Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Micron Only Forum -- Ignore unavailable to you. Want to Upgrade?

To: TREND1 who wrote (31970)	4/11/1998 1:38:00 AM
From: DJBEINO	Respond to of 53903

DRAMs face tough times in PC/100 era by Mark Ellsberry Dram vendors are jockeying for a piece of the PC/100 bus-compatible DRAM business. The PC/100 standard, initiated by Intel, establishes parameters for the next-generation memory bus operating at 100 MHz. Achieving this speed with the current construction of PC motherboards requires careful coordination among memory, chip-set and processor designers. There are several major issues to confront before fully launching into PC/100 system designs. To set this stage, it's important to understand that DRAMs complying with the PC/100 standard represent the third step of the PC's five-year performance road map for DRAM development. (The first step was extended-data-out; the second is synchronous DRAM.) There is considerable confusion in the industry over what the clock speed of an SDRAM means in a real memory application. A PC/100-compliant SDRAM is specified at 100 MHz, but some DRAM vendors already have 100- and 125-MHz SDRAMs on the market. Some of the more critical aspects of the current PC/66 and the new PC/100 SDRAM specifications include setup, hold and clock-to-output timings. Clock-to-data delay in PC/66 is 9 ns, whereas it is 6 ns in the PC/100 spec. Setup time is 3 ns in PC/66 and 2 ns in PC/100; hold time after clock is 1.5 ns and 1 ns respectively. Thus, even if an SDRAM operates at 100 or even 125 MHz, it still may not comply with PC/100 specs. For some DRAM suppliers that up to now have focused on manufacturing, the PC/100 era will mean staffing up with system-engineering and applications talent. This level of engineering support will be needed to deal with such system-design issues as timing margins and signal integrity, among others. Though these tightly defined parameters bode well for performance, they will also trig-ger some problems. For starters, there likely will be a shortage of PC/100-compliant SDRAMs, chips that are difficult to build. An adjunct issue is device incompatibility. These DRAMs will vary from one vendor to the next, with required features and functions interpreted in different ways. Moreover, there is a lack of general understanding of how to utilize SDRAMs in a system-especially in consumer applications like set-top boxes. All too often, an SDRAM is seen as a garden-variety DRAM, with little thought given to the implications for other aspects of the design. System makers are qualifying the 16-Mbit version of the PC/100-compliant SDRAM now, and 64-Mbit versions will be qualified this year. We expect the 64-Mbit SDRAM to be the more popular, and to significantly cut into 16-Mbit demand. This makes sense, because the number of memory slots in a PC is being reduced. Most vendors are shipping 32 Mbytes as the base memory, in the form of a 4-x-64 module-an entry point where a 64-Mbit SDRAM can be effective. What's next after PC/100? There are two schools of thought: Rambus and the double-data-rate SDRAM, seen by many as a long shot. The DDR specification is undergoing slow and tedious creation, with the negotiations dealing with functions split along lines of competitive self-interest of various vendors. Regardless of which approach is adopted, the system engineer is guaranteed another round of related systems-design issues. -Mark Ellsberry is vice president of marketing for the semiconductor division of Hyundai Electronics America (San Jose, Calif.). Copyright c 1998 CMP Media Inc.

To: TREND1 who wrote (31970)	4/11/1998 8:52:00 PM
From: yard_man	Respond to of 53903

>>This would end up obsoleting the sub-$1000 PC.<< With something a lot more expensive? May be a tough sell.

To: TREND1 who wrote (31970)	4/11/1998 9:00:00 PM
From: Skeeter Bug	Respond to of 53903

>>This would end up obsoleting the sub-$1000 PC<< larry, this guy is a dolt. i bet 8 months ago he was paying $55A+ for mu expecting them to mint money this year ;-) the sub $1k pc will not go away. period. the value of this price/performance trade off is by far the best thing going. anybody that moves away from it will not sell boxes. everyone will hunker down with their p233 sub $1k box and flip intel the bird by not buying any boxes. i'm amazed that rank amatuers in basic economics get published. this guy is wishing intel's wish. it ain't gonna happen. just like mu ain't gonna make money. just like many said when estimates were for $3+ right now! mu made about $1.50 from 1984 to 1993 as mike pointed out on his thread. oh, that was not per year, either. that was total. seems 1994-2003 has shown a regression towards mu's eps mean. barnum was right...