On Jun 4, 14:14, Joe R. wrote: 2708. Maybe. If so it must have been identified before the leads were added, because the Intel Data Catalogue 1976 lists both 2704 and 2708 on the same page, and the only difference is pin 22. On a 2708, that's A9, on a 2704 it's always 0V (always 0V, cf 2716/2758 below). No pictures, unfortunately, so no immediate way to tell if the dies look the same. Interestingly, the 1976 book doesn't list anything bigger than a 2708. So a 2708 obviously wasn't a half-bad 2716 (at least, Intel ones weren't). The 1979 Data Catalogue lists the 2704 only as a footnote to the 2708 description, and the 2716 appears only as a single-rail version. I thought they did both, but I must have been thinking of another manufacturer. The 1979 book lists 2716, 2732, and 2578. The 2758 has exactly the same power consumption, access times, programming, and pinout as the 2716, except that pin 19 is A10 on a 2716 and is AR on a 2758. AR is "select reference input voltage". The tables show this as always Vil (ie, 0V) but in one place only, the small print says that it's always 0V *except* for devices labelled "2758 S1865", when it needs to be Vih (+5V). Hmm, that sounds exactly like Dwight's description of a half bad chip to me :-) -- Pete Peter Turnbull Network Manager University of York