After the 10GHz milestone is reached in mainstream PCs, how will the next speeds look like? Will they slowly evolve 1GHz at a time, or will we see a very rapid approach to 20GHz? How long after 10GHz do you think 20GHz will come out? I would guess it would be similar to the time it took from the 1GHz milestone to reaching 2GHz. There is probably a mathematical formula that would answer these questions :?
Moore's law does not really apply to the speed of the processors, rather than its transistor density and its performance in MIPS. New technologies exist, but are not implemented yet for economical reasons. If the companies had infinite budgets we would have had 100GHz CPU in the mainstream market by now. As soon as the market puts the pressure on the CPU manufacturers to bring out the new tech CPUs, there will be a boom in technology, speeds etc, for a few years again. Then a flat period again as other walls will be hit, then another boom, etc etc...
It goes in cycles like everything in life.
1mhz CPUs that have an IPC of like a billion. Thats what i would do, start over, knock down the speed of a proc and improve the arch.
Yeah, i doubt we'll see mainstream 10GHz(single-core, or..per core) DT processors. I'd bet that we won't see >7GHz before something like the 80core Intel comes out. :?
After the 10GHz milestone is reached in mainstream PCs, how will the next speeds look like? Will they slowly evolve 1GHz at a time, or will we see a very rapid approach to 20GHz? How long after 10GHz do you think 20GHz will come out? I would guess it would be similar to the time it took from the 1GHz milestone to reaching 2GHz. There is probably a mathematical formula that would answer these questions :?
Ummm.. when they finally make a flux capacitor (back to the future), and Dish Satellite TV has more capacity then... cable.
. o O (I hate that commercial)
1mhz CPUs that have an IPC of like a billion. Thats what i would do, start over, knock down the speed of a proc and improve the arch.
I was thinking about something like that, and the biggest issue I could see is that programs would have a starved bandwidth. IPC , speed(GHZ), size(nm) and use of cash have yet to reach a balance...
Once software becomes multi threaded, it is easier to add more threads.
More and more mainstream software is using at least some multi threading.
Personally, I think this time next year all new game releases will be multi threaded. At this point, we will have quad core processors being pushed at us by both AMD and Intel. We'll be looking at octo core soon after, and so on.
We will never hit 10GHz. Maybe 5GHz, after which we will just add more and more cores untill we have massively parallel CPUs with hundreds, even thousands, of smaller more specialised cores.
Im not disagreeing with anyone that says we won't hit 10Ghz, but 10 years ago did we think there would be dual core processors. Someone may come up with a unique way of dissipating heat through using a material that may just be in the experimental stage right now. 10 years from now they may be using Bio Processors. I mean they are trying to store data on bacteria as stated in a post called Wack on THG. There is an article too.
I think many people have not read their books before replying here...
There are already CPUs that run at more than 200GHz. They are made primarily of synthetic diamond or carbon nanotubes, BUT they are too expensive yet to manufacture for the mainstream market. Hence, CPUs with mainstream technology may never reach 10GHz, but that does not mean that there are not big GHz CPUs out there.
After the 10GHz milestone is reached in mainstream PCs, how will the next speeds look like? Will they slowly evolve 1GHz at a time, or will we see a very rapid approach to 20GHz? How long after 10GHz do you think 20GHz will come out? I would guess it would be similar to the time it took from the 1GHz milestone to reaching 2GHz. There is probably a mathematical formula that would answer these questions :?
as soon as my local quickmart gets a resupply of plutonioum
i,ll recharge the time machine and let you know.
Didn't Bill Gates didn't think we would need more than like 512kb memory either. And no one thought you would need all the information of a DVD either. Now we have Blue-Ray disks and 4gb RAM with 300gb hard drives easily accessable.
10ghz will come and go before you know it... but not before 10 core cpu's do.
The funny thing is, UNIX and LINUX can run it all.... like 5 years ago.
getting to 10ghz will take a long time but we will get here eventually. Atm focus is on more cores + more IPC.
They would never bother building a processor with 1mhz and 1 billion IPCs when you could probably build a processor with 2ghz and 100 million ipc that would probably be smaller.
10ghz may or may not come mainstream, but i suspect it will just needs a change of technology aint gonna happen with current tech they have maybe the new hafnium gates will make higher frequencies more accessable.
Didn't Bill Gates didn't think we would need more than like 512kb memory either. And no one thought you would need all the information of a DVD either. Now we have Blue-Ray disks and 4gb RAM with 300gb hard drives easily accessable.
10ghz will come and go before you know it... but not before 10 core cpu's do.
The funny thing is, UNIX and LINUX can run it all.... like 5 years ago.
one thing is. i cant see the need for a 500gig hd.
at my work (automotive) we have alldata.
it is around 100gigs(anywhere from9 to11 dvd,s)
by the time i could use 500gigs the hdd would give out
Neither. Progress is marching along just as it had before, but it's just being manifested in different ways. Before, it was more or less a mad dash to get the highest clock speed, with a little architectural improvement along the way. Now that we're at speeds that CPUs can get seriously hot at, the effort has gone into making CPUs that do more per clock tick. However, the throughput change is pretty steady.
Clock speed isn't the main concern anymore. I don't think we'll see a core go over 5ghz. We'll see more and more cores on a single die and maybe even a revamp of the CPU architeture that will yeild better performance without the need for high clock speeds.
that should be de answer, if you study frecuency response of electronic circuits youll see that the capacitances created by frecuency are to limit their bandwidht
10ghz processor is not relevant now, because with 2 cores you can do more aritmetical process than 1 single core @ 10ghz. but thats true we wont have limits on the clock speed on the thing that we will have limits is on the graphics because when we hit the real life graphic barrier we will reach the limit
You are about to answer a thread that has been inactive for more than 6 months. If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.