Path: utzoo!attcan!uunet!husc6!linus!mbunix!eachus From: eachus@mitre-bedford.ARPA (Robert Eachus) Newsgroups: comp.sys.amiga Subject: Re: Ram Chips... Summary: Forget the chip agreement, it's supply and demand. Keywords: economics, DRAM, 256K Message-ID: <32871@linus.UUCP> Date: 25 May 88 19:31:25 GMT References: <2996@crash.cts.com> Sender: news@linus.UUCP Reply-To: eachus@mitre-bedford.arpa (Robert I. Eachus) Organization: The MITRE Corporation, Bedford, Mass. Lines: 125 In article <2996@crash.cts.com> tsouth@pro-pac.CTS.COM (Todd South) writes: -- Some stuff fed to the line eater for brevity. RIE >Welcome to Reaganomics! :-) Yes, for a few measley hundred dollars, you >too can have one meg of memory! > >Don't you find it rather funny that one can buy a board with one meg >of memory already on it, strip the board of the memory to put on another >board, and pay less! Doesn't this smell of someone's stockpiling? > >IMHO, I truly think that a lot of stockpiling went on during the low priced >floods. But with this stupid trade agreement, Japan has to sell at the >same rate as American companies. American companies? Well, they could care >less about you, since its all a tax writeoff to them anyway (if they are >buyers) and you don;t need that much memory! (in the eyes of most producers, >and again, IMHO). I still remember when I had a chance to buy a couple of >megs of 1024x1 memory at $18.50 a chip. Too bad I just didn't buy it then >because the damn chips (along with every chip on the market today) has truly >escalated in price. > >Come on, you chip producers! I know that a number of you are reading this >feed! Explain to us why 256K chips have risen 500% in the last two months. >They sure as hell weren't that much before the flooding began! This is one >case where I think they will really screw themselves in the long run. Too >bad the people that immediately suffer are the consumers. > >Todd South ***Sitting back with Flame Shields on!*** The chip market normally is supply and demand. The Japanese tried (successfully for a while) to change that. What we are seeing now is partially a complex reaction to the Japanese grab, but for the most part is almost textbook free market. Several years ago, when the DRAM market was predictable, what would happen was that a new technology (they came out about every two years or so) would start out at low yield and high price. The only buyers would be people developing new generation boards. When the price point reached six times the previous generation (as lines and designs were debugged and yields improved) the products using the new chips would start to ship. The learning curve would continue, and soon the new chips would be about twice the price of the previous generation and old products would disappear as uneconomic. The Japanese who with the 64K DRAM became a factor in the international market, decided to beat the world (read US companies) to market with the 256K RAM. As a result 256K RAMs appeared at competative prices before most manufacturers were ready for them. The Japanese continued to ramp up production (as new lines came on stream, yields improved, etc.), and started developing 1Meg DRAMs as follow on products. The price plummeted (more production than demand, as most manufacturers were not ready to use the new parts. Eighteen months ago 256K DRAMs cost about as much as 64K DRAMs and the price (around $2.00) was uneconomical for chip makers. The responses to this oversupply (which was really underdemand) was threefold. Chip manufacturers converted DRAM lines to more profitable products. In the US this tended to be ASICs, in Japan it was 1Meg DRAMs. User demand increased, and the US manufacturers screamed foul. Not so much because the Japanese had cornered the 256K DRAM market (at uneconomical prices), but because they were threatening to do the same thing with the 1Meg chips. (Sell at a price that kept them in business in Japan, and dump elsewhere. Since US manufacturers were frozen out of the Japanese market, they would not be able to compete worldwide.) The trade agreement attempted to stop this, and had the effect that one US manufacturer stayed in the 256K market and several were able to enter the 1Meg game. Notice that the main effect of the agreement was to increase the number of manufacturers, while several Japanese companies curtailed their manufacturing of 256K DRAMs. Also computer manufacturers, looking at the tariffs, ordered large quantities in advance and started moving to 1Meg parts. The company I was at introduced a 1Meg chip memory board less than nine months after the 256K chip board. So one year ago, mainframe manufacturers are switching to 1Meg parts, chip makers are switching to 1Meg parts, and no one is adding 256K DRAM capacity. But demand is increasing, and in some markets 256K RAMs are as far as you want to go. (How many 1Meg DRAMs can you use in a 808 based PC? If you use 4x parts -- four, and two 256K parts for parity. One bit wide parts are useless.) Add one last factor, that 1Meg DRAM lines are taking longer to debug than expected because the generation arrived early, and you get to where we are now. End users screaming for 256K parts, and screaming at the prices. One Meg parts available, but the price is higher than six months ago (but only 50%, nothing like what's happening in the 256K market). If some fortune-teller told me two years ago that 1Meg DRAMs would cost $30 this month, I would have said, "So tell me something I didn't know!" When she told me that 256K DRAMs would be over fourteen dollars if you could get them, I would have called her a fake. The market that is out of joint is the 256K market, and it is affecting the prices for other DRAMs. The recent increase in 1Meg part prices is due to the partial substitutability, and the inability of manufacturers to ramp up as fast as the would like. What will happen in the future? The price of 256K parts may continue to rise for a while, but not significantly since it will be capped by the 1Meg price. The 1Meg price will probably be lower by the end of the year, but it won't drop too fast, again because of the 256K situation. Some manufacturers may start putting bad 1Meg parts in 256K wrappers to ease the situation. If you buy a memory board for your machine, try to get one that can take 1Meg parts, since two Meg using one Meg parts may soon cost less than 1 Meg using 256K parts. Two years from now, 1Meg parts will be cheap, but 4Meg parts may still be in the sampling stage. Sorry to be so long winded, but I do have a degree in Economics even if my job title says Software Engineer. (Actually, I'm a computer language lawyer, but that's another story.) These details are needed to understand that the Japanese, not the trade pact or the Reagan adminsitration, are resposible for the high 256K prices, AND THAT NO ONE CAN OR WILL BRING 256K PRICES DOWN IN THE NEAR FUTURE. The situation will only get worse as computer companies use up the chips they bought on long term contracts, and go into the spot market to get more. Robert I. Eachus with STANDARD_DISCLAIMER; use STANDARD_DISCLAIMER; function MESSAGE (TEXT: in CLEVER_IDEAS) return BETTER_IDEAS is...