"You may be right about intel, my area is software. We have slowed down considerably since the bubble, the companies have downsized and new products are minimal."
Fair enough. Thank you for acknowledging that you area is software, not chips. I hope you will think deeply about this, and stop making simply incorrect claims such as you have been making recently:
" that was different hueyone- in Andy Groves time we had no global competition"
A claim that is silly. Intel faced _substantial_ global competition during Grove's tenure, more so than it does now, actually. Do I need to list the companies in Japan, Korea, and Europe that Intel was competing with?)
And your claims, made here in connection with Intel, that hiring in India is what tech companies should do:
"I am not a big fan of outsourcing but when you look into it really, most of the issues are communicative due to management being in the US with core R&D in india. If there is a way to deal with the IP issues (a big issue I know) then I would move the entire division including VP to Bangalore. I have worked with lots of indian development talent over the years and there is NO DIFFERENCE in terms of quality vs. what you can get here. The one big difference is the creative environment, the marketing side and visionary climate. That stuff is much better in silicon valley- but thats the beta or prototype. But large engineering depts in big stable US companies to me sound like a luxury for those firms who have patents or some other profit protector."
While this may work for certain kinds of software companies, the issues I brought up in my post about 90 nm, 300 mm, SOI, automation, etc., depend critically on a close connection with actual factories. It isn't going to work to have some Indian software guy in Bangalore working on the 90 nm shrink. It may someday, but not now. There are currently (insofar as I know) no reasonably modern wafer fabs in all of India.
(Personally, and this is a reach I admit, I think India's nearly single-minded focus on becoming the software capital of the world is headed for problems. The kind of software they are tending to concentrate on, software without much other technology and without a local base of entrepreneurs innovating new products, is exactly the kind of software development which is most likely to be automated or make redundant by powerful applications (the way Excel and Quicken took away a lot of the need for custom accounting packages) and frameworks. Sort of like a country deciding in 1975 that it "intends to become the keypunch entry capital of the world." And, in fact, routine software generation is in many ways following the trends we saw in the 1960s: "At the Control Data Institute, we will train you for an exciting and profitable career in data entry..." India needs to diversify, to have a richer ecology of software and hardware. China is, I think, doing things more cleverly.)
"There is one thing supporting my position that innovation has slowed though- the fact that I as a consumer don't need to upgrade to a new PC every 2 years like I used to. That statistic speaks volumes from my pov. It could be that intel is doing just as much as they did then but it is more under the covers, so to speak. But I know that enterprise software is doing half or less than what we did in say, 1998, and I haven't upgraded my PC at work for 3 years. "
Yes, we have discussed this here several times. Amy J. and I have had a couple of exchanges on just this topic.
This actually makes my point, I believe. Is the lack of "killer apps" (which is what this issue really is) caused by not enough jobs being in India? Of course not.
The big upgrade cycles for PC users has, experts tell us and I believe them, caused by a) new Windows releases needing new hardware, and b) new apps which people want to run.
Examples abound, and our earlier discussions on this covered many of them: VisiCalc in 1980, Lotus 1-2-3 in 1982-3, desktop publishing/PageMaker on the Mac in 1986-7, Excel on the Mac in 1989, Windows 3.0/3.1 on the PC in 1990-91, Windows 95, browsers and the Web in 1995-6, Windows 98, desktop video on Macs and PCs, and games on various platforms.
Many users are just like you (and me, and most I know): the 500-1000 MHz machines bought a few years ago are perfectly fine for surfing the Web (where the limitations are on the bandwidth), preparing business presentations, writing documents, casually playing games or running various apps. Very few of us need Radeon 9800/2.8 GHz P4s.
(Myself, I only a month ago upgraded my 1999-purchased 400 MHz G4 Macintosh to a new iMac with 17-inch 1440x900 LCD and DVD-R Superdrive. My older machine handled my Web needs quite well, and would have even had I had a faster Net connection. I only got the newer machine because I have been upgrading my desktop pretty regularly every 3-4 for the past 25 years, with laptops bought at about the same intervals but bought in-between, and only for the past 12 years.)
I expect a lot of businesses will be replacing the machines they bought during the Bubble (and for Y2K compliance) in the 1998-99 period. There was a _lot_ of inventory bought during those days: servers, routers, desktops. A lot of this inventory has been written off, even crushed in landfills. The desktops mostly have CRTs and not LCDs, mostly processors in the 300 - 1300 MHz range.
What will be a new "killer app" that businesses figure they need to upgrade employee computers to exploit? (I'm talking about the bulk of PCs on business desktops, in Boise and Schenectady and Peoria, not the CAD systems and software development workstations that engineers and designers have.)
Desktop video? Unlikely, as most employees obviously will not be making home movies.
Video conferencing? Not on the bulk of desktops, for various reasons. It is not compelling to give J. Random Cubedweller the ability to have video conferences with other cuboids.
"B-2-B"? (E.g., buying and selling over the Net.) Sure, businesses are doing this, but this is a perfect example of where CPU power is NOT needed on the desktop.
Games? Not at businesses, though home sales will perhaps spike as PC game machines come down in price.
Lower energy consumption machines? (Don't laugh...some experts are saying companies will replace their desktops with more energy-efficient models. Doing the ROI, I don't see this happening on a wide scale, at least not for this reason alone.)
Wireless? Yep, this is a hot(spot) area. Been growing for several years (Metricom modems abounded at meetings I was involved in a few years ago). My own home has 802.11b/WiFi/AirPort connections between my iMac, my Titanium Powerbook, and my older iBook. (And schools are adopting wireless networks and laptops widely. A friend of mine, a Windows fan, has a son in a school where every student gets an Apple iBook and these are wirelessly linked to the teacher and school and other kids.
This is why Intel is pushing Centrino so hard. It's likely that most of the "desktop replacement" will be with laptops the employees can carry from spot to spot, creating temporary workspaces (a la Sun).
There are other major trends coming, but I think wireless laptops and PDAs will be the big one for the next upgrade cycle. Which means Bluetooth, 802.11g, the "last mile problem" (RF, optical), and so on.
Many pundits have called 2003 "The Year of the Laptop."
I agree.
--Tim May |
