Less new projects?

Non-technical related questions should go here.
User avatar
Bianco
XCore Expert
Posts: 754
Joined: Thu Dec 10, 2009 6:56 pm

Re: Less new projects?

Postby Bianco » Tue Sep 27, 2011 2:36 pm

What are the sales prospects for a DIP package? Think about it.
The investments for a multi-chip-module DIP package including flash, voltage regulator etc is huge.
You would need to sell at least 200k to get ROI. And that's not easy: arduino even sells less than 100K a year i believe.

It would be a unicum to have a 400-500MHz processor in a DIP package. I think we should already be happy to have a 400-500 MHz processor in QFP packages. Even that is already almost a unicum. I only know a few in the same class (ADI Blackfin has one, and there are a few vendors (at least 2) with ARM9/ARM11 in QFP (be it with 200+ pins)).

You do not need to connect the ground pad in the QFP-128. It has enough ground pins to ensure a proper return path and heat isn't much of an issue. I see that the QFP-64 and 48 do not have ground pins. That makes things a bit harder :).

I believe that XMOS should be focussing on the more advanced users for the upcoming years.
The only downside from not supporting arduino-level users is that there is less "spreading of the word".
But that is also a tradeoff between investments. Are these users that want to use DIP packages likely to generate massive revenues by integrating XMOS chips in mass products? I doubt it.
User avatar
Folknology
XCore Legend
Posts: 1274
Joined: Thu Dec 10, 2009 10:20 pm

Postby Folknology » Tue Sep 27, 2011 2:52 pm

What are the sales prospects for a DIP package? Think about it.
The investments for a multi-chip-module DIP package including flash, voltage regulator etc is huge.
You would need to sell at least 200k to get ROI. And that's not easy: arduino even sells less than 100K a year i believe.
I'm not dictating if it is financially viable or not it would clearly have to be a long term investment, sticking with the safe and small expert market is much cheaper and it is obviously where Xmos currently generates income. I assume that you also feel that way as do many other experienced Xmos folk.

Frankly I don't understand why Xmos want such a magical technology confied to a tiny bunch of Xmos ninjas, if it were my organisation I would want everybody to use it for a force of good to change the world, but hey that's just me!!

And besides getting more users and projects and making the whole thing more accessible was the theme of this thread hence my comments, maybe I'm alone on this crusade ;-)

P.S. I would also add that you should not be aiming to sell less than Arduino but rather significantly more :twisted:

regards
Al
User avatar
phalt
Respected Member
Posts: 298
Joined: Thu May 12, 2011 11:14 am

Postby phalt » Tue Sep 27, 2011 3:05 pm

Folknology wrote: Frankly I don't understand why Xmos want such a magical technology confied to a tiny bunch of Xmos ninjas, if it were my organisation I would want everybody to use it for a force of good to change the world, but hey that's just me!!
I think everyone could benefit from being an XMOS ninja!
And besides getting more users and projects and making the whole thing more accessible was the theme of this thread hence my comments, maybe I'm alone on this crusade ;-)
Oh I want this too, that is why the thread started! Your input was really handy as was everyone elses. I'm writing all this down for when I work on building the new competition and updating the site so people understand what sections are used for.

As for new dev kits and stuff, my powers are pretty weak, but we've had people with those magical powers reading your thoughts.
believe that XMOS should be focussing on the more advanced users for the upcoming years. But that is also a tradeoff between investments. Are these users that want to use DIP packages likely to generate massive revenues by integrating XMOS chips in mass products? I doubt it.
XMOS overall goal is to have our chips in products world wide, so there is tendency to target just the big companies, which as you quoted is why there is:
less "spreading of the word".
between hobbyists for XMOS and probably why our community isn't as big as others like arduino because they target hobbyists.
User avatar
Interactive_Matter
XCore Addict
Posts: 216
Joined: Wed Feb 10, 2010 10:26 am

Postby Interactive_Matter » Tue Sep 27, 2011 3:42 pm

phalt wrote: XMOS overall goal is to have our chips in products world wide, so there is tendency to target just the big companies, which as you quoted is why there is:
less "spreading of the word".
between hobbyists for XMOS and probably why our community isn't as big as others like arduino because they target hobbyists.
I think this is a point where I think you (as in XMOS) are terribly wrong.

This community isn't here for the fun (OK, it is fun too, but there are also other aspects):
  • a bigger community can spread the word of mouth how cool XMOS technology is - free marketing
  • you can get more people in touch with XMOS technology more easily - convincing people to convince their employer to use XMOS chips
  • the community has some vital interest in enhancing and improving the code base for the XMOS building blocks, may even add new building blocks - good arguments for your sales stuff, faster time to market for your customers
  • perhaps some of us build a successful product based on XMOS technology - migrating from a hobbyist to a customer (in the target market)
And I think, sorry folknology, that there is no real reason/market to develop a PDIP version of the XMOS chips. I really feel the need to have a small XMOS module for my projects since the layout & power supply is rather complicated. I personally think a prefabbed module would do the trick. There have been some efforts to realize this here and in the XMOS company - but somehow this never turned into something I can buy somewhere. The XK1-A is from my point of view too big and wastes too much precious pins on buttons & LEDs - and still a tad too expensive.
User avatar
Folknology
XCore Legend
Posts: 1274
Joined: Thu Dec 10, 2009 10:20 pm

Postby Folknology » Tue Sep 27, 2011 3:53 pm

And I think, sorry folknology, that there is no real reason/market to develop a PDIP version of the XMOS chips. I really feel the need to have a small XMOS module for my projects since the layout & power supply is rather complicated. I personally think a prefabbed module would do the trick. There have been some efforts to realize this here and in the XMOS company - but somehow this never turned into something I can buy somewhere. The XK1-A is from my point of view too big and wastes too much precious pins on buttons & LEDs - and still a tad too expensive.
It will be interesting to see the size and how much such a module would be to purchase, but it will be many magnitudes more expensive than a PDIP or SOIC from the end user POV. What do you consider a good enough price out of interest and who is it marketed at?

PS I prefer SOIC (no holes required and smaller) to PDIP for prototyping (I rarely use PDIPs) and SOICs can be soldered easily by hand
PPS the trouble is with with current Xmos TQFP/LQFP with Paddle chips is that one often has to go for stencils making prototyping expensive and time consuming, that often leads to less experimentation (compared to other MCUs like say Pics,PSOCs,Stellaris etc..)

regards
Al
User avatar
Interactive_Matter
XCore Addict
Posts: 216
Joined: Wed Feb 10, 2010 10:26 am

Postby Interactive_Matter » Tue Sep 27, 2011 4:16 pm

Folknology wrote:
It will be interesting to see the size and how much such a module would be to purchase, but it will be many magnitudes more expensive than a PDIP or SOIC from the end user POV. What do you consider a good enough price out of interest and who is it marketed at?
I think it should cast less than 20$ - better less than 10$ and I would try to market it a anybody wanting to throw an XMOS into a project. Something very similar to the minimalist xmos board - even smaller and produced completely (nothing you assemble at home).
Folknology wrote: PS I prefer SOIC (no holes required and smaller) to PDIP for prototyping (I rarely use PDIPs) and SOICs can be soldered easily by hand
PPS the trouble is with with current Xmos TQFP/LQFP with Paddle chips is that one often has to go for stencils making prototyping expensive and timeconsuming
TQFP is doable … but often a PITA - and the bottom pad makes it more or less unusable.
I have seen some nice boards with half holes at the edges (don't know the correct name for this) which can be either soldered to pads or to pins.
User avatar
Bianco
XCore Expert
Posts: 754
Joined: Thu Dec 10, 2009 6:56 pm

Postby Bianco » Tue Sep 27, 2011 4:17 pm

Would there be any interest in a barebone pcb module?
You will need to solder it yourself and source the components yourself.
Although i might be able to supply a part or all of the components, depends on the interest.
User avatar
Interactive_Matter
XCore Addict
Posts: 216
Joined: Wed Feb 10, 2010 10:26 am

Postby Interactive_Matter » Tue Sep 27, 2011 4:41 pm

Bianco wrote:Would there be any interest in a barebone pcb module?
You will need to solder it yourself and source the components yourself.
Although i might be able to supply a part or all of the components, depends on the interest.
For me personally: yes (send PM)

General with a permissive license as building block to integrate it into my projects as 'copy this to have an XMOS in you pcb': yes (have you published it as project yet?)

As an offering for the general public: It should be smaller (ok, have not seen it) and come completely assembled and for cheap in a dip like package (to be used on a breadboard as well) for next to nothing ;)
User avatar
Folknology
XCore Legend
Posts: 1274
Joined: Thu Dec 10, 2009 10:20 pm

Postby Folknology » Tue Sep 27, 2011 4:45 pm

Well I did design Amino for my own use based around the L1-64. The idea was to also build suitable plug in cards for it. However even this proved overkill in many cases and difficult to integrate with my other projects. I am now working on a second version that plays much better with some of the other components and peripherals that I am working with (including other platforms).

Do you have a design we can look at Bianco?

PS I don't really use breadboard stuff myself as it often causes to many problems at these higher clock speeds.

regards
Al
User avatar
dan
Experienced Member
Posts: 102
Joined: Mon Feb 22, 2010 2:30 pm

Postby dan » Tue Sep 27, 2011 4:55 pm

Interactive_Matter wrote:Moved over from here: forum/viewtopic.php?p=10066#p10066

In my opinion some additional information (wiki page/blog post - somewhere prominent) could help:

- what is avail on github
http://github.xcore.com/community_doc/repo-index.html
- why is it there
What do you mean with this? The stuff is there because someone wanted to put it there - there isn't a grand plan, this being a community repository. For individual components the reason for its existence ought to be in the README
- how to use the new build structure
If documentation is insufficient for your needs then please file an issue with the community repo.
- how to participate
- how to share so that your code is usable for others too
Are the above not covered here?

http://github.xcore.com/community_doc/

If not, please file issues against the community repo.
some central page listing and explaining all the different modules out there on github could help a lot too. I think a lot of people are perhaps still relying on the software listings on xmos.com http://www.xmos.com/applications/software-components
I can confirm we are working on that. We definitely recognise the absolute need for and value of, an intuitive way to browse the existing solution space in terms of IP and capabilities for all our users, new and experienced, community focussed and commercial.
I find especially the new make file structure hard to understand (it works great and is a good idea - no question about it).
Again, issue an issue...

We are keen to make xcore.com and github as accessible as reasonably possible, but as a community site my feeling is that this should be driven by you guys. For specific, and key, things like the build system that we have contributed I think the onus is definitely on us to document it properly but if there are holes in the community documentation more generally then the community should act to fix that , starting by pointing out specific shortcomings that are actionable by busy xmos engineers or busy community members (whomever gets there first).

This should permit improvements to the community to be made while we at xmos are mainly focussing on creating new or improved tools, chips and IP that makes code run faster while being easier to write.

Who is online

Users browsing this forum: No registered users and 0 guests