Well, the pads/pad drivers do consume a very significant quantity of the total die power. A dual-die solution has more pads that run at the same speed as a single-die and so should consume more power. In *theory*, it should be possible to leverage those additional I/O for increased performance since you do have more bandwidth. In practise, it might be damn difficult.
All of this is implementation dependent. Did they do a true MCM with reduced driver sizes? What sort of package are they using? How is the interconnect between the two dies managed? You'd have to see it and benchmark it to know for sure -- everything else is fairly idle speculation. Including this post.
Well, it seems to me that having a black hole eat the moon wouldn't be *so* bad. I'll miss the thing, but the resulting singularity shouldn't cause massive gravitational changes since it will have the same mass as the moon and the same orbital velocity. Might even be sorta handy as a bottomless garbage pit.
Actually, my understanding is that the moon *is* getting further away, just very slowly. It sucks out the rotational energy of the Earth to do it. Destroy the moon! It's stealing our precious rotational energy!
Here is the problem I have. The "data source" is not a "known one". They are not listening to "His line" they are listening to the whole ISP. Even if its just a header grep...they have NO RIGHT to recieve and look through ANY data except that which comes from or two who they are looking at...even if it is JUST a gheader grep.
But it's unreasonable to require FBI agents to wear blinders and earplugs in case they observe a crime that they weren't directly investigating!
The big problem isn't really resistance (although that is certainly a part of it) -- its the RC constant and the drive of the transistors connected to a wire. Making some gross assumptions we have: R proportional to L/A or since Z doesn't change, L/W C proportional to L*W RC proportional to L^2
transistor drive proportional to feature size squared.
So, when we move to a new process with a smaller feature size (lets say 1/2 of the current), the wires reduce in length by 1/2 and the RC reduces to 1/4. This is good since the drive of our transistors has also dropped to 1/4. But of course, the whole reason for moving to a smaller process is to get more transistors on an economical die, so the die size remains fairly constant -- we just cram four times as many transistors into it! And the wire length remains the same so our poor 1/4 drive transistors have to charge up a wire with an unchanged RC.
Notice that all of this ignores the effect of having to reduce voltage to avoid hot-electron effects -- these are quite important.
It seems rather likely that all known transuranic elements occur *somewhere* in nature, if only briefly at a result of a supernova. As for radiation, its radiation -- source is irrelevant. No highspeed heavy nuclei are going to escape the sun anyway.
Slashdot Mirror
Imp. gal == 4.55 l
A. gal == 3.78 l
1 gal == 4 quarts
1 quart = 2 pints.
20 Imp oz == 1 Imp pint
16 A. oz == 1 A pint
therefore:
12 A oz bottle = 355 ml
12 Imp oz bottle = 341 ml
and, more importantly
472.5 ml == 1 A pint of beer
569 ml == 1 Imp. pint of beer.
Just clearing things up.
Well, the pads/pad drivers do consume a very significant quantity of the total die power. A dual-die solution has more pads that run at the same speed as a single-die and so should consume more power. In *theory*, it should be possible to leverage those additional I/O for increased performance since you do have more bandwidth. In practise, it might be damn difficult.
All of this is implementation dependent. Did they do a true MCM with reduced driver sizes? What sort of package are they using? How is the interconnect between the two dies managed? You'd have to see it and benchmark it to know for sure -- everything else is fairly idle speculation. Including this post.
And presumably iOK as well .....
Well, it seems to me that having a black hole eat the moon wouldn't be *so* bad. I'll miss the thing, but the resulting singularity shouldn't cause massive gravitational changes since it will have the same mass as the moon and the same orbital velocity. Might even be sorta handy as a bottomless garbage pit.
Actually, my understanding is that the moon *is* getting further away, just very slowly. It sucks out the rotational energy of the Earth to do it. Destroy the moon! It's stealing our precious rotational energy!
The big problem isn't really resistance (although that is certainly a part of it) -- its the RC constant and the drive of the transistors connected to a wire. Making some gross assumptions we have:
R proportional to L/A or since Z doesn't change, L/W
C proportional to L*W
RC proportional to L^2
transistor drive proportional to feature size squared.
So, when we move to a new process with a smaller feature size (lets say 1/2 of the current), the wires reduce in length by 1/2 and the RC reduces to 1/4. This is good since the drive of our transistors has also dropped to 1/4. But of course, the whole reason for moving to a smaller process is to get more transistors on an economical die, so the die size remains fairly constant -- we just cram four times as many transistors into it! And the wire length remains the same so our poor 1/4 drive transistors have to charge up a wire with an unchanged RC.
Notice that all of this ignores the effect of having to reduce voltage to avoid hot-electron effects -- these are quite important.
It seems rather likely that all known transuranic elements occur *somewhere* in nature, if only briefly at a result of a supernova. As for radiation, its radiation -- source is irrelevant. No highspeed heavy nuclei are going to escape the sun anyway.