July 15, 2009 – Jan Vardaman discusses alternatives to 3D/TSV technologies for cost-sensitive applications, and makes the case for additional work to develop design guidelines and software, and test methodologies for 3D/TSV technologies.
There’s still a lot of work to be done in design, thermal, and test, she says. There’s been a lot of work done and progress made recently in the thermal field, she noted. But more needs to be done in the design area, so that software can be available to anyone who wants to use the technology (e.g. fabless companies who want design guidelines from their foundry). TSMC and others have made progress here (and IDMs like Intel and IBM have their own internal capabilities), but more needs to be done, she said.
As with any new technology (including 3D TSV), it has to be placed in the context of a cost/benefit analysis to be judged vs. current technologies, and only when determined to be a compelling reason to shift to a technology will it happen, she says. E.g., what can TSV offer that wire bonding cannot for cost-sensitive flash makers (who can thin a die down to ~30 microns with bonding)? DRAMs for server applications, on the other hand, probably will need to explore new technologies (TSV as well as things like vertical circuits) as wire bonding runs out of gas there. Decisions will be made on a case-by-case basis, she noted.