In my previous blogs, we discussed how the 8051 microcontroller managed to become one of the most popular controllers ever. We also focused on its architecture and some of the architectural improvements that have been implemented to tailor 8051-based solutions to the 21st century. (See: Is the 8051 MCU Still in the Game? and Redesigning the Architecture of the 8051.)
All of these discussions touched upon another interesting topic -- that of intellectual property (IP) cores in general. As we all know, an IP core is a complex functional block of electronic circuitry. Some IP cores are developed internally by the system design house, while others are acquired from third-party suppliers. In this latter case, the use of the IP core is licensed to other companies by the original designer (there are several sites that provide more information on this, such as WiseGeek.com).
When we peer back into the mists of time to the early ages of electronics (but not as far back as Sven's trousers or Max's ZZ Top haircut), it's easy to see that the IP core licensing approach has changed dramatically over the course of the last decade or two. Ten or 15 years ago, almost everybody was leery about the concept of third-party IP -- for example, engineers simply were not sure if they could trust outside suppliers.
As the years passed by, bellbottom trousers were overtaken by the "slim fit" approach, and things changed in the electronics world also. The combination of shorter product cycles, reduced time-to-market, demands for cheaper products with increased functionality, and the general rapid acceleration in the world of electronics product development increased the pressure on design teams. As a result, the use of third-party IP cores became the new reality for almost every design company. Why "reinvent the wheel" when you can easily use IP that has been successfully implemented in similar projects. Thus, designers started to go outside for processors, peripheral functions such as USB interfaces, and so forth.
These days, of course, a large proportion of the majority of new design projects is implemented in the form of third-party IP, leaving the designers free to concentrate on the "secret sauce" that will differentiate their product from their competition.
OK, so you've decided to take an IP Core from a third-party vendor. And? The use of IP cores can bring you measurable benefits in a variety of ways, including saving time, reducing costs, and minimizing risks. This is if you choose the right cores from the right vendors. The alternative -- if you acquire "bad IP" -- is to waste time, increase costs, and maximize risks. Which option would you prefer?
In order to lay your hand on good IP, there are a number of questions you should ask of your potential IP supplier. Based on more than 12 years of experience in the world of IP cores, please let me be your guide. Some of the questions I think you should be asking are as follows:
What are the deliverables with your IP cores? Will I get appropriate documentation?
Can I evaluate your IP cores before buying?
Do you provide hardware-assisted debugging capabilities?
How long have you been creating IP cores?
Can you show me your IP core portfolio?
Can you show me your customer and success story portfolio?
Are your IP cores silicon proven?
What level of support can I get from you?
How long will it take to tailor your IP to my needs?
Do I know you?
The answers to these questions will be much more helpful than those to the standard question that has been asked since the Jurassic Era: "Do you meet my specifications?"
So what do you think about the topic of IP? Do you find it interesting? Have you already tried to license an IP core from a third-party vendor? If so, what has been your experience?
Jacek Hanke 8/23/2012 2:50:33 AM User Rank Blogger
Re: Build vs. buy
"I can do this as no one is my master any more, other than the wife." I think we all understand what you mean. My friend once told me something very wise: God forgives, wife does never... :)
And re the outsource for 3rd party IP. Some of the stuff will be published in my next blogs (Max? :)) The chain is important thing - as you mentioned, is great if all suppliers share the same high values. As from my poin of view and experience, we're trying to make the chain as concise as possible. Eg when our customer asks for 8051, I can offer him eg DQ8051XP (let's say: All Inclusive version) and bundle it with DoCD hardware debugger. Thanks to it the customer who sahre the same high values, gets everything from one vendor.
But still, you know from your experience, that it doesn't always look like that at different vendors. That's why I decided to start the blog with our "10 commandments" which is based on your, mine and other APP users experiences.
"Come on we're all learning our whole life :)" on that I can agree.
"Do you have the time and desire to solve interesting puzzles".
My whole life has been a puzzle to me. Luckily a lot of my employers let me have my head and allowed me to solve their puzzles.
However when the time frame and cost are collapsing on you and the deadline can not be shifted, then yes it's get the job done.
A further thought occurred to me about the outsource side of Buy, that of secondary and tertiary outsourcing in the outsource chain. Does your primary supplier have as high values for his outsource supply as you do? As part of my 3rd party user role, i was an accredited BS compliance auditor and found, not always, that the mission statement and deliverables got diluted as they went further down the chain. This was often because the secondary outsource had been procured on price, with a rubbish specification from the primary outsource.
I now again have the time and desire to learn what I want to know and play with puzzles that interest me, I can do this as no one is my master any more, other than the wife. I am now able to tinker in my attic work-shop, after these aims.
I am also for hire at an exorbitant daily rate, as I have at best 30 years more before my electron transport chain collapses.
Jacek Hanke 8/22/2012 6:52:49 AM User Rank Blogger
Re: Build vs. buy
Come on we're all learning our whole life :) I just noticed that you're experienced "3rd party user". And re Duane - let's cite classic: "Do you have the time and desire to solve interesting puzzles? Or do you just need to get the job done." ;]
"4. How long have you been creating IP cores? 5. Can you show me your IP core portfolio?"
Sorry to disappoint you, but I am way behind Duane in the learning cycle with FPGA and I only have at the present one failed core to my credit, so failed I am ashamed to show it to my mentor.
Most of my comment comes from being the intelligent? client for the last 30 years, during which I specified original equipment, project managed, outsourced, development & production, ran acceptance testing, then installed, commissioned and maintained delivered equipment.
Jacek Hanke 8/22/2012 5:45:23 AM User Rank Blogger
Re: Build vs. buy
@ Pradeep. There is a lot of web sources which yuu could ask "IP supermarket". For example you can look at the design-reuse.com or chipestimate.com. Just please remember about the factors we discussed here. Some of them will be clarified in my following blogs.
Another way to look at benefits of Buy vs. Build is the user base. If the vendor has a big user base then the errors lurking in the module are more likely to be found and dealt with over a short duration and reach a mature stable condition sooner.
A one off build will typically have only one user exercising the component in one way only, so any inherent errors hang around for the right or wrong time to manifest, the obvious errors come out at the testing stage. The subtle errors usually appear in the field and are a nightmare to track down especially for the field engineer and designer who now have a client breathing down their neck while the fault is found and rectified.
I think this applies to all hardware, software and firmware that I have had to use so far.
Like you I am enjoying every minute in developing my knowledge of FPGA technology, but at the moment if it were mission critical, the expert company with a satisfied user base would get my money.
Duane, the problem is even though we have the time and passion to do the things; company managements are more interested in outsourcing the works to outside. I know some of my friends had faced similar problems in their company and forced to change the employer for challenging works.
Pradeep, now companies have started outsourcing the FPGA/ASIC design works. I know some of the companies like Cadence, creative chips, Nallatech etc are some of the companies doing similar outsourced works. If you are able to contact with them, you might be able to gather more information.
pradeep029 8/22/2012 2:33:51 AM User Rank Beginner
Re: Build vs. buy
Thanks for sharing of the knowledgeable content. I have been hearing alot that there are many companies going for outsourcing of the FPGA/ASIC/other EDA design works. Is there any platform where you can find a list of companies who offloads their design works? I am associated with a new startup and heading the Bussiness Development. We have presence in India but want to go global.
In order to implement the components of a new chip design, any IP core should be delivered with everything needed to design, test, and validate that core in the target product.
If you are interested in licensing a third-party IP core, it's a good idea to ensure that the vendor can also supply you with the debugger. The ideal situation is to obtain the debugger together with the IP core -- all from one vendor.
To save this item to your list of favorite All Programmable Planet content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.
Tweet This
