Mark Wickens wrote: Whilst I'm thinking about it, does anyone have any Alphaserver 1000A memory kicking around. I started getting a bit parity error a couple of days ago (which the box helpfully pinpointed to an individual SIMM) but had to pull all four SIMMs in the bank. I also had a problem with memory in an Alphaserver 1000A. The firmware did indeed narrow it down to an individual SIMM. However the diagram inside the lid did not identify which SIMM was which, only the banks. In the course of figuring out which SIMM was which by moving them around, I found that the firmware sometimes lies about which SIMM is faulty :-( If I put the SIMM which I eventually determined to be faulty into slot 2, it got reported as a problem with slot 3 and despite trying all sorts of permutations, I never found a way to make it report a problem with slot 2. Problems with slots 0, 1 and 3 were correctly identified. Someone kindly gave me a pdf of the owners guide since then which does have a diagram of the memory layout with the individual SIMMs labelled and I was able to confirm that I had correctly figured out which slot was which. The part number I'm looking for is: 50-21276-01-02, it's a 8MB x 36 SIMM 60ns EDO FPM, also listed as being for an AlphaServer 200 4/233. My set of 3 spare working SIMMs arising from my one failed SIMM are marked: "54-24123-AA 16MX36.60 Z" which doesn't seem to be what you need :-( (My experience with ES40s is that they are very picky about memory and don't like memory being mixed and matched. I don't know if this also applies to the Alphaserver 1000A - hopefully not.) Regards, Peter Coghlan.