From the comments to http://olimex.wordpress.com/2013/12/09/a10-olinuxino-lime-oshw-linux-computer-developer-edition-is-in-stock/
in March-April A20-LIME2 with 1GB RAM and Gigabit Ethernet
I'm sure that will catch the interest of various people here 
From the comments to http://olimex.wordpress.com/2013/12/09/a10-olinuxino-lime-oshw-linux-computer-developer-edition-is-in-stock/
OLIMEX Ltd
Dec 10, 2013 @ 11:32:03in March-April A20-LIME2 with 1GB RAM and Gigabit Ethernet
I'm sure that will catch the interest of various people here 
selsinork wrote:
I'm sure that will catch the interest of various people here
Even the plain base-level 30-euro A10-OLinuXino-LIME is already interesting because of its SATA and ready LiPo socket, which aren't common in this price range. Each extra facility just adds more interest, and this could be seen as the basis for an entire range of boards from Olimex:
Come to think of it ... are Olimex basing their business plan on my posts? 


This is great news.
Morgaine.
Morgaine Dinova wrote:
- + 4GB on-board flash -- probably +5 euro
If you look at the A20-OLinuXino-Micro, then the 4GB nand flash seems to add 10 euro. I'm not sure I think that's worth it.
selsinork wrote:
If you look at the A20-OLinuXino-Micro, then the 4GB nand flash seems to add 10 euro. I'm not sure I think that's worth it.
I agree, if it turns out to be +10 euro, this would be pushing the price envelope far beyond the limit of reasonable on a 30-euro product. Flash chips don't cost that much. Hell, BBB managed to fit 2GB eMMC into its very small price budget, and that was a long time ago in this industry.
I hope Olimex recognize that before the 4GB model appears and shave some euros off. +10 would undermine their highly attractive LIME pricing.
Morgaine.
Morgaine Dinova wrote:
I hope Olimex recognize that before the 4GB model appears and shave some euros off. +10 would undermine their highly attractive LIME pricing.
From some chat in #olimex on freenode yesterday:
<Tsvetan> oliv3r is there any other way to write to the NAND exept with the phoenix suite?
<Tsvetan> right now testing the boards with NAND is pain in the @@@
<Tsvetan> as phoenix suite write Android image for about 20 minutes, and there is no way to run 10 Phoenix suite instances
<Tsvetan> on same PC
<Tsvetan> but I think for something like this
<Tsvetan> SD card which I plug in the SD_card slot and power the board
<Tsvetan> it boots and writes the NAND then blink a LED that its finished
<Tsvetan> this way we can make farm with 20 board slots and we just plug the cards and after 20 minutes we have 20 ready to test boards
<Tsvetan> no need to run Phoenix suites and bla bla windows tools
<Tsvetan> if we are going to release LIME with NAND we have to find a way to parallel program it or we will have no capacity to test the boards in volume
<Tsvetan> right now all our boards with NAND cost EUR 10 more for this reason - slow programming and test of the boards with NAND
So less to do with the cost of the NAND part, much more due to needing to use a slow windows based tool as part of the testing process.
Then earlier this morning:
<Tsvetan> hno I just got reply from Allwinner that there is SD card upgrade possibility, i.e. boot the board and the image written in the SD card is written to the NAND Im waiting more info on this
There seems to be some other things to sort out, for example they also talked about u-boot/mtd support needing additional work.
For me, the nand is relatively un-interesting when the board has SATA, I'd mostly want to use it to store SPL & u-boot that was able to then boot linux off a sata drive without needing a uSD. Research so far leads me to believe that the allwinner stuff can't boot directly from sata without help - unlike the i.MX6 which can.
selsinork wrote:
So less to do with the cost of the NAND part, much more due to needing to use a slow windows based tool as part of the testing process.
On the positive side, the fact that they realize the effect of this on their pricing and sales scalability is very good news, since it's the start of resolving it.
Morgaine.
selsinork wrote:
So less to do with the cost of the NAND part, much more due to needing to use a slow windows based tool as part of the testing process.
On the positive side, the fact that they realize the effect of this on their pricing and sales scalability is very good news, since it's the start of resolving it.
Morgaine.