Actions
Forum Thread Details
  • Replies 58 replies
  • Subscribers 59 subscribers
  • Views 5005 views
  • Users 0 members are here
Related

Another manufacturer for BBB

bwelsby

Just seen Embest (a Farnell company ) is producing their own clone of the  BeagleBone Black

Parents

  • Looks like a price increase is happening:

     

    http://article.gmane.org/gmane.comp.hardware.beagleboard.user/64166

     

    Since it seems to be aimed at getting more manufacturers involved in order to increase supply, I'm surprised that e14 isn't using the embest one for this.

     

    Still, a price increase is dubious at best. The BBB is already more expensive than the more capable OLinuXino-LIME and it won't take much to push it into head to head competition with e14's RIoT board.

  • in reply to Former Member

    Looks like we might be getting the Embest one outside China after all, noticed this the other day

     

    image

     

    Mfg is e14, country of origin China and unfortunately includes the previously mentioned price increase image

     

    The RIoT board is £46.55 here in the UK, it has twice the RAM, twice the eMMC, gigabit network, four usb, audio connectors, better CPU, better GPU... the list goes on of ways it offers more than the BBB.  So while I'm sure there will be a certain amount of RPi effect where people will choose the BBB simply because it's a BBB, personally I don't think it'll be quite so simple a choice. 

    The BBB is really good at the original price, but it's competing against a different set of boards at this price.

  • in reply to Former Member

    While it might make financial sense to offer it for this price while supply is still low, I really hope this is not now the permanent price for it for this version. The only people who continue to benefit are those putting the boards to commercial uses, because they would not be able to manufacture it themselves at the higher price anyway, as we saw from that snippet (below) from the link you posted earlier. Individuals and educational users will be heavily impacted. It's a shame, it could have been prevented through better T&C's preventing commercial use for these boards without (say) a license cost.

     

    <ohararp <at> ohararp.com> wrote:

     

    Gerald, I have looked into both producing our own boards and System on Module type platforms with the AM3559 chipset.  The system on module setups are about $125 in quantity with a 12 week leadtime.  To produce our boards it would be a minimum of about $100 pre board to do our own production.  The latter does not factor in parts availibility, capital required, etc.  Finally, if you just look at single chip costs for the AM2559 you are looking at $34 - http://www.ti.com/product/AM3359!  The $45 pricing for the BBB is really unbelieveable and quite possibly unsustainable?  I would have no problem paying $75 for this board if we could guarantee availibility and continued support (ie being able to stay in business).

  • in reply to shabaz

    Somewhere earlier in that rather long thread Gerald had suggested increasing the price to $75  http://article.gmane.org/gmane.comp.hardware.beagleboard.user/62794

     

    xe.com today shows $75 = 44.75 GBP

    Fullscreen contentimage_176599.html Download 
    <html><head><title>Jive SBS</title></head>
<body><font face="arial,helvetica,sans-serif">
<b>Error</b><br><font size="-1">
An general error occurred while processing your request.
</font></font></body></html>

     

    So I'd not rule it out until we see something definitive on exactly what they're going to do.

     

    Either way, if we assume this is the Embest built version rather than the CircuitCo one, then it's all up for grabs. Not being an 'official' version I think lets e14 charge what they want, and people will pay if the official one is unobtainable.

    You are of course right that it's the non-commercial users that suffer most, but that's the trade-off for OSHW when priced at a point that commercial users building their own clones can't come close to matching.

    There was a discussion here a few months back talking about SBC's not being the right answer and that a purpose built design should be used instead. While there may be merit in that point of view, it doesn't work out economically when you can buy a pre-built SBC and use it as a module at a third of the cost of your own design.  That appears to be the problem we're seeing, the BBB hardware is just too good to be true at $45 and the people using it commercially know it.

  • in reply to Former Member

    Price has been updated on the Embest version:

    image

     

    So it now matches the CircuitCo version.

     

    There's also this http://article.gmane.org/gmane.comp.hardware.beagleboard.user/64767 from Gerald that suggests a move to 4GB eMMC sometime soon.

  • in reply to fustini

    Interesting.. I suspect the Newark stock will be short lived now that you've pointed it out image

     

    I'm curious what e14/embest will do when the CircuitCo rev C arrives.  If they can continue to offer their version at $45 with 2GB eMMC it might take the pressure off.. Or do they match price/features with the CircuitCo version?

  • in reply to Former Member

    I hope it will track fairly closely as new revisions come out, since the Embest BeagleBone Black is part of the BeagleBoard.org Logo program: BeagleBoard.org - logo

  • in reply to fustini

    That comment on the wiki about eMMC wearing out....  people talking about wear levelling and worrying about lifetimes should go and read http://en.wikipedia.org/wiki/Write_amplification While that page is written with SSD's in mind, the principles are similar for any sort of flash memory wear levelling and things like eMMC and uSD possibly may not have the equivalent of TRIM.

  • in reply to fustini

    Drew Fustini wrote:

     

    since the Embest BeagleBone Black is part of the BeagleBoard.org Logo program

    That's interesting as the only photos of the embest version I've seen don't have the logo..

  • in reply to fustini

    fyi - photo from @beagleboardorg on twitter from Newark element14 booth at EELive.  Thanks !

    https://twitter.com/beagleboardorg/status/451792077177618432


    image

Reply Children
  • in reply to fustini

    Pity that e14 didn't choose some different colour of LEDs image

  • in reply to Former Member

    Hehe, should be blue and orange bi color LEDs per our design guidelines image

     

    just posted a pic of the element14 BeagleBone Black box! image

    https://twitter.com/beagleboardorg/status/451800453085270016

    image

    Still processing that I work for a company that makes an Open Source Hardware Linux computer!!!

  • in reply to fustini

    The RIoTboard is supposed to be open source too...

  • in reply to Former Member

    Ah, good point... I'll qualify with Open Source Hardware Linux computer with cute mascot and commitment to mainline kernel image

  • in reply to fustini

    I have to say that I'm surprised at the seeming lack of interest in mainline linux support for RIoT.  It would seem to make commercial sense to open up the platform to a wider audience, and they're bound to understand how well these could sell given their involvement with RPi & BBB.  As e14 is in full control of the platform it seems odd that they don't seem to want to take advantage of the BBB supply shortage with their own board and are instead going with a BBB clone.

  • in reply to Former Member

    selsinork: I think BeagleBone has an advantage as the BeagleBoard started in 2008 as a way to get more Linux developers working on ARM.  They seem to have carried the interest of kernel hackers on to present day.  The impression I get from other boards with impressive specs is they often lack that critical mass of developers.  BeagleBoard.org community seems to make a consistent effort to work with the kernel maintainers.  I imagine this can be difficult to learn for other projects/boards.

     

    I'm really excited to be going to San Jose in a few weeks for the 10th annual Embedded Linux Conference... my first!  I figure I'll more about how all of this works and the relationships between chip makers, board makers, kernel maintainers and community projects.

  • in reply to fustini

    Freescale have made a big effort with the i.MX6 though, that's visible in the support that Sabre-Lite, Wandboard, UDOO and a host of other i.MX6 boards are benefiting from. So much so that virtually everything on the RIoT (and MarS) are already supportable. Like with Beagleboard, all it really takes is a bit of interest from the manufacturer (e14 in this case) to bring things together. Once that's done, you'd be surprised at how easy it'll be to maintain the momentum as it'll benefit from the work all of the other i.MX6 board manufacturers are doing.

    There seems to be two cohesive communities that have critical mass these days, centered around the Freescale & Allwinner devices, and the interest seems to be largely due to there being lots of manufacturers using both SoC families.

    The impression you get is that Beagleboard has Robert Nelson and that's it. Sure there are various TI employees doing bits of OMAP support but none of it looks particularly related to Beagleboard and things like the BBB's capemgr look like they'll never make it. Indeed it looks a lot like Beagleboard and mainline TI are completely orthogonal.

     

    Either way, when the manufacturers show no interest at all then any community effort may be doomed to failure regardless.  Interesting then that when you look at arm devicetree support in current kernels it's probably Marvell that has the widest support...

  • in reply to Former Member

    Good to know - I'm glad that i.MX6 is doing well with support in general as I backed the Novena (iMX6 laptop).

     

    Marvell, hmmm, that was used in a lot of the tiny ARM file servers like pogoplug, right?  Are there any OSHW SBCs you know that use it?

  • in reply to fustini

    Yes, sheevaplug is the one I know the marvell stuff from. I think their chips are also used in various NAS type systems as well judging by the names.  Can't say I've seen any SBCs using them.

     

    Looking at the dts files in arch/arm/boot/dts/ in the 3.13 kernel, out of 286 the top 3 are

    SoC manufacturer
    		Boards
    Freescale49
    Marvell42
    TI30

    followed by Samsung, Allwinner, Atmel, Nvidia all in the 10-20 range.  The Freescale number is interesting as there are lots of older imx2* boards while we're much more likely to be interested in imx53 or imx6 today.

    The overall impression is that while there's lots of Arm devices, comparitively few are SBCs and even fewer OSHW.

  • in reply to fustini

    image

    image

     

    Help... can't read this image

    image

    Interesting that a BBB is now 'Exclusively from element14'

    image

    image

    image

    Yes,its really an AM3358..

    Fullscreen 6378.contentimage_176609.html Download 
    <html><head><title>Jive SBS</title></head>
<body><font face="arial,helvetica,sans-serif">
<b>Error</b><br><font size="-1">
An general error occurred while processing your request.
</font></font></body></html>

     

    Spot the differences...

    Fullscreen 3225.contentimage_176610.html Download 
    <html><head><title>Jive SBS</title></head>
<body><font face="arial,helvetica,sans-serif">
<b>Error</b><br><font size="-1">
An general error occurred while processing your request.
</font></font></body></html>

    Fullscreen 1172.contentimage_176611.html Download 
    <html><head><title>Jive SBS</title></head>
<body><font face="arial,helvetica,sans-serif">
<b>Error</b><br><font size="-1">
An general error occurred while processing your request.
</font></font></body></html>

     

    One interesting thing to note, and that's already been noticed in the beagleboard google group is that the e14 boards don't have the A5B, A5C, A6 etc marking, so we don't really know what version they are and what fixes are incorporated.

    Fullscreen 8037.contentimage_176612.html Download 
    <html><head><title>Jive SBS</title></head>
<body><font face="arial,helvetica,sans-serif">
<b>Error</b><br><font size="-1">
An general error occurred while processing your request.
</font></font></body></html>