editor's blog
Subscribe Now

A New Spin on Logic

Way back in 2008, we took a look at MRAM technology. As a brief review, you may recall that Crocus in particular takes advantage of tunneling magneto-resistance between two magnetic layers. The bottom layer is fixed, or “pinned” and acts as a reference layer. The top one – also referred to as the “free” or “storage” layer – can have its magnetic polarity (or, more accurately, moment) reversed. Selectivity can be improved by engineering the materials so that a current during the write operation will heat the cell and lower the “coercivity” of the material – meaning that you switch that storage layer’s cell without disturbing any other cell. Crocus refers to this as thermally-assisted switching.

With that background (and recommending you to the original article for the details), Crocus has announced what they call a “magnetic logic unit” (MLU). They claim this capability lets them implement a NOR memory architecture, a NAND architecture, or an XOR cell.

They’re still being a bit cautious about the details of how this works, but Crocus’s Barry Hoberman took me though the XOR scenario. Before we can go all the way there, we should take one intermediate step by changing how a cell is read.

Originally, we had a pinned reference layer, and we read the cell by measuring the resistance through the cell. Relatively lower resistance means both layers magnetized alike (or in “alignment”); higher resistance meaning they’re magnetized oppositely (or in “anti-alignment”). So the first step we’re going to take is to remove the pinning. Now the reference layer – also called the “sense” layer, since it helps sense the state of the cell – is magnetically “floating”. Then add some metal lines so that you can magnetize the sense layer as north or south. (To pick arbitrary names for two magnetic states).

To read the cell, first set the sense layer to north and do a resistance read; then, very quickly, switch the sense layer to south and do another read. This is a differential-mode read; whichever resistance is higher establishes the polarity of the storage node.

But here’s where the XOR characteristic comes in: you can ignore the specific northness or southness of the fields. If the two layers – sense and storage – have the same polarity (regardless of what it is), they will run lower resistance; if they have opposite polarity, they’ll have higher resistance. That’s the very definition of the exclusive-OR function (assuming low resistance means 1).

Exactly where all of this will lead product-wise isn’t clear yet. They discuss a number of applications of the NAND and XOR capability, but right now it’s just a technology story. Presumably, staying tuned will give us the rest of the story at some point.

More details in Crocus’s release

Leave a Reply

featured blogs
Aug 14, 2018
I worked at HP in Ft. Collins, Colorado back in the 1970s. It was a heady experience. We were designing and building early, pre-PC desktop computers and we owned the market back then. The division I worked for eventually migrated to 32-bit workstations, chased from the deskto...
Aug 14, 2018
'€œPrediction is difficult, especially the future.'€ '€” Niels Bohr Okay, in my post last week , I revealed that I was a deterministic Newtonian, and my reasoning was about two hundred years old. I posited, '€œIf I could identify all the forces and weights and measur...
Aug 14, 2018
Introducing the culmination of months of handwork and collaboration. The Hitchhikers Guide to PCB Design is a play off the original Douglas Adams novel and contains over 100 pages of contains......
Aug 9, 2018
In July we rolled out several new content updates to the website, as well as a brand new streamlined checkout experience. We also made some updates to the recently released FSE locator tool to make it far easier to find your local Samtec FSE. Here are the major web updates fo...
Jul 30, 2018
As discussed in part 1 of this blog post, each instance of an Achronix Speedcore eFPGA in your ASIC or SoC design must be configured after the system powers up because Speedcore eFPGAs employ nonvolatile SRAM technology to store its configuration bits. The time required to pr...