Decomposition Performance Explorer™ (DPE™)

The Decomposition Performance Explorer™ (DPE)™ lets everyone see how M1™’s SEEj™ RjDj Decomposition measurement will perform on real-life jitter.

dpeslices Since the first RjDj decompositions methods showed up in industry standard specifications, there has been no standard way of measuring RjDj. Even worse, the attempts to measure the performance of RjDj algorithms has been woefully inadequate. For more background information on RjDj measurements, click here.

ASA Corp. has performed the largest jitter validation study in the industry: DAVE v2. This RjDj validation study breaks Jitter Space into ~300,000 parts. DAVE v2 shows that the representative behavior of the M1 SEEj™ decomposition performance to be better than 5%/0.002UI on Pj, 0.0005UI on Rj and 0.001UI on DCD in the vast majority of the ~300,000 jitter cases that were studied. This information was taken from actual runs of data through the algorithm.

But decomposition performance is a complicated issue. To understand if a particular effort in the SEEj™ correlation process represented an improvement or a step backwards, ASA invented a way to visualize results across the enormous jitter space that we are working with. Decomposition Performance Explorer (DPE, or DPExplorer™) is a software application that runs on your PC or scope. Anyone can use it to understand the performance bounds of SEEj across a huge portion of multi-dimensional Jitter Space. It is not only an electronic datasheet for SEEj (to describe it before you purchase), but soon it will be tied into M1 to actually deliver to M1 users a "you are here" mode after each RjDj measurement to let you understand the performance of that measurement.

With DPE, you can identify the characteristic performance of the M1 SEEj RjDj decomposition for many jitter conditions under test. DPE lets you graphically navigate through Jitter Space to determine the representative accuracy that can be expected of SEEj in a given jitter condition. Error plots can be shown for one or all of: Rj, Pj, DCD, or (coming soon) ISI. Each of these can be shown over a very wide range of typical values. You can also use DPE to determine how SEEj can be expected to perform when different kinds of input waveforms are being used. For example, you might be using SEEj on clock signals, on short (20-40 bits long) to long (1,000s of bits long) repeating patterns or even live traffic (random data) waveforms.

Nobody but ASA describes their RjDj decomposition algorithm's characteristic performance over that wide of a range of Jitter Space. ASA Corp. has decided to release the results of its current validation work on SEEj™ to any one. Download the application installer now!

Screen shot of Decomposition Performance Explorer

"I am comfortable saying that ASA has done more work and has provided more innovation in this space than every other competitor combined. We don't just look at a small handful of cases and then decide we're done. It's an ongoing quest for the very best possible solution. Calibration... neural nets... visualization tools... wide-area validation... NOBODY but ASA is giving you this kind of science or innovation. And NOBODY but ASA is giving you a way to get THE SAME method on all your scopes, no matter who made them. THAT is a genuine correlation strategy and it's only available from one place."

 Mike Williams, President of ASA Corp.