So thanks to a couple tweets from Jim Prior (@cavemanjim) who’s blog post was included in last week’s Hot Chips curation, I was pointed to a couple of posts on SemiAccurate written by Charlie Demerjian, who covered the Xilinx tutorial featuring AMD, Amkor, Xilinx Qualcom and UMC in two parts: Hot Chips Talks about Chip Stacking, Good and Bad and Die Stacking has Promise, and Problems.
When I started to read Part 1, I was turned off by Demerjian’s negative tone. I quote “Unfortunately the problems are as severe as the benefits are positive, if not worse. In fact the most important problem, we can’t actually make a 3D package with high wattage silicon, is quite the show topper. The near complete lack of design tools, non existent standards, and no industry knowledge base is nearly as crushing.” Seriously? Crushing?
I think this is really a gross exaggeration for editorial impact (oh look, it worked!) and Cadence, SEMI, TSMC, Global Foundries, ASE, Amkor and host of others would beg to differ. (Charlie, I don’t think you’ve been paying close attention! I suggest you watch this video , and read this press release from SEMI , and this post by me.)
Editorializing aside, these posts are both worth the read purely for the analysis of Qualcomm and Xilinx presentations. (But don’t expect any mention of AMD, Amkor, and UMC, because there isn’t any). So I guess what we really have here is coverage that is, as indicated by the website’s name, semi-accurate. What did I expect? ~ F.v.T.
