Is there any kind of cross-over here, with GreenArrays processors? Is it feasible to have an Erlang VM running on a GA chip, and would it have any advantages over the current (typically x86) hardware?
(From the Erlang website (http://erlang.org/faq/implementations.html): "Getting Erlang to run on, say, an 8 bit CPU with 32kByte of RAM is not feasible. People successfully run the Ericsson implementation of Erlang on systems with as little as 16MByte of RAM. It is reasonably straightforward to fit Erlang itself into 2MByte of persistant storage (e.g. a flash disk).")
Another question: the computing demands of self-driving cars are somewhat self-defeating (https://www.bloomberg.com/news/articles/2017-10-11/driverles...). Isn't low-power computer vision a perfect match for GA chips...?
Sure, I can understand why GA went with 144 cores. Because of their insistence on software defined i/o, you need them. Smarter move would have been GA4+I2C allowing any number of them to be glued using their asynchronous i/o.
Whatever. Their fate is not my problem. Would have loved for them to be insanely successful.
Computing is so ing boring these days.
It's way to expensive to put into a consumer product (adding another cell is way cheaper than shaving 100uA).
For niche products (can afford a $20 uC on the BOM), development cost is non-trivial, so the excellent MSP430 documentation and support negates any advantage the GreenArray product might have.
In essence, it's too expensive for the low and high end (albeit in different units of cost) so who do they go after?
If they want to fix #2, don't use a 10 chip MOQ ($200) and charge $450 for the demo board. Hand out a few at cost to some ECE departments or at a conference and let people do cool things with your tech. Their trials an tribulations are free publicity and support material.
So much of the stuff out there is just a variation on a well-worn theme. A multi-core stack machine with 18-bit words implemented in asynchronous logic is definitely a road not taken. It's nice to see people sticking their necks out--especially in the space of tangibles where starting costs are usually much higher.
The advantage of the GA144 is not that it has slightly better specs. It is a completely new paradigm. This is basically a new zeppelin that can take you from one continent to another in a comparable time to a commercial flight, and you are complaining because the price of the ticket is much higher.
I do not think the point of these new chips is not make a dent in the market (either for consumer products or enthusiasts), but to show that radically different computing systems are possible. Maybe it is not directly profitable, but it surely is mind-opening.
My wording is used.
If the GA144 can actually be better than existing solution, there’s thousands of engineers who would happily try it. But until GreenArray can demonstrate a compelling use case, an unmet need, something — the product won’t make it very far.
MSP430's aren't trying to be energy efficient per operation - they are really bad at processing. What they are good at is running complex circuits/peripherals at a low total energy consumption over a long time.
The fact that they're running at 8Mhz is telling... that's far from the "low power" level. As far as I recall, you generally run using internal low-frequency RC oscillator for low power, and then just boot up the high speed crystal oscillator if you really need it.
These really are completely different type of computers - I believe if you can apply the MSP430 successfully to your application then the GA144 is probably the wrong chip to use.
But what if you need real-time nano-second reaction times on many separate pins? What if you need to process a 30 Mhz signal? While controlling a display and accepting input? All at the same time?
Then you might need the GA144, which can do all those things at the same time without needing to worry about interrupts or waking up from low power sleep modes or any of the other complex mechanism computers employ to minimize power loss.
I can't imagine the utility of low-power processors that
"aren't trying to be energy efficient per operation"
10 seconds a week running a WiFi radio, TCP/IP, SSL and all that. 604,790 seconds a week waiting for a button to be pressed. Battery powered.
If you can monitor a button on 1 microamp, and run WiFi on 60 milliamps, 50% of your battery capacity will go on sleeping and 50% on waking.
And wake-state power consumption is dominated by the radio module, so the best way to cut down on wake state power consumption is to make the wake as short as possible.
GreenArrays: You can use our chips if you unlearn EVERYTHING YOU KNOW ABOUT SOFTWARE AND HARDWARE.
That said, it's amazing hardware