A Tale of Three Start-ups

Actually it was three stories of three semi-conductor start-ups being told at the Cadence VC Forum. The third annual occurrence of this event was held at the Institute of Physics in London.

Jed Hurwitz, CTO of Gigle Semiconductor spoke of how you had to be ‘Just Mad Enough’ – too much and no-one takes you seriously, too little and no-one wants to invest in you. I particularly enjoyed hearing about how he was able to take his experience in one field (video and imaging) and use it to great benefit in another area- home networking. Gigle have just successfully launched a Gigabit powerline modem in June 2009, and Jed spoke about the path that led them there, including advice for up and coming start-ups. He reminded all when coming up with their product ideas to ask themselves ‘Could someone else do it?’ – he answered this well saying yes anybody could do it, but ‘Can someone do it?’ is a different matter. No-one else may be in the right position today to do what you can.

Next we had ‘The Challenges of a 21st Century Start-up’ presented by Mirics Semiconductor’s CEO Simon Atkinson. He talked about Mirics road to success and the course corrections that they had to go through to ultimately be successful. Today Mirics have the only demodulator that covers all the worlds standards for broadcast TV – they do this with a combination of dedicated hardware and software running on the main PC or laptop CPU. Also interesting was hearing about the hardware acceleration for the next generation by using the GPU. Afterwards we chatted to Simon and discussed power utilization due to using non-dedicated hardware. While the power consumption is higher with the general purpose CPU approach, it is nearly as low as dedicated hardware (only a couple of minutes less TV watching). However, more importantly, the power consumption fits within the threshold that a consumer is happy with.

The last start-up story was about Phyworks – their CEO Stephen King talked about the ups and downs of the company. Going through many iterations and lots of fund-raising in difficult years, they also eventually found their way. Stephen King also shared his dos and don’ts of which the most memorable was his sentiment of ‘pushing people’ to get the most out of them, whether it was employees, partners or customers!

It was a well organized event, in a great location. The food on offer this time was a huge improvement over last year. The only downside was that the talks did all overrun which resulted in a very long intense session, and less time for talking over dinner afterwards. All in all well worthwhile and looking forward to next year’s event!

Keynotes & Top Three at ESC

The Embedded Systems Conference, UK was held in a new location of FIVE at Farnborough, home to the famous airshow. The conference keynotes featured two Q&A panels. The first was a set of CEO’s discussing ‘The State of Microelectronics’. They discussed high level issues such as the current state of education in delivering the next-generation embedded engineers. One interesting aspect highlighted by Mark Robson, CEO of Freescale Semiconductors was how the new breed of engineers graduating will be much more suited to collaboration and working remotely thanks to the influence of social networking. Interesting side benefit from the likes of facebook, twitter & co!

The second keynote, ‘The State of Embedded’ was focused more on engineering trends including the old language debate of C vs C++. Niall Cooling of Feabhas mused on the point that there should be more uptake of Java for embedded but there are still barriers out there, particularly with issues such as the non-deterministic behaviour or garbage collection. You can’t very well have your deeply embedded device pause for a moment while GC runs.

Both the keynotes highlighted some interesting discussion points. On the whole, I felt they could have benefited from being a bit more structured and was a bit disappointed that there was not more talk about the future direction of the embedded space.

I attended quite a few of the conference sessions. Some of the sessions suffered due to the presenters attempting too cover too much rather than just sticking to the remit of the talk (after all who wants to keep hearing about waterfall models). Three of the best sessions I attended were:

1. “Implemening a Memory Manager for Small Footprint Embedded Systems” by Michel de Champlain of Deep Object Knowledge.
Michel gave an interesting, technical talk about a low footprint deterministic memory allocator for small embedded processors. I could immediately see how this could be very useful to our clients.

2. “Understanding Quality” by Glennan Carnie of Feabhas.
This was a very well presented session of which the key take-away was the ‘Quality Triangle’. The triangle consists of three measures of quality: Customer perceived, Intrinsic and Compliance-based. Quality of products are measured using one or more of these measures depending on the situation. In our experience, many embedded companies rely on customer-perceived quality. This is okay until the product gets used by a larger set of customers or used for a slightly different purpose. Having good intrinsic quality allows quick and less expensive adaptation to continue to achieve the desired customer-perceived quality.

3. “Tips and Tricks for Debugging In The Trenches” by Greg Davis of Green Hills Software.
This talk included a series of advanced debugging features that are not often used, but could be very effective in the right situation. It was presented by the head of the compiler group so very relevant to me as a compiler-writer. The most interesting feature was the advanced ‘scripted breakpoints’ which allow you to automatically run a few commands every time a certain breakpoint is hit. Its nice to see that GDB is keeping up with the commercial debuggers, now that Python is integrated into GDB.

In addition to those sessions there was also the latest and greatest from many exhibitors to partake in. On a less positive note, while the new location was convenient to travel to, unfortunately the centre did not provide the best venue for the conference talks. The talks suffered from a high level of ambient noise from the exhibition floor. It made listening to the talks difficult and a poor environment for interacting. Fortunately the organizers are well aware of this and will aim to improve on this for upcoming conferences. All in all, ESC UK was worthwhile, and we look forward to a new improved ESC UK next year.

Improving Code – The Open Source Way

The latest CDT offering has just been released. Today CDT is generally acknowledged as the de-facto C/C++ IDE in the industry. The extensive code navigation features are amongst the highlights that the framework has to offer. Not wanting to rest on their laurels, work is already well underway on the next release, CDT 6.1. In a recent blog article, CDT lead Doug Shaefer highlights that there will be a big focus on fixing up the scanner discovery mechanism.

Scanner discovery mechanism is the feature in CDT that supports code navigation by automatically discovering build information such as include paths and build macros. The build information is normally retrieved in one of two ways. The first involves running a fixed gcc command to query the known include paths and macros. The second way involves parsing the build output to identify relevant build information.

This is a great feature that enhances code navigation, and in particular creates a very good start up impression. It means users can navigate to code libraries even before they build a project. In our experience this has been a very desirable thing, particularly when our clients provide specialist libraries such as peripheral driver code. Scanner discovery means their customers can easily navigate the code and link to header files to browse the api for the peripherals. However, as it stands today the scanner discovery mechanism is both architecturally flawed and buggy. We have spent significant time fixing it up for clients in order to give the desired experience. However, there’s only so many times we can patch it up before wanting to help fix it up once and for all.

As a result we will be working with the CDT team to help fix scanner discovery for the next release. In fact we will be part of the ‘rag-tag group’ Doug refers to in his blog. The first step was to have an initial kick-off meeting, followed by all capturing the problem and trying to define the architecture. This has been done using the CDT wiki. While one or two CDT committers have put down details of what the high-level architecture should be, we have added our thoughts on the specific issues we have seen with the implementation and low-level architecture based on our experiences.

The next step will be the interesting one as we work out how best to proceed. In the meantime, Kichwa Coders will also be working on some low-hanging fruit by working on scanner-related bug fixes such as using variables in dialog boxes. All in all, it will be an interesting experience working on a relatively far-reaching feature with several other developers spread out across the globe, yet ultimately very rewarding to be part of fixing up this key feature of the CDT. In particular for us, it means yet another feature our clients can take for granted and keep focusing on their core competences.