That's been a feature in the T-series IBM ThinkPads since early 2003. A gyroscope and accelerometer.3dcgi said:Coming up with the idea to put a gyroscope in a controller is not difficult. Years ago I thought it would make a neat add on for a game boy. The difficult part is finding ways to use it that aren't gimmicky. So yes, it comes back to the games.PC-Engine said:If PS3 controller has gyro then we definitely know who copied who in the console space.
In a related note some laptops already contain gyroscopes. I saw someone wrote a program that uses the hard drive parking feature of recent Apple laptops so you can control games by tilting the laptop.
http://laptops.engadget.com/entry/1234000680037005/
Revolution GPU and CPU STILL in development.
- Thread starter SoVos20
- Start date
- Status
Gyroscopes did feature on GB in at least one game (forget the name). Also MS released a Gyro joystick. There have been several.
Description of piezoelectric pump and cooling apparatus.
http://www.labs.nec.co.jp/Eng/brochure/data/NECLabo_E07.pdf
http://www.labs.nec.co.jp/Eng/brochure/data/NECLabo_E07.pdf
Ooh-videogames, your dream has come true.
http://www.sapphiretech.com/vga/blizzard.asp
Looks like Revolution won't have any problems with cooling considering the plethora of advanced low profile cooling solutions available. 8)
http://www.sapphiretech.com/vga/blizzard.asp
Looks like Revolution won't have any problems with cooling considering the plethora of advanced low profile cooling solutions available. 8)
MechanizedDeath
Regular
PC-Engine said:Ooh-videogames, your dream has come true.
Looks like Revolution won't have any problems with cooling considering the plethora of advanced low profile cooling solutions available. 8)
Heat pipes, fansinks...it doesn't matter. All these cooling solutions do is move heat from the chip to a place off the chip. That heat doesn't disappear (conservation of energy). All this thing does is lower the profile, which we've seen since the days of the blorb. Those were low-pro fansinks. Heat pipes are ancient, and peltiers are actually much better IMO. But it doesn't ignore the fact that they'll still be exhausing heat either into a really small case, or outside of it. I don't think this is a solution for the heat. Hollywood and Broadway will have to be low-voltage or low-clock to keep heat down. AND they'll still need a good cooling solution. And then the case, being small, needs good ventilation, hopefully without being very loud. PEACE.
MechanizedDeath said:PC-Engine said:Ooh-videogames, your dream has come true.
Looks like Revolution won't have any problems with cooling considering the plethora of advanced low profile cooling solutions available. 8)
Heat pipes, fansinks...it doesn't matter. All these cooling solutions do is move heat from the chip to a place off the chip. That heat doesn't disappear (conservation of energy). All this thing does is lower the profile, which we've seen since the days of the blorb. Those were low-pro fansinks. Heat pipes are ancient, and peltiers are actually much better IMO. But it doesn't ignore the fact that they'll still be exhausing heat either into a really small case, or outside of it. I don't think this is a solution for the heat. Hollywood and Broadway will have to be low-voltage or low-clock to keep heat down. AND they'll still need a good cooling solution. And then the case, being small, needs good ventilation, hopefully without being very loud. PEACE.
Well thanks for pointing out the obvious.
The point is it's low profile which fits in nicely with the small form factor of Revolution. Take a close look at that card. Do you see a huge heatsink? How thick do you think that card is?
heh using a tec / peliter is a horrible thought. They dump out more heat than the cpu would dump out .
Anyway the trick is to move the heat away from the die as fast as possible . With air cooling there is only so much u can do . However with good air flow in a small case like the x360 or ns5 you can push enough air through the case that the air inside doesn't become saturated with heat .
THat is the problem with current pc cases there is so many dead zones that by the time the air reaches the cpu and gpu its already hot and some airs retain the hot air (like around the back side of the gpu that is typicly a dead zone )
You don't have all these problems in a case . You can set it up where the front fan is pushing air across the pcb and the back fan is pulling it out with all the chips on the pcb and very little dead zones .
That is why btx is a good idea as the air is taken in directly over where its needed the most across the cpu and gpu . It also flips the gpu to give it better cooling ability as heat rises and its much easier to take the heat off the heatsink and up into the case instead of down into another pci card
Anyway the trick is to move the heat away from the die as fast as possible . With air cooling there is only so much u can do . However with good air flow in a small case like the x360 or ns5 you can push enough air through the case that the air inside doesn't become saturated with heat .
THat is the problem with current pc cases there is so many dead zones that by the time the air reaches the cpu and gpu its already hot and some airs retain the hot air (like around the back side of the gpu that is typicly a dead zone )
You don't have all these problems in a case . You can set it up where the front fan is pushing air across the pcb and the back fan is pulling it out with all the chips on the pcb and very little dead zones .
That is why btx is a good idea as the air is taken in directly over where its needed the most across the cpu and gpu . It also flips the gpu to give it better cooling ability as heat rises and its much easier to take the heat off the heatsink and up into the case instead of down into another pci card
jvd said:
The ATI cooler isn't Peltier (which has problems with thermal runaway if I remember correctly).
jvd said:
Because air itself is a relatively poor conductor of heat.
jvd said:
A large amount of airflow also means that the amount of noise goes up assuming that there are some constrictions (i.e. dust grates) in the way.
So the trick is to balance the amount of airflow (heat dissipation) vs. noise.
I don't see why a small case would have worse airflow than a larger one. In fact I think the opposite would be true. The real issue with a small case is the restrictive size of the cooler.
I cant remember where, but someone dug up the Nintendo R&D for the past few years and the spending was a bit more signifigant than beforehand, if I recall correctly that is, though having 3 different gameboys that play the same thing(GBA, GBASP, GBA Retro) and a new handheld(DS) may change reflect on that a bit.
It's not obvious at all.PC-Engine said:MechanizedDeath said:PC-Engine said:Ooh-videogames, your dream has come true.
Looks like Revolution won't have any problems with cooling considering the plethora of advanced low profile cooling solutions available. 8)
Heat pipes, fansinks...it doesn't matter. All these cooling solutions do is move heat from the chip to a place off the chip. That heat doesn't disappear (conservation of energy). All this thing does is lower the profile, which we've seen since the days of the blorb. Those were low-pro fansinks. Heat pipes are ancient, and peltiers are actually much better IMO. But it doesn't ignore the fact that they'll still be exhausing heat either into a really small case, or outside of it. I don't think this is a solution for the heat. Hollywood and Broadway will have to be low-voltage or low-clock to keep heat down. AND they'll still need a good cooling solution. And then the case, being small, needs good ventilation, hopefully without being very loud. PEACE.
Well thanks for pointing out the obvious.
The point is it's low profile which fits in nicely with the small form factor of Revolution. Take a close look at that card. Do you see a huge heatsink? How thick do you think that card is?
It would seem that most people are either confused about the issues of cooling, or simply ignorant. The cooling solution is indeed low profile, but what does that say other than that the ingenious pumps aren't very large? Nothing. The large size of some coolers is because they want a large surface area for dissipation. The liquid metal is just a (good, presumably) way of transferring the heat, just as heatpipes can be. The dissipation part of the job is unrelated, and that's where the dual fans and cooling fins, i. e. the size of the cooler, come into the picture.
No, there are two main avenues for Nintendo to drop power use.
1. Use a better manufacturing process than their competitors, allowing them to cram more transistors onto the dies, while keeping voltages and power draw lower. They would presumably use this advantage to reduce their power draw relative to their competitors, rather than increase performance.
2. Remove redundant computing elements. Take the triple core XBox360 CPU as an example. How often will for instance the least used core be used at all? How often will it run at anything close to full utilization? How much would it really cost you in terms of performance to simply drop down to a dual core chip? My guess - typically very little indeed. Witness the moaning we hear from developers (and reviewers) about even getting reasonable utilization from dual core desktop chips, and that's only a grand total of two load balanced threads....
Amdahls law applies mercilessly.
So by waiting for a process with better thermal properties, and reducing parallellism that will largely go underutilized, Nintendo could offer good performance within a much smaller thermal envelope. This would allow them to fit into a much sleeker package, and completely sidestep the size, noise, and reliability issues of a higher power draw design.
(Even on the same process, they could simply say - "To hell with the numbers game, we'll do the best we can within our design envelope", drop voltages and redundancy, and get large savings even with fairly small compromises in performance, due to how power draw increases with frequency at a given process. Bút if I were Nintendo, I'd try to hit 65nm out of the gate if at all possible.)
yes i know i was responed to someones post above mine
Its not even thermal runaway thats the problem. Last peilter i used was a 135watt peltier to cool a athlon xp 2000+ . You would need a powersupply just for the tec then not to mention that you have to cool a 135 watt heat source .
yes i know .... see my points
correct . I really don't see why you responed to my post . Your just saying the same thing i am .
However your missing one thing . airflow does not equal noise . You can hae good air flow from larger fans at less dba than a smaller fan. So for instance a 80mm fan at 3000rpm can move the same amount of air as a 60mm fan at 6000rpm (just numbers it depends on the make of the fan and whats used for the motor) however the dead spot in the center is bigger so its not ideal for heatsinks but a system it would be . Like in the back of a x360
which is exactly what i was talking about . In a pc case you have alot of dead spots from pci add in carts , hardrive bays , cdrom drives whatever .
In the console you can have the optical drive away from the main pcb with the chips. and then have a fan in the front of the case blowing over the pcb and a fan at the back of the case taking the air .
Since its a smaller case you need less air flow than a bigger case to move the air out of the case. Couple that with less dead spots and the real question is what is the best way to get the heat from the chip to the air . the faster you can get the heat off the xgpu the cooler you can keep it and with proper case air flow you might not need a fan on the xenos or xcpu . Or you might need a very quiet one .
The faster the heat transfer the farther away you can mount the main radiator/fan giving you more routing flexibility and hence better space utilization. With a simple HSF, you'd have to have a very big heatsink and fan over the chip because it is too slow to transfer heat to a far away location.
True.PC-Engine said:
But "far away location"? The entire Revolution console is going to be tiny. There is no far away location. And the innards are going to be packed. In fact, if they can go low enough in total power draw, they might use heatpipes (to gain a bit of flexibility) and connect dirctly to the external housing and use that as a heatsink.
No, the Revolution is probably not going to achieve its ergonomics through innovative cooling. I'd say it will achieve them by not needing exotic cooling in the first place. So what will it have to give up in order to fit within a narrower power envolope? As per my reasoning above - not a whole lot really.
This thread is depressing, it's sad to see all the Nintendofans come riding to the rescue when it is so blindingly obvious already that rev WILL be the weakest design (probably by far) of all three stationary consoles.
Where do all you oddball people come from? In what world is a massively parallel out-of-order Power5 core inferior to a 33% higher clocked, narrow, cost-cut, in-order core?
Take a look at a die shot of a Power5 some day. It's neither small, nor cheap, nor frugal in the electricity department either, and at 276 MILLION transistors/core, well lets just say even if L2 cache is unified for all three cores, it's still going to be a far far bigger and more expensive and baggage-heavy chip than xCPU. It has features that will NEVER find use in a console - which isn't surprising, seeing as it's a SERVER chip...
Let me ask you something... When Nintendo put 24MB of main RAM in the GC and MS put 64MB of DDR in XB, what do you think is the main reason for that discrepancy? The ELEGANCY of the number 24 perhaps, or maybe something else?
You are aware that 1T SRAM is much more expensive than regular memory, right?
It's going to be used eventually, don't worry. There will most likely be APIs in the SDK for both physics and AI for devs to use, as well as third-party stuff as well (LUA, Havok etc).StarFox said:
WTF? This is crazy.
Where do all you oddball people come from? In what world is a massively parallel out-of-order Power5 core inferior to a 33% higher clocked, narrow, cost-cut, in-order core?
Take a look at a die shot of a Power5 some day. It's neither small, nor cheap, nor frugal in the electricity department either, and at 276 MILLION transistors/core, well lets just say even if L2 cache is unified for all three cores, it's still going to be a far far bigger and more expensive and baggage-heavy chip than xCPU. It has features that will NEVER find use in a console - which isn't surprising, seeing as it's a SERVER chip...
You assume there will be threads to actually feed four cores (which may or may not be hyperthreaded, so that would be up to eight virtual CPUs). Without heavy use of that which you doubted devs would not be able to afford, namely physics and AI, there won't be use for more than ONE core really.
You're just speculating! You have no way of knowing if any of that is even slightly true!
What we're witnessing here is the mating-dance of the fanbird, or in other words, the explanation of how my preferred brand can do what everybody else does, but much better and cheaper, because my preferred brand is so much smarter than them and they'll launch a few months later too so that fixes any issues with reality-distortion. It works wonderfully elegantly in the minds of the fanbird, though less so in the real world.
Let me ask you something... When Nintendo put 24MB of main RAM in the GC and MS put 64MB of DDR in XB, what do you think is the main reason for that discrepancy? The ELEGANCY of the number 24 perhaps, or maybe something else?
You are aware that 1T SRAM is much more expensive than regular memory, right?
As I mentioned in the ArsTechnica article on XB360, it's apparent that all those GFlops are going to be used for a lot of stuff. It's not just game code, AI and physics (incluing fluid dynamics character control, not just barrels and boxes and axels), but procedural world creation as well, plus fancy 3D sound generation (for those that don't have a sound processor) and possibly post-processing and other techniques we havn't even dreamed of yet.Guden Oden said:
Any console falling short on sheer muscle power is going to show it, certainly a few years down the line.
jvd said:
Perhaps I worded it poorly (though I did mention constriction) - it's not the amount of airflow alone that determines noise level.
Let me put it another way. It's the amount of turbulence. The more airflow you have the greater chance of you creating more turbulence. The more turbulence, the more noise.
- Status
