Ever since I moved to the left coast, UC Berkeley has become the most frequent destination of my research outings (it used to be MIT when I lived in Boston). I’m a regular guest at their RADlab retreats. Yesterday, I joined the 1-day Berkeley EECS Annual Research Symposium (BEARS). The morning was packed with four first-rate keynotes and a panel:

The future of devices, Elad Alon. Nano-electromechanical relays are a promising alternative to CMOS-based technologies and their unavoidable energy leakage. Like any other relay, nano-relays are leakage-free albeit much slower than CMOS and not as reliable. To mitigate these side effects, Elad is looking into more complex logic circuits and the opportunity to exploit parallelism (like in a N-bit adder or an ADC/DAC).

The future of computation, Kurt Keutzer. Deeper pipelining is not sustainable, parallelism is the saving grace. For this, Intel Larrabee and Nvidia Fermi are hugely exciting new processors. But how do we change the code to leverage the new silicon? There is early indication that algorithm/code conversion pays off with up to 100x improvements to time-to-result (teams started off from commodity software, like public domain support vector machines libraries – libsvm). Kurt did a great job at describing the whole ecosystem of parallel and show why/how it’s labor intensive. We need more/better frameworks to absorb these costs.

The future of Mobile, Eric Brewer. iPhone has converged dozen gadgets into just one (and more so every day). Inside, there are many discrete HW components taking up space and power, hinting that smartphones can either shrink further or carry more logic into them. Access is the smartphone’s killer app. Increasingly, mobile is a key factor in developing countries. There, it can save lives (e.g, a cellphone “microscope” contraption to detect malaria in the field; a diagnostic device connecting heart monitor and other sensors via the headset jack). The SIM card may become a good, universal place to store a private key. In developing countries, this setup actually works quite well because it’s already common practice for folks to own a SIM card and share a physical phone. Within every country, there’s a growing digital divide between urban and rural connectivity, with impact to just as many aspects of life as mobile touches.

The future of the Cloud, Michael Franklin. Cloud momentum will continue to be fueled by these value props: variable cost, cost associativity (1000 CPUs for 1 hr same as 1 CPU for 1000 hrs), risk transfer, and get the IT gatekeepers out of the way. There will be more devices and more virtual resources joining the cloud, including mechanical turks seamlessly blended in. Quite fittingly, there will be a new program at UCB to best harmonize Algorithms, Machines, People (AMP). It will launch in Jan 2011 when RADlab wraps up.

Energy panel hosted by Greg Papadopoulos. Can we innovate in energy the same way we innovated in technology? Three principles that served us really well in EECS and are worth cross-pollinating into energy are: a) layer decoupling, b) distributed innovation, and c) best equip for en-masse customization. A smart power grid is a dumb grid with many different smart endpoints. Some food for thoughts: Make solar panels become as cheap as a sheet of glass; Do nothing well (i.e., energy proportionality); Don’t recycle, up cycle.

The day was nicely complemented by open houses in the various departments, with plenty posters and demos. For ease of tech transfer to my children, I single out the demo of the software-intensive Starmac quadrorotor flying machines by the Berkeley Sensor and Actuator Center (see really cool videos 1, 2, 3 … heck, thou shalt see cool toys, green grass and the blue sky, once you’ve survived those pesky 3D Fourier transforms

This week, slashdot called my attention to EFF’s effort to level set the community on web tracking — how unique (and traceable) does my browser make me look when I visit a web site?  This new EFF site returns my overall score along with the break down of its factors (like plugin details, screen size, system fonts, cookie handling). For instance, it tells me that the Safari fingerprint generated off of my Mac is still unique among the half-million fingerprints on file at the EFF.

This is a great example of crowd-sourcing at work. The more participants, the better the study. EFF’s work gets a huge boost from being slashdotted. Moreover, EFF is no .com and doesn’t  have the halo of big-brother or world domination.

How does one know when the samples have hit a critical mass leading to a reasonably accurate model? It’s a recurring conundrum for both frequentists and Bayesians.

I agree with EFF’s view that a smartphone’s browser is due to show lesser entropy. That kind of browser is less likely to veer from stock config. To witness, my iPhone browser scored 1 in 1,442 uniqueness (10.49-bit entropy) and my Android browser scored 1 in 8,513 uniqueness (13.06-bit entropy). To the previous point, it’s unclear how many smartphones have hit the EFF site altogether.

This smartphone/browser early conclusion should not be generalized to native apps running on a smartphone. These native apps can yield the richest fingerprint features yet. They can draw upon some sophisticated UUID and TPM schema in system software, with the SDKs exposing programmatic access, resulting in stronger software/hardware linkages than their desktop/laptop equivalents. Today, the limiting factors here have to do with policy – e.g., a vendor’s authorization to export off-device the UUID material that is key to its own DRM.