Redmond (WA) – Less than a year from his previously announced departure from the company he co-founded, the Bill Gates today outlined his vision and opportunities for Microsoft.
The last time I check they where having trouble getting over 3ghz with out over clocking.
So how in the world are they going to get to 10ghz...
Someone who is much more computer experienced then me please explain..? does he mean overclocked or is he saying with a Quad. because The fastest quad I have seen, ran at about 3ghz for each processor. Is it possible to get one chip to run at 10ghz. really?
I think I just read the whole thing wrong because he said parallel so maybe I just didn't understand how this 10ghz is actually reached.
technology evolves. for instance take a look at the power5 from IBM or the cell processors which can go higher than 4Ghz.
Technology will change and new materials will be used. The transition will be slow because things have shifted to the number of cores per chip and that's the only thing that matter now.
Also, speed was relinquish years ago when AMD published the opteron chips.
Size matters. Moving from 90nm to 45nm(?) will create a lot less heat and allow for speed improvements. Also, shrinking components allows for more transistors and what not to fit on a chip. I think the limiting factor is that an electron must not stray outside the physical paths and so they can only shrink things so much. Cheers.
ya the main reason why we aren't sitting at 4.0 or 5ghz is because the chips were creating so much heat around 3.5 ghz that they would either overheat and literally melt or the CPU would need a 200 dollar fan, a top of the line heat sink gell, and a liquid cooling system.
I managed to crank up my E6600 Dual core 2.4 ghz cpu to 3.4Ghz with watercooling, albeit the temp rose from 35 celcius to 48 celcius. I've read that they've managed to get a 3ghz to 5ghz with subzero temperatures. Maybe in future pc's with 10ghz Cpus come with their own refridgation units. I know Bill's a visionary but he's the one who said we wont require more than 640kb ram. Damn my washing machine has more ram.
Edit: Oops dint read the above post which already mentioned his famous quote.
I am running an x6800 at 3.6 with air cooling and no voltage tweaks so hitting above 4 GHz doesn’t seem very hard to achieve for the core2 series of chips. Hitting 4 GHz with 4 cores will happen by this time next year for sure. The core2 series is currently at 45nm and has plenty of head room. More than a few users have their x6800s and x6850s running at 3.9ghz on air cooling. MHz speed isn’t the way of the future multi cores are and that is why the trend of MHz shooting up each year has tapered back.
Odion I have a 35 dollar fan and I am not breaking 50c at full load.
This $35 fan you bought from Airforce? Maybe a part from an F16 air intake even after watercooling the temp of E6600 fluctuates between 55c-59c on full load like coverting a movie or playing BF2142 for hours on. The ambient temp based on mobo is 35c celcius. Maybe E6600 is on 65nm and X6800 is 45nm?
Edit: @ Darthbillboy
LOL maybe you should try watercooling (if you are handy enough to fiddle with coolant and plastic hoses inside the pc) its pretty silent except for the radiator fan which you can place either inside or outside at the back of the pc case. I have put 2 additional 6" fans inside the case and also kept one side pannel off.
"Hitting 4 GHz with 4 cores will happen by this time next year for sure." Presently hell no way, quads are literally hot hot, except maybe if you use something like Vapochill, which cost a grand to set up. Its the reason I went for PS3 instead of upgrading my comp.
"The core2 series is currently at 45nm and has plenty of head room." Nope, its 65nm now, 45 when Penryn comes later this year.
"More than a few users have their x6800s and x6850s running at 3.9ghz on air cooling." ahh well they got to be living in dreamworld or they are lucky as hell.
@sak500 "This $35 fan you bought from Airforce?" What a coincidence? I just got an 80mm Vantec fan. I installed it, got a terrible headache from the noise(10-15 times louder than a 360, imagine!!!), cursed a storm, and threw the mofo blowdrier out. Sure it does great at 89CFM, but not worth 2-5 degrees from my processor. And X6800 is 65nm.
So what the hell happened to the cooler in xb360? 3 cores running at 3.2ghz should be generating quite a bit of heat. And they used passive cooling instead of active one.
Darth you are right it is 65nm my mistake. If you do a Google search you will find tons of users in the 3.8, 3.9 ranges with the x6800 when I get mine up to 3.8 this weekend I can send you picks if you like. It isn’t a dream like you so arrogantly suggest. What CPU do you have? Like I said above I am running at 3.6 with out even tweaking the voltage and my heat is still not going above 50c.
Darth I think you need to do some more reading. Extremetech and a few other sites have overclocked the qx6800 to 3.6 on air. So I am not sure where you are getting the idea that quads get so hot.
LOL I like that but your computer was probably outdated by the time you got to the car and put it in the trunk and the funny thing is your computer was probably the most advanced when you bought it. I bought my computer a year and a half ago for $2,500 and it's now worth about $400 on ebay.
great idea! lets take advice from someone thats alot more experienced like you...
shrinking the process helps out, but also leads to other deamons such as tunneling. Increasing the speed causes such problems because as the frequency increases the capacitance between the gate of the transistor becomes more like a short increasing the amount of current flowing through the system causing heat.
Sak500 I am running the Rosewill RCX-Z775-SL 92mm 2 Ball CPU Cooler in an antec 900. I had the e6600 (best bang for you buck by a long shot) before i got the x6800 (i got it for 500!!!) last week and i was running it at 3ghz and it didnt even break a sweat. The antec 900 might be a big part of my temps; i have it loaded down with fans and i put a ton of time into getting the proper flow to keep everthing cool.
I believe x6800 doesn't have the multiplier locked right? Anyway my FSB is at 379Mhz so maybe that's the reason for the whole cpu heating upto 50+ deg.
LOL I guess it would for some, bubbles up bro that was good laugh for being honost. Tech talk can get a little hard to follow somtimes, I'm fairly tech savy and I get lost more than I'm found.
IBM has seems to have found the key for this already. IBM's Power6 Processor can be up to 4.7 Ghz. Also, the Cell processor has been clocked up to 4.6 Ghz. That is without an overclock. :P IBM said it will go even higher with the next-gen. Cell processor and their Power7 PC Processor up to 6+ Ghz. I believe that they will indeed do so.
Intel already has some plans to go back up with Ghz with some of its new processors being released into next year. They are taking their time with this though. Their next quad core CPU should hit 3.3 Ghz. I doubt Intel cares to go past 4 Ghz, but they might do so. Their new processors can overclock pretty close to it if not beyond.
IBM has used some sort of silicon-germanium fusion for transistors that enabled them to reach a 500 Ghz speed. Yes, it was in a controlled laboratory inside a really cold refrigerator... but still, that gives hope that in a not so distant future we might reach blazing fast speeds nearing those.
Furthermore, another laboratory in japan unveiled a 1000 core CPU, running at 500 Mhz.
Now, imagine combining both breakthroughs.... yeah 10 Ghz, just like a few kilobytes were enough for an HDD.
The last time I check they where having trouble getting over 3ghz with out over clocking.
So how in the world are they going to get to 10ghz...
Someone who is much more computer experienced then me please explain..? does he mean overclocked or is he saying with a Quad. because The fastest quad I have seen, ran at about 3ghz for each processor. Is it possible to get one chip to run at 10ghz. really?
I think I just read the whole thing wrong because he said parallel so maybe I just didn't understand how this 10ghz is actually reached.
Size matters. Moving from 90nm to 45nm(?) will create a lot less heat and allow for speed improvements. Also, shrinking components allows for more transistors and what not to fit on a chip. I think the limiting factor is that an electron must not stray outside the physical paths and so they can only shrink things so much. Cheers.
I think we would see this on a Single core, before a multi-core system.
I am thinking that for that it would take MAJOR hardware changes for Mobo's, Mem, and Harddrives to even utilize that type of speed.
It would be interesting to see none the less.
ya the main reason why we aren't sitting at 4.0 or 5ghz is because the chips were creating so much heat around 3.5 ghz that they would either overheat and literally melt or the CPU would need a 200 dollar fan, a top of the line heat sink gell, and a liquid cooling system.
bah Bill Gates can be an idiot sometimes. Anyone remember his famous other quote "640kb ought to be enough for anybody".