Looks nice, a little worried about how low some of the funding options were.
I do wish someone would build an ITX sized motherboard with SATA, actual ram slots, and gigabyte network ports. Perhaps when the 64-bit version is more available someone will.
Yeah.
My little pandaboard, if i'm not powering the USB ports at full bore, doesn't take more than ~7 watts while doing full 1080p video decode of a movie over the network.
The same on one of these ITX boards would take probably 25-30 watts, minimum.
One interesting thing about the Cubieboard is the SATA port, otherwise it's very similar to a lot of other ARM boards out there (e.g. Beagleboard, Pandaboard, etc). This makes it much easier to stick a low power SSD on the device, rather than rely on a hugely slow SD card.
What I'm looking forward to next are Cortex A15 boards, the first CPUs are shipping (In the new Chromebook), so hopefully we'll start seeing them on small hackable boards soon.
I think the final price of the Cubieboard is $50, the cheapest SSDs start at around $50 too [1][2], so yeah it's going to add to the price, but it's also going to make it useful for a lot more projects than an SD card.
Aaargh, that looked really good until I saw there were no GPIOs. You almost solved my problem. with GPIOs being standard in so many SOCs I am surprised how some just don't bother placings SILs or DILs so we can easily use them
I have a few Mali devices next up on the chopping block after I finish reversing the VideoCore IV
Hehe. Reversing a GPU is a lot of work. From my experience it usually involves zillions of commands, state bits that subtly change handling, and different instruction sets for different kinds of shaders and other sub-processors... (even figuring texture formats can be lots of work) I can't even imagine starting such a project again, though it was fun, wish I'd still have the time...
Wladimir, I respect your view here - I agree with everything you write. My field of expertise is more ISA encodings, overall functional block architecture and decompilers. Once I hold the encodings, I point at the blob and say there is your source & specification. ie Binary blobs + ISA encodings = specification to me.
Absolutely, there's all sorts of things that can be done which aren't really workable without some better performance/capacity storage than an SD card (even the nice new 64GB ones)
On my TS-7800 [1] which has GigE and a 500MHz (!) CPU, Samba outperforms the Raspberry Pi (15MB/s vs 7MB/s). Note that on the RPi not even the whole 100Mb link is utilized despite the CPU being slightly faster in general (yes, I did some benchmarks) - and the CPU on the TS-7800 was actually designed for storage applications. So there may be something on to your claim, and it would appear that CPUs that include graphics and stuff are not that good at I/O.
...it would appear that CPUs that include graphics and stuff are not that good at I/O.
I made the same observation on my blog[0]. It's a shame that decent I/O and fast CPUs/GPUs are rarely if ever combined in ARM systems. What good is a quad core CPU if you can't get any data to it?
I don't think so. Bulk pricing usually lists the per unit price. The quantity is denoted since you can get better pricing for larger quantity orders. For example, I may get a microcontroller for $4 when I order 10k, but you'd pay $25 if you just ordered one from DigiKey.
I didn't realize there was a significant speed difference between SD and SSD technology! After some googling, I found a good video that gives the relevant figures:
In short, the class 10 SD card I got for my Raspberry Pi is rated for 10MB/s. An SSD drive can do 1GB/S (I imagine the numbers vary for reads vs writes). Also, SD cards seem to be able to take a lot fewer writes per block before they wear out.
To be fair, a class 10 SD card could be ten times faster than the figure you quoted. Still a big difference, of course, and I imagine there are some write queueing feature differences as well.
It's quite significant. SD card is fine for the hacker community though. They have different intentions I suppose. SD cards weren't really designed for OS (particularly if you forget to disable a pagefile which will constantly write old memory to "disk" and drastically shorten its life)
1GB/s is a bit much for an SSD, since SATA currently only goes up to 6Gbit/s. PCIe-based SSDs can do it, but for a good SATA SSD, you can expect speeds ranging from 100MB/s to 500MB/s depending on the workload.
Most of the speed advantage SSDs have comes from having a dozen or more flash chips that are internally treated as something like a RAID array. Memory cards and USB sticks usually have only one or two chips.
I had been looking at the ODROID-X (hardkernel.com) and various other single-board machines for a while. Whilst this isn't particularly powerful (ODROID is quad core exynos), it does have an advantage I haven't seen anywhere else - SATA.
Does anyone know of any other boards that have a SATA connector?
The Marvell-based stuff (e.g. SheevaPlug, DreamPlug, D2Plug, OpenRD, CuBox) mostly has SATA (not SheevaPlug) and gigabit Ethernet (though for some reason the GuruPlug Display only has 100mbit). It just tends to be a bit pricier, and mostly lacks an FPU or GPU.
I look forward to the day when I can get a Cortex A9 or A15 board with a USB host good for the full 480mbit/s, SATA, and gigabit Ethernet.
If they are this late in the cycle, they should've gone with the Cortex A7 CPU and Mali 450 GPU. They might even be cheaper than A8 and Mali400, since they are targeted at the mass-market.
I'm sort of amused that an outfit from Shenzhen China, perhaps the worlds great manufacturing centre for options like this, is asking for US dollars to make 1000 boards, when firms like ARMJISHU can knock ARM boards out in no time and sell them cheap as chips.
There's nothing that special about this that I can see. And giving loaned Chinese money back for free when they have such a massive comparative advantage seems an exercise in market distortion to me.
Small innovative Chinese outfits lack capital. There's big, government affiliated outfits with lots of capital, and outsourcing companies (like Foxconn, which is Taiwan run), but not a lot of capital for bootstrapped companies.
OK, I've heard that's changing, but it's not Silicon Valley.
I'm not even sure what "compatibility" means in this context. You can run executables for one on the other? But you can swap out the OS on the Pi, so... I'm just confused.
This kind of thing is what has really annoyed me about the Pi 'movement'. "Raspberry Pi compatible" means nothing. It's a totally standard ARM based linux computer, of which there are a huge (and growing) number.
Intel are just about to release the "NUC" Next Unit of Computing. Looks pretty reasonable, but significantly more powerful and thus expensive than a PI.
There's a few companies that make "pico-ITX" boards that run x86 processors [1]. Mostly low powered chips like Atom, but you can get Core 2 Duo boards if you've got the cash. Prices started around $300 last time I checked.
There's tons of book-sized systems available, typically with lower-end AMD or Intel chips. Throw in a DDR2/3 SODIMM and a SATA drive and you're good to go. These start around $100.
This is good news— the rpi is really an embarrassing system. It is shockingly slow compared to other small arm systems clock per clock. A lot of people seem to be trying things on it that would be fine on almost any other arm SBC but fail on rpi…
There are many alternatives which are much faster but none are within a factor of two of the rpi's price, so people keep buying the rpi and walking away disappointed.
I completely disagree with you calling it an "embarrassing system". I've had my Pi for a couple of weeks now and been happily tinkering with it. For the price I am utterly amazed with it (in fairness, I'm a bit of an old git and amazed that you can buy a 2TB HD for less than £100 so I may not be the best judge here).
Whilst it may not be the quickest thing out there, it is a start, and I think this is the most important thing about the Pi - it is a statr. It is so cheap it will get people interested in and tinkering with this kind of technology. As a result more people will move into this space, surely the Cubieboard is testament to that. (As well as the Parallella[1] which sadly I don't think will make it's Kickstarter goal)
Somehow people keep missing the point: Raspberry Pi is meant to be a nearly-disposable kids' learning toy first and foremost. The object of the game, from the beginning, was to get a programming sandbox into as many small hands as possible—in families that couldn't afford a "real" computer, or couldn't/wouldn't allow kids to try potentially damaging things on the family machine.
It's nice that it can, incidentally, play other roles, but those other roles are not it's purpose. It's not an "embarrassing system"; low cost, "universal" peripherals (common televisions as the "monitor", for example), enough portability to get the unit to and from school are its primary concerns. It's an OLPC for the western world, a BBC Micro reborn for the modern age. Y'all rabbits might like these Trix, but they're for kids.
"in families that couldn't afford a 'real' computer, or couldn't/wouldn't allow kids to try potentially damaging things on the family machine."
How many of those families are going to buy a rpi, have the knowledge to teach set it up and teach their kids, etc? What's going to get kids programming is a really cheap "normal" computer, just like it always has.
Schools were and are the primary target. Just go back and read the story behind Raspberry Pi[1]. The point always was to get back to the days of the BBC Micro and a general introduction to computing fundamentals (as opposed to mere "using the applications" training) for British kids (although kids in other places can benefit as well), based on the notion that broad exposure would mean a larger cohort of people to draw upon as adults in the tech sector (commercial and academic) in later life. Full-featured computers in schools are expensive for this sort of task, and not everybody can take their work home with them to expand upon it—or even just to do homework. At around the price of a textbook, the Raspberry Pi makes that at least possible.
The low price, suggested curriculum, etc., are a big part of making it work. Or is it your contention that computing should be restricted to people who can afford a "proper" computer?
It's amazing how many people totally miss the point about R-Pi. There's so much hate (here and in other forums) because the Pi is not fast enough, or open enough, or supported enough, or available enough.
People don't realize that it was never intended to be a replacement to your desktop computer, or a completely open hardware platform, or widely available at retailers, or a commercial system widely supported. The primary target are schools and young developers, interested in learning. Honestly, kids don't really care if the GPU is open sourced or not. But they'd love things like this: http://www.cl.cam.ac.uk/freshers/raspberrypi/tutorials/os/.
Realize that the A10 does not have mainline kernel support and part of that reason is because the code is so bad no one wants to even try to get Linus to add it.
The Allwinner A10 is a SoC that uses an ARM Cortex A8 as the CPU, a Mali 400 GPU, Ethernet, SATA, etc. Yes it's confusing. It's like calling a hypothetical Intel Core based SoC the Foobar Core 10.
Apple called their ARM SoC the A4 and then Allwinner probably designated their SoC the A10 to play off that marketing.
I've looked at the Cubieboard in the past, its an interesting take.
What strikes me though about this, the Pandaboard, the Chumby, Etc, is that one of the effects of the 'post PC' momentum shift is that people who were building general purpose computers have switched to building these things which are tailored more for browsing and content consumption, and perhaps business content (documents, spreadsheets, reports) generation.
That shift has once again opened up the market for a general purpose hobbiest computer. That is pretty refreshing to see.
Reflecting about all this, I find the rpi design curious. Most of the SoC is wasted and impose heavy side effects (lan over capricious usb). The graphical applications (video processing, etc) seems to be limited by what is already in the GPU blob.
Honest question, isn't there more simple SoC (armv6 + basic 2d gpu) with sane default and i/o chipsets ?
How much of the closed source binary is due to the GPU ?
With a simpler and open source system, I think it would have provided a better learning substrate than what it is now.
Sorry if slightly OT but we organized a "Raspberry Pi's Impact on Hacking" Web Show with forensics expert DJ Palombo. Thought might be of interest to y'all!
82 comments
[ 3.7 ms ] story [ 174 ms ] threadI do wish someone would build an ITX sized motherboard with SATA, actual ram slots, and gigabyte network ports. Perhaps when the 64-bit version is more available someone will.
Or am I missing something? (Ah, you probably want ARM rather than x86[_64]).
The same on one of these ITX boards would take probably 25-30 watts, minimum.
What I'm looking forward to next are Cortex A15 boards, the first CPUs are shipping (In the new Chromebook), so hopefully we'll start seeing them on small hackable boards soon.
1 - http://www.newegg.com/Product/Product.aspx?Item=N82E16820227...
2- http://www.crucial.com/store/partspecs.aspx?IMODULE=CT064V4S...
With WiFi. Other alternatives are at the same site.
I have a few Mali devices next up on the chopping block after I finish reversing the VideoCore IV.
I've only just stumbled across details of your efforts in the past couple of days.
I wish you every success, if you were to be successful, you would open the door to impressively cheap possibilities.
Hehe. Reversing a GPU is a lot of work. From my experience it usually involves zillions of commands, state bits that subtly change handling, and different instruction sets for different kinds of shaders and other sub-processors... (even figuring texture formats can be lots of work) I can't even imagine starting such a project again, though it was fun, wish I'd still have the time...
I wish you the best luck of course!
[1] http://www.embeddedarm.com/products/board-detail.php?product...
I made the same observation on my blog[0]. It's a shame that decent I/O and fast CPUs/GPUs are rarely if ever combined in ARM systems. What good is a quad core CPU if you can't get any data to it?
[0] http://nitrogen.posterous.com/the-ideal-arm-platform
The TS-7800 has a dedicated Ethernet controller, which presumably explains some of the difference.
[0] http://www.smsc.com/index.php?tid=300&pid=135
FTFY ;-)
http://www.youtube.com/watch?v=5fxAQv_KDV0
In short, the class 10 SD card I got for my Raspberry Pi is rated for 10MB/s. An SSD drive can do 1GB/S (I imagine the numbers vary for reads vs writes). Also, SD cards seem to be able to take a lot fewer writes per block before they wear out.
Most of the speed advantage SSDs have comes from having a dozen or more flash chips that are internally treated as something like a RAID array. Memory cards and USB sticks usually have only one or two chips.
I had been looking at the ODROID-X (hardkernel.com) and various other single-board machines for a while. Whilst this isn't particularly powerful (ODROID is quad core exynos), it does have an advantage I haven't seen anywhere else - SATA.
Does anyone know of any other boards that have a SATA connector?
I look forward to the day when I can get a Cortex A9 or A15 board with a USB host good for the full 480mbit/s, SATA, and gigabit Ethernet.
I also look forward to that day. We seem to be at the stage right now that we can have some but not all of those things. Soon, soon...
Wandaboard (http://www.wandboard.org/) should have a SATA connector on future configurations, but nothing concrete has been announced.
There's nothing that special about this that I can see. And giving loaned Chinese money back for free when they have such a massive comparative advantage seems an exercise in market distortion to me.
OK, I've heard that's changing, but it's not Silicon Valley.
It surprises me that Pi compatibility is a criteria for stuff, though I know it's popular. Are we talking pin-compatibility?
AFAICT the Pi is just another linux on a single board device. It's low cost but otherwise fairly unremarkable.
I'm just wondering why compatibility would be an issue. It's open source baby, we can rebuild it!
I recommend reading this page and its links: http://wiki.xbmc.org/index.php?title=Allwinner_A10_devices and http://linux-sunxi.org/Main_Page
The specs seem to match the dime-in-a-dozen Allwinner A10 Chinese tablets/phones/set-top-boxes/...
I have 3 of these A10 devices, and the performance isn't that bad. It's also not great, but it's sufficient for all the things I tried with it.
http://www.redferret.net/?p=33596
[1] http://www2.warwick.ac.uk/fac/sci/eng/meng/wmr/projects/uav/...
Example: http://www.newegg.com/Product/Product.aspx?Item=N82E16856205...
There are many alternatives which are much faster but none are within a factor of two of the rpi's price, so people keep buying the rpi and walking away disappointed.
Whilst it may not be the quickest thing out there, it is a start, and I think this is the most important thing about the Pi - it is a statr. It is so cheap it will get people interested in and tinkering with this kind of technology. As a result more people will move into this space, surely the Cubieboard is testament to that. (As well as the Parallella[1] which sadly I don't think will make it's Kickstarter goal)
[1] http://www.kickstarter.com/projects/adapteva/parallella-a-su...
It's nice that it can, incidentally, play other roles, but those other roles are not it's purpose. It's not an "embarrassing system"; low cost, "universal" peripherals (common televisions as the "monitor", for example), enough portability to get the unit to and from school are its primary concerns. It's an OLPC for the western world, a BBC Micro reborn for the modern age. Y'all rabbits might like these Trix, but they're for kids.
How many of those families are going to buy a rpi, have the knowledge to teach set it up and teach their kids, etc? What's going to get kids programming is a really cheap "normal" computer, just like it always has.
The low price, suggested curriculum, etc., are a big part of making it work. Or is it your contention that computing should be restricted to people who can afford a "proper" computer?
[1] http://www.raspberrypi.org/about
It's amazing how many people totally miss the point about R-Pi. There's so much hate (here and in other forums) because the Pi is not fast enough, or open enough, or supported enough, or available enough.
People don't realize that it was never intended to be a replacement to your desktop computer, or a completely open hardware platform, or widely available at retailers, or a commercial system widely supported. The primary target are schools and young developers, interested in learning. Honestly, kids don't really care if the GPU is open sourced or not. But they'd love things like this: http://www.cl.cam.ac.uk/freshers/raspberrypi/tutorials/os/.
Apple called their ARM SoC the A4 and then Allwinner probably designated their SoC the A10 to play off that marketing.
What strikes me though about this, the Pandaboard, the Chumby, Etc, is that one of the effects of the 'post PC' momentum shift is that people who were building general purpose computers have switched to building these things which are tailored more for browsing and content consumption, and perhaps business content (documents, spreadsheets, reports) generation.
That shift has once again opened up the market for a general purpose hobbiest computer. That is pretty refreshing to see.
Honest question, isn't there more simple SoC (armv6 + basic 2d gpu) with sane default and i/o chipsets ?
How much of the closed source binary is due to the GPU ?
With a simpler and open source system, I think it would have provided a better learning substrate than what it is now.
The Pi team addresses this on their site. I think the answer to your question is "a lot."
Here is the link - http://www.concise-courses.com/infosec/20121106/
Thanks Henry