Have you had a chance to try these tools in comparison to a pure HDL approach?
Saying "You can get much higher clock rates, using much less FPGA real estate, if you avoid the C-based tools and stick to real HDLs." is a pretty bold statement.
Without a doubt there are cases where an expert HDL designer will be able (or even needed) to optimise code, perhaps get critical timings right or to force a design onto a particular target device etc. But such HDL experts are in short supply in even the best resourced teams, letting the software/firmware types in on the design to break the back of it with support for fine tweeks from a HDL guru can produce good results very fast. This is especailly the case where you are trying to get a balance between what you do on a micro and what you do in FPGA logic.
From what I have seen, including the results of head to head coding by parallel development teams, the C based approach stacks up well. Ultimately it depends on your application and alogrithms - and of course on what your most imprtant design constrains are, speed, area used etc.
As for not touching it with a 10 foot pole in my view that's a mistake, embrace it, even if just to do the grunt work, let then use it to tweek and experiment faster...then take the glory!
I'm not saying that the C based tools are a panecea just that, used right, they can be a very powerful asset.
Celoxica's main product is a development environment that basically lets you write C code and drop it onto an FPGA. (Yes it runs on Linux.)
Its a lot faster to develop this way than more traditional methods (HDL's) as its so easy to iterate, for example being able to drag code back and forth to optimise the flow between a processor on your board and an FPGA being used as a custom parallell coprocessor is pretty cool.
As for the demos, that ray trace one is pretty cool, but I did like the space invaders demo - I think the game code was from a ROM dump - you even got an insert coin prompt!
Its pretty handy when you want to ditch cables for laptop / pda to phone connections and for in car use with a BT car kit ot head set - you can upgrade you phone without dumping the car kit for a start! Can be a pain in the ass to set up with some devices though.
"There is no way you're going to be able to stuff an 802.11b/a transceiver into an already high priced, low battery life phone."
Hmm right. Never heard of Moores law then?
Take a look at this:
An IP company who are already well on the way to doing this. I'd put money on a combined 802.11 / GPRS (or 3G) smartphone being avaliable for about £300 before the US has a nationwide cellphone system.
Anyway GPRS / GSM / 802.11 PC cards are ALREADY in the market.
"If we had a set protocol for doing all things wireless, then it wouldn't be a matter of what physical network you're on, even what type of network you're using or who owns it."
I can't help thinking a protocol that was all things to all men would be complete overkill for many wireless applications - for example a lot of non critical telemetry applications don't even need bi-directional communication. On the other side of the coin a safety critical applications would require lots of error checking, redundancy of data etc. which would be complete over kill for internet type applications where the whole system is set up to just retry if packets are lost - if there were multiply redundant copies of packets specified in the standard that could get really slow.
If this sort of scheme were to be implemented it owuyld make more sense to do it whenever 2am to 4am is local to the site's country code. (I know that that is not necessarily its location or area served.)
This would also allow the spiders to work 24/7 as opposed to a few hours a day as you suggest.
How long after this becomes avaliable will the first supeona for full access be issued - for example in a divorce court, patent dispute antitrust case...
Lynux Works has just announced that it will support the Open Source Development Lab's Carrier-Grade Linux® specification (CGL) v1.0 and that it plans to features of CGL v1.0 into BlueCat® Linux, in early 2003. The details here (Press release).
Nice fact, all the defenses were designed to keep the general public out in times of strife and the one time it was fully stocked and ready to be buttoned up.... during our miners strike in the 1980's
Now as a Brit this is not my field, but I thought that NASA was the "National AERONAUTICAL and Space Administration".
I think the clue to "Why is NASA spending money on this." might be in their name.
Wether the research is worthwhile I can't say, but it does look like it is within the remit and it has certainly got them plenty of PR bang for thier buck.
It sends data at very low power across a very wide range of frequencies, avoiding all of the used frequencies, or even just many of them would kind of defeat the objective.
Slashdot Mirror
Another new design called B'ngo,a sub $200 handheld console / phone is covered on The Register.
"Why Jupiter, who's size compared to Earth is immense, has not had a large asteroid annihalate it yet? (ok, maybe it has, i don't know)"
It has been hit, very spectacularly, had it been Earth that was hit I doubt we would be discussing this.
See:
Comet Shoemaker-Levy Collision with Jupiter
Have you had a chance to try these tools in comparison to a pure HDL approach?
Saying "You can get much higher clock rates, using much less FPGA real estate, if you avoid the C-based tools and stick to real HDLs." is a pretty bold statement.
Without a doubt there are cases where an expert HDL designer will be able (or even needed) to optimise code, perhaps get critical timings right or to force a design onto a particular target device etc. But such HDL experts are in short supply in even the best resourced teams, letting the software/firmware types in on the design to break the back of it with support for fine tweeks from a HDL guru can produce good results very fast. This is especailly the case where you are trying to get a balance between what you do on a micro and what you do in FPGA logic.
From what I have seen, including the results of head to head coding by parallel development teams, the C based approach stacks up well. Ultimately it depends on your application and alogrithms - and of course on what your most imprtant design constrains are, speed, area used etc.
As for not touching it with a 10 foot pole in my view that's a mistake, embrace it, even if just to do the grunt work, let then use it to tweek and experiment faster...then take the glory!
I'm not saying that the C based tools are a panecea just that, used right, they can be a very powerful asset.
Celoxica's main product is a development environment that basically lets you write C code and drop it onto an FPGA. (Yes it runs on Linux.)
Its a lot faster to develop this way than more traditional methods (HDL's) as its so easy to iterate, for example being able to drag code back and forth to optimise the flow between a processor on your board and an FPGA being used as a custom parallell coprocessor is pretty cool.
As for the demos, that ray trace one is pretty cool, but I did like the space invaders demo - I think the game code was from a ROM dump - you even got an insert coin prompt!
So if you want a new one that'll be $250k for a single user licence.
Licence agreement:
No licence transplants allowed.
Separate work and recreational licence required.
This design is not tested for mission critical applications, any usage in mission critiacal applications is at the licencee's own risk.
If you work at Boeing or Airbus that could be in the same building. *Grin*
Just don't play counter strike.
Getting caught by the crew discussing mission orders and weaponry could be bad news...
A user of the service on the flight was posting to the Professional Pilot Rumour Network PPRUNE www.pprune.com , the thread is here.
Bluetooth is no substitite for a cellular network but as for the range try 100m range i.e. 300ft as a maximum with 10m 30ft as the lower bound.
It all depends on the class of the device.
See Blue tooth specs.
or
article discussing bluetooth range.
Its pretty handy when you want to ditch cables for laptop / pda to phone connections and for in car use with a BT car kit ot head set - you can upgrade you phone without dumping the car kit for a start! Can be a pain in the ass to set up with some devices though.
ooops for got to paste the link! 802.11 / telecoms IP story.
"There is no way you're going to be able to stuff an 802.11b/a transceiver into an already high priced, low battery life phone."
Hmm right. Never heard of Moores law then?
Take a look at this:
An IP company who are already well on the way to doing this. I'd put money on a combined 802.11 / GPRS (or 3G) smartphone being avaliable for about £300 before the US has a nationwide cellphone system.
Anyway GPRS / GSM / 802.11 PC cards are ALREADY in the market.
"If we had a set protocol for doing all things wireless, then it wouldn't be a matter of what physical network you're on, even what type of network you're using or who owns it."
I can't help thinking a protocol that was all things to all men would be complete overkill for many wireless applications - for example a lot of non critical telemetry applications don't even need bi-directional communication. On the other side of the coin a safety critical applications would require lots of error checking, redundancy of data etc. which would be complete over kill for internet type applications where the whole system is set up to just retry if packets are lost - if there were multiply redundant copies of packets specified in the standard that could get really slow.
A fixed time for this would be dumb.
If this sort of scheme were to be implemented it owuyld make more sense to do it whenever 2am to 4am is local to the site's country code. (I know that that is not necessarily its location or area served.)
This would also allow the spiders to work 24/7 as opposed to a few hours a day as you suggest.
to get /.'ed
How long after this becomes avaliable will the first supeona for full access be issued - for example in a divorce court, patent dispute antitrust case...
Do you have clams that slide open in the US?
You must have otherwise saying "clamshell-case PDA as embodied by the Zaurus 5600" would be complety wrong.
Lynux Works has just announced that it will support the Open Source Development Lab's Carrier-Grade Linux® specification (CGL) v1.0 and that it plans to features of CGL v1.0 into BlueCat® Linux, in early 2003. The details here (Press release).
So you are saying:
747 = Linux = suitable for mass usage (i.e. transporting millions of passengers every week.) now an established part of every day life.
Shuttle = Windows = tempramental, specialist technology,costs millions of $ per use and is only operable by an elite highly trained few.
Er... ok if you say so, not the ususal argumnent I have heard from either side of the lin/win battle, but I'm all for free speech...
It's not a micorwave relay site, that's just part of it!
It was to be the seat of government for a post WWIII UK.
The Bunker Website here has loads of info.
Nice fact, all the defenses were designed to keep the general public out in times of strife and the one time it was fully stocked and ready to be buttoned up.... during our miners strike in the 1980's
Very good...
So would you forsee them using flooding of the Maskable Furniture Interupt or the Multiplexed Furniture Interface?
I'll stop now.
...yell and fight the packer if they try and ship it without all it's screws?
And if so will MFI (My Furniture is Incomplete) change their name?
Now as a Brit this is not my field, but I thought that NASA was the "National AERONAUTICAL and Space Administration".
I think the clue to "Why is NASA spending money on this." might be in their name.
Wether the research is worthwhile I can't say, but it does look like it is within the remit and it has certainly got them plenty of PR bang for thier buck.
UWB means Ultra Wide Band.
It sends data at very low power across a very wide range of frequencies, avoiding all of the used frequencies, or even just many of them would kind of defeat the objective.
Ridiculous, banning headphones will not solve this, they are an innocent party, we need to get rid of the passengers....
I asked as it has never been my experience, while flying on Go, Ryan Air, BA, Deutche BA and KLM in the past two years.
Mind you as the Captian has the final say, it could well be variable by flight - or perhaps I was sleeping!
Personally I usually use a Psion when flying, much smaller and a 30hour usable battery life!
Er really?
Can you list those airlines you know this is the case for?