Open Lunchbox Forums

AMD Carrizo APU With Excavator Core Architecture Targets Higher Efficiency, Longer Battery Life

http://hothardware.com/news/amd-carrizo-apu-with-excavator-core-architecture-targets-higher-efficiency-longer-battery-life

http://www.anandtech.com/show/8995/amd-at-isscc-2015-carrizo-and-excavator-details

http://blogs.coreboot.org/blog/2015/02/23/the-truth-about-purism-why-librem-is-not-the-same-as-libre/

SATA M.2 Card
https://www.sata-io.org/sata-m2-card

https://github.com/openlunchbox/openlaptop

https://github.com/openlunchbox/openlaptop.git

All the non-final files are in the Ernie branch

git clone https://github.com/openlunchbox/openlaptop.git

SATA IGES Model

Moved to https://github.com/openlunchbox/openlaptop  in the ernie branch

https://github.com/openlunchbox/openlaptop.git

Two simple speakers integrated into the keyboard module.

How important is it to you to have an internal DVD drive?

Open Firmware - No Blobs

Long battery life and bigger batteries

Non-proprietary batteries

More USB ports than just 2-3

PCIe connectors for cables and expansion boards and external GPU's

Replaceable anything and everything

Desktop CPU performance

>8GB of RAM, 32GB

Areas for custom sensors, small boards, devices etc DIY

The cost of tooling for a typical laptop enclosure is over $250K

The enclosures may be printed using rapid SLA with engineering photopolymers. They can be entirely made from photopolymers or partially photopolymer with an aluminum alloy chassis similar to the T60 thinkpads.

Another option is to use aluminum alloy and machine the entire enclosure or again a combination of aluminum alloy shell along with photopolymer components.

Short to medium production runs of various case and display sizes is easy as well as custom one offs since we aren't making tools.

The costs for tooling are >$100K for injection molding. To keep initial costs down until there's enough demand or interest we'll be printing the enclosures vs injection molding. Printing also allows you to make custom cases quite easily that would require new tools for injection molding.

3D printed enclosure using engineering resins (not FDM). Everyone seems to be stuck at this step, even Google.  Not sure why since the tech is decades old.

https://www.youtube.com/watch?v=KVhhZd1p8FI#t=56m10s

Youtube doesn't always like starting so far into a long video. The conversation about 3d printing starts 56 minutes and 10 seconds in.

"3D Systems Will Not Be 3D Printing Google’s Project Ara Smartphone Modules"
http://3dprint.com/33331/google-project-ara-3d-print/

~$3 Mediatek MT6260 GSM/GPRS/EDGE-RX ARM SOC

http://www.datasheet4u.com/datasheet/M/T/6/MT6260_MEDIATEK.pdf.html


32b ARM7EJ-S RISC processor
GSM/GPRS/EDGE-Rx radio
LCD controller
Supports QPI and SPI serial flash
Bluetooth
5-row x 5-column keypad controller
2 UARTs with hardware flow control and supports baud rate up to 921,600 bps
FS/LS USB 1.1 device controller
Multimedia card, secure digital Memory Card, host controller with flexible I/O voltage power
Supports SDIO interface for SDIO peripherals
DAI/PCM and I2S interface for audio
I2C master interface for peripheral management
SPI master interface for peripheral management
Li-ion battery charger and 14 LDOs for the power supply

http://www.u-blox.com/en/wireless-modules.html

LEON-G100    
Quad-band GSM/GPRS module
http://www.u-blox.com/en/wireless-modules/gsm-gprs-modules/leon-gsm-module-family.html

LISA-U2 3.75G UMTS/HSPA(+) module series
http://www.u-blox.com/en/wireless-modules/umtshsdpa-modules/lisa-u2-series.html

Are there any LCD's that we might consider using that still use LVDS?

How many Watt Hours should the minimum battery capacity be?

Modular was decided to be able to cover as many different options that people ask for. Some want high performance, long battery life, rugged keyboard and a 12" width. Others want the same with a 15" or 17"display, and others only want a multicore ARM SOC and a durable case with rugged keyboard.

Having the mainboard, batteries, display and keyboards as modules allows the end user to  mix and match the features they want now. This also allows them to upgrade or replace these features in the future.

coreboot is the firmware of choice.

http://www.coreboot.org/

The firmware for the EC will be based on http://www.chromium.org/chromium-os/ec-development

Possible EC's are https://chromium.googlesource.com/chromiumos/platform/ec/+/master/chip/, but most likely a STM32.

http://www.st.com/web/en/catalog/mmc/FM141/SC1169?sc=stm32

AMD has been a supporter of coreboot for the past several years. AMD has also made made BIOS register docs available and many comments are allowed in the coreboot source.

The problem with Intel is that even though Google has provided coreboot source for a few Intel chipsets the source contains no comments for what those magic settings are for and what the other settings in those registers will do. Intel also makes it very difficult for anyone but a chosen few to obtain BIOS register docs.  If this situation were to change then I'd see less problems with designing Intel mainboards for this project.

We could make an exact copy of the Google laptop boards and use the coreboot source but then why not just buy a Chromebook?

T60 Keyboards

Chicony FRU 39T7178
ALPS FRU 39T7118
NMB FRU 39T0958

Have any other part numbers or models that you'd like to see made available?