Interesting "Competitors" for the Raspberry Pi

Got a quick email update from the Cubieboard folks ...

Intial 200 prototypes sold out. 1,000 users in the forum, 4,000 subscribed to the notification list, growing about 100 per day.

To meet demand they are crowd funding via http://www.indiegogo.com/cubieboard

Estimated shipping was end of Nov.

-J

Hey jamodio

I liked the sound of cubieboard's specs - so I just "contributed". It could be fun since I have several sata drives.

Peg

Peg Wagner wrote:

 

Hey jamodio

I liked the sound of cubieboard's specs - so I just "contributed". It could be fun since I have several sata drives.

Peg

I totally agree, SATA is very attractive as well as having LVDS +HDMI so I'll probably throw something at them but I think I'll watch for a little bit to see what happens since the campaign isn't ending for another month. To be honest though, I'm a little nervous about something coming out of Shenzhen and what the support or dare I say quality, will be like.

just read this news post which is a little concerning http://semiaccurate.com/2012/11/01/freescale-layoffs-target-one-particular-product/

Well (gulp) it won't be the first time I've jumped at something like this - won't be the last I'm afraid. I do get lucky though - so we'll see what happens. That video was the deal maker, for sure )!!

George Ioakimedes wrote:

 

just read this news post which is a little concerning http://semiaccurate.com/2012/11/01/freescale-layoffs-target-one-particular-product/

I wonder how semi-accurate this news is?  Freescale seemed to be promoting i.MX6 enthusiastically at ARM TechCon yesterday.  It doesn't make sense to abandon the whole i.MX team, or even move the team to another site.  You've got to keep institutional knowledge around, particularly for a brand-new product.

IMO Freescale has had a long-term problem with having way too many chips and product lines.  Personally, I'd be delighted to see them concentrate on Coldfire and PowerPC instead of trying to compete with TI, NXP, ST, and a bunch of other small players in ARM.  Coldfire and PowerPC are nice differentiators.  I rather liked the MPC5125 and was sorry to see that line discontinued just when it started to get interesting.

I could be wrong, but I think I recently heard that TI is planning to ease out of the OMAP biz.  It may be that TI and Freescale (and ST-Ericsson, IIRC) have decided that its too hard to complete with the likes of Rockchip and AllWinner in cell phones and tablets -- better to go for automotive and industrial.  Consumer electronics is really cut-throat.

JMO/YMMV

If I were a Freescale shareholder, I think I'd be confronting the dear CEO with a point blank question about whether  i.MX has a future at Freescale, to be answered in English rather than in CEO doublespeak.

John Beetem wrote:

 

I could be wrong, but I think I recently heard that TI is planning to ease out of the OMAP biz.  It may be that TI and Freescale (and ST-Ericsson, IIRC) have decided that its too hard to complete with the likes of Rockchip and AllWinner in cell phones and tablets -- better to go for automotive and industrial.  Consumer electronics is really cut-throat.

You're right, the rumors have been flying around that Amazon will buy OMAP (specifically http://www.lab126.com). I doubt they'll tell me anything but I'm going to ask about the future of i.MX. No sense jumping into it if the future is uncertain. I certainly hope that the future doesn't revolve around Rockchip and ALLWinner..

I was just looking at the schematic of the Cubieboard (http://linux-sunxi.org/Cubieboard) and one thing I see that I like is that there's direct access to a 2-channel LVDS interface which means you can connect a FHD LDC Panel. That's the good news, the bad news is that the only I/O I see is a single SPI, I don't se any free GPIO pins.

George Ioakimedes wrote:

 

John Beetem wrote:

 

I could be wrong, but I think I recently heard that TI is planning to ease out of the OMAP biz.  It may be that TI and Freescale (and ST-Ericsson, IIRC) have decided that its too hard to complete with the likes of Rockchip and AllWinner in cell phones and tablets -- better to go for automotive and industrial.  Consumer electronics is really cut-throat.

You're right, the rumors have been flying around that Amazon will buy OMAP (specifically http://www.lab126.com). I doubt they'll tell me anything but I'm going to ask about the future of i.MX. No sense jumping into it if the future is uncertain. I certainly hope that the future doesn't revolve around Rockchip and ALLWinner..

I see Rockchip and AllWinner and perhaps a few others at the low-end phone and tablet space, i.e., devices sold for under US$100.  Samsung, QualComm, and nVidia will fight it out for the high end: devices over US$200 unsubsidized.  Dunno about the no-man's land in between.  If Amazon buys OMAP, I don't know what that does to B&N.  Probably doesn't matter.  Amazon can afford to have a higher-cost device since they make their profits from the "razor blades".

Samsung has the advantage that it uses its own ICs and displays in its end products, eliminating middle-men.  They seem incredibly well run for a huge company.  They probably don't understand any of the jokes in Dilbert.

George Ioakimedes wrote:

 

John Beetem wrote:

 

I could be wrong, but I think I recently heard that TI is planning to ease out of the OMAP biz.  It may be that TI and Freescale (and ST-Ericsson, IIRC) have decided that its too hard to complete with the likes of Rockchip and AllWinner in cell phones and tablets -- better to go for automotive and industrial.  Consumer electronics is really cut-throat.

You're right, the rumors have been flying around that Amazon will buy OMAP (specifically http://www.lab126.com). I doubt they'll tell me anything but I'm going to ask about the future of i.MX. No sense jumping into it if the future is uncertain. I certainly hope that the future doesn't revolve around Rockchip and ALLWinner..

I see Rockchip and AllWinner and perhaps a few others at the low-end phone and tablet space, i.e., devices sold for under US$100.  Samsung, QualComm, and nVidia will fight it out for the high end: devices over US$200 unsubsidized.  Dunno about the no-man's land in between.  If Amazon buys OMAP, I don't know what that does to B&N.  Probably doesn't matter.  Amazon can afford to have a higher-cost device since they make their profits from the "razor blades".

Samsung has the advantage that it uses its own ICs and displays in its end products, eliminating middle-men.  They seem incredibly well run for a huge company.  They probably don't understand any of the jokes in Dilbert.

John Beetem wrote:

 

Samsung has the advantage that it uses its own ICs and displays in its end products, eliminating middle-men.  They seem incredibly well run for a huge company.  They probably don't understand any of the jokes in Dilbert.

Awesome line.

John Beetem wrote:

 

Samsung has the advantage that it uses its own ICs and displays in its end products, eliminating middle-men.  They seem incredibly well run for a huge company.  They probably don't understand any of the jokes in Dilbert.

I passed this information on to some friends I have at the top level in Samsung. They thanked me for the information and said. "We as a company were unaware of any prior jokes in Dilbert. We will with immediate effect find any and all Dilberts (whatever they are) and scan and remove any jokes that are in it." "We shall then patent all such legally acquired jokewarez forthwith before that company of fruits in the US do it." Sighhh an ora.

Ray

OK, the NDAs with Freescale are signed and I have the datasheets. I've fired off my first questions to them but I'm wondering if anyone has anything specific they want to know.

I assume that the first thing I need to do is see if it is possible to make a board that competes in cost with the RPi. I figure if that's not possible then interest is not going to be as great.

Hmm, is there really no interest in this project? I got some budgetary pricing back and it looks possible to compete with the RPi.

George Ioakimedes wrote:

 

Hmm, is there really no interest in this project? I got some budgetary pricing back and it looks possible to compete with the RPi.

Don't know about everyone else, but I'm interested.. Was mostly waiting to see what sort of spec you'd come up with and what sort of cost..  I'd be happy with a $50 price as long as some of the more dubious design decisions the RPF made are not followed.

I'm trying to schedule a conference call with them but can we put together a list of questions/requests from the community and I can address them with the Freescale folks.

George Ioakimedes wrote:

 

Hmm, is there really no interest in this project? I got some budgetary pricing back and it looks possible to compete with the RPi.

I think you can rest assured that interest in a board that competes on price with RPi but is based on i.MX6 would be colossal.  After all, it would provide a substantial increase in ARM capability from its more modern core, a fully working USB system, and probably a more openly documented platform.

What's not to like?  I don't see how anyone with any interest in ARM boards could fail be interested, unless they are so deep into Pi fanboism that better specs don't matter to them.

I'd not had a chance to read this site for the past few days, but I'd be interested in half a dozen at that price ; )

George, as I wrote earlier, interest in an i.MX6 board that price-competes with RPi is a foregone conclusion I think, but my interest also extends further to low-power products without such a guaranteed market but more interesting architectures.

Over in the ARM server clusters thread, I've been describing meshes of interconnected CPU cores as pioneered way back by the transputer, a concept which is rather foreign in computing today and does not currently exist in the commercial ARM product space, AFAIK.  Because of the lack of on-chip integrated interconnects, such a concept is not simple to implement on a board for more than a small handful of ARM devices in a way that is simultaneously compact, fast, and cheap.  So much glue logic would be required for interconnects that an 8x8 array of devices on a single board is out of the question today, and even 4x4 would be pushing at the limits of sensible board form factor.

Perhaps a modular approach would be more manageable though, and at the same time might be of interest to more people.  How does a quad-i.MX6 board strike you as a product?  I'm just thinking aloud about this at this stage, but it seems to me that a stackable board bearing four i.MX6 SoCs without peripheral connectors would allow construction of a 4x4 array fairly easily.  A fifth board stacked on top could provide external connections.

This doesn't address the interconnect controllers problem of course, but as I said, it's just thinking aloud for now.

Any thoughts or suggestions?

Morgaine.

Continuing this train of thought, just for fun imagine a zero'th board stacked on the bottom, providing fanout to 4 other such boards.  The 0+4x4 stack taken as a unit could then become the tip of a pyramid of such 0+4x4 units.  Such an N-ary tree is one desirable topology for such systems because the logical distance from the top to one of the N devices in the bottom layer is only O(log N) so communication latency scales very well.

Such an expanded system would probably be of interest to relatively few people, but its building block of a quad ARM board could be the basis of more down-to-earth quad machines for those who aren't interested in constructing larger systems.

And of course, should they outgrow their little quad board, it just so happens that infinite expandability is built in ...

Morgaine.