MikroElektronika Forum

Anouncing the development of FT90x compilers

MikroElektronika starts development of compilers for FTDI Chip's new 32-bit FT90x family of microcontrollers.



After the award-winning FT800 graphics controller, FTDI Chip's team of engineers have now added another piece
of the puzzle to their portfolio—a powerful 32-bit microcontroller called FT900. And boy, they did one hell of a job!

FT90x's 32-bit RISC architecture is built completely from scratch. The innovative IP enabled our friends from FTDI
to get the very best performance in the industry—a staggering 2.93 DMIPS/MHz! Just to give you a hint—that's
more than double the performance of ARM Cortex-M4. First chips feature 256KB of Flash and 64KB of RAM, with
true zero wait-states operation up to 100MHz. FTDI guys are specialists when it comes to USB and fast data throughput
so they've supported the highest data rates in all integrated interfaces. One of the conveniences is a parallel camera
interface and a host SD controller.

Fred Dart, FTDI Chip's CEO and Lee Chee Ee, Director of Engineering have visited us in Belgrade a few weeks ago
and we really enjoyed their company and being their hosts. We've shown them couple of Belgrade landmark sites,
and had a good time tasting traditional domestic food and drinks. On Saturday afternoon, after cheering with delicious
beer from our own beer tap, Fred and our CEO, Neb, signed an agreement on our terrace, accepting the deal that
MikroElektronika develops mikroC, mikroBasic and mikroPascal compilers for FT90x microcontrollers.



Be prepared to experience the fastest microcontroller on the market and see demos at Embedded World 2015 in
Nuremberg, Germany. We'll build the entire ecosystem, including development boards, examples for click boards
and support in Visual TFT software. We can't wait to get started. Make sure to follow our news and catch on the
updates we're going to share with you in the process. It will be one exciting autumn!

You can read the official press release here

Yours sincerely
MikroElektronika

106 views and no reply?!?
What's going on?!? ... seriously

Okay ... I'll write a reply
These news are like ... like ... the bomb!
Although I feel very sorry for my poor wallet (pun intended), I gotta say that I'm very excited about the products to be released!

Congratulations to MikroElektronika on this huuuuuuuge deal!
I believe that both parties are winners
However, both of them aren't the biggest winners here. We, the costumers, are the biggest winners!
We'll get to use these awesome new MCU's while they're mounted on even more awesome development systems, while writing the programs using compilers that are so awesome that you think they're not real!

Man! It's going to be one exciting autumn indeed
I will be following these news because they're way too exciting to be missed!
Keep at it MikroElektronika! You're ... You're ... too awesome that words cannot fully describe you!

Best Regards

Speed isn't everything. Remember Ubicom (Scenix). They made a 100MHz PIC compatible MCU which was 10 times faster than PICs of the time. Ten years or so later, where are they today (bought over by another small company, Parallax Inc., developers of the STAMP).

Well it all good and fine however in my experience with mE over the last 6 years is that they cannot even support properly the compilers they have and then they keep on adding new ones
with the result that the support for the existing compilers go backwards even more
any case good luck

I also feel the same as Peter. 2014 is by far the worst year in compiler update at mE. Only 1 compiler (PIC) has been improved very slightly during all the year. No new chips were supported by mikroProg although almost everyday request were made for adding new chip to the existing compilers.

And now they announce why all the team have left the existing product, it's for working on something new. Unfortunately this seems to be a corporate culture at mE that repeats over and over again.

I have to agree with Peter and Toley. I guess now we can expect updates to MikroProg and MicroC on a 3 year timetable? I hope they hire more devlopers to handle the addtional workload obviously the existing individual is overwhelmed.

Again, its been almost 8 months since you updated MikroProg for PIC, numerous request throughout the forum to add PICs, some of which have been out over two years. When exactly do you plan on doing something about this? If you are going to be charging top dollar for the programer the least you can do is keep it up to date.

If it isn't obvious, no I won't be buying the new FTDI compiler, I no longer trust any of your products.

Toley wrote:I also feel the same as Peter. 2014 is by far the worst year in compiler update at mE. Only 1 compiler (PIC) has been improved very slightly during all the year. No new chips were supported by mikroProg although almost everyday request were made for adding new chip to the existing compilers.

The MikroC for PIC only had one update in 14 months as was already pointed out to you here

http://www.mikroe.com/forum/viewtopic.php?f=88&t=60649

Here is the relevent snip.

filip wrote:Hi,
The time gap between the last two updates of the PIC compilers was more than a year, and the number of ARM compiler updates in the 2013 was 6.
Regards,
Filip.

Hi All.

Yes I am excited by this announcement particularly that the FTDI website states that the FT900 will be faster than the ARM Cortex-A9. I have a large project plan that will require video image processing and this might me suitable in future. I am currently working on a big project with the FT800 graphics controller and this product works great!

I am however concern as to whether this is a good business decision for mE though. The FT900 Brochure is boasting a Eclipse plug-in. Will this be free or too much of a cheaper alternative to mE compiler? Will FTDI be producing libraries for this plug-in? Also FTDI is marketing that the FreeRTOS will be easily integrated and we know that this has not been ported on mE compilers except for mikroC STM32 M3. http://www.libstock.com/project_categor ... nt-systems
For this processor power you'll more than likely be considering an RTOS or uClinux.

It is a worry of a lot of us that mE is drifting from core products and spreading itself too thin. Maybe a slightly different business model will work, like a fee for existing compiler plug-ins if there is an issue of hiring additional engineering resources... Anyway a more important point to be made; is why is everyone at the table drinking beer except the mE manager? Coke is bad for you, beer is good. Trust me I'm Australian.

Andrew

Hello Andrew

I totally agree with you on one point. "Which one?" you may ask. Well ... I mean this one:

AndyIvan wrote:why is everyone at the table drinking beer except the mE manager? Coke is bad for you, beer is good.

This isn't always the case. Their CEO doesn't always drink bad stuff such as coke, he also drinks healthy stuff as "tea" too
You can see it at the very end of this video:
http://www.youtube.com/watch?v=YiG6tzkxE3U

By the way, I've just realized that the chosen colour of the new compilers makes them look like they're related to V.TFT software

Best Regards

FT900 is one of those MCU where we want to be there from the start, so we are happy to work together with FTDI and it will have
just plus effects on the SW of mikroE and all rest compilers, ofcourse there will be small drawback but it will help us grow.

Things that were and are affecting SW department so there is smaller number of release this year are:

1. We are working (for some time) on two major upgrades that will be included in all compilers this are longterm projects.
We have and will employ more people to cover this and ongoing activity, but we must not add to many new people because in start they are slowing things down.


We had plan to have in best case 2 release of compilers this year we will probably have just one per architecture

For now there were dsPIC, minor 8051 release and PIC release.

This is the plan for rest of the year

we have prepared PIC32 release with support for PIC32MZ (but errata for MCU is rather big so it slowded things down), it is not always all up to us.
we are about to relase ARM with support for about 130 new MCUs.

At the end

Simply there was no choice, if we wanted to make big step forward we had to put major part of strength in those long therm projects.
We are aware that for our customers time in between seems like gap so we will employ more pepople and focus on constant updating, until
mayor updates are ready.

And we want to keep it quite about that not to spoil suprise what and when it will be.

fred85 wrote:I have to agree with Peter and Toley. I guess now we can expect updates to MikroProg and MicroC on a 3 year timetable? I hope they hire more devlopers to handle the addtional workload obviously the existing individual is overwhelmed.

Again, its been almost 8 months since you updated MikroProg for PIC, numerous request throughout the forum to add PICs, some of which have been out over two years. When exactly do you plan on doing something about this? If you are going to be charging top dollar for the programer the least you can do is keep it up to date.

If it isn't obvious, no I won't be buying the new FTDI compiler, I no longer trust any of your products.

The MikroC for PIC only had one update in 14 months as was already pointed out to you here
http://www.mikroe.com/forum/viewtopic.php?f=88&t=60649
Here is the relevent snip.

Yes, I agree but is there others solutions ?
I sometimes start looking for other Pic dev products...

Hello Rajkovic
First of all, I thank you so much for taking the time to make your post
I think we all know that the team is working very hard, so both of your hard work and your time are very appreciated

I'm quite satisfied with the performance of your softwares, but I can understand why some people don't feel the same way since they need to use latest MCU's on the market in their work.
Adding the FT900 compilers to your collection is absolutely a very huge move!
So you cannot blame the people if they started saying that you are slowing down on supporting their favourite MCU's to support another set of MCU's.

I have a suggestion here that might solve this issue.
I know that the costumers will be very satisfied if there were very small and very frequent updates instead of the long waited huge updates, so please consider dividing your single update into several updates. I'm sure the results will be much better than you think they would

And Rajkovic, please remember that I'm not saying I'm not happy with the way things are done, I'm only saying that there might be a better way to consider

Best Regards

Kalain wrote: Yes, I agree but is there others solutions ?
I sometimes start looking for other Pic dev products...

Look at any mainline distributor like Newark, Digikey or Farnell also Tindie or if you want really cheap, Asia like dealextreme or tme.edu etc.. fleaBay. Their is no shortage of DEV boards out there. You could also just custom make your own on vero board, home etch or use a cheap Asian PCB house like ITEAD.

Right now ARM is a really competitive market so all the manufacturers have really cheap DEV boards and some form of either free or cheap tool chain. If you go to EEVBLOG forum and search the microcontroller section there are lists of compilers, dev boards and such for ARM.

MaGiK wrote:Hello Rajkovic
I'm quite satisfied with the performance of your softwares, but I can understand why some people don't feel the same way since they need to use latest MCU's on the market in their work.

Some of the chips being requested have been out for almost two years maybe more.

As far as 8b PICs I’m not interested in 1628,887 type chips they are old and more expensive then newer 1708-9 and 1713 and a host of others in the enhanced portfolio. MC seems to have corrected most of the bugs they were having with their analog peripherals on their PSMC series, in these newer chips. The large offset errors present in the PSMC chips opamps isn’t present on the newer chips mentioned.

I was at one point considering the ME ARM compiler, but with the experience with the 8b PIC compiler and MikroProg over the last 3 years or so, I think I’ll pass and stick either with CoCoox or the manufactures free but limited compilers. Now that I see they added another compiler, I think I made the right choice. What happened to their PIC compiler once they started an ARM compiler is probably now going to be happening with their ARM compiler now that they are starting an FTDI compiler?

fred85 wrote: What happened to their PIC compiler once they started an ARM compiler is probably now going to be happening with their ARM compiler now that they are starting an FTDI compiler?

Exactly the ratio last year was 6 to 1 in the favor of the ARM compiler PIC32 did not see 1 upgrade and now mE says they will release and PIC32 upgrade and an ARM upgrade this the ratio is 7 to 1 what will the ratio be for the PIC compilers with FTDI included

The story op employing more people we hear in this forum since 2007 and still not done

MaGiK wrote:Hello Rajkovic

I have a suggestion here that might solve this issue.
I know that the costumers will be very satisfied if there were very small and very frequent updates instead of the long waited huge updates, so please consider dividing your single update into several updates. I'm sure the results will be much better than you think they would

Best Regards

Another idea would be to discontinue Compilers and languages that are not paying for themselves and turning a profit. I look at the forum and all the languages and compilers and I can’t help but notice a lot of them get little traffic. Right now there are maybe three compilers that actually seem to be popular, these are likely subsidising the other less used ones. What you could do is offer an incentive to transition to a more popular compiler and then gradually get out of the older less active ones. When you get spread to thin you wind up doing things half ass. Jack of all trades and master of none comes to mind.

To maintain all those compilers is obviously way too much for your company in its current state. Add the flash programmers and that’s quite the workload.

I also wouldn’t be against paying a subscription in order to keep it up to date. Something along the lines of buy the compiler for the current price include 3 years or so of free updates then maybe charge 50 bucks or so a year to be able to update. I wouldn’t be willing to go much more then that, to many other options out there now, but something along those lines.

My 2 cents