SI
SI
discoversearch

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 : Rambus (RMBS) - Eagle or Penguin -- Ignore unavailable to you. Want to Upgrade?


To: Dave B who wrote (26237)8/3/1999 1:23:00 AM
From: Don Green  Read Replies (2) | Respond to of 93625
 
Dave

So you actually believe there is such a thing as the Random Walk?

I guess that is where we differ..

Beware of the Darkside of "the Force"

BTW When you sell it and actually get that gain Congrats. You will have certainly earned it.

regards
Don



To: Dave B who wrote (26237)8/3/1999 4:07:00 AM
From: Tenchusatsu  Respond to of 93625
 
To Dave and others who enjoy technical discussions,

I have to correct myself from earlier posts here. Previously, I thought that the minimum granularity of SDRAM was eight chips, i.e. you can't build a memory system with less than eight chips without hurting bandwidth. This was because I thought all SDRAM chips were 8-bit, and that you needed eight of them for a 64-bit wide data path. (DRDRAM, on the other hand, allows for a minimum of one chip in the memory system. This is good for low-cost solutions like Intel's upcoming Timna CPU.)

I was wrong. Apparently, there are also 16-bit and 32-bit SDRAM chips out there. These chips allow for four-chip and two-chip memory configurations, respectively. For example, with 64 Mbit technology, 32 MByte SDRAM DIMMs are being sold in the market right now, and each DIMM holds four 64 Mbit chips.

So it seems that you can build SDRAM memory systems with less than eight SDRAM chips. However, the pin count will still be higher than DRDRAM. For instance, two DRDRAM chips both sit on one 16-bit data bus. In comparison, two 32-bit SDRAM chips must work in parallel to drive a 64-bit data bus. You can guess which solution saves on pins, and therefore on cost.

For those of you who want to see pin counts per chip:

4-bit, 8-bit, 16-bit SDRAM chip: 54 pins
32-bit SDRAM chip: 86 pins
16-bit DDR SDRAM chip: 66 pins
32-bit DDR SDRAM chip: 100 pins

DRDRAM chip (all are 16-bit): 54 pins

Tenchusatsu



To: Dave B who wrote (26237)8/3/1999 5:50:00 AM
From: unclewest  Read Replies (1) | Respond to of 93625
 
I see downswings that suddenly reverse themselves

dave,
how true...and how enlightening...
you must reside on the "bright" side.

a former next door neighbor and great friend, i have since lost track of, a very smart rabbi, wrote a book the title of which describes your efforts well..."enlightening the darkness"

talking about predictions...i have one.

i am going to do a bit of spekulatin agin...

i have posted several times in the past 2 weeks that i have been getting vibes about a mid august announcement. specifically august 16.
i thought it would have something to do with the hot chips conference at stanford and it may. i now speculate that this is what it is about.

some of us were concerned yesterday about sunw's new product the majc.
the wsj talked about animation over the net in real time. i believe it has to use rdram. afterall what other choices are there. this is precisely the kind of bandwidth demanding app that we have long predicted on the thread.

discussing this last night with a friend, he wrote...

"The significance to me is that MAJC is a net empowering processor...meaning graphics and streaming video crunching.
In fact, if you are right this will be the most positive and significant development regarding RMBS...it would be the beginning of the net enabling applications which is the point of RMBS...not the standard computer apps."

i have heard nearly the same words used to describe sony's new pII which we now know will use rambus as well.

my prediction is majc will use the rambus memory system and we will hear the details this month.
unclewest

this will drive intel nuts...to get their rambus products on the street. :o)



To: Dave B who wrote (26237)8/3/1999 6:07:00 AM
From: unclewest  Read Replies (2) | Respond to of 93625
 
read...
developer.intel.com

then read...

developer.intel.com

then...

hamsterdance.com