On The Move

 

MIPI Interoperability and Conformance: Interview with David Woolf (Part 2)

We are back with part#2 of the interview with Mr. David Woolf, Senior Engineer at UNH-IOL, which actively engages in interoperability and conformance tests for various MIPI protocols. If you missed the first part of the interview you can find it here.


Question: David, what could go wrong if a cell phone integrator uses a device that did not pass MIPI protocol interoperability?

Answer: An untested device is an unknown. The goal of interoperability workshops is to be a frontline for finding interoperability problems and erase some of those unknowns. I think that the biggest problems would crop up during ‘in house’ interoperability testing in the labs at the integrator companies. When they come across an interoperability problem they need to double check it, go back to the supplier, ask for a fix, try again. If they are constantly dealing with interoperability problems during integration, that’s going to slow down the validation process.


Question: Please explain why MIPI protocol conformance is important? Can you give an example?

Answer: Although our ultimate goal is interoperability, conformance is the foundation that interoperability is built on. If products ignore key aspects of the specification, we won’t have interoperability. Conformance also is what can provide margin in a design. For example, in the lab a receiver may be able to decode a non-conformant signal correctly. Maybe the amplitude is too low, or there is some other conformance violation. However if that non-conformant signal is sent in a noisy design across a degraded channel, interoperability slips away. So, it is key that both sides of a link are conformant.


Question: Can you comment on the growing Importance of hardware testing? and give an example of costs and what could happen if you don’t do it right.

Answer: Everyone knows that SoC designs are getting more complex, with more interfaces. Statistically, the chances of a bug popping up in the field are increasing, unless we correspondingly ramp up our test efforts. A bug being found by a customer has massive costs not only in re-spinning a design but in damage to the companies reputation. I’ve had several customers come through our lab just a few days or weeks before tape out, eager to find any bugs because of the cost of trying to fix them after tape out. At the same time, I’ve had the pleasure of seeing products come through the lab several times with incremental improvements. When I see that, I know that the testing we’re providing is helping to refine products early on, and ultimately save time.


Question: As you interact with test engineers from leading mobile IC vendors, can you comment on the general care abouts and challenges they face?

Answer: We’ve heard it said that time is money. But there are times lately when I’ve seen that for many companies time is more important than money. More and more people care about speeding up the test time and getting the testing done faster. Delays are costly.
Next, they care about automation and reproducibility. That makes sense, because often companies are trying to reuse IP. Automation that they create for one design will save time when they reuse that IP.


Question: Interesting point, can you elaborate on cases where time is more important than money? Do you mean time to design the IC ?

Answer: Time to validate. Once a design is ready, it needs to be tested thoroughly as possible and as fast as possible. When you think of all of the interfaces that are on a mobile application processor, the matrix of possible tests is massive (supply voltage, temperature, process variation). Because of this, companies are investing in ways to get this testing done faster. That could be by buying designs that are already tested from IP companies, working with 3rd party test labs, and automating testing.


Stay tuned to the next post with more Answers from David.
If you have further questions you can send to both of us using this page and we’ll try to answer within a reasonable time.


If you like to read this or other previous posts, send this URL to your friends and tell them to Subscribe to this Blog.
To SUBSCRIBE
One option to subscribe is as follows:
• Go into Outlook
• Right click on “RSS Feeds”
• Click on “Add a new RSS Feed”
• Paste in the following “http://feeds.feedburner.com/synopsysoc/sqLy?format=xml”
• Click on “Accept” or “Yes” or whatever the dialogue box says.

  • del.icio.us
  • Digg
  • Facebook
  • Google Bookmarks
  • Print
  • Twitter
  • StumbleUpon