A little over a year ago we briefly discussed through-silicon vias (TSV). We examined some potential advantages and provided a link to a little TSV primer. Back then, the projection was that we ought to expect chips featuring TSV by the beginning of 2008. And while a few ICs here and there have been manufactured using TSV, a general standard was lacking thus slowing adoption across the industry. To overcome this problem, the 3D-IC Alliance, whose founding members include Tezzaron Semiconductor and Ziptronix, has released a first standard for 3D integration, dubbed the Intimate Memory Interface Standard (IMIS) – could they really not come up with a better name? While complaining, what is up with the 3D-IC website? It is truly an eyesore, and bags for a makeover.

Anyhow, as with most standards, the specification is quite extensive coming in at over thirty pages, but keep in mind that tables and diagrams take up a lot of that space.  The contents cover pin specifications and pin usage direction for most common memory types and their variations. Following the pins, the actual surface and target requirements are discussed at length. This section is split into three categories: Direct Bond Interconnect (DBI) which is championed by Ziptronix, Copper to Copper Bonding which is backed by Tezzaron under the FaStack brand, and a third category which at this point is undefined and reserved for future use. The final section is a short discussion of footprint diagrams and their variations.

As mentioned in our prior post, there are some immediate advantages that come to mind: shorter interconnect paths and more compact floor plans that no longer need to account for large caches, however, what we missed are the possible implications for Integrated Circuit (IC) security. On this topic, Tezzaron has a short paper which within the space of a couple pages, discusses on a high-level some of the reverse engineering techniques, as well as the two major security advantages for 3D ICs. To summarize, the first advantage is that if one of the layers in the stack is face-up while the connecting one is face-down, the outside surfaces would consist only of I/O pads, making etching and de-layering difficult. The second advantage is the fact that with a well defined interface, separate components can now be manufactured at different foundries, further reducing the risk that the final functionality of the part might be exposed. Obviously, the paper is intended as promotional material for 3D ICs, nevertheless it makes a few interesting points and is worth checking out for those interested in the subject.

Last fall we profiled Black Sand Technologies when they successfully completed raising $8.2 Million in Series A funding. Other than a few appointments of individuals to key positions, there really has not been much news regarding the company since. Needless to say, their web-site has also not been updated with any new product information. But thanks to Forbes.com and Clair Cain Miller who wrote the actual article, we get a little insight on Black Sand’s chief technologist Susanne Paul and her thoughts regarding silicon based power amplifiers.

If you ever wondered what it takes to be a chief technologist, here are a few clues: you ought to be able to work 80-hour weeks, while taking care of several children, and fixing your own car troubles! Anyhow, here is why she believes that silicon based power amplifiers are they way to go in the future. First of all, gallium arsenide, which is currently used for power amplifier design, is much more expensive to manufacture than silicon. The article points out that the cost per wafer for gallium arsenide has been constant over the last 25 years at about $400. In stark contrast, the same waver costs only about $45 for silicon. Additionally, silicon based amplifiers ought to consumer significantly less power then their gallium arsenide counterparts, assuming that engineers will be able to develop algorithms allowing them to vary their broadcasting strength based on transmission distance. They should also be less susceptible to interference, such as tall buildings, and more flexible in accommodating different networks when people are roaming.

Sounds pretty impressive, but should you believe what Susanne is saying? Well how about this quote from the aforementioned article: “She is uniquely qualified, as the only human being in the world ever to have built and put into production a silicon power amp - Venu Shamapant, Austin Ventures” What else do we learn from the article? Well, it seems that Black Sand is expecting to have their first silicon samples later this summer and believe their design will be incorporated into cell phone products in the 2009 time frame. This of course assumes that none of the other startups pursuing similar solutions, or some of the established players in the cellular filed will be able beat them to the punch.

Dean Takahashi, a regular contributor to VentureBeat.com, posted yesterday a very interesting article over at the Industry Standard regarding Montalvo Systems failure and what it might mean for future x86 startups. We wrote about Sun’s acquisition of Montalvo Systems a couple of months ago, highlighting the more than $70 million dollars that the company was able to raise and the rather paltry sum for which it was eventually acquired. Dean’s postmortem gives us a concise history about how Memory Logix evolved from their initial synthesizable cores charter to eventually become Montalvo. We get a good look at some of the major players behind Montalvo including Vinod Dham, the co-founding partner of New Path Ventures, Mike Yamamura, Greg Favor, Peter Glaskowsky, Laurent Moll, and of course Peter Song who started Memory Logix.

It seems that from the start Montalvo had a hard time deciding on what to build. Ideas contemplated were: a processor incorporating a graphics chip, a partial x86 compatible chip only capable of running Linux, a non-x86 cell phone chip, and finally a four-core chip with asymmetric cores for the mobile space. The last one, being quite different from what AMD and Intel offered at that time, was probably what allowed the company to obtain the initial round of venture capital funding between the years of 2004 to 2006. However, it seems that around 2007 things started falling apart. The company was expanding rapidly and burning through cash very quickly while at the same time falling behind schedule. An interesting observation that Dean makes is that Bangalore, where the company hired a lot of talent, become such a hot area of recruitment that even there engineers were no longer cheap, further not helping the company. Soon, the company dropped the asymmetric core idea and decider on a simpler core with four identical cores. This might have been the major strategic mistake that the company made. Suddenly, the chip became indistinguishable from what the big guys would offer over the next few years. On the performance side it would have to compete against Intel’s latest Core 2 Duo processor while on the power side the recently introduced Intel Atom would have surely become a nail in Montalvo’s coffin.

Interestingly, Dean reveals that several investors were nevertheless willing to throw another $150 million dollars at the company to let them finish the initial chip and maybe a second version after that, but one of the major investors advised against it. Too bad in a way, since it would have been really interesting to see what the Montalvo team could have delivered – some small teams are capable or extraordinary achievement. On the business side of things, the lesson to be learned here is that when market conditions change, the product has to be modified to address the new circumstances, but one has to be really careful not to change it such that it looses the uniqueness and with it any advantage or niche that it was targeting.  Sun Tzu said that good warriors prevailed when it was easy to prevail because they knew themselves and properly assessed the situation, thus positioning themselves on ground where they would surely win. Montalvo on the other hand had a identity crisis right from the beginning, followed by poor product positioning in the end, thus severely diminishing any chances of victory that the company might have had – which explains why the money eventually dried up.
Ironically, just the other day Montalvo was granted its first patent, according to Peter Glaskowsky, titled Adaptive Computing Ensemble Microprocessor Architecture. Of course, given the recent acquisition of the company, the patent assignee is now Sun Microsystems.