posted by Bryon Moyer
One of the defining characteristics of an embedded system is that you should have no expectations about what it’s made of or how it’s arranged. There are no architecture standards, and that’s how everyone likes it.
Well, ok; not everyone: the poor dudes writing tools for embedded systems have a heck of a challenge dealing with all the variety. And, frankly, some of those tools come full circle and help architects decide how to optimize their systems. But if each variant takes a major project to configure the tools, then that’s not going to work.
Of course, we could try and standardize hardware architectures…
Yeah, good luck getting that one to go anywhere.
Instead, there’s a middle ground being explored by the Multicore Association: it’s called SHIM, which stands for Software-Hardware Interface for Multi-many-core. The idea is to give software tools a way to discover the hardware configuration via an XML file.
This is one of those projects where “restraint” is the name of the game. It would be really easy for something like this to get out of control and far exceed its scope, but the folks driving this – in particular Masaki Gondo of eSOL – are taking great pains to define what this is and isn’t.
For instance, it’s not a complete hardware description of everything in the system. It’s restricted to documenting hardware that matters to software, and it describes the hardware in a manner that makes sense to software (unlike IP-XACT, which is intended for hardware designers). Things like defining the type and number of processor cores, synchronization mechanisms, inter-core communications, memory architecture, interconnect, and virtualization scheme.
They also take pains to ensure that this is not a functional hardware model – you’re not going to plug it into some simulator and have it work. It’s just a description. It’s also not a tool in and of itself; it’s a format for data that can be consumed by tools that others create. So there’s really no threat to anyone in the ecosystem.
It’s partly intended to allow performance estimation of a given architecture, but it won’t be 100% cycle-accurate. It will help with the creation of – but will not auto-create – hardware abstraction layers.
The specific news here is that a working group is starting up to define the details; the first spec should be out sometime next year. You can find more info on the effort and how to participate in their release.
posted by Bryon Moyer
We’ve seen a number of different ways in which magnetic interactions with electron current can be put to use thanks to the concept of spin. Those magnets are also conductors, so electrons are moving through materials having various (or no) magnetic polarization.
You might wonder why I went through the trouble to specify “electron” current. I mean, that’s what current is: a flow of electrons. Right? Well, it turns out there’s another more subtle current. Or perhaps better to say pseudo-current, since no actual object is moving. You can have a spin current too. Who knew!
This is really more of an “influencing” thing: the spin of one electron can be transferred to its neighbor, thence to another neighbor, and so forth. This is spin (or spin torque) transfer. So you’ve got this spin alignment thing going on, and it flows out from wherever it started. Think of it as spin going viral. So it acts like a current, even though it’s only the influence that’s moving; the electrons themselves aren’t.
Actually, the electrons can be moving while this happens; they’re just not moving in the same direction that the spin is. This came up in an article about work done at Japan’s Tohoku University to identify yet another type of spin interaction, which they call Spin Hall Magnetoresistance. While the previous types of magnetoresistance we’ve seen involve currents going through the magnets, this involves insulating magnets adjacent to an electron-current-carrying metal.
The fundamental idea is that spin from the current can transfer into the magnet even though the electrons themselves can’t move into the magnet. When that happens, it’s sort of like an energy leak from the wire, and it reduces the current in the metal, which, given a constant potential driving the wire, makes it feel like a higher resistance.
There are apparently two things going on that make this work. One is a polarization of spins in the wire, analogous to the charge polarization due to the normal Hall effect. In this case, instead of getting opposing charges on each side of the wire, you get opposing spins. So the side that abuts the magnet will have an accumulation of electrons with a particular spin.
The second piece has to do with what they call scattering, but fundamentally, they found that the spin at the metal/magnet boundary can only transfer into the magnet if the magnet is polarized perpendicular to the accumulated spin. In other words, you can, in theory, modulate the apparent resistance of the wire by changing the field direction of the adjacent insulating magnet.
Granted, it’s a small effect: the change is around 0.01%. But it’s yet another mechanism that might have promise for… something. And, fundamentally, it’s the first I’ve run up against this concept of a spin current. So it piqued my interest on that score too.
If it’s piqued yours, you can get more detail in this Physics article.
posted by Bryon Moyer
Epson has recently made a series of announcements in the IMU space, including a new V-series that they claim features the “world’s smallest IMU” (defined as “The smallest IMU among high-performance IMUs having gyro bias instability of 10 dph or less (as of the beginning of August 2013, according to Epson's research)”). That would be 10x12x4 mm.
Why are they not comparing themselves to the silicon guys? Because their fundamental sense element material isn’t silicon; it’s quartz, branded as QMEMS. At least for the gyroscope, which they make. The accelerator in the IMU comes from someone else; they’ve not disclosed that partner. (And they say that no one has yet successfully made a quartz-based accelerator.)
So what’s with this quartz stuff, anyway? And isn’t quartz simply SiO2, like the SiO2 on silicon chips? Well, yes and no. It is SiO2, but it’s crystalline – the dielectric layers on chips are not. They grow a specific quartz crystal – the fabrication process includes an anneal that lasts for months in a 30-m-high autoclave. The cut they make off the crystal also determines the temperature stability.
The benefit is a high-quality (i.e., high Q), stable output that requires less conditioning than a silicon element would need. In fact, they say that silicon folks sometimes use redundant elements and average them to stabilize the result. This is also different from Qualtré’s approach, which uses a resonating disk: Epson’s element vibrates, but not at resonance.
Of course, this isn’t going to compete with the gyro that’s going into your phone when it comes to price. This is a more upscale version for use where accuracy commands a premium. One application they mentioned that can take advantage is SATCOM OTM – “on the move” – used for maintaining a satellite linkup while moving. The IMU is needed for fine alignment.
They’ve announced a G-series of other industrial IMUs as well that use their older technology. The difference with the V series is the new sense element – they say they can fit eight of the new ones in the space required by one of the old ones (which they claim was already pretty small). They also don’t need mechanical isolation or vacuum chambers like some larger high-performance IMUs need.