Re: Larger memory option for VAXstations

[Johnny Billquist <bqt%update.uu.se@localhost>]

On 2015-08-19 14:50, Felix Deichmann wrote:
> 2015-08-19 14:29 GMT+02:00 Johnny Billquist <bqt%update.uu.se@localhost>:
> > That don't make sense. 36 x 4mx1 would mean 32 megs of memory. My VX4000/90
> > have 128M of memory, and that was DEC original. So obviously DEC also made
> > 16Mx1 based memory modules.
>
> Pardon, but 36 x 4Mx1 means 16 MB of usable memory (considering
> parity, i. e. 9 bit per byte).

Yeah, I was also assuming parity on each byte.

> 36 x 4Mx1 = 150994944 bits, and 150994944 bits / (9 bit / byte) = 16
> MB. Or am I completely wrong now?

You are right. I don't know how I managed to count to 32M. Sorry.

> And your VS4000/90 has the max. 128 MB from using 8 DEC original
> modules with 16 MB each, right? At least my /96 is configured that
> way.
> I thought that for the VS4000, the 16 MB module was the largest
> original one available.

I think so too. And to correct the math, the 128M means that they must 
be using 32Mx1.

> > For 512 meg, why not locate 64Mx1 memory chips? Or did those not exist?
>
> Might be even rarer than 4Mx1 or 16Mx1. And maybe the larger (and
> therefore later) ones are only available as EDO, which could introduce
> new trouble when the machine expects "plain" DRAM and its timing.
> Don't know.

Well, correcting my math. If we want 512MB with 36 chips means that each 
chip needs to be 128Mx1.

> But this whole idea will require serious effort, fiddling and money...
> Why not just buy a MicroVAX 3100/88 or /98, which is NVAX and seems to
> support 512 MB by design? :)

The CPU do, but does the motherboard? That was what Nemonix was replacing...

	Johnny