atari email archive

a collection of messages sent at Atari from 1983 to 1992.

VADRS chip in prototype fabrication and 256k VDRAM shortfall

(1 / 3)


VADRS and 256k VDRAM's:
	Purchasing's difficulty in getting 256k VDRAMS for the "DRIVER" project
in Q3, Q4 88 has alerted me to the analogous potential problem for Sys3d in '89.
The word from TI's Jon Henricks (Field Sales Engineer) is that fab lines are
being converted from 256k DRAM and VDRAM to 1M DRAM and VDRAM, that the parts
are hard to get now and will be harder to get in the future. This is just one 
vendor's plan.
	The VADRs chip is currently on the fabrication path.  It is useless
without 256k 120ns Video RAMS.  It will not work with 1M Video RAMS.  Using 1M
parts would require redesign starting at the TTL prototype stage and would take
an estimated 9 monthes to get to engineering samples.  The other chips in the
system will work with 256k or 1M or ... video RAMS (Sorter, Gobot, Zippy,
Clippy, Maggie). Dave and I are contacting vendors (Hitachi, TI, Fujitsu, Mits,
NEC, Toshiba) and working with purchasing to get a feel for 256k VDRAM
availability in '89, '90. 

	Consequently, I asked Robert Pescosolido to investigate 
	what we could do to suspend the VADRS.  He has put the VADRS 
	fabrication on hold just prior to mask making. 

Suspension of the job incurs no additional cost, avoids the mask
charges (10 - 15K) and will not impact the Sys3d video schedule if the hold is 
release before GOBOT CIF is released (expected at the end of August).
	I expect/hope that once the VDRAM vendor information is in, and the tea
leaves are read, we will feel adequately comfortable in releasing the VADRS for
fab. 
	Also, Mary Burnias has offered to invite the vendors in individually to 
discuss their plans for the 256k VDRAM's.  Who wants to attend these sessions?  
I'd like to get a head count by the end of Thurs  6/9.  

----------

Rick and Jed, I think only TI has come in to discuss long term 256k VDRAM plans 
with you and that the suggested vendor sessions would not duplicate your prior efforts.  
Is that correct?
					Morgan

Mult-sync and VRAM's

(2 / 3)


A flag has been raised that 256K VRAMS as used in  Multi-sync are in short
supply now, and may become even scarcer as VRAM manufacturers convert 256K Fab
lines to 1MEG VRAM fab lines. Just in case this really happens, I would like
to know what contingencies we have, if any. 

1) Can Multi-sync be re-designed to use (1) 1Meg VRAM for (4) 256k VRAMs?
    if not...

2) Can Multi-sync be re-designed to use (1) 1Meg VRAM for (1) 256k VRAM?
	(and then only use 1/4 of the 1Meg part...ugh)

3) Are there software implications if the VRAM size changes? If so,
   what software changes and to what extent?

I am not really going to worry about this in a serious way until about January
'89. StunRunner production is tentatively planned in June '89. We would have
to start a PCB re-design in a January timeframe to have it ready for June 
production. 

RE: Mult-sync and VRAM's

(3 / 3)


1. No
2. Yes/No
3. No
Message 1 of 3

Jun 08, 1988