go out and get those things to play those games. There just has to reach a critical mass of games that <i>require</i> this before it will happen. Since most DX8 game developers write their games to run on DX7 systems (and sometimes even less) by disabling some of the effects and offering lower quality graphics, no one has actually needed to upgrade just to play a game.

    Yet develop just fifteen awesome looking games (that look like Doom3's graphics) that flat out <b>require</b> a DX9 card and a P4 2GHz/AXP2000+ or higher and you'll see a swarm of upgrades in just six months and in a year the PC industry will be booming again. It may be stupid, but that is what is <i>really</i> needed to save the PC industry from the funk it is currently in.

    * Do PCs need to be faster? No. Do we want them to be faster anyway? <b>HELL YES!!!</b> ;) We could all drive Yugos. Yet we still lust after Audi, BMW, Mercedes, etc. And we still purchase as nice of a car as we can possible afford. (And sometimes more than we can actually afford.) Why? Because human nature isn't to buy just what we need to get by. Human nature is to buy what we think will make us feel happy, proud, and cool. And darn it, we <i>deserve</i> to feel good about our purchases. :)

    <A HREF="" target="_new"><font color=red>Join</font color=red> <font color=blue>the</font color=blue> <font color=green>THGC</font color=green> <font color=orange>LAN</font color=orange> <font color=purple>Party</font color=purple>!</A>
  15. My first post doesn't appear to have turned up, pity, I don't think I want to spend an hour re-writing it. Crutch is that the hammer is late, it has had a long development cycle, like many non-existent 3D card companies did compared to Nvidia. Trying to extend product life and therefore minimise costs is very unpredictable and likely to lead to massive failures as has happened in the Video Game industry crashes and the PDA market (Pocket PC is catching up). In a perfect world you may be able to accuratelly predict when you should introduce a new product before customers become dienchanted with the company and the old one, but in reality trying to extend product life can lead to "too late" product replacement. Intel could pull ahead if they had secretly developed a superior core to the K7 and crushed them. AMD needed to introduce a replacement by the time the P4 could match the Mhz that the Athalon would max out on (given AMD a clearly marketable 25% performance lead).



    * Do PCs need to be faster? No. Do we want them to be faster anyway? HELL YES!!! ;) We could all drive Yugos. Yet we still lust after Audi, BMW, Mercedes, etc. And we still purchase as nice of a car as we can possible afford. (And sometimes more than we can actually afford.) Why? Because human nature isn't to buy just what we need to get by. Human nature is to buy what we think will make us feel happy, proud, and cool. And darn it, we deserve to feel good about our purchases. :)

    [End Quote]

    As you pointed out, 3D games require more power, as well as intelliegent agents with server like database engine searching will require more, HD and virtual reality. So a VIA C3 based system will suffice for most over things.

    This is where AMD and Intel are facing very strong competition. The Playstation 3 and its Cell processor design is offering a completely superior future compared to current micrprocessor road maps. By combining realistic massive parrallel processing abilities in what may turn out to be a cheap Linux Box, we may have the first true alternative to the PC for a while. This could pull AMD down, and Intel down a notch. The only thing that may save them and take Sony down is talk of a 2005 PS3 release while maintaining the aged PS2 against the X-box and X-box2.

    I had an idea for virtual reality environments based on an 3d Matrix of elements (unlike normal VR), and did a few draft calculations. From the calculations and Moore's Law, I estimated it would take at least around 30 years before the technology could produce a sizable VR environment at maxium resolution, but that a smaller 300dpi environment could be made in around 10 years. So we are going to more need computing power for some things for some time. Sobering to think that the first article I wrote was on a Commodore 64, so lets hope that we don't have to use Windows to render each VR pixel in future ;)

  16. AMD will hopefully survive or else Intel will be free to charge high prices and limit inovation.

    AMD was in bad shape when all they had to offer was the K6 line of processors that had a hard time keeping up with the Celeron of the time. Then came their savior of the time - the Athlon processor. They finally had a processor that was better than what Intel had to offer but THEY FAILED TO TAKE ADVANTAGE OF IT! They needed to have a higher selling price on their top speeds so that they could generate cash. then they needed to use some of this cash for marketing (i.e. tv commercials) so they could generate more sales. they also could have used some of this cash for r&d which would have helped them get hammer to market on time. Also some money for production capital would have been good... they are betting the farm on the Dresden plant. Where would they be if something happened to that plant. It also seems to me that the Dresden plant isnt able to ramp up clockspeed and production as well as the US plant was able to in the past.

    The other problem that AMD had was the back-stabbing activities of VIA. All of VIA's newest chipsets were first targeted for Intel processors and support for AMD followed a few months later. How did intel return the favor to VIA... no license for the P4 bus!!! Maybe if VIA would have released the AMD chipsets first then maybe AMD would have gained a larger market share (due to less value chipsets for the Intel market) and then maybe VIA wouldnt care so much that they didnt get a license to the P4 bus.

    I think the Athlon processor with the faster FSB, 512KB cache, and SSE-2 instruction support could give the P4 a run for its money. AMD has no license for SSE-2 until next year so that wont do them any good for the holiday buying season. It may be enough to let them hang in there for the next few months though. Their only hope to be around in 5+ years is the success of the SledgeHammer in corporate servers. This would let AMD get their foot in the door for desktops as well. Being the easier (not necessarily better) path to 64bit processing, they may just have a chance.
Ask a new question

Read More