I'm long mostly Gilder picks, but I am hanging on to the CSCO I bought before I saw the light.Saw the "light," I get it.Having big old-tech positions is something that many investors must be facing. If years ago you bought into the MSFT, INTC, DELL, CSCO "microcosm" and let your positions run, then these old giants dominate your portfolio. The issue, now, is whether to reduce - or eliminate - the positions in search of more fertile ground, or even in favor of cash.It's a bit OT, but humor me for a moment while I worry about these old-tech positions.My current feeling on old tech:INTC, DELL - gradually eliminateMSFT - reduce and holdCSCO - hold or accumulatedjackn, I agree INTC has uncontested power in the PC platform business, and that the PC business is here to stay for a long while. INTC is the company in the old ecosystem that is facing the fewest direct threats from the new tech wave. All its important battles coninue to be fought in the PC arena, and here it knows how to play to win. (AMD, Transmeta - cheap PCs, low-power PCs, fast PCs - it's all PCs, and it's Intel's turf.)However, INTC depends on the PC ecosystem not just for revenues, but for growth and rising margins. If or when capacity in the PC business greatly overshoots customer demand, growth will stall, margins will vanish, and there will have been many years for other companies to mature and dominate in new customer segments in which customers still need more than they can buy. If you believe Christensen, it's because of the strength of the current PC business that INTC is likely to find itself unprepared.For example, we're already passing the point where very few people really need a faster x86 chip. But it's unclear what new set of customers is emerging. Perhaps people need a portable, ultra-lower-power chip. Perhaps a cheaper one, one with an integrated cell phone, or one with a hardware MPEG decoder.The management at INTC has read Gilder and Christensen, too. You can see they're trying to buy pieces of the puzzle in order to save themselves:INTC acquisitions in 1999 alone -DSPC - DSP for cellular wirelessDialogic - internet telephonyIpivot - web traffic "director" hardwareStanford telecom - cable and wireless broadbandNetBoost - firewalls, VPNsSoftcom - routing softwareShiva - networkingLevel One - SONET/ATM/DSL/telephony switchingLook at it this way: they're not buying the technology. They're buying the new customer groups. Its current customers are the PC makers - the big OEM business. INTC knows they need a set of "new wave" customers (it looks like it thinks these might be cable and telephone companies, or ISPs, perhaps) to drive their business strategy in a new direction in the future.The problem is, INTC has no choice but to try and make a few bets with high-volume or high-margin customers, people who need faster processing on their routers and switches, for example. INTC can make a few bets, but it is leaving must of the landscape untouched.Meanwhile fierce little hardware-platform companies are carving out niches in a large number of other "little" side shows - Xilinx/Altera, ARM/MIPS, Analog Devices/Texas Instruments, Broadcom, etc. Their customers are asking for things like $5 Bluetooth chips, or low-power MP3 chips that can be made in small quantities. INTC can compete in some of these areas, but it cannot play in them all. As long as nobody knows which of these approaches will become important, INTC will have trouble focusing its energy. It is difficult, and inappropriate, to squeeze a big company into a little niche.What do you think of MSFT? MSFT is a puzzle to me. I can't figure out if it is in the same boat. Is the "internet software" business fragmented into a large number of perilous niches? Or is the new business consolidated and lucrative enough that MSFT can successfully compete?And are the economics of software - infinitely flexible, freely replicable - sufficiently bizarre that Christensen's "disruptive technology attacking from below" doesn't apply? What I have in mind is the trend for software to be aggregated into bigger and bigger pieces, at little or no cost to the customer. Yahoo, Ask, Amazon, and so on continue to aggregate features into their websites: the Internet "software" business is consolidating rather than fragmenting. Unify enough software, and you will find that you're in the same market, competing directly for the same customers, as Microsoft. And when focused on the same customers as its competitors, Microsoft knows how to play to win. Maybe?