Identifying Wafers - How? Where?

[Vlasta]

Hi - I have a bunch of wafers that I just cannot identify no matter how much googling I seem to do. I am sure there is a simpler way and I hate to keep bugging you guys - for example I just looked at an AMD 6" wafer that is marked "sembar c rev. c" and with the AMD logo and '5924A' marked on the die next to the logo. What is the best resource for finding out just what this wafer is? I have lots of others too and they are a mystery to me most of them - do I keep 'em all? not possible as I invested time and money to get them to trade - certainly I will keep some like the MIPS wafers that I can easily identify. Any help appreciated.

[Birdman.]

Complete die shots could help. I could then check whether I have found some similar dies elsewhere and at least they would give an idea about what kind of chips they are and from what era.

I'm pretty sure the year is written in the first image, but it's too blurry to read. Information about the year could also help a lot in identification.

[Vlasta]

about the best I can do by hand....

[Birdman.]

OK, that is clearly AMD 486DX2 or DX4 because I have found many die shots of them and they are just like this die. Check chip10 at http://www.pbase.com/fotoopa/inside, at least it is marked as 5924B, so pretty close. Thus, I'm pretty sure this wafer has AMD 486DX4 dies with earlier revision than the chip10 in fotoopa's collection.

You can compare the edges of the two DX2 versions and that DX4 in fotoopa's collection to your wafer because there seems to be clear differences between those dies. DX4 die is also clearly smaller than those DX2 ones, so die size measurement is also useful there.

BTW, those fotoopa's die shots are the best ones I've found for AMD 486DX2 and DX4, so you can try to take better one now that you have them on your wafer. I actually happened to find fotoopa's collection yesterday, and it helped to identify this wafer today

[Vlasta]

Yes!!! - clearly the same die - thanks a lot.

[CPUShack]

a very nice find then!

[Guest]

[Vlasta]

Birdman - any idea what this is?

[Birdman.]

It was not in my die shot collection, but after a bit of thinking it started to look like some MIPS R4600-R5000 with on-die L2 cache. I then looked at Wikipedia to see which MIPS had on-die L2 cache, and the only one was RM7000.

So, there it was:
http://commons.wikimedia.org/wiki/File:QED_RM7000_diephoto.jpg
http://www.csd.uoc.gr/~hy225/2013a/images/rm7000.jpg
Looks like exactly the same die, so you have got a RM7000 wafer there!

Again another die that would be great in my die shot collection with best possible quality! Thank You a lot for the die shot, it is already good enough to get into my collection after some sharpening and cleaning. However, maybe I'll wait for possible better version before trying to edit this die shot.

[Vlasta]

Cool!!! - Now I have MIPS 3000, 3010 FPU, 4000 SR71000 and now RM7000 wafers - getting to be quite a MIPS collection even though that's not how I started out. My goal is a bunch of Intel wafers!!!!

I have 5 of these wafers and one got broken - sooooo I am going to treat it to an acetone bath to try and remove the yellow coating it has on it - maybe get a better and clearer die shot after that.

[Vlasta]

OK - Acetone won't remove the coating (conformal coating or what?) and paint stripper did not budge it either - any ideas?

[CPUShack]

Xylene, or if its silicon based, conformal coating remover (tetrahydrofuran or CW3500 by Circuit Works)

[Vlasta]

[Birdman.]

Does the coating blur the details? If it doesn't, then it doesn't really matter in die shots because you can adjust the colors afterwards if you don't like that yellow. A sharper and cleaned version of the RM7000 die shot without those white spots would already be really beautiful. Is the sharpness limited by the equipment (scanner?) used or by the yellow coating?

[Vlasta]

Yes unfortunately it blurs the details. I ordered a small bottle of xylene from eBay so lets see.

The big die shot above was taken by John (CPUShack) on his scanner I think. I will try it on my scanner later.