Sandberg, Finding NIMo, Finding NIMo v10.
The Board of Governors of the Federal Reserve System, via the Comprehensive Capital Analysis and Review (CCAR) program, has opened up a $100+ billion annual market in optimizing the Net Interest Margin at significant banks in expected case scenarios. This note describes a novel Control Theory implementation of the Net Interest Margin Optimization (NIMo) algorithm that is most relevant to banks with 1+ trillion dollar balance sheets and subject to CCAR regulatory reporting. Broadly, the idea is to automate the Bank’s discretionary capital allocation plan to optimize the new business revenue (e.g., funding loans with deposits) relative to dynamic operational and market constraints. We are working with Cray Inc. to host this computation on a large-scale, parallel, competitive, contemporary hardware platform transitioning from AVX2 on Haswell to AVX 3.2 on Skylake.
FinMath.net, Libor Market Model, here. Java implementation zipped up here
This page describes the implementation of a simple LIBOR Market Model. The model is single-currency, single-curves (although extensions are straight forward).
Brigo, Libor and Swap market model lecture notes, here.
Andrew Lesniewski, Libor Market Model 2008, here.
Joel Hruska, ExtremeTech, IBM breakthrough improves carbon nanotube scaling below 10nm, here.
The new approach involves welding — nanowelding — a nanotube with molybdenum before they are self-aligned as transistor channels. The final step is to heat the assembly to 850C, melting the molybdenum off and creating carbide. According to Richard Doherty, of Envision Engineering, this solution gives IBM a unique advantage in scaling all the way down to 1.8nm. According to EETimes, IBM may be prepping this technology to be ready at the 5nm node, for introduction at 3nm and below. With the method already proven in theory at 9nm, there seems to be little barrier to further scaling.
R. Colin Johnson, EE Times, IBM Nanotubes May Redefine Future of Moore’s Law, here. Get your xlc compilers out.
If a breakthrough in carbon nanotube transistors from IBM Research pans out, the hard stop of 2028 in International Technology Roadmap for Semiconductors (ITRS) is about to get extended. IBM says it has found a way to scale down the channel length to the 1.8 nanometer node (four technology generations away) and beyond to the angstrom level eventually. If they are right, Moore’s Law may now be extended to the sub-nanometer angstrom (1/10th of a nanometer) levels using the same extreme-ultraviolet (EUV) complementary metal oxide semiconductor (CMOS) process technologies already in place.
“The 1.2 nanometer wide carbon nanotube channel is already proven,” Shu-Jen Han, IBM manager of nanoscale science and technology at its T.J. Watson Research Center (Yorktown, Heights) told EE Times in an exclusive interview. “The major issue for scaling, not only for carbon nanotubes, but for silicon and III-V materials [indium, gallium, arsenide] is the contact—which is no longer scaling.”