ctrl.altered.mind

A friend asked me for my brief opinion regarding solid-state drives, particularly in the event of actual reviews of these devices pouring in (such as this one by Walt Mossberg of the Wall Street Journal). For the uninitiated, these solid-state drives are drop-in replacements for conventional hard disc drives (or HDDs), but employ flash storage instead of storing data on rotating magnetic media.

While this technology may be an interesting alternative for some, I do not see SSDs denting the expansive HDD market at least over the next decade. There will be some low storage capacity markets that would be taken over by solid-state storage, such as handhelds and special-purpose laptops. SDDs differ from flash storage employed by portable devices such as digital cameras and music players in two aspects: the type of flash memory used, and the interface. Current generation SDDs use the SATA interface and have a HDD-like form factor; and are therefore designed as a drop-in replacement for conventional hard drives.

My comparative reasoning between SDD and HDD technologies, is as follows:

Conventional hard drives are a mature technology and are continually seeing impressive gains in storage density (due to better materials, advances in physics, and better recording techniques). It is a just a matter of time until we start seeing multiple terabyte storage in hard drives. With a tremendous shift in media consumption towards high-definition audio and video formats, the quantity of available storage will be a perennial question.

While they do involve moving parts, the energy usage to operate a hard drive is a tiny fraction of that used for other computer parts. In todays laptops, powering the screen consumes the most energy, although this is being mitigated by the use of LED backlighting in some newer devices. In fact, in a comparative test recently performed on the Apple Macbook Air by Ars Technica, it was seen that a solid state drive presented no significant benefit in terms of battery life, even in an LED-backlit laptop computer.

The data access time benefit from flash storage is usually mentioned, but there are two sides to this story. Flash beats HDDs when it comes to random access times (due to spin delays in HDDs). However, HDDs have far superior sequential access times than SDD storage, which is of great importance in video playback and retrieving of any sequentially placed data. Moreover, with a new breed of hybrid HDDs that employ larger flash memory caches for frequently accessed information, users would be able to benefit from the best of both worlds (with greatly improved random/sequential access times as well as higher storage capacity).

A prime disadvantage of SSDs is the limited number of rewrites to memory. This makes it mandatory to perform wear-leveling within the SSD to prevent repeated rewrites to the same sections. This issue is expected to make recovering lost data from SSDs far more difficult than is the case with HDDs.

So far, the biggest advantage that I see favoring SDDs in portable computers, is the resilience from shock and damage arising from dropping the device. Therefore, computers built for rugged use and/or harsh environments would strongly benefit from using SDDs, if the compromise on storage capacity is acceptable.

The M.A.R.S. (Multi-Appendage Robot System) is a robotic instrument designed in conjunction with the NASA JPL to be used in zero gravity to walk outside the space station (to perform maintenance tasks), or next-generation space exploration tasks.

For traction, the robot cannot use magnets (because of aluminum or ceramic surfaces) or suction cups (due to lack of air). Therefore, its feet are covered with microscopic hair with adhesive properties not too different from the appendage system of geckos. According to this BBC article, the little lizards have tiny hairs and pads on their feet that produce electrical attractions, literally gluing them down to any kind of surface - even polished glass.

The hexapod robot can perform six-dimensional range space calculations in order to maintain balance in space. LabVIEW RealTime software enables the robot to perform this control analysis in the vicinity of 25 nanoseconds.

Click here to download a video (37MB) of the M.A.R.S. robot demonstration.

The video also briefly talks about contact force calculation for a three-legged, winch-supported robot that can tackle adverse vertical terrain.

Here is a paper (pdf) by Dr. Dennis W Hong (who also presents the video demonstration) that goes into further detail about the NASA JPL LEMUR IIa design (that the MARS robot is based on), along with references to other robotic gait systems.

(via EngineeringTV)

‘String Ducky’ is the winning submission for a recently concluded user-generated-video contest to present String Theory in Two Minutes or Less. The official winning video (chosen by Columbia University physicist Brian Greene) was created by Sandy Chase, a Science TV producer from New York city.

I caught this significant bit of news on MPR while driving home today - the 2007 Nobel Prize in Physics has been awarded to Albert Fert of the Université Paris-Sud in Orsay, France, and Peter Grünberg of the Forschungszentrum in Jülich, Germany, for their discovery of the phenomenon known as giant magnetoresistance (GMR).

GMR was coincidentally and independently discovered by the two physicists in 1988, which rapidly led to the science of spin-based electronics, or ‘spintronics‘. Spin-valve GMR is the physical phenomenon that makes gigabyte hard drives a reality today, and the prospect of Magnetoresistive RAM (MRAM) more promising.

According to this Ars Technica article:

While the GMR phenomenon was only discovered 20 years ago, it has already found many practical applications, mainly in the read heads used in high density computer storage. Other uses are still in the development phase; nonvolatile, low-power, high-density magnetic random access memory (MRAM) that is based on GMR materials may be the successor to DRAM that is found in most PCs today. The applications of this technology are still in their infancy, but some suggest that materials that exploit this phenomenon could eventually lead us to practical optical or quantum computers.