Also you must learn your synthesis tool. While i'm sure there are cases where code synthisises without any hint as to things being done wrong there are also many cases were it is user error. You really do need to learn how to use timing constraints and the timing analyser that comes with your tool and which warnings are important.
Even if this is true (would be good to have a citation so we can see what assumptions are made), flying cars have a couple of big problems.
The first is fuel, any kind of non-winged craft is going to require huge ammounts of it. Even a folding light aircraft is probablly going to guzzle a lot more fuel than your average car.
The second is training and maintinace. If something goes wrong with your car you either stop where you are or coast to the hard shoulder and stop. If something goes wrong in a winged aircraft you can possiblly glide if there is a usable landing site near enough. If something goes wrong in a non-winged craft you are likely screwed.
Whether or not that "dumb" approach is effective depends entirely where you are on the current exponential curve of processing power. And whether that exponential trend continues and how close the complexity estimates are to reality.
VHDL has an incrediablly anal-retentive type system and has some silly ideas like the splitting of entities and architectures. It's also old meaning you need a lot of boilerplate to manually import STD_LOGIC stuff (which is what the synthisis tool vendors tell you to use for everything)
Whichever you use be aware that both VHDL and Verilog weredesigned as hardware simulation languages not hardware synthisis languages. This means it is vital to get to know the synthisis tool you will be using. It is vital to know what warnings matter and what can be safely ignored and which are important, what parts of the language must be avoided, how to assign clocks to clock nets, how to use the timing analyser (without timing alalysis you don't know if your code will actually work) and so on if you are going to use the languages for FPGA programming.
And I don't buy the limited lab time either. The basic lab sciences are always empty, you just go there out of class time. I'm from an EEE not a science background but I know that in my department first and second years are only allowed in the labs during thier official lab sessions. Third and fourth years are allowed in the general labs to do project work at other times but still don't have the access to equiment that would be needed to repeat thier lab experiments (most lab experiments are done with experiment specific equipement and/or in more specialist labs)
So if you get bad results here in an undergrad lab and don't notice them until you come to writeup/analyse (which is not unlikely because a lot of the equipment is old and unreliable and three hours is often only barely enough to complete the experiment) your options are to speculate as to why things went wrong or fudge your results. The former is the honest thing to do but the latter will probablly get you better marks.
Desktops, of any kind, now make up around 40% of total computer sales. So less than half but still a significant ammount.
Not only is this not a huge market On what do you base this assertion? Afaict when you look at the desktop PC market towers dominate with all-in-ones and small form factor boxes being niche markets.
I can see why apple doesn't release an ordinary desktop (releasing one would take away many peoples excuse to spend thier employers money on a mac pro) but that won't stop people snubbed by it from finding other (legally questionable) soloutions.
Because Macs tend to be a trifle more costly than your average Windows box? Look, I'm a Mac user. I love MacOS X. But I can build a PC with better specs (and better expandability) than the Mac mini for about the same cost Heck you don't even have to build it, you can buy one from dell for considerablly less than a mac mini which has lower specs. see my post at http://yro.slashdot.org/comments.pl?sid=1245805&cid=28115689 for a detailed comparison.
If you want OS X that bad why not just buy a Mac? Because the mac pro costs more than most geeks are prepared to spend and the mini and imac have no expansion room.
By Microsoft doing this, the increased sales to businesses would more than offset the losses due to piracy. This also would allow Windows 7 on networks that just don't connect to the Internet at all, and where a reimage of a box doesn't mean calling MS for a manual activation for each non connected machine. According to MS you can telephone activate a KMS and then use that KMS to activate all the boxes on the non-internet connected network.
MS knows they have to release a rock bottom price version to remain competivive in the netbook/nettop market. At the same time they don't want people using the netbook version on higher end machines. Those requirements look to me mainly like an attempt to draw a line between netbooks and regular laptops.
I'm sure some netbook manufacturers will offer netbooks with specs that exceed theese bundled with either linux or a higher version of windows (HP already do this, you can get the mini 2140 with 2G of ram but only if you don't choose XP home).
Maybe some did but not all, I have a friend with a highish end card that took over a gigabyte of address space pushing usable ram under XP down to 2.5G.
only if you live in a country with horriblly wimpy power circuits;)
but in general I agree the power consumption of higher end PCs is getting kinda crazy. Just as CPU power consumption stabalised and even dropped slightly (with the move from P4 to core 2) graphics power consumption started going thorugh the roof.
if your talking a cluster of 100kva diesel gensets then sure it might well take a few months to have it all up and running. lets say a 1U server uses 100W and aircon and UPS overheads push that up to 200W. You can fit about 40 1U servers in a standard rack (with say a network switch and console/kvm server in the remaining two slots) so that is about 8KW per rack.
So assuming your UPS's present a unity power factor to the supply (modern large ones are pretty close afaict) a 100KVA generator would do about 12.5 racks. I don't have figures handy but I bet large datacenters have a lot more racks than that.
Afaict the issue is thet all the fitting out of the building (power, aircon, raised floors etc) and ordering of things like massive power feeds from the local powerco, massive network infrastructure from the local communication provider(s) and massive generators takes time. You can't just buy a generic building and have a high grade datacenter in a month or two.
I don't use Java, but I hear you can marshal Java data for use by C libraries and vice versa. If the security manager lets you then you can. If you are introducing untrusted code to your VM then you set up the security manager not to allow the untrusted code to do that (as for example the java plugin does for untrusted applets).
As serial ports begin to diappear from laptops, and only ONE is available on most modern PCs rather than the two of old, I've long thought we are over-due for some other standard to develop for headless comms. Of the vaious applications for serial ports terminal emulation is one of the most ameniable to USB to serial converters since it doesn't require any precise timing and just generally deals in two streams of data in opposite directions.
Some kit (e.g. the sheevaplug) even puts the USB to serial chip inside the device.
The great thing about serial ports as console ports go is they are simple to drive. That means they can be activated very early in the boot process without needing complex dedicated hardware.
Wow. What an enormous difference. Paying a "service fee" instead of a "monthly rental fee" Afaict the way phone service used to work (at least here in the UK and I get the impression it was pretty similar in the US) is you paid a rental for the line, a rental for each and every phone you had on that line (you weren't allowed to connect your own) and then paid for your calls on top of that.
You've missed this one. When CDROM and DVD-ROM manufacturers faced this problem they came up with the obvious answer: Read multiple tracks at once. They certainly tried it, I don't remember it being much of a success though. They used some optical tricks though rather than multiple read heads.
The problem with reading multiple tracks at once on a hard drive is you would need multiple head assemblies and control circuits for them. This would put up the price hugely and decrease reliability.
PC cases tend to be pretty cramped environments and since sata have a mixture of connections with very different force requirements. This a recipie for stuff getting broken if you aren't VERY carefull. It's especially bad in mixed SATA/PATA systems.
Whats so bad about scart? it's big enough to be easy to hand solder, Reasonablly good at staying in it's socket, doesn't tend to break if it gets yanked out.
It's a bit bulky but thats only really a big deal with smaller/thinner flatpanels, not with the CRTs that were the norm when scart was introduced.
As compatibility with other stuff. On a windows netbook you can run the same software on your desktop (resources permitting of course but office should be fine as should quite a bit of internal and specialist software and some older games). Without too much messing arround you can use any cheapshit printer you pick up in a big chain, any mobile broadband dongle and so on.
Luckily RAM is so cheap now, that if you really want a RAM drive, you just buy 16GB and create a software RAM drive. Ram may be cheap but server grade motherboards and CPUs aren't. And even when you get the server grade stuff there is still a hard limit on how much you can put in (and usually the largest size of module has a FAR higher cost per gigabyte than the one half the size)
HDDs, SSDs and ram drives with SATA interfaces and one can easilly connect 50 or so drives to a machine with multiport cards and port multipliers.
Also software ram drives are usually volatile while hardware ones are often battery backed.
Slashdot Mirror
Also you must learn your synthesis tool. While i'm sure there are cases where code synthisises without any hint as to things being done wrong there are also many cases were it is user error. You really do need to learn how to use timing constraints and the timing analyser that comes with your tool and which warnings are important.
Even if this is true (would be good to have a citation so we can see what assumptions are made), flying cars have a couple of big problems.
The first is fuel, any kind of non-winged craft is going to require huge ammounts of it. Even a folding light aircraft is probablly going to guzzle a lot more fuel than your average car.
The second is training and maintinace. If something goes wrong with your car you either stop where you are or coast to the hard shoulder and stop. If something goes wrong in a winged aircraft you can possiblly glide if there is a usable landing site near enough. If something goes wrong in a non-winged craft you are likely screwed.
Whether or not that "dumb" approach is effective depends entirely where you are on the current exponential curve of processing power.
And whether that exponential trend continues and how close the complexity estimates are to reality.
VHDL has an incrediablly anal-retentive type system and has some silly ideas like the splitting of entities and architectures. It's also old meaning you need a lot of boilerplate to manually import STD_LOGIC stuff (which is what the synthisis tool vendors tell you to use for everything)
Whichever you use be aware that both VHDL and Verilog weredesigned as hardware simulation languages not hardware synthisis languages. This means it is vital to get to know the synthisis tool you will be using. It is vital to know what warnings matter and what can be safely ignored and which are important, what parts of the language must be avoided, how to assign clocks to clock nets, how to use the timing analyser (without timing alalysis you don't know if your code will actually work) and so on if you are going to use the languages for FPGA programming.
I think you clicked reply on the wrong post.
And I don't buy the limited lab time either. The basic lab sciences are always empty, you just go there out of class time.
I'm from an EEE not a science background but I know that in my department first and second years are only allowed in the labs during thier official lab sessions. Third and fourth years are allowed in the general labs to do project work at other times but still don't have the access to equiment that would be needed to repeat thier lab experiments (most lab experiments are done with experiment specific equipement and/or in more specialist labs)
So if you get bad results here in an undergrad lab and don't notice them until you come to writeup/analyse (which is not unlikely because a lot of the equipment is old and unreliable and three hours is often only barely enough to complete the experiment) your options are to speculate as to why things went wrong or fudge your results. The former is the honest thing to do but the latter will probablly get you better marks.
Desktops, of any kind, now make up around 40% of total computer sales.
So less than half but still a significant ammount.
Not only is this not a huge market
On what do you base this assertion? Afaict when you look at the desktop PC market towers dominate with all-in-ones and small form factor boxes being niche markets.
I can see why apple doesn't release an ordinary desktop (releasing one would take away many peoples excuse to spend thier employers money on a mac pro) but that won't stop people snubbed by it from finding other (legally questionable) soloutions.
Because Macs tend to be a trifle more costly than your average Windows box? Look, I'm a Mac user. I love MacOS X. But I can build a PC with better specs (and better expandability) than the Mac mini for about the same cost
Heck you don't even have to build it, you can buy one from dell for considerablly less than a mac mini which has lower specs. see my post at http://yro.slashdot.org/comments.pl?sid=1245805&cid=28115689 for a detailed comparison.
If you want OS X that bad why not just buy a Mac?
Because the mac pro costs more than most geeks are prepared to spend and the mini and imac have no expansion room.
By Microsoft doing this, the increased sales to businesses would more than offset the losses due to piracy. This also would allow Windows 7 on networks that just don't connect to the Internet at all, and where a reimage of a box doesn't mean calling MS for a manual activation for each non connected machine.
According to MS you can telephone activate a KMS and then use that KMS to activate all the boxes on the non-internet connected network.
MS knows they have to release a rock bottom price version to remain competivive in the netbook/nettop market. At the same time they don't want people using the netbook version on higher end machines. Those requirements look to me mainly like an attempt to draw a line between netbooks and regular laptops.
I'm sure some netbook manufacturers will offer netbooks with specs that exceed theese bundled with either linux or a higher version of windows (HP already do this, you can get the mini 2140 with 2G of ram but only if you don't choose XP home).
Maybe some did but not all, I have a friend with a highish end card that took over a gigabyte of address space pushing usable ram under XP down to 2.5G.
only if you live in a country with horriblly wimpy power circuits ;)
but in general I agree the power consumption of higher end PCs is getting kinda crazy. Just as CPU power consumption stabalised and even dropped slightly (with the move from P4 to core 2) graphics power consumption started going thorugh the roof.
if your talking a cluster of 100kva diesel gensets then sure it might well take a few months to have it all up and running.
lets say a 1U server uses 100W and aircon and UPS overheads push that up to 200W. You can fit about 40 1U servers in a standard rack (with say a network switch and console/kvm server in the remaining two slots) so that is about 8KW per rack.
So assuming your UPS's present a unity power factor to the supply (modern large ones are pretty close afaict) a 100KVA generator would do about 12.5 racks. I don't have figures handy but I bet large datacenters have a lot more racks than that.
Afaict the issue is thet all the fitting out of the building (power, aircon, raised floors etc) and ordering of things like massive power feeds from the local powerco, massive network infrastructure from the local communication provider(s) and massive generators takes time. You can't just buy a generic building and have a high grade datacenter in a month or two.
I don't use Java, but I hear you can marshal Java data for use by C libraries and vice versa.
If the security manager lets you then you can. If you are introducing untrusted code to your VM then you set up the security manager not to allow the untrusted code to do that (as for example the java plugin does for untrusted applets).
As serial ports begin to diappear from laptops, and only ONE is available on most modern PCs rather than the two of old, I've long thought we are over-due for some other standard to develop for headless comms.
Of the vaious applications for serial ports terminal emulation is one of the most ameniable to USB to serial converters since it doesn't require any precise timing and just generally deals in two streams of data in opposite directions.
Some kit (e.g. the sheevaplug) even puts the USB to serial chip inside the device.
The great thing about serial ports as console ports go is they are simple to drive. That means they can be activated very early in the boot process without needing complex dedicated hardware.
And how much does it cost those big companies in lost productivity when an unpaid supplier deciedes they have had enough and stops shipments?
Wow. What an enormous difference. Paying a "service fee" instead of a "monthly rental fee"
Afaict the way phone service used to work (at least here in the UK and I get the impression it was pretty similar in the US) is you paid a rental for the line, a rental for each and every phone you had on that line (you weren't allowed to connect your own) and then paid for your calls on top of that.
You've missed this one. When CDROM and DVD-ROM manufacturers faced this problem they came up with the obvious answer: Read multiple tracks at once.
They certainly tried it, I don't remember it being much of a success though. They used some optical tricks though rather than multiple read heads.
The problem with reading multiple tracks at once on a hard drive is you would need multiple head assemblies and control circuits for them. This would put up the price hugely and decrease reliability.
PC cases tend to be pretty cramped environments and since sata have a mixture of connections with very different force requirements. This a recipie for stuff getting broken if you aren't VERY carefull. It's especially bad in mixed SATA/PATA systems.
Whats so bad about scart? it's big enough to be easy to hand solder, Reasonablly good at staying in it's socket, doesn't tend to break if it gets yanked out.
It's a bit bulky but thats only really a big deal with smaller/thinner flatpanels, not with the CRTs that were the norm when scart was introduced.
What kind of drive is this? unless it is a SSD I highly doubt the SATA connection is the bottleneck.
As compatibility with other stuff. On a windows netbook you can run the same software on your desktop (resources permitting of course but office should be fine as should quite a bit of internal and specialist software and some older games). Without too much messing arround you can use any cheapshit printer you pick up in a big chain, any mobile broadband dongle and so on.
Luckily RAM is so cheap now, that if you really want a RAM drive, you just buy 16GB and create a software RAM drive.
Ram may be cheap but server grade motherboards and CPUs aren't. And even when you get the server grade stuff there is still a hard limit on how much you can put in (and usually the largest size of module has a FAR higher cost per gigabyte than the one half the size)
HDDs, SSDs and ram drives with SATA interfaces and one can easilly connect 50 or so drives to a machine with multiport cards and port multipliers.
Also software ram drives are usually volatile while hardware ones are often battery backed.