because that whole "game designers know most about the game universe and can thus create greate movies" concept worked so greate for wing commander or final fantasy?
Well IMHO your config is suboptimal because you wast to fingers for farward/backward, plus the fact that giving spave to right isnt that good on most keyboards (because space usually has worse feedback when hitting non-centred).
With wsad, middle finger can easily hit both forward and backward, a and d for left and right each have their finger, the thump is resting on space ready to jump and the base if the little finger rests on control, ready to crouch, while the tip can activate shift for running without moving the finger a bit.
The "other" keys i have mapped on the buttons of my mouse (the 2 thump keys of the mx700 are perfect for reload/activate)
Slashdot should rename itself to "news for computer kiddies and layed of cynical IT-veterans who lost touch with technology".
Both this story and the last one (the quad core one) were nice technical stuff, perfect for nerds. And lets take a look here. at the time of that posting , only 2 or 3 comments are even remotely touching the subject. The rest is stupid jokes and dumb ranting. The quad core article is even worse, were the only non-joke posters are to stupid to tell apart SMT and dualcore.
Also it seems to be a sad trend that the initial reaction to ANYTHING even slightly technical/scientific seems to be a self preservation (" im not stupid, this stuff is just ununderstandable !!!11") joke posting.
Re:Multicore is great, but not for the obvious rea
on
AMD Quad Cores, Oh My
·
· Score: 1
(i dont know in which century you are, but there is nothing like "banks" that have to be filled completely. In fact, your opteron will even run with only one dimm installed, although with rather limited memory bandwith. For optimal use, 2Dimms per processor). A Dual socket opteron has a NUMA architecture. No bank-shit or anything. each processor has a simply dual channel memory config like ANY OTHER MAINBOARD out there (im sure you have seen p4s or AMD64 mainboards without any socket filled). (with the added benefit of being able to share memory bandwith between the cpus)
So 4GB ram is a modest amount... Ok, than use 4*1GB Dimm. Complete memory bandwith realisation for a dual cpu system.
You dont understand: HYPERTHREADING =! MULTICORE. These are 2 complete cpus+a crossbar switch on one die. No shareing of execution units/registers,no sharing of anything but the ram bandwith.
Amd dual core cpus are FASTER than 2 single core cpus in dual socket boards (with the exception of extremely bandwith demanding streaming applications) simply because of much faster on-die cache coherence communication.
A quad core cpu will most likely see more bandwith problems, but could (with ddr-2, ect) still be very well in the same class as a 4 single-core machine.
but looking back, the alpha wasnt powered by magic either.... It was way faster than anything else, but it bought that kind of dominance by using something that now limits x86: A massive power budget. Alphas used 80W+ back in times when 25W of a pentium2 seemed horrendious, so its not that miraculous that they got more performance out of it.
Well, i think maybe the main reason is fear of fabbing problems. Being burned with IBMs problems, they may have some paranoia about another cpu supplier with only a single cutting edge fab (even if its a megafab)
i know that the american voting system is prone to create a situation where one can only chose between bad and worse...
But the sad thing is, there were still more people voting for worse... So voting would still help, at least compared to bitching about corrupt politicans and not going to vote because "it doesnt matter anyway"...
But dont forget that our galaxy is in chaotic movement within the local cluster, which is moving towards to Great Wall.
But there is a simple way: Cosmic Background radiation Anisotrophy. If you look at the temperate of the background, there is an almost perfect doppler shift corresponding to a speed of (IIRC) about 600-700Km/s for earth. Absolute. Total.
Its not that easy. Just for your information: the classical picture of a laser trap always presented actually is only a so called optical molasse, it will slow down molecules, but cannot trap them. (its actually quite obvious, think about the balance of force on the particle as its speed aproaches 0... it can be descriped by a modification of the stokes formula for objects in fluids) For real confinement, you need futher stuff, for example quadrupol coils for a penning type trap, ect. I dont know what exactly they have working, but its quite possible that they have some kind of feedback loop for their coil management or something like that...
No, current heads ARENT bigger because they have to move around. IN fact, being big is quite a hinderance when moving. The simply size-limiting factor of current heads is the writing coil. You cant just shrink it down to 500nm or something.
b)Again: i now see what you have in mind (wasnt that clear before). Ok, now you have the hyothetical disc with all 20,000 heads in a row (20,000 singly points of failure, btw). you already spoke of the thermal problems/vibration.
But take a look at the electrical side. Signal processing wise: If you rally want to use all those heads at once, every head needs the signal processing pipeline backing it that now a single HD has to at its disposal... So we are talking about supercomputer class dsp power for a single drive here. The alternative would be just switching heads, but that would cut down STR to normal hd levels, and the access time wouldnt improve that much (at least considering the cost of the design) because modern high end HD alrady have positioning times in the ballpark of the spin latency.
Also you simply disregard problems with stiktion (you cant park such a head assembly outside the data layers) and keeping the damn thing floating above the surface (the old drum memories hadnd those problems). We are talking about sub um pressision in lateral and horizontal direction over the whole assembly. Thats not cheap, even if it were possible.
Sorry, was a bit overreacting. But it can get very annoying if somebody who has litteraly no idea what he is talking about has an idea and presents it like he alone found the philosophers stone and everybody else is stupid.
Its even worse with all those free energy nuts that claim infinite forces and stuff only because they cannot comprehent that even an integral from -inf to +inf can converge...
And about the harddiscs: its incredible enough that those things are working even with one head, given all the density improvements and the low cost manufactoring.
Do you know just how many tracks a disk has? Hint: its many many thousand. You cant put so many heads over the disc because the head is orders of magnitudes wider than the track. Plus if you were only doing sparse head placement (like 100 heads evenly spaced), you could not keep them calibrated, plus even IF you could, seek times would get MUCH worse because of resonanz modes of the arm complex, increased inertia, ect. Add to this the problems of sticktion, disturbance of the bernoulli effect because of the high head density, ect,ect, and the idea becomes braindead. (there were dual head hds years ago, but calibration was a bitch).
You know, the harddisc industry (in fact any industry with x000 employees ) HAS people that are WAY smarter than you. So whenever you encouter something that doesnt make sense to you, maybe YOU are wrong.
Thats not cold fusion.... For "cold" atoms, the tunnel barrier is way to high and broad... thats not a few orders of magnitutes too low, its hundreds...
The only way ever to get fusion (and not even in the break even way of things, just like 1 reaction per mol per century) is the very far end of the bolzmann tail, which isnt cool anymore.
Just think about it: There are detectors for neutrino experiments using thousands of tons of water or other stuff (like that in japan, nakasomthing), under EXTREME survilance by photodetectors &co. And in all those years, not a SINGLE spontanous fusion process ever happened with all those 10^30 or so atoms...
There IS NO SUCH THING als cold fusion actually happening in the natural world.
And what THEY (in the article) do is just a cheap trick in order to remain "cold". Their particles are not in thermal equilibrium, so temperature has no meaning, and their kinetic energy would correlate to a VERY high temperatur. Its no rocket science to to use a particle accelerator and get fusion. Scale it down to the minimum possible, sacrificing production rate to a few 100 per second, and you can even put it in a breadbox.
You can stop it now. The reality distortion field was officially shut down, so welcome back to the real world.
the g5: - greater clockspeed increase. yeah right. With watercooling 700Mhz. Including die-shrink (which completely failed). - Better vectore engine: never running at peak performance because of sucky chipset/memory interface. YOu cant MAC without data, even if you have the execution units.
its not about the 3Ghz. with 130nm, the g5 was so hot that the 2.5 needed watercooling (and the real amount was never dicloused), and the die shrink didnt help. So no laptops (did you see the story that laptops sales surpased desktop ones? or that the powerbooks are a apple stronghold?). No g5-mini.
And if IBM were so superiour with their fabbing, why does the result suck that much?
While i too agree that AMD should have been prefered, it is understandable because of the Pentium M. Apple needs cool chips for notebooks, and while the k8 core is catching up with power-now&co, its still behind p-m. But that doesnt matter, because they are compatible....
Slashdot Mirror
because that whole "game designers know most about the game universe and can thus create greate movies" concept worked so greate for wing commander or final fantasy?
Put a dongle on the motherboard?
I mean, they BUILD those computers, do you really think its so hard to check for some hardware details?
Of course you could crack it somehow, but still..
Well IMHO your config is suboptimal because you wast to fingers for farward/backward, plus the fact that giving spave to right isnt that good on most keyboards (because space usually has worse feedback when hitting non-centred).
With wsad, middle finger can easily hit both forward and backward, a and d for left and right each have their finger, the thump is resting on space ready to jump and the base if the little finger rests on control, ready to crouch, while the tip can activate shift for running without moving the finger a bit.
The "other" keys i have mapped on the buttons of my mouse (the 2 thump keys of the mx700 are perfect for reload/activate)
SO its like car racing with EVERYBODY stearing with their buttcheeks. :)
I see. Now thats MUCH more sensible
Slashdot should rename itself to "news for computer kiddies and layed of cynical IT-veterans who lost touch with technology".
Both this story and the last one (the quad core one) were nice technical stuff, perfect for nerds.
And lets take a look here. at the time of that posting , only 2 or 3 comments are even remotely touching the subject. The rest is stupid jokes and dumb ranting.
The quad core article is even worse, were the only non-joke posters are to stupid to tell apart SMT and dualcore.
Also it seems to be a sad trend that the initial reaction to ANYTHING even slightly technical/scientific seems to be a self preservation (" im not stupid, this stuff is just ununderstandable !!!11") joke posting.
(i dont know in which century you are, but there is nothing like "banks" that have to be filled completely. In fact, your opteron will even run with only one dimm installed, although with rather limited memory bandwith. For optimal use, 2Dimms per processor).
A Dual socket opteron has a NUMA architecture. No bank-shit or anything. each processor has a simply dual channel memory config like ANY OTHER MAINBOARD out there (im sure you have seen p4s or AMD64 mainboards without any socket filled).
(with the added benefit of being able to share memory bandwith between the cpus)
So 4GB ram is a modest amount...
Ok, than use 4*1GB Dimm. Complete memory bandwith realisation for a dual cpu system.
You dont understand:
HYPERTHREADING =! MULTICORE.
These are 2 complete cpus+a crossbar switch on one die. No shareing of execution units/registers,no sharing of anything but the ram bandwith.
Amd dual core cpus are FASTER than 2 single core cpus in dual socket boards (with the exception of extremely bandwith demanding streaming applications) simply because of much faster on-die cache coherence communication.
A quad core cpu will most likely see more bandwith problems, but could (with ddr-2, ect) still be very well in the same class as a 4 single-core machine.
but looking back, the alpha wasnt powered by magic either....
It was way faster than anything else, but it bought that kind of dominance by using something that now limits x86: A massive power budget.
Alphas used 80W+ back in times when 25W of a pentium2 seemed horrendious, so its not that miraculous that they got more performance out of it.
no, it can only run omega fighter.
Most likely because they only ported one of the many many emulation kernels of the "real" mame.
Well, i think maybe the main reason is fear of fabbing problems. Being burned with IBMs problems, they may have some paranoia about another cpu supplier with only a single cutting edge fab (even if its a megafab)
i know that the american voting system is prone to create a situation where one can only chose between bad and worse...
But the sad thing is, there were still more people voting for worse...
So voting would still help, at least compared to bitching about corrupt politicans and not going to vote because "it doesnt matter anyway"...
If your ega emmits the capital letters everywhere, maybe its no wonder you were left behind, Mr. Linux Administrator.
Vote?
But dont forget that our galaxy is in chaotic movement within the local cluster, which is moving towards to Great Wall.
But there is a simple way: Cosmic Background radiation Anisotrophy. If you look at the temperate of the background, there is an almost perfect doppler shift corresponding to a speed of (IIRC) about 600-700Km/s for earth. Absolute. Total.
Well, by posting your password sheme here, you give any atacker trying to brute force them 3-4 orders of magnitude lower problem compexity.
Which helps SO much here on slashdot where every link usually starts with RE:...
Wow.
Thats a really neat "case".
One question: Typing error on your side? Do you mean 10ms or 10 us?
Because 10ms is hardly a latency to brag about... for any os...
http://www.w3sh.com/archives/2005/05/enfin_un_bon
Just as stupid and non-technical as the one the story is about, but at least i dont claim that its something cool or ingenius (like that summary of this story)
Its not that easy.
Just for your information: the classical picture of a laser trap always presented actually is only a so called optical molasse, it will slow down molecules, but cannot trap them. (its actually quite obvious, think about the balance of force on the particle as its speed aproaches 0... it can be descriped by a modification of the stokes formula for objects in fluids)
For real confinement, you need futher stuff, for example quadrupol coils for a penning type trap, ect.
I dont know what exactly they have working, but its quite possible that they have some kind of feedback loop for their coil management or something like that...
No, current heads ARENT bigger because they have to move around. IN fact, being big is quite a hinderance when moving.
The simply size-limiting factor of current heads is the writing coil. You cant just shrink it down to 500nm or something.
b)Again: i now see what you have in mind (wasnt that clear before). Ok, now you have the hyothetical disc with all 20,000 heads in a row (20,000 singly points of failure, btw). you already spoke of the thermal problems/vibration.
But take a look at the electrical side. Signal processing wise: If you rally want to use all those heads at once, every head needs the signal processing pipeline backing it that now a single HD has to at its disposal... So we are talking about supercomputer class dsp power for a single drive here.
The alternative would be just switching heads, but that would cut down STR to normal hd levels, and the access time wouldnt improve that much (at least considering the cost of the design) because modern high end HD alrady have positioning times in the ballpark of the spin latency.
Also you simply disregard problems with stiktion (you cant park such a head assembly outside the data layers) and keeping the damn thing floating above the surface (the old drum memories hadnd those problems). We are talking about sub um pressision in lateral and horizontal direction over the whole assembly. Thats not cheap, even if it were possible.
Sorry, was a bit overreacting.
But it can get very annoying if somebody who has litteraly no idea what he is talking about has an idea and presents it like he alone found the philosophers stone and everybody else is stupid.
Its even worse with all those free energy nuts that claim infinite forces and stuff only because they cannot comprehent that even an integral from -inf to +inf can converge...
And about the harddiscs: its incredible enough that those things are working even with one head, given all the density improvements and the low cost manufactoring.
Sorry, but i call bullshit.
Do you know just how many tracks a disk has? Hint: its many many thousand. You cant put so many heads over the disc because the head is orders of magnitudes wider than the track. Plus if you were only doing sparse head placement (like 100 heads evenly spaced), you could not keep them calibrated, plus even IF you could, seek times would get MUCH worse because of resonanz modes of the arm complex, increased inertia, ect.
Add to this the problems of sticktion, disturbance of the bernoulli effect because of the high head density, ect,ect, and the idea becomes braindead.
(there were dual head hds years ago, but calibration was a bitch).
You know, the harddisc industry (in fact any industry with x000 employees ) HAS people that are WAY smarter than you. So whenever you encouter something that doesnt make sense to you, maybe YOU are wrong.
Thats not cold fusion....
For "cold" atoms, the tunnel barrier is way to high and broad... thats not a few orders of magnitutes too low, its hundreds...
The only way ever to get fusion (and not even in the break even way of things, just like 1 reaction per mol per century) is the very far end of the bolzmann tail, which isnt cool anymore.
Just think about it: There are detectors for neutrino experiments using thousands of tons of water or other stuff (like that in japan, nakasomthing), under EXTREME survilance by photodetectors &co. And in all those years, not a SINGLE spontanous fusion process ever happened with all those 10^30 or so atoms...
There IS NO SUCH THING als cold fusion actually happening in the natural world.
And what THEY (in the article) do is just a cheap trick in order to remain "cold". Their particles are not in thermal equilibrium, so temperature has no meaning, and their kinetic energy would correlate to a VERY high temperatur.
Its no rocket science to to use a particle accelerator and get fusion. Scale it down to the minimum possible, sacrificing production rate to a few 100 per second, and you can even put it in a breadbox.
You can stop it now. The reality distortion field was officially shut down, so welcome back to the real world.
the g5:
- greater clockspeed increase. yeah right. With watercooling 700Mhz. Including die-shrink (which completely failed).
- Better vectore engine: never running at peak performance because of sucky chipset/memory interface. YOu cant MAC without data, even if you have the execution units.
its not about the 3Ghz. with 130nm, the g5 was so hot that the 2.5 needed watercooling (and the real amount was never dicloused), and the die shrink didnt help. So no laptops (did you see the story that laptops sales surpased desktop ones? or that the powerbooks are a apple stronghold?). No g5-mini.
And if IBM were so superiour with their fabbing, why does the result suck that much?
While i too agree that AMD should have been prefered, it is understandable because of the Pentium M. Apple needs cool chips for notebooks, and while the k8 core is catching up with power-now&co, its still behind p-m. But that doesnt matter, because they are compatible....