Oh - I was more thinking like it was a ARM board running a full Linux stack, that had the headers you need to plug random bits and bobs in, to so stuff... my mistake.
Let me just duct-tape an Arduino to my Cubieboard and jumper the power and serial ports together...
As FPGA Arcade are conversions of hardware to run in an FPGA each conversion requires a lot more work than software emulation - but it can be done! Drag down some arcade schematics of the Web, get a cheap FPGA board and get started - see this post
For each different hardware platform I guess it takes three months of spare time to convert - as long as usable implementations for the major ICs can be found!
I've got a US$60 pcDuino, which is close to this. Sort of like a Raspberry Pi with Arduino(ish) I/O headers - they just lack the same spacing so an interposer is required. Runs Andriod, Linux and XBMC just fine...
The devil is in the detail. I wonder how good their Arduino work-a-like API library will be... with the ADCs have the same resolution? Will timing sensitive bit-banged I/O still work OK? Will PWM be the same?
Flying drones outside of visual range is illegal where I live, so I would not be able to say anything if I was to have have seen HD footage filmed from a home-built fixed wing drone that has flown on autopilot for 60 minutes, at 60km/hr and then landed back at the starting point entirely on autopilot.
FWIW I have no problem playing 1080p content on my CubieBoard using the Mali400 hardware acceleration on either Linux (the XBMC image) or Andriod, but that is based around the AllWinner A10 SoC, a completely different chip to this (but it is still a Dual Core ARM)
What is more likely? an explosive device, loaded with a cell phone that required a working/registered/traceable SIM, or a generic clockwork egg-timer or other mechanical clock, or if you are really high tech maybe a small micro like an Arduino, or just an RTC clock module, using the alarm signal?
After all, the device didn't have to explode the second that a motorcade drove past, just sometime in the next few hours would do fine
What I don't understand is why why would you time it when most of the field have been through? Surely for maximum impact you would want to have exploded it earlier?
This change of packaging allows greater memory density, and maybe higher transfer bandwidths. It will not alter the "first word" latency much, if at al.
Signal propagation over the wires isn't the problem, it is the way all DRAM works is.
- The DRAM arrays have "sense amplifiers", used to recover data data from the memory cell. The are much like op-amps, To start the cycle both inputs on the sense amplifier are charged to a middle level, - The row is opened, dumping any stored charge into one side of the sense amplifier. - The sense amplifiers are then saturate the signal to recover either a high or low level. - At this point the data is ready to be accessed and transferred to the host (for a read), or values updated (for a write). It is this part that the memory interconnect performance really matters (e.g. Fast Page mode DRAM, DDR, DDR2, DDR3). - One the read back and updates are completed then the row is closed, capturing the saturated voltage levels back in the cells.
And then the next memory cycle can begin again. On top of that you have have to add in refresh cycles, the rows are opened and closed on a schedule to ensure that the stored charge doesn't leak away, consuming time and adding to uneven memory latency.
To actually build a memory controller is another step up again - RAM chips have configuration registers that need to be set, and modules have a serial flash on them that holds device parameters. With high speed DDR memory you have to even make allowances for the different lengths in the PCB traces, and that is just the starting point - the devices still need to perform real-time calibrate to accurately capture the returning bits.
Forgot to say, the pcDuino is US$59 + shipping, the Cubieboard is US$65+shipping (and that included cables and case). Both came in about a week to New Zealand.
This website is the center of the world for the Allwinner A10 SoC that these boards and quite a few other systems are based on, and individual boards have their own board specific forums too (e.g. http://www.pcduino.com/) . As the A10 SoC is used in a lot of low-end Android tablets I am pretty sure that it will have shipped more units than Raspberry Pi, (although not as many in hardware hacker firendly boards like this).
I am playing with a pcDuino from Gadget Factory, and they are also available from Sparkfun dealers.
At US$59 it is a bit more expensive, but does have a few advantages
- Onboard 2GB of Flash to hold an OS - no SD card needed! - Decent Power supply system for USB ports - 1GHz A8 CPU - More than 2x as fast - 1GB RAM - Mali 400 graphics accelerator - Android image available - No need for an I/O expander for hardware hacking
Once you take off the cost of a powered USB hub and an SD card for the making the Pi usable I think it is price neutral between the two.
I've been using Android on it to play back 1080p files (at 720p) without a problem, and play Angry Birds Star Wars... can't do that on a Pi.
The Australian company Miniand make similar boards, some with cases and so on. Their Cubieboard even has SATA header on it, and I've got one with a 2.5" 120GB disk hanging off it...
You should look at the other ARM boards out there e.g. pcDuino. More memory, more I/O, onboard flash, Linux or Android.
However there are some things that you can do with a micro-controller that can not be done with a full OS - e.g. bit-banging I/O to one-wire temperature sensors. I've even used a full USB 1.1. HID driver implemented completely in software, which would be impossible with an full OS running!
Every other year I have to remove fried mice out of my in-wall stove's wiring, In autumn they try to come in side and look for a nice warm place for winter. I guess they find the oven before they find the mousetraps.
This link is a really interesting info on some of the SKA signal processing.
The SAK's power budget is 58MW for signal processing - this is such a high running cost that by spending 30 Million Euro on developing a few custom ASICs to halve that power usage will pay off in 9 months!
* 1GHz ARM Cortex A8 * 1GB RAM * 2GB Onboard Flash (no SD card required for software, and faster!). * Mali 400 graphics core * Way more on-board GPIO, including analogue inputs
If you truly believe that then you should get an FPGA development kit and start your revolution. The are excellent at evolving a large state vector over time,
A large FPGA may have 2.5 megabits of state and clock rates of 500MHz, so you can update over 1,000,000,000,000,000 bits of state every second - about 1000 times what a CPU can do., They also have 300 pins of I/O, and a dozen interfaces of > 10Gb/s and the logic transform can be dynamically reprogrammed. Oh, can have 5,000 GigaMACs (multiply + accumulates per second) - about 4,000x that of a CPU core.
Maybe get yourself a small one - small development boards are under $100. Have a go - let me know how you get on. The same board can implement a system equivalent to a 486 CPU - 32 bits, 100MHz. Should be enough to get started with.
Here is another idea for you. What if some basic unit of the universe is in some way a signed binary number. If you fill a block of bits with random values and then average them as if they are n-bit signed integers then you always end up with an average of -0.5 (as the range of values is from -2^(n-1) to 2^(n-1)-1).
So in a uniformly,random field which has an underlying signed binary number representation there will always be an energy imbalance....
In his book "Cycles of time", Roger Penrose tries to look before and beyond the Big Bang.
The general idea is that when the universe will suffer a cold death - all the atoms have decayed away, all the black holes evaporated and all the photons have redshifted, The universe looses then loses all dimension, and becomes an very small, uniform, very hot thing, containing all the energy of the prior universe. A new big bang.
We may be the inflationary period of the next Universe!
The interesting thing is that cosmic events in the old universe should have left traces on the distribution of energy in the new one.
In his book Cycles of Time, Roger Penrose attempts to look before the Big Bang, and after the end of our Universe.
The general idea is that in the far future the universe is so uniform and cold that it becomes completely uniform, with no sense of scale. All the block holes have evaporated, all the sub atomic particles have decayed away into photons. At this point the universe undergoes spontaneous rescaling, into a very compact, bounded, hot uniform object, busting with all the energy that existed in the original universe.
If I read it correctly, this could be interpreted as the cold death of our universe is the inflationary period of the following one, and the rescaling event is the big bang.
The interesting thing is that he makes testable predictions. The ghosts of energy ripples of cosmic events the old universe should be imprinted on the structure of the following genesis.
Slashdot Mirror
Niggard : (noun) an excessively parsimonious, miserly, or stingy person. See here
Oh - I was more thinking like it was a ARM board running a full Linux stack, that had the headers you need to plug random bits and bobs in, to so stuff... my mistake.
Let me just duct-tape an Arduino to my Cubieboard and jumper the power and serial ports together...
As FPGA Arcade are conversions of hardware to run in an FPGA each conversion requires a lot more work than software emulation - but it can be done! Drag down some arcade schematics of the Web, get a cheap FPGA board and get started - see this post
For each different hardware platform I guess it takes three months of spare time to convert - as long as usable implementations for the major ICs can be found!
I've got a US$60 pcDuino, which is close to this. Sort of like a Raspberry Pi with Arduino(ish) I/O headers - they just lack the same spacing so an interposer is required. Runs Andriod, Linux and XBMC just fine...
The devil is in the detail. I wonder how good their Arduino work-a-like API library will be... with the ADCs have the same resolution? Will timing sensitive bit-banged I/O still work OK? Will PWM be the same?
Flying drones outside of visual range is illegal where I live, so I would not be able to say anything if I was to have have seen HD footage filmed from a home-built fixed wing drone that has flown on autopilot for 60 minutes, at 60km/hr and then landed back at the starting point entirely on autopilot.
Have a look at the Cubieboard - US$49 with SATA, 1GB RAM and 4GB Flash - both twice that of this board. It also has an IR sensor, and 96 pins of I/O
FWIW I have no problem playing 1080p content on my CubieBoard using the Mali400 hardware acceleration on either Linux (the XBMC image) or Andriod, but that is based around the AllWinner A10 SoC, a completely different chip to this (but it is still a Dual Core ARM)
All I can think is "How do you like your eggs? Fried of fertilized?"
What is more likely? an explosive device, loaded with a cell phone that required a working/registered/traceable SIM, or a generic clockwork egg-timer or other mechanical clock, or if you are really high tech maybe a small micro like an Arduino, or just an RTC clock module, using the alarm signal?
After all, the device didn't have to explode the second that a motorcade drove past, just sometime in the next few hours would do fine
What I don't understand is why why would you time it when most of the field have been through? Surely for maximum impact you would want to have exploded it earlier?
This change of packaging allows greater memory density, and maybe higher transfer bandwidths. It will not alter the "first word" latency much, if at al.
Signal propagation over the wires isn't the problem, it is the way all DRAM works is.
- The DRAM arrays have "sense amplifiers", used to recover data data from the memory cell. The are much like op-amps, To start the cycle both inputs on the sense amplifier are charged to a middle level,
- The row is opened, dumping any stored charge into one side of the sense amplifier.
- The sense amplifiers are then saturate the signal to recover either a high or low level.
- At this point the data is ready to be accessed and transferred to the host (for a read), or values updated (for a write). It is this part that the memory interconnect performance really matters (e.g. Fast Page mode DRAM, DDR, DDR2, DDR3).
- One the read back and updates are completed then the row is closed, capturing the saturated voltage levels back in the cells.
And then the next memory cycle can begin again. On top of that you have have to add in refresh cycles, the rows are opened and closed on a schedule to ensure that the stored charge doesn't leak away, consuming time and adding to uneven memory latency.
If you think that modern memory is simple send an address and read or write the data you are much mistaken.
Have a read of What every programmer should know about memory and get a simplified overview of what is going on. This too is only a simplification of what is really going on.
To actually build a memory controller is another step up again - RAM chips have configuration registers that need to be set, and modules have a serial flash on them that holds device parameters. With high speed DDR memory you have to even make allowances for the different lengths in the PCB traces, and that is just the starting point - the devices still need to perform real-time calibrate to accurately capture the returning bits.
Roll Serial Port Memory Technology!
Wow! I never new of that discovery - p00kiethebear is going to be feeling pretty foolish about now!
He has a lot of flowers to buy, and taking a trip out there is a great idea!
Forgot to say, the pcDuino is US$59 + shipping, the Cubieboard is US$65+shipping (and that included cables and case). Both came in about a week to New Zealand.
This website is the center of the world for the Allwinner A10 SoC that these boards and quite a few other systems are based on, and individual boards have their own board specific forums too (e.g. http://www.pcduino.com/) . As the A10 SoC is used in a lot of low-end Android tablets I am pretty sure that it will have shipped more units than Raspberry Pi, (although not as many in hardware hacker firendly boards like this).
I am playing with a pcDuino from Gadget Factory, and they are also available from Sparkfun dealers.
At US$59 it is a bit more expensive, but does have a few advantages
- Onboard 2GB of Flash to hold an OS - no SD card needed!
- Decent Power supply system for USB ports
- 1GHz A8 CPU - More than 2x as fast
- 1GB RAM
- Mali 400 graphics accelerator
- Android image available
- No need for an I/O expander for hardware hacking
Once you take off the cost of a powered USB hub and an SD card for the making the Pi usable I think it is price neutral between the two.
I've been using Android on it to play back 1080p files (at 720p) without a problem, and play Angry Birds Star Wars... can't do that on a Pi.
The Australian company Miniand make similar boards, some with cases and so on. Their Cubieboard even has SATA header on it, and I've got one with a 2.5" 120GB disk hanging off it...
Or maybe a cheap small ARM DC powered board like a pcDuino or a cubieboard, run Linux on it, do whatever you want.
You should look at the other ARM boards out there e.g. pcDuino. More memory, more I/O, onboard flash, Linux or Android.
However there are some things that you can do with a micro-controller that can not be done with a full OS - e.g. bit-banging I/O to one-wire temperature sensors. I've even used a full USB 1.1. HID driver implemented completely in software, which would be impossible with an full OS running!
Every other year I have to remove fried mice out of my in-wall stove's wiring, In autumn they try to come in side and look for a nice warm place for winter. I guess they find the oven before they find the mousetraps.
This, however, never makes it to Slashdot...
This link is a really interesting info on some of the SKA signal processing.
The SAK's power budget is 58MW for signal processing - this is such a high running cost that by spending 30 Million Euro on developing a few custom ASICs to halve that power usage will pay off in 9 months!
The pcDunio is just like a Raspberry Pi, but...
* 1GHz ARM Cortex A8
* 1GB RAM
* 2GB Onboard Flash (no SD card required for software, and faster!).
* Mali 400 graphics core
* Way more on-board GPIO, including analogue inputs
Home page is here
Ordered one today from Sparkfun... US$59. A little bit more expensive, but no GertBoard or SD card required.
If you truly believe that then you should get an FPGA development kit and start your revolution. The are excellent at evolving a large state vector over time,
A large FPGA may have 2.5 megabits of state and clock rates of 500MHz, so you can update over 1,000,000,000,000,000 bits of state every second - about 1000 times what a CPU can do., They also have 300 pins of I/O, and a dozen interfaces of > 10Gb/s and the logic transform can be dynamically reprogrammed. Oh, can have 5,000 GigaMACs (multiply + accumulates per second) - about 4,000x that of a CPU core.
Maybe get yourself a small one - small development boards are under $100. Have a go - let me know how you get on. The same board can implement a system equivalent to a 486 CPU - 32 bits, 100MHz. Should be enough to get started with.
I've read the links, and that is an awfully long bow to draw.
The use of encryption is to try and limit information to those that are intended to see it.
None of the ideas on your blogs address how to "end the Turing madness" in a way that will still allow you to post on Slashdot.
Here is another idea for you. What if some basic unit of the universe is in some way a signed binary number. If you fill a block of bits with random values and then average them as if they are n-bit signed integers then you always end up with an average of -0.5 (as the range of values is from -2^(n-1) to 2^(n-1)-1).
So in a uniformly,random field which has an underlying signed binary number representation there will always be an energy imbalance....
In his book "Cycles of time", Roger Penrose tries to look before and beyond the Big Bang.
The general idea is that when the universe will suffer a cold death - all the atoms have decayed away, all the black holes evaporated and all the photons have redshifted, The universe looses then loses all dimension, and becomes an very small, uniform, very hot thing, containing all the energy of the prior universe. A new big bang.
We may be the inflationary period of the next Universe!
The interesting thing is that cosmic events in the old universe should have left traces on the distribution of energy in the new one.
So the idea is testable.
In his book Cycles of Time, Roger Penrose attempts to look before the Big Bang, and after the end of our Universe.
The general idea is that in the far future the universe is so uniform and cold that it becomes completely uniform, with no sense of scale. All the block holes have evaporated, all the sub atomic particles have decayed away into photons. At this point the universe undergoes spontaneous rescaling, into a very compact, bounded, hot uniform object, busting with all the energy that existed in the original universe.
If I read it correctly, this could be interpreted as the cold death of our universe is the inflationary period of the following one, and the rescaling event is the big bang.
The interesting thing is that he makes testable predictions. The ghosts of energy ripples of cosmic events the old universe should be imprinted on the structure of the following genesis.
...you cycle.