LONDON – Warren East, CEO of processor IP licensor ARM, says the company stands ready to help STMicroelectronics make a success of its fully-depleted silicon-on-insulator (FDSOI) chip manufacturing process, but that it is up to ST to make the process more widely available.
Speaking to EE Times in a discussion of ARM's 4Q12 and full year financial results East said of FDSOI: "We think it is pretty good technology and we would encourage ST to proliferate it. The physical IP we need to create is essentially the same [as for bulk CMOS]."
FDSOI has emerged at the 28-nm node as a potential chip manufacturing alternative to bulk planar CMOS, which is being pushed to 20-nm by foundries such as Taiwan Semiconductor Manufacturing Co. Ltd. and Globalfoundries Inc. Both those foundries then propose to move rapidly to FinFET-based processes at nodes labeled 16-nm and 14-nm, respectively.
Some observers have argued that 20-nm bulk planar CMOS will not produce significant power savings while being hard to manufacture. The FinFET processes, where transistors are created in fins standing proud of the silicon surface, could be even more complicated to make and harder to yield.
However, the FDSOI process developed by STMicroelectronics does not yet have any high volume or foundry manufacturing capacity in place. The earliest that any could come on-stream would be the fourth quarter of 2013. ST does have an understanding that Globalfoundries would license the process and provide volume manufacturing.
ARM has been working extensively with leading EDA companies and foundries developing physical design kits, design flows and test chips to prove out both bulk CMOS at 28-nm and 20-nm, and on FinFET designs with TSMC and Globalfoundries at below 20-nm. A similar body of work has not been publicized around the FDSOI process.
"At the moment it is effectively a proprietaty technology. We can help ST if they can proliferate the technology," said East.
East declined to be drawn on whether ARM's own "big-little" technology – where a power-optimized processor core is paired with a performance-optimized core as part of a dynamic voltage and frequency scaling regime – is a good fit with FDSOI. FDSOI can save power but also achieve-leading edge clock frequency through an extended operation voltage range.
"Big-little is a technique for achieving power efficiency. FDSOI is another technique. There's room for lots of techniques. Our people have seen the demos. We don't need to amend our IP. The ST demos are based on ARM anyway," said East.
I first saw ST present on fully-depleted planar SOI technology over 10 years ago -- it's been a long, meticulous journey to reach this point, with a steady stream of papers at the major conferences. The folks at Leti, as well as IBM, ARM, GF, Hitachi, UCBerkeley, Soitec, UCL, Cadence and more have been a major part of this effort, too. This is not a rabbit they've pulled out of a hat. They're getting awesome, silicon-proven results -- especially in terms of cost & power. It looks to me like they've got the right technology at the right time...and now time will tell if they are right.
Yes, from design side would want to reduce well area and switch small area wells.
That would still leave parasitic C under drain junction (for thin BOX) and that would need to be improved from process side (perhaps make box thinner under drain).
To be clear I do think all this is fundamentally solvable.... I just point this out since I think all these type of work needs to be done if FDSOI ever goes mainstream and FDSOI shows its full potential (low variation and low power). FDSOI is a solid concept (perhaps better than bulk 20SOC or bulk FinFET) but it will take an open debate and a few fixes to move concept forward vs. today.
thanks for the discussion
I can confirm your point is correct and identified by TSMC 5 years ago as a flaw in FDSOI. STI trench is needed (cost) and wafer non-planarity would make "hybridization" 0 yield.
See slide 15 for both points
Lastly, the other major issue is strain for high mobility is ineffective in FDSOI.
This limits FDSOI to low performance.
Also has implications on physical IP porting. N/P ratio goes from ~1.3 at 28nm back to ~2. Makes all my physical digital and analog IP non usable without major redesign.
see slide 12 on strain in FDSOI
Adele does not seem to admit problem thin box approach.
can you comment on parasitic capacitance from 3V on UTBB (Asele's recommended 25nm thin buried oxide).
It adds too much extra C and wasted power to be viable.
I would like to thank Mr. Cesana for pointing out: 6nm is 60A, not 0.6A. Since MR. Cesana’s comments are mostly on UTBB, I will respond to FDUTBB. Remember FD UTBB and FDSOI are not the same.
ST video claims that its UTBB behaves like a vertical double gate. It doesn’t. The double gate is an ideal transistor structure having common gates and common source /drain, thus good control of electrostatics and doubling the transistor on-current, Ion. ST’s UTBB has common source and drain, but has two independent gates consisting of two transistors, the top transistor having the proven HK metal gate very reliable used today in semiconductor industry but the bottom transistor having the Si substrate for a gate and the 25-nm thick buried oxide for gate oxide, sharing 7nm channel is totally new and unproven in reliability, performance, and not adopted by semiconductor industry. During UTBB operation a positive 3V is applied to the bottom gate to control Vt of the top gate. How much the transistor I-on is improved by the positive 3V applied to the bottom gate is not shown. Furthermore, some of channel electrons could drift toward the buried oxide and become trapped inside under the 3V positive bias field during UTBB operation, especially near the source region where electron velocity is very slow. Also, a number of interface states could be generated at the thin Si channel-the buried oxide interfaces, and the channel electron mobility could be degraded due to enhanced scattering at the channel-buried oxide interface, resulting in reduced I-on. These could adversely impact UTBB reliability and performance. These phenomena are unique to FD-UTBB because planer bulk, FinlFET, and FDSOI are not substrate biased or grounded during device operation.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.