Jason,
I find myself rather ignorant of the intended target users of this board. I ask this because the projected price is rather hefty ("$239") and the board has some features (E.g. 2nd Ethernet port) that are not interesting to some folks.
Since I probably missed the target audience discussion, I'd appreciate it if someone could direct me .... thanks.
-Richard
P.S. Finally, someone realized that 2GB RAM is overdue on an SoC in order to run a desktop! Good job.
Note that I never saw a response, so I don’t think I sent a board to this individual.
I believe BeagleBoard-X15 is good for robotics, media centers,
interactive art, machine vision, home security, industrial automation
and build servers among other tasks. I call it a no-barriers "what-if"
board. It lets you prototype things that would be out of reach with
other hobbyist-level boards based on performance, connectivity and
(soon to be) availability (ie., a readily available board used by many
different people in a large community). Things that require ultra-low
latency (ie., needs the PRUs) or really high MAC/signal-processing
like functions (ie., needs the DSPs) are especially well-suited, but
things that simply need a fast ARM CPU and high-speed I/O (like build
servers) are also a good target.
More of my marketing spiel:
"I believe BeagleBoard-X15 is good for robotics, media centers,
interactive art, machine vision, home security, industrial automation
and build servers among other tasks. I call it a no-barriers "what-if"
board. It lets you prototype things that would be out of reach with
other hobbyist-level boards based on performance, connectivity and
(soon to be) availability (ie., a readily available board used by many
different people in a large community). Things that require ultra-low
latency (ie., needs the PRUs) or really high MAC/signal-processing
like functions (ie., needs the DSPs) are especially well-suited, but
things that simply need a fast ARM CPU and high-speed I/O (like build
servers) are also a good target."
I'm curious how the IPU's compare to the PRUs. I'm assuming since the "IPUs" are mentioned as Cortex M4's, that they have a pipeline like most / all Cortex M based processors. Anyway, I think it would be interesting to see the differences in determinism, between these two different core types.
As far as cool uses for this board. I've been imagining a power control / monitoring system with a real time web interface. Now, by "power control / monitoring" I mean high power( kW range ) control / monitoring. Context: We're off-grid, and completely solar powered. With a ~6kW peak solar array, and 1050aH battery bank at 24v nominal. DC-DC switching solar input here with this board surely would be fun !
Is there any chance of getting one of these boards for testing alternative OS's we already have a port working on the 5432 evm and agepv5
Hello there,
1) creating a system that collects high-speed 8-Bit parallel data (Multi-camera array) at about 40 MHz, creating files out of it and provide the files via 1 Gbit Ethernet
2) https://github.com/UltraFX/BBoard-Parallel
3) I am student with a Bachelor's degree in EE and working on my Master's Thesis. I have a lot of experience in Embedded Systems (mainly 8-Bit and 32-Bit MCUs) and a bit into Linux Kernel Hacking.
For my current Master Thesis I need a development board that provides very fast GPIOs, a big buffer memory and 1 Gbit Ethernet. The BeagleBoard-x15 would be perfect for this (PRUs, Gbit Ethernet). I need the BeagleBoard-x15 ASAP because my Thesis has a deadline until March 2016!
I would share all my experiences with the PRU modules with the public.
Regards,
Nicolas Dammin
I believe BeagleBoard-X15 is good for robotics, media centers,
interactive art, machine vision, home security, industrial automation
and build servers among other tasks. I call it a no-barriers "what-if"
board. It lets you prototype things that would be out of reach with
other hobbyist-level boards based on performance, connectivity and
(soon to be) availability (ie., a readily available board used by many
different people in a large community). Things that require ultra-low
latency (ie., needs the PRUs) or really high MAC/signal-processing
like functions (ie., needs the DSPs) are especially well-suited, but
things that simply need a fast ARM CPU and high-speed I/O (like build
servers) are also a good target.More of my marketing spiel:
BeagleBoard-X15 - BeagleBoard> Jason,
>
> I find myself rather ignorant of the intended target users of this board. I ask this because the projected price is rather hefty ("$239") and the board has some features (E.g. 2nd Ethernet port) that are not interesting to some folks.
>
> Since I probably missed the target audience discussion, I'd appreciate it if someone could direct me .... thanks.
>
> -Richard
>
> P.S. Finally, someone realized that 2GB RAM is overdue on an SoC in order to run a desktop! Good job.
>
> --
> You received this message because you are subscribed to the Google Groups "beagleboard-x15" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard-x15+unsubscribe@googlegroups.com.
> To post to this group, send email to beagleboard-x15@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/beagleboard-x15/ea7999c4-3602-472a-8b7d-e82246c2cfd6%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.1) What your plans for the boards would be.
I am working one a real-time video detection project for harbor crane safety.
OpenCL, OpenCV and more will be used. Since the most important part of this project is performance, image resolution and fps and accuracy need to be satisfactory, I will squeeze the potential of the DSP to make that happen from optimizing structure and algorithm.2) What open source repositories (URLs please) will host your contributions.
https://github.com/fire3280/ I will share my work here once I have improvement from XM
Waiting until you "have improvement" is not inspiring and not at all
in the spirit of the community. Share now and engage the community
regarding ways to improve and make more useful to everyone.
3) What expertise you have to accomplish your goals.
I have been working on XM board for a year now. As we know, XM is a good board for video processing, but when it comes to Full HD video processing, XM can't really perform that well. I was using OpenCV and DSPlink on XM for my camera project. Also, I have work experience (on F280x DSP) on I2C CAN UART ePWM and almost all functions on F280x DSP. I am currently doing maintenance on a industrial electric-powered vehicle drive board, which is mostly dsp programming in CCS and Visual Studio for UI. Also, I have side project on BBB for home surveillance, it's more like a smart home all based on BBB.
Sound like you could use some help from the community. OpenCL should
make your C66 integration easier. I don't think the OpenCV support for
OpenCL is very mature. Contributing directly to the mainline or a
highly visible fork to OpenCV the OpenCL support for X15 would be what
would be most beneficial to the community. It might "just work", but
publishing any required build steps or bug fixes into the mainline has
a lot of value.
In addition to that, I will be publishing my paper to IEEE(IEEE Transactions on Industrial Electronics | About Journal | IEEE Xplore). Beagleboard x15 is new and so powerful to be used in real-time video DSP, as there is not many board available in the market for entry to medium developers, what I will be publishing is so beneficial to our community.
I agree.
Hey, sorry I'm not sure if my last message was sent properly so I write it again.
1) I wanna use the board to do some high-speed data transfer. I get parallel data from a camera array with 40 MHz per line. I wanna use the PRU to receive that data and send it to one A15 Core via RPmsg. Then I want to let linux create image files and videos from the data and store it on SD Card. The other core shall provide an FTP Server via Gigabit Ethernet.
2) I would share everything I learn using the PRU on my github: https://github.com/UltraFX/BBoard-Parallel
3) I played around with the Beagleboard-xM and the BeagleBone white some time ago. But my expertise is mostly 32-bit Cortex-M3/M4 microcontrollers and a bit of Linux Kernel hacking.
I need a Beaglebone-x15 very soon for my Master's Thesis. So I'd really appreciate getting one of the early beta versions.
1) What your plans for the boards would be.
As academic researchers on operating systems, we are particularly interested in offloading OS workloads (such as light network and filesystem services) to “weak cores” for energy efficiency. To this end, the AM57xx SoC, which embraces the efficient Cortex-M4 cores, is an ideal platform.
In the past, we have created K2 [1], an award-winning OS research project, atop Pandaboard to demonstrate that such offloading is both desirable and tractable. Now we are looking at porting K2 to Beagleboard-x15, and testing an even wider range of OS services.
Note that BBB does not work for us -- its Cortex-M3 core is confined in L4, limiting the use.
2) What open source repositories (URLs please) will host your contributions.
Like other top research groups, we will opensource the code on our group website [2], which already hosts multiple successfully completed projects.
3) What expertise you have to accomplish your goals.
We have a long tradition of conducting system software research on the Texas Instruments platforms. Our research work, as exemplified by [1][3][4], has been published on premier venues. We publish our code and documentation (see these links).
http://k2os.org
http://xsel.rocks
http://www.ruf.rice.edu/~mobile/downloads/ASPLOS2015/
https://engineering.purdue.edu/~xzl/xsel/projects/memif/memif.html
Hi,
I am writing on behalf of our small startup Robotics company in The Netherlands. We are currently working on a hard-realtime robotics motion control system based on Linux and EtherCAT (we currently use EtherLAB. For best hard-realtime performance we need a multi-core system and we require two Ethernet Ports (one for real-time EtherCAT and one for non-realtime communication to higher level controls.
1) What your plans for the boards would be.
We would like to evaluate the board to see if it meets our performance requirements (we are aiming for 5kHz update rate with maximum 10% jitter). We are connecting to industrial components that support EtherCAT and want to control systems like robots, motion systems for flight simulators or industrial equipment. We are currently developing a low-cost industrial robot for education an small enterprises, so price is an important issue.
2) What open source repositories (URLs please) will host your contributions.
We are posting our updates to EtherLAB to improve running realtime-servo control: ethercat/etherlab_master/src/etherlab.cpp at master · redheli/ethercat · GitHub
Also we would like to share information on how to get the best realtime performance from our hardware on our website (website will be renewed in 2016, including a public git repository)
3) What expertise you have to accomplish your goals.
We have 4 robotics controls people on board, working on this project with vast experience in realtime systems programming (Linux RTAI, PREEMPT) and Controls Algorithms. We work on real projects worldwide, like flight simulators, giant Ferris wheels and robotic systems.
Since we are moving in fast-forward we need the evaluation board pretty soon. I am guessing the X15 would be perfect for our Motion Engine.
Best regards and happy Holidays,
Philippe
Vectioneer
I've had especially good fortune getting hardware through FCC. I'd be happy to give your test reports and layouts a look if that would be helpful and it's "FCC" is still an issue.
Please provide a specific proposal of what code you’d contribute back per the questions. We aren’t looking for people to just do testing right now.
1) developing my own version of hardware (not a cape) for cnc-ing, e.g. 3d printing and others like machinekit
2) i plan to update this one https://github.com/Noobman/RDB-STPTS-002-DIY
Note: this is not a cape, it currently uses wires of breaboard type, and can be used by any developer any board for either 5v or 3v3 boards, has features like optoisolation and mosfet drivers; not really end hardware more like hardware for other devs;
3) very few things to this area http://reprap.org/wiki/User:NoobMan also i did previous version
What i am planning to update: -not sure if lvds or other logic level would be required, but i plan to make it work with 1,8v logic, mosfets could do easy, the step signal would have to check options -planning to optoisolate endstop switches too -etc minor other stuff - will keep on diy and gplv2
I know this is probably not smth that should be on the list here but well, here it is anyway. You said anxious not worthwhile :). Also note i would do this anyway with time so im probably not a priority. Just a bit anxious about this board, expecting a good development around it coz of prus and mcus. At very least i think many ppls waiting for it.
Dear all,
I realize I am late to this thread. I found this thread after ordering my x-15 many months ago, and coming here to find out news about availability :-).
My primary goal in obtaining a X-15 is porting RTEMS to it. If I can get access to a beta board, I can get started sooner. So, if there are any boards still or any beta items coming in the future:
Ideally I also want it to boot using the bsd-compatible-licensed bootloader umon (https://git.rtems.org/umon/), and also have the x15 running in rtems smp mode. Those I consider stretch goals, as it were, however.
1) eBPF byte-code compiler kernel driver for the HW packet filter engine on the AM5728.
Packet Filter Engine:
"AM572x ARM Processor Silicon Revision V1.1" section 24.11.4.9
http://www.ti.com/lit/er/sprz429g/sprz429g.pdf
2) https://github.com/zebop/x15-pkt-engine
3) Many years of OS hacking in C and assembler. Recently ported Nuttx to the TI CC3200 ARM Cortex-m4 ( https://hackpad.com/ep/pad/static/Rrol11xo7NQ )
Late to the party, but here goes anyway…
If a board were to ever fall into my lap, I’d work on getting the the Adafruit BBIO Python library working on it along with other people who’ve contributed to the library. I’d be eager to get this library working with X15 ASAP after release.
https://github.com/PeteLawler/adafruit-beaglebone-io-python (via elsewhere forked from https://github.com/adafruit/adafruit-beaglebone-io-python )
Not much, though I did help get the Adafruit BBIO Python library updated to work with 4.1+ kernels, so there’s that.
Should this thread be "un-pinned" now that Rev C has been released?