Trying to sell solutions like a "Physics chip" or an "AI chip" is the joke. They are trying to sell you what could be a very generic add-on product but they are shoe-horning into doing a very limited set of things. They do this so they can charge developers an arm and a leg for the API, and so that they can make you buy three or four accelerator cards to use up all those empty PCI-e slots (thanks to motherboard integration of everything).
No, the answer is simple multithreading and taking advantage of dual core machines (or quad core when that day comes around).
I could maybe see the use of a parallel/vector processing add-in card maybe coupled with an FPGA that developers could use to tackle specific embarrasingly parallel problems for the game code. The card would be multi-purpose and it'd create a new market for developers to come up with FPGA code that does whatever hot new thing might need accelerating:
1) Water and cloth simulations 2) Procedural texture generation 3) Radiosity calculations 4) Swarming/agent-based modeling
You don't need seperate cards to do this stuff. One card is fine. Two for extra oomph. Or in the case of AMD w/hypertransport, you could have special purpose plug-in coprocessors.
I said VirtualDub when I meant DScaler. The motion compensation algorithm is based off Tom's Mocomp and it's been in tvtime for as long as I've been using it (at least 18 months).
... it was in VirtualDub and in TvTime since like forever.
TVTime had the absolute best motion adaptive de-interlace filter I'd ever seen... and my CPU usage was decent. I only wanted to watch cable TV on my PC the picture looked so good.:-)
I could code DNF in 10 _months_ all by myself in Perl. Not that I'd want to. I mean it's pretty clear how you should do it and it wouldn't be as bad as you think. Especially since they aren't writing their own graphics/physics engines or anything... it's all just tying together APIs and interpreting scripts (hey... PERL's good at that).
It'd be slow, but it'd work.
But at this point I'm not the least bit interested in Duke Nukem anymore. The man's a joke. No one wants to play his games anymore. 3DRealms has missed the boat. In fact, they missed the dock completely and instead attempted to board the log flume ride at Sea World.
Man, that tool is old. I know some people who work in the department where it was created and I think the consesus is that no one has thought about that tool in a long time (as there are much better ones now).
I think if there was some pressure from users of Tesseract... it would be quite possible to push through a request to re-release it under the Apache license.
I think the biggest hurdle there would be the paperwork and explaining why it'd be a nice gesture to the people who have to sign the forms (managers, corporate, etc.). We have a pretty hefty PR/licensing process since most of our work is delivered to our sponsors (government).
But its not like anyone considers it some kinda asset. In fact, if anyone asked for support for it there'd be groaning because no one has touched it in over a DECADE.
But yeah, I encourage users of Tesseract to send snail-mail letters explaining the issue to the Neuroscience folks there at the Washington location.
That's the Windows XP recovery console. Which is Windows XP running a minimal set of drivers (just enough to load the OS), and then giving you a CMD.EXE
It has absolutely nothing to do with DOS. The only real-mode component of Windows XP is the beginning portion of NTLDR before it switches to protected mode.
I was rather fond of DJGPP's environment with RHIDE as an IDE. You could do real-modish realtime peeks and pokes with it while simultaneously allocating large arrays and doing fancy stuff with DMA.
I would think it'd be easier to interface an industrial controller DAC/ADC board that way... but my only experience was with sound cards. But don't similar principles apply?
Diebold ATMs aren't exactly the most secure devices either. It's the bank it's attached to that implements all the security, from where it's installed to how the device indicates transactions that have occured.
Besides, you're FDIC insured. You'll never know if an ATM has been successfully hacked because it's no skin off your back (although an OUT OF SERVICE sign may get taped to the front of it).
In the past few years Dell has made an effort to make sure that you don't need to drop to DOS to flash the firmware in a piece of hardware. If your PC is ACPI compliant there's some device presented to the OS that can be used so that Windows/Linux/FreeBSD can flash the BIOS. http://linux.dell.com/libsmbios/main/dellBiosUpdat e.html
The newer LSI and 3Ware cards have a feature like this too.
... beeps when PCMCIA devices or USB devices are being added or removed. Motherboards that can give this feedback aurally are nice because then you know you've got a good connection.
This is especially infuriating when dealing with USB memory sticks that are bulky or oddly shaped being plugged into a recessed front panel in a computer case. Without sound feedback, you don't know whether its plugged in correctly or if you need an extension cable to give it physical clearence.
... from being released on a Nintendo system. Have you seen the number of non-US release "Girl-Get" games for the GBA and DS? In the US the thing companies are afraid of is the AO rating they will inevitably get, which means you won't find it at GameStop, Target, or WalMart.
Which for some reason, is preventing game manufacturers from doing anything that might earn that rating. I suppose it's because most big development houses with good relationships with Nintendo are concerned about retail channels in a big way, and getting blacklisted from WalMart is never good.
No, if a third party made a killer adult game for the Wii that was actually entertaining and sexy, they could market it. But they'd have to pony up the money themselves for getting Nintendo licensing and stuff, and hope that they can make it on Internet sales, catalog order, adult chains, etc.
Maybe the cheap ones do (the ones that take a long time to turn on or hum) But the newer, "fast on", fit-in-a-ceiling-fan ones don't have that problem. Especially if they are dimmable. They use a solid state transformer that operates at a much higher frequency than 60Hz so you won't notice the flicker. (They need to be solid-state to dim; they use something like pulse width modulation to emulate the dimming effect, and you need a fast base frequency in that ballast in order to fake that effect convincingly).
When I first saw the NForce 4 SLI boards for LGA775 it was on NewEgg, and I thought, that can't be right. But apparently the NForce MCPs were quite readily converted into an SPP and are giving the stock Intel chipsets a run for their money (while being essentially compatible with your AMD64 Nforce drivers).
Your idea sounds pretty good. I'd probably aim for a 6300 or 6400 though, I hear they can hit 2.8, 3GHz without a voltage increase or water cooling. And while I'm too young to have experienced the Celeron 300/450 era directly (no job, no money), I read all the articles about it and that's what got me interested in this stuff.
It's too bad. I was all set to get a Denmark Opt 165 with some of that crazy ass +500 DDR and overclock the everloving shit out if it. I heard people were hitting 3.2GHz and it was getting me all wet. @_@
There isn't a decent board for the Conroe that's under $250. Either they don't support DDR2800 (anything less is a waste), or they don't have SLI, or they're missing amenities like firewire or decent onboard sound. A "budget" Conroe system is difficult to spec since unless you go DDR2800 you aren't going to have much over a DDR400/DDR500-based AMD K8 system (and I'm not talking AM2, but the same logic applies). Memory bandwidth is a bottleneck for performance and usability. Despite Conroe's advances in CPU power, most situations where you wait for the computer are not CPU bound (unless you are heavy into movie/music decoding/compression). An bus-overclocked low power K8 (like the Opteron Denmark) can still beat a Conroe system in memory throughput. DDR2800 brings this to parity but then you are not talking about a cheap system anymore; it's everything EXCEPT for the CPU that costs too much.
Hopefully in the next few months we'll see price drops in DDR2 memory and more competetion in the Core-2 Duo compatible motherboards. This should make them more affordable and help to shake out the gold implementations.
I've got a dual Quadro NVS 440 setup. In both Windows, and Linux, I can have 8 monitors. EIGHT MONITORS. In any configuration I want, fully 3d accelerated at 1600x1200 per screen.
I currently am using it for a 6144x768 sized desktop for an AV switching system demo.
There is nothing in the ATI camp that can do this save (possibly) the FireMV line. And do you know what chip they use in that internally? A 9200. A friggin 9200. That only works in linux using the open source driver! Absolutely ridiculous.
Also, the iRiver U10 is more interesting than this silly Sandisk model. The U10 supports OGG with the default firmware, plays flash, and functions as a remote control (nice additional touch).
No one is going to kill the iPod. The iPod is the iPod. But they can certainly put out more interesting/useful products. Sandisk's isn't one of them.
with the exception of our planets and asteroids, all of the information we have about stuff outside our solar system is based on spectra that is radiated from or filtered through the object in question.
Dark matter is just that. Dark. No light is coming out of it, and unlike interstellar gas/dust, it doesn't seem to affect the light passing through it either (except for graviational lensing, which is how we know something is there).
Dark != doesn't fit into standard model. Dark == doesn't radiate photons. That's all. Dark matter fits into the standard model. We just don't which particles it's likely to be. We do know it's made of WIMPs. But that could be a lot of stuff. Dark energy doesn't fit into the standard model. But the standard model isn't complete anyway so there are lots of way to cast it in there along with gravitons, the Higgs boson, and the rest.
Dark Matter consists of WIMPs. Meanwhile Dark Energy only manifests itself in gravitational interactions. They don't like to interact with other fields, and they are ESPECIALLY not likely to interact with electromagnetic fields. Hence, any device that involves magnets or electric fields that claims to harness ZPE or Dark Matter/Energy is just REACHING.
Meanwhile, there are tons of theories currently to try to explain Cold Fusion. It's becoming rapidly accepted as a valid phenomenon (but only after years of experimental testing and making sure its not some other phenomenon but masked). It's probably just Hot Fusion, but accomplished by situational lowering of the energy barrier... its statistically unlikely in isolation but you take a sufficiently large catalyst and a bunch of fuel and then you start observing it.
There is no conspiracy. But there is a lot of rapidly changing stuff in particle physics right now; it's in a state of flux. So when you start talking about Cold and Hot fusion, it's hard to talk about that when we still can't commit to whether protons decay or not.
That's all you have to say in his defense? Mod this down on principle, thanks.
I'd like to see the field equations where they show you being able to end up with more potential energy than you started with. You know, a time-parameterized finite element analysis in three-dimensional space with suitable boundary conditions. They say they accomplished this on paper "in software". WELL THEY COULD JUST VERY WELL RELEASE THOSE RESULTS
But no. No. They want to do a "demo" with a "jury".
If its true, the current energy corporations will do anything and everything possible to make sure it never sees the light of day
That's a load of bullshit.
About 50% of the business of an oil company is automotive fuel. The next 35 goes to diesel and kerosene for trucks, ships and planes. The last 15 is used for plastics and other derivative products. The latter two markets are not threatened in the least by an alternative energy source since there's currently no way to have a large battery powered aircraft or oceanliner. Furthermore it will take at least 25 years to transition away from gasoline automobiles even if the product just appeared on the market today, magically. Not everyone can afford to (or would prefer to) purchase a new, electric car.
Furthermore, energy companies spend hundreds of millions of dollars a year on energy costs themselves! If they came across a free-energy technology, it would be financial idiocy to NOT use it at least internally to get a leg up on their competitors!
At best, the coal industry would be threatened by free energy. EVERY OTHER INDUSTRY AND GOVERNMENT WOULD BE ALL OVER IT
Slashdot Mirror
Bet you 10 bucks that this AI accelerator is a PowerPC mated with a Xilinx FPGA.
Same thing for the PhysX processor.
Hmmm... i wonder if you could make one into the other with a firmware flash and FPGA update.
Trying to sell solutions like a "Physics chip" or an "AI chip" is the joke. They are trying to sell you what could be a very generic add-on product but they are shoe-horning into doing a very limited set of things. They do this so they can charge developers an arm and a leg for the API, and so that they can make you buy three or four accelerator cards to use up all those empty PCI-e slots (thanks to motherboard integration of everything).
No, the answer is simple multithreading and taking advantage of dual core machines (or quad core when that day comes around).
I could maybe see the use of a parallel/vector processing add-in card maybe coupled with an FPGA that developers could use to tackle specific embarrasingly parallel problems for the game code. The card would be multi-purpose and it'd create a new market for developers to come up with FPGA code that does whatever hot new thing might need accelerating:
1) Water and cloth simulations
2) Procedural texture generation
3) Radiosity calculations
4) Swarming/agent-based modeling
You don't need seperate cards to do this stuff. One card is fine. Two for extra oomph. Or in the case of AMD w/hypertransport, you could have special purpose plug-in coprocessors.
cp == b&
I said VirtualDub when I meant DScaler.
The motion compensation algorithm is based off Tom's Mocomp and it's been in tvtime for as long as I've been using it (at least 18 months).
... it was in VirtualDub and in TvTime since like forever.
:-)
TVTime had the absolute best motion adaptive de-interlace filter I'd ever seen... and my CPU usage was decent. I only wanted to watch cable TV on my PC the picture looked so good.
I could code DNF in 10 _months_ all by myself in Perl. Not that I'd want to.
I mean it's pretty clear how you should do it and it wouldn't be as bad as you think. Especially since they aren't writing their own graphics/physics engines or anything... it's all just tying together APIs and interpreting scripts (hey... PERL's good at that).
It'd be slow, but it'd work.
But at this point I'm not the least bit interested in Duke Nukem anymore. The man's a joke. No one wants to play his games anymore. 3DRealms has missed the boat. In fact, they missed the dock completely and instead attempted to board the log flume ride at Sea World.
Man, that tool is old.
... it would be quite possible to push through a request to re-release it under the Apache license.
I know some people who work in the department where it was created and I think the consesus is that no one has thought about that tool in a long time (as there are much better ones now).
I think if there was some pressure from users of Tesseract
I think the biggest hurdle there would be the paperwork and explaining why it'd be a nice gesture to the people who have to sign the forms (managers, corporate, etc.). We have a pretty hefty PR/licensing process since most of our work is delivered to our sponsors (government).
But its not like anyone considers it some kinda asset. In fact, if anyone asked for support for it there'd be groaning because no one has touched it in over a DECADE.
But yeah, I encourage users of Tesseract to send snail-mail letters explaining the issue to the Neuroscience folks there at the Washington location.
That's the Windows XP recovery console.
Which is Windows XP running a minimal set of drivers (just enough to load the OS), and then giving you a CMD.EXE
It has absolutely nothing to do with DOS. The only real-mode component of Windows XP is the beginning portion of NTLDR before it switches to protected mode.
I was rather fond of DJGPP's environment with RHIDE as an IDE. You could do real-modish realtime peeks and pokes with it while simultaneously allocating large arrays and doing fancy stuff with DMA.
... but my only experience was with sound cards. But don't similar principles apply?
I would think it'd be easier to interface an industrial controller DAC/ADC board that way
Diebold ATMs aren't exactly the most secure devices either. It's the bank it's attached to that implements all the security, from where it's installed to how the device indicates transactions that have occured.
Besides, you're FDIC insured. You'll never know if an ATM has been successfully hacked because it's no skin off your back (although an OUT OF SERVICE sign may get taped to the front of it).
In the past few years Dell has made an effort to make sure that you don't need to drop to DOS to flash the firmware in a piece of hardware. If your PC is ACPI compliant there's some device presented to the OS that can be used so that Windows/Linux/FreeBSD can flash the BIOS.t e.html
http://linux.dell.com/libsmbios/main/dellBiosUpda
The newer LSI and 3Ware cards have a feature like this too.
... beeps when PCMCIA devices or USB devices are being added or removed. Motherboards that can give this feedback aurally are nice because then you know you've got a good connection.
This is especially infuriating when dealing with USB memory sticks that are bulky or oddly shaped being plugged into a recessed front panel in a computer case. Without sound feedback, you don't know whether its plugged in correctly or if you need an extension cable to give it physical clearence.
... from being released on a Nintendo system. Have you seen the number of non-US release "Girl-Get" games for the GBA and DS?
In the US the thing companies are afraid of is the AO rating they will inevitably get, which means you won't find it at GameStop, Target, or WalMart.
Which for some reason, is preventing game manufacturers from doing anything that might earn that rating. I suppose it's because most big development houses with good relationships with Nintendo are concerned about retail channels in a big way, and getting blacklisted from WalMart is never good.
No, if a third party made a killer adult game for the Wii that was actually entertaining and sexy, they could market it. But they'd have to pony up the money themselves for getting Nintendo licensing and stuff, and hope that they can make it on Internet sales, catalog order, adult chains, etc.
...the disenchanted NetBSD developers should jump ship to OpenBSD and bring their toys with them.
Maybe the cheap ones do (the ones that take a long time to turn on or hum)
But the newer, "fast on", fit-in-a-ceiling-fan ones don't have that problem. Especially if they are dimmable. They use a solid state transformer that operates at a much higher frequency than 60Hz so you won't notice the flicker. (They need to be solid-state to dim; they use something like pulse width modulation to emulate the dimming effect, and you need a fast base frequency in that ballast in order to fake that effect convincingly).
When I first saw the NForce 4 SLI boards for LGA775 it was on NewEgg, and I thought, that can't be right.
But apparently the NForce MCPs were quite readily converted into an SPP and are giving the stock Intel chipsets a run for their money (while being essentially compatible with your AMD64 Nforce drivers).
Your idea sounds pretty good. I'd probably aim for a 6300 or 6400 though, I hear they can hit 2.8, 3GHz without a voltage increase or water cooling. And while I'm too young to have experienced the Celeron 300/450 era directly (no job, no money), I read all the articles about it and that's what got me interested in this stuff.
It's too bad. I was all set to get a Denmark Opt 165 with some of that crazy ass +500 DDR and overclock the everloving shit out if it. I heard people were hitting 3.2GHz and it was getting me all wet. @_@
There isn't a decent board for the Conroe that's under $250.
Either they don't support DDR2800 (anything less is a waste), or they don't have SLI, or they're missing amenities like firewire or decent onboard sound.
A "budget" Conroe system is difficult to spec since unless you go DDR2800 you aren't going to have much over a DDR400/DDR500-based AMD K8 system (and I'm not talking AM2, but the same logic applies). Memory bandwidth is a bottleneck for performance and usability. Despite Conroe's advances in CPU power, most situations where you wait for the computer are not CPU bound (unless you are heavy into movie/music decoding/compression). An bus-overclocked low power K8 (like the Opteron Denmark) can still beat a Conroe system in memory throughput.
DDR2800 brings this to parity but then you are not talking about a cheap system anymore; it's everything EXCEPT for the CPU that costs too much.
Hopefully in the next few months we'll see price drops in DDR2 memory and more competetion in the Core-2 Duo compatible motherboards. This should make them more affordable and help to shake out the gold implementations.
I've got a dual Quadro NVS 440 setup.
In both Windows, and Linux, I can have 8 monitors. EIGHT MONITORS. In any configuration I want, fully 3d accelerated at 1600x1200 per screen.
I currently am using it for a 6144x768 sized desktop for an AV switching system demo.
There is nothing in the ATI camp that can do this save (possibly) the FireMV line. And do you know what chip they use in that internally? A 9200. A friggin 9200.
That only works in linux using the open source driver! Absolutely ridiculous.
Also, the iRiver U10 is more interesting than this silly Sandisk model.
The U10 supports OGG with the default firmware, plays flash, and functions as a remote control (nice additional touch).
No one is going to kill the iPod. The iPod is the iPod.
But they can certainly put out more interesting/useful products. Sandisk's isn't one of them.
Desu desu??!
Desu desu, desu desu desu.
Desu desu desu , desu desu desu desu -- desu desu desu desu desu desu desu. Desu desu, desu desu desu desu!!! ^_^
with the exception of our planets and asteroids, all of the information we have about stuff outside our solar system is based on spectra that is radiated from or filtered through the object in question.
Dark matter is just that. Dark. No light is coming out of it, and unlike interstellar gas/dust, it doesn't seem to affect the light passing through it either (except for graviational lensing, which is how we know something is there).
Dark != doesn't fit into standard model.
... its statistically unlikely in isolation but you take a sufficiently large catalyst and a bunch of fuel and then you start observing it.
Dark == doesn't radiate photons. That's all.
Dark matter fits into the standard model. We just don't which particles it's likely to be. We do know it's made of WIMPs. But that could be a lot of stuff.
Dark energy doesn't fit into the standard model. But the standard model isn't complete anyway so there are lots of way to cast it in there along with gravitons, the Higgs boson, and the rest.
Dark Matter consists of WIMPs. Meanwhile Dark Energy only manifests itself in gravitational interactions.
They don't like to interact with other fields, and they are ESPECIALLY not likely to interact with electromagnetic fields. Hence, any device that involves magnets or electric fields that claims to harness ZPE or Dark Matter/Energy is just REACHING.
Meanwhile, there are tons of theories currently to try to explain Cold Fusion. It's becoming rapidly accepted as a valid phenomenon (but only after years of experimental testing and making sure its not some other phenomenon but masked). It's probably just Hot Fusion, but accomplished by situational lowering of the energy barrier
There is no conspiracy. But there is a lot of rapidly changing stuff in particle physics right now; it's in a state of flux. So when you start talking about Cold and Hot fusion, it's hard to talk about that when we still can't commit to whether protons decay or not.
That's all you have to say in his defense?
Mod this down on principle, thanks.
I'd like to see the field equations where they show you being able to end up with more potential energy than you started with. You know, a time-parameterized finite element analysis in three-dimensional space with suitable boundary conditions. They say they accomplished this on paper "in software".
WELL THEY COULD JUST VERY WELL RELEASE THOSE RESULTS
But no. No. They want to do a "demo" with a "jury".
That's what magicians do in Vegas.
Utter bullshit. MOD THIS DOWN.
If its true, the current energy corporations will do anything and everything possible to make sure it never sees the light of day
That's a load of bullshit.
About 50% of the business of an oil company is automotive fuel. The next 35 goes to diesel and kerosene for trucks, ships and planes. The last 15 is used for plastics and other derivative products.
The latter two markets are not threatened in the least by an alternative energy source since there's currently no way to have a large battery powered aircraft or oceanliner. Furthermore it will take at least 25 years to transition away from gasoline automobiles even if the product just appeared on the market today, magically. Not everyone can afford to (or would prefer to) purchase a new, electric car.
Furthermore, energy companies spend hundreds of millions of dollars a year on energy costs themselves! If they came across a free-energy technology, it would be financial idiocy to NOT use it at least internally to get a leg up on their competitors!
At best, the coal industry would be threatened by free energy. EVERY OTHER INDUSTRY AND GOVERNMENT WOULD BE ALL OVER IT
So yeah, complete and utter BS