Who is working with the SAM9263?

I do not know how many devices they do support, but it looks like they should be able to support most of the ARM MCUs from Atmel, NXP, ST etc. by only changing their CPU board. Their prices do not compare with the cheap boards from companies like olimex, but they do compete with boards from the semiconductor manufacturers themselves.

I do believe if Atmel did something similar, one might have to pay US$500 for the first development board, but after that if they made compatible daughter boards available, they should be able to sell them at a much reduced price. Also the daughter boards might actually be suitable for some types of final product boards, and there is scope for third parties to develop different mother boards to take advantage of daughter boards. I am sure that many hobbiests would love to be able to buy a small CPU module that can plug into a dev kit, but can be used on one's own board. Especially for the BGA only devices.

Regards Anton Erasmus

Reply to
Anton Erasmus
Loading thread data ...

I could not get some of these links to work. But the others gave some good results. Cogent has a 520 MHz PXA270 module for only $50 more than the SAM9260 module. Way better deal and a *HUGH* increase in speed! I have been looking for something like this.

Yeah, but there is one thing wrong with it... it uses an AVR32!

Reply to
rickman

AVR32 Board Schematic is here, I believe :

formatting link

-jg

Reply to
Jim Granville

Very good summary of part of my thinking. There are a few more twists and turns though.

--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

Funny, I remember communicating with a guy recently which balked at $600 price for a development kit with Ethernet and $975 for DevKit with Ethernet and LCD screen.

The PXA270 DIMM is $399 DIMM in single qty. If you want to get a single DIMM with baseboard, it is $1499. With QVGA LCD = $1799..

If you need VGA or better resolution, then a dual bus system like the

formatting link
module and the AT91RM9200 is probably faster than the PXA, since you do not steal bandwidth from the CPU and you have several interesting accelerators. The VGA controller, they have implemented in the FPGA collects windows on the fly, so no heavy bitblt in the processor just to move windows, or handle overlapping windows.

The AT91SAM9263, which is likely to soon be available from a number of module companies, also has a dual bus structure so you can have pretty large screens without affecting the CPU performance.

The Ronetix module has a complete frame buffer inside, and if you make all redraws in this for a QVGA screen, you have up to 800 MB/s bandwidth to the memory. The effect is about 2 x performance compared to doing draws to SDRAM.

MHz is not everything.

--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

Funny, I remember communicating with a guy recently which balked at $600 price for a development kit with Ethernet and $975 for DevKit with Ethernet and LCD screen.

The PXA270 DIMM is $399 DIMM in single qty. If you want to get a single DIMM with baseboard, it is $1499. With QVGA LCD = $1799..

If you need VGA or better resolution, then a dual bus system like the

formatting link
module and the AT91RM9200 is probably faster than the PXA, since you do not steal bandwidth from the CPU and you have several interesting accelerators. The VGA controller, they have implemented in the FPGA collects windows on the fly, so no heavy bitblt in the processor just to move windows, or handle overlapping windows.

The AT91SAM9263, which is likely to soon be available from a number of module companies, also has a dual bus structure so you can have pretty large screens without affecting the CPU performance.

The Ronetix module has a complete frame buffer inside, and if you make all redraws in this for a QVGA screen, you have up to 800 MB/s bandwidth to the memory. The effect is about 2 x performance compared to doing draws to SDRAM.

MHz is not everything.

It is a pity that the gateway has an AVR32 at such an outrageous price going up to almost 5% of the price of the PXA270 kit (single quantity). One would expect to get at least a QVGA LCD included for that price, I agree ;-)

--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

Yes, that is $50 more than *their* price, not yours. It is $600 less than your board.

Your LCD is a tiny thing that is of no value to me. I can connect whatever I want to their PXA module with a starting price of $400! That's $600 less than your 9263 board. I can get lots of good LCDs for that price, and they fit *my* needs, not a arbitrary choice. If the LCD is what is driving up the price of your board, maybe you should consider selling the LCD separately.

I am not sure what application you are thinking about, but I don't expect this to be a general purpose computer that has to solve spread sheets while managing a full size display. I seem to remember 386 AT computers managing Windows very adequately as long as you didn't ask it to swap out memory. So I don't think I will need special FPGA hardware to manage a display in a handheld device.

I would like to consider that. But how long will I have to wait for an eval board in a reasonable price range?

No, but it is very important. What happens with the 9263 when I go to a full VGA?

Reply to
rickman

What can you do with a DIMM board without connectors? The equivalent board is the $1799. If you skip the LCD, and buy your own, then it is still $1499. If you are prepared to design your own baseboard, then you can get a chip for free. If you burn a PXA module as part of the baseboard design, then you are close to the cost of the 9263EK Time to market can't be a concern if you are prepared to wait for a working baseboard before you start.

IIRC, the 386 class processors only visibly moved the outline of a window and not the complete window in itself, maybe even with WinCE, the performance needs have increased since then. I know for sure a number of guys starting with PXA, and ending up with an external display controller when going past the QVGA.

If QVGA is OK, then SAM9261 will excel, due to its internal 160 kB SRAM.

Since the AT91RM9200 does not have a display controller, it needs something outside and the FPGA will support large screens like SVGA/XVGA, something I doubt the PXA will handle.

You can't do zilch with the DIMM module, unless you have a baseboard, so you will have to develop your own board. If you dont value time, why not start with a chip?

Since the part is just about to enter production, you will not find anything, but I know there are DIMM modules coming as soon as they can get production volumes. My best guess is May/June.

How fast you can turn around a baseboard for the PXA? What will be the long term cost of using the Cogcomp module? How will that compare to doing your own design?

The part was optimized for a VGA screen 24 bit per pixel. A typical design would use one 32 bit bus (EBI0) for O/S and application and a 16 bit bus (EBI1) for display refresh.

IMHO, The way a decent driver should work is that the 9263 will nmaintain a single buffer to draw the screen in a 32 bit SDRAM on EBI0, with the help of the 2D accelerator (Line drawing, Fill etc.)

Once a screen is ready, it can be copied from EBI0 to EBI1 by the scather/gather memory to memory DMA. In EBI1 it should be double buffered.

If you use SDRAM for EBI1 at 100 MHz /16 bit, your max throughput is 200 MB/s Don't know sustainable throughput but I hope 120 MB/s is not overly optimistic. VGA at 24 bit per pixel is 921600 bytes per frame @ 60 Hz < 60 MB, so you are using less than half the bandwidth to maintain screen refresh and leaving half the bandwidth for DMA. With 16 bit VGA, then you need about 40 MB/s and it becomes realistic to use a PSRAM, with which you get more I/O pins, some of them real useable.

Typically you would not need to update the screen at the same rate as display refresh, and every time this happens, you gain bandwidth.

VGA or no QVGA, the display refresh does not affect the performance of the chip, except for the cost of DMA; and the small latency coming from the need to copy before you display. You could of course at the cost of bandwidth always do display refesh from the

32 bit SDRAM the until the frame is in the EBI1.
--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

no value to me. I can connect

You are such a trip! Again you miss the point. If your board is $600 more because it has an LCD and the baseboard does not, then you need to think about leaving the LCD off and charging $600 less, or are you now saying that it is the connectors that are running up the cost of the module? The PXA is a *much* faster chip for nearly any purpose. You can argue that the 9263 will do this and will do that special thing, but that still does not put it in the same class as the PXA devices. In the end there is no way to justify the high prices of your ARM9 evaluation boards. Remember, Cogent was a company that

*you* recommended in your post, not me. If the DIMM connector board is such a bad idea, why did you suggest it for the SAM92xx devices?

My Windows 2000 running on a 1.4 GHz processor *still* moves the outline, not the actual window, until you let go of the button! I have not used PDAs other than the Palm, but the Palm does not move windows. I think this is a poor choice for a critical function that needs to be accellerated.

To me that makes it a poor choice when using a display. FPGAs are very poor on power consumption and they don't come in anything remotely like a small package. We will see what works, at least we will get the chance to evaluate some processors on boards that don't cost quite so much.

Are you saying the 9263 will only be available on a DIMM? I thought you were producing a compete eval board???

Why are you focusing on the Cogent module? I am asking about *your* eval board.

I hadn't considered the power issues of driving a larger display. This all sounds very power intensive. Driving a VGA type display may well use more power than the rest of the system.

Reply to
rickman

As I am a hobbyist (at least when it comes to electronics), I will comment what I personally look for in CPU modules.

When it comes to electronics, my software skills far exceed my hardware skills, ie: while I have absolutely no problems writing, say, device drivers for a new piece of hardware, I still use PDIP components when I build my own circuits.

Cost is also very much an issue in a hobbyist world. For example, if I had a choice between (1) a board costing, 70-80 UKP, but I had to put a great deal of effort into building my own software development toolkit, or (2) buying a fully complete, ready-to-go package for 250-300 UKP, I would choose the cheaper option (1) everytime. In a business environment, you would likely choose option (2) every time instead.

This means that I would like to see a CPU module that had it's own power point, a USB host connector, some flash/ram and a method of programming without requiring a baseboard, but that could still be plugged into a baseboard to take out the rest of the signals from the module.

The baseboard capability is to make it of interest to those people with the skills to design a baseboard (and hence, the increased volumes make it cheaper).

Have a look at the Digi ConnectCore 9C as an example of the kind of thing that I am thinking of. Unfortunately, with the Digi, it appears that you have to buy a full package to get started.

BTW, Ulf may like to know that the _only_ reason that I don't use AVR in my own circuits is because of a lack of a AVR with USB device in a PDIP packaging. Yes, I know that there's various AVR development boards from Atmel as well a GPL USB device software emulation, but what would be nice would be a AVR PDIP version of Freescale's HC08 JB8 that I can drop into my own circuits.

Simon.

--
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980's technology to a 21st century world
Reply to
Simon Clubley

[snip]

Baseboard + 1 x CPU Module = $1499 Without QVGA LCD Baseboard + 2 x CPU Module = $1899 Without QVGA LCD Baseboard + 3 x CPU Module = $2299 Without QVGA LCD

------------------------------------------- SAM9260EK = ~$600 SAM9261EK =~$1000 With QVGA LCD SAM9263EK =~$1000 With QVGA LCD

Just to get your math right.

--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

If you want portable, yes, FPGAs are not a good choice. If you want to have 1280 x 1024 resultion with instant reaction time, and have power to spare, then this is a good solution.

If you want low power consumption and rethink about VGA in favour of QVGA solution, then you

No I am talking about DIMM modules in a generic sense. You seem to believe that you can do anything useful after you buy a $399 DIMM module... You can't do anything. You cant power it up, you can't connect a serial port, you cant load linux, you can't do ANY THING...

If you want to do ANY THING, with the Cogent Module you need the Cogent Baseboard which is $1100.

You keep saying that the Cogent Module is less expensive than a complete kit. That is comparing Apples with Pears.

You need to move 55 MBytes per second just to refresh the screen.

--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

If I am using an FPGA to get speed, I won't be using a 200 MHz part. I'll use the 500+ MHz part.

I will be looking at that, but I am not convinced that I can't have low power at VGA resolution. Your architecture may not support it well, but there are many ways to solve the problem.

Maybe *you* can't use it, but I am sure I can hook it up without blowing $1000.

You were the one who brought up the Cogent modules as viable alternatives to your in-house eval boards. Yes, it is apples and oranges, but I don't need apples to make orange juice.

Yes, it needs to move a lot of data, but the question is how far? I don't need to run that through the CPU all the time. I only need to run updates through the CPU. You suggested an FPGA, I expect there are VGA controller chips that provide a simple interface between the CPU, ram and the screen or I can select a screen with a controller on board and a good interface.

Reply to
rickman

It is all a matter who's paying. If you are hooking up a DIMM module for fun and do not count working hours as expenditure, I am sure that you can hook up something for less than $600. ´ If you could do some consultancy work that earned you more than $600 in the same time as it takes you to create your own baseboard replacement, then your "low cost" solution is the more expensive one.

--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

"Simon Clubley" skrev i meddelandet news: snipped-for-privacy@eisner.encompasserve.org...

That is my belief as well.

Module cost is a key criteria

Key is to have a low cost baseboard. I think a good approach is to have a voume board for 8 bit controllers as a baseboard and then you soup it up with a 32 bitter which has the Ethernet and USB connectors.

While this is OK for hobbyists, it is not OK for OEMs. It is in conflict with the module cost criteria.

If I got my way, then an STK500 lookalike for the AVR would be a nice baseboard

but I doubt I got my way

--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

"Simon Clubley" skrev i meddelandet news: snipped-for-privacy@eisner.encompasserve.org...

That is my belief as well.

Module cost is a key criteria

Key is to have a low cost baseboard. I think a good approach is to have a volume board for 8 bit controllers as a baseboard and then you soup it up with a 32 bitter which has the Ethernet and USB connectors.

I think that your type of CPU module which fits on top of the STK500 which many people already have would be a nice entry point.

If you start to get serious about modules for OEM volumes, then I doubth that you want to have the connectors there.

It should be possible to make a small module with the QFN version of the AT90USB1287. That would be appropriate for hobbyists I do have a feeling that PDIP is going to become more and more scarce. The new 28 pin AVR with USB will not come in DIP.

--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

The baseboard is not a good concept for widely diverse products. The differences in the three ARM9 chips you make are enough that it would drive up the parts cost for each of them with otherwise unneeded parts and the cost to two boards with only a doubling or tripling of the volume. That is not enough to offset the added costs.

On the other hand, if you sold 20 times as many boards, you could easily cut your parts costs. Will your idea increase the sales by 20 fold? I don't think a baseboard approach will provide the numbers you need.

Reply to
rickman

Again, that depends on who is doing the work. Some people are better at it than others.

Reply to
rickman

With most parts SMD, the main way to keep prices down, is to make the PCBs as small as possible.

Daughter cards add connectors, increase PCB area, and dont save much.

To me, the Atmel ATNGW100 & EVK1100 look pretty good - just swap in SAM9xx ?

I'd fine tune the EVK1100 a little to allow the Educator sector to plug it into a project, but that's mainly judicious placemet of

0.1" pin headers. They have a kludge area, but forgot to bring the uC pins to a 0.1" dual row header.

-jg

Reply to
Jim Granville

The AT91SAM926x are not widely diverse products in my opinion. Most of the communication peripherals are the same. You have three basic choices,

  • with Ethernet, Without LCD
  • without Ethernet, With LCD
  • with Ethernet, With LCD.

otherwise they are quite similar.

If you use the same components as on a high volume board like the STK500, then you win. Atmel has sold > 100,000 STK500's so there is decent volume there.

The Gateway will go in decent volume as well which would bring down cost on advanced connectors like Ethernet and USB Device.

The AT90USB1287 development kit will bring down price on USB host.

--
Best Regards,
Ulf Samuelsson
 Click to see the full signature
Reply to
Ulf Samuelsson

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.