I think we hit 3Ghz before hitting the wall. IBM is going to have to utilize some of the more expensive processes to eek out the performance(Ghz) and then scale horizontally. I'm not worried I'd take a quad 2.8Ghz any day.
I think we hit 3Ghz before hitting the wall. IBM is going to have to utilize some of the more expensive processes to eek out the performance(Ghz) and then scale horizontally. I'm not worried I'd take a quad 2.8Ghz any day.
I agree. I think it'll probably hit 3.2GHz at some point, but probably not by MWSF. I would rather see Apple make motherboard modifications, and open the graphics system to the outside rather than a speed increase right now. Nobody is getting much faster, so in the mean time try, and address some of the other areas that prevent Mac sales to the greater public.
Kickaha and Amorph couldn't moderate themselves out of a paper bag. Abdicate responsibility and succumb to idiocy. Two years of letting a member make personal attacks against others, then stepping aside when someone won't put up with it. Not only that but go ahead and shut down my posting priviledges but not the one making the attacks. Not even the common decency to abide by their warning (afer three days of absorbing personal attacks with no mods in sight), just shut my posting down and then say it might happen later if a certian line is crossed. Bullshit flag is flying, I won't abide by lying and coddling of liars who go off-site, create accounts differing in a single letter from my handle with the express purpose to decieve and then claim here that I did it. Everyone be warned, kim kap sol is a lying, deceitful poster.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
I agree. I think it'll probably hit 3.2GHz at some point, but probably not by MWSF. I would rather see Apple make motherboard modifications, and open the graphics system to the outside rather than a speed increase right now. Nobody is getting much faster, so in the mean time try, and address some of the other areas that prevent Mac sales to the greater public.
Opening the "graphics system" to the outside isn't going to happen. Apple uses Open Firmware, and there's no way in hell they're going to implement an ATX style BIOS. . . ever. This isn't a bad thing. The only reason why there aren't a ton of graphics cards out there for mac is because Apple has to supply the firmware for them, and the mac isn't a gaming platform, so the manufacturers have no incentive to make there own mac-compatible firmware.
Splinemodel, as your nick would suggest I thought you would be more susceptible to what I was I was referring to.
I'm not interested in gaming on the Mac nor are most of the people that I know that would like more graphics options from Apple. Highend 3D cards is what we are after. Not the kind intended for use with games.
Kickaha and Amorph couldn't moderate themselves out of a paper bag. Abdicate responsibility and succumb to idiocy. Two years of letting a member make personal attacks against others, then stepping aside when someone won't put up with it. Not only that but go ahead and shut down my posting priviledges but not the one making the attacks. Not even the common decency to abide by their warning (afer three days of absorbing personal attacks with no mods in sight), just shut my posting down and then say it might happen later if a certian line is crossed. Bullshit flag is flying, I won't abide by lying and coddling of liars who go off-site, create accounts differing in a single letter from my handle with the express purpose to decieve and then claim here that I did it. Everyone be warned, kim kap sol is a lying, deceitful poster.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Notice the article doesn't say Intel will never produce higher Hz chips ever, just not higher Hz P4's because the multi-core version will be out before they could get proper ROI for a 4GHz single core P4.
I'm not holding out alot of hope that Intel can bring to market a fast dual core chip either. The problem with the P4 is thermal and it won't get that much better with a dual core chip.
Sure they could lower the clock rate to manage thermal issues but that won't satisfy everyone. I certainly wouln't complain about a dual core P4 running at 2.6GHz or so, but that doesn't mean everybody can make good use of slower cores. SMP support in Linux is really good now, it would be nice to have low cost hardware to take advantage of that software.
Quote:
And that for now they are working on a wide number of options to increase performance--just as IBM has been saying and designing for over a year now. Those are huge differences in the business details from the main thesis of that painfully expired thread.
There are still options floating about for improving 90nm and smaller processes. So I will not be throwing in the towel on faster processors any time soon. I would not be surprised at all to see Intel though come out with a faster processor that does away with alot of the baggage that P4 carries around. Intel may very well build dual processor chips on Dothan type cores, this could happen much faster that trying to squeze 2 P4's on a die. Dothan would allow them the physical space along with the possibility of scaling frequency in the future. That is my take anyways Dothan could be the basis of Intels dual core efforts.
Sure they could lower the clock rate to manage thermal issues but that won't satisfy everyone. I certainly wouln't complain about a dual core P4 running at 2.6GHz or so, but that doesn't mean everybody can make good use of slower cores. SMP support in Linux is really good now, it would be nice to have low cost hardware to take advantage of that software.
They'd have to break the megahertz myth.. which they started in the first place.
"bleh, this xeon 3.6ghz is like way faster than this dual core 2.8Ghz 64bit IBM wuzamajibbit... hah! it even costs more!.. morons.. What's this fancy 'dual core' thing anyway, I bet it means the processor is painted purple or something."
Gaming card revenue drives the R&D budget for supposed high end Pro cards, there's your tie-in. Not to mention that most of those high end pro cards are just non-crippled versions of the same hardware the gamers get. Then the requisite QA and Cert process is done--I'll grant the manufacturers need to be paid for that, but they could go about the whole vertical lineup much more effectively and reduce those outlandish prices by spreading the QA/Cert over a larger number of shipped cards--which don't get crippled--even though they are just Gamer boxes they are shipped in..
That is total speculation from point 1. And crippled is not anywhere near an accurate way to describe the distinct differences in Pro 3D, and gaming card drivers.
While there was a kernel of truth deeply buried in what Nr9 said, the rest of the histrionics around his prediction separate it enough from this that it bears too little resemblance to give him much if any credit.
Notice the article doesn't say Intel will never produce higher Hz chips ever, just not higher Hz P4's because the multi-core version will be out before they could get proper ROI for a 4GHz single core P4. And that for now they are working on a wide number of options to increase performance--just as IBM has been saying and designing for over a year now. Those are huge differences in the business details from the main thesis of that painfully expired thread.
Kickaha and Amorph couldn't moderate themselves out of a paper bag. Abdicate responsibility and succumb to idiocy. Two years of letting a member make personal attacks against others, then stepping aside when someone won't put up with it. Not only that but go ahead and shut down my posting priviledges but not the one making the attacks. Not even the common decency to abide by their warning (afer three days of absorbing personal attacks with no mods in sight), just shut my posting down and then say it might happen later if a certian line is crossed. Bullshit flag is flying, I won't abide by lying and coddling of liars who go off-site, create accounts differing in a single letter from my handle with the express purpose to decieve and then claim here that I did it. Everyone be warned, kim kap sol is a lying, deceitful poster.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
I think there is another GHz left for each processor architecture. There is still the 65 nm node and possibly 45/50 nm node to eek out additional GHz.
Nr9's multi-core PPC 440 predictions is also looking to be true, though probably not for Apple machines. The BlueGene/L (dual-core 440) will be the heart, well 65k hearts, of IBM's upcoming 360 TFLOP supercomputer.
Also remember that there are no free lunches. Multi-core is nice, but it also has diminishing returns with increasing cores. They will become more and more inefficient due to network traffic (communication between cores) as more cores are added. And only certain applicatioins will benifit fully from it. The 700 MHz BlueGene/L supercomputer won't run many applications much faster than an iMac G5.
So there is pressure for companies to both increase single threaded performance (clock rate) and multi-threaded performance for a long while yet.
I think scaling has largely died, that doesn't mean we wont see faster processors, like THT said, there is still a gig left in maturing the process.
There will come a point when cores catch up with the hertz, and the two will increase slowly side by side, but for the next few years it seems that cores will be favourable to speed.
I think there will come a time in about 10 years, when personal computer will become so powerful that there just won't be a need for the rapid development we've seen so far.
If you have an 8 core chip running at 10ghz with a 128 pipe graphics card, that can render 200fps of exceptional photorealistic graphics on a 21"screen with 300 dpi resolution, with 128 channels of sound, realword physics and particles systems, 32 GB of memory, then I don't think the demand will be there to push the development at anywhere near the speed we have seen.
Some people will want more of course, but they'll be buying the 32 way Power8's - that will cost about the same as a kitted out Powermac 2.5 today.
Prediction: There will be faster processors....maybe not copper based switching, but perhaps optics. Anyway, the wall is temporary and will be brought down. Yawn.
Now I thought AMD and IBM where teamed up at this process node. Actually this would be pretty significant as it would indicate to me that AMD probally can get another 500MHz at will out of their Athlon64/Opteron hardware. They are most likely enjoying the fact that they have whiped Intels behind in processor design.
That extra 500MHz comes from the fact that IBM got that much out of the 970FX. Now AMd is actually in the postion of supplying the low power processor so one has to wonder how much they will be willing to crank up the current design.
As to IBM and our supplier of 970's I do wonder when the low power variants will be on line. With the rumors of the iBook updates coming soon there is a remote possibility that the new low power 970FX is ready to hit the street. What would be even better is if this low power 970 was the GX variant with the larger cache. Larger cache is now a popular way to help control power demands in portable chips.
I'm sure we'll see significant shrinkage, but each subsequent step will require more and more problems to be surmounted. That's clear to everybody. I'm no chip expert, but making statements about absolute speed caps is still a little sketchy at best.
Eventually, a viability wall will be hit and the quantum phenomena (which will be responsible for the eventual downfall of current technologies) will prove to be our saving grace. Processing technology built on quantum states will make current computing technology look like vacuum tubes. http://www.sciencedaily.com/releases...1007082711.htm
Comments
Originally posted by the cool gut
Intel Kills Plans for 4GHZ
Guess we'll have to hear from everyone else now ...
this is not news to me, and nr9 had nothing to do with it.
Originally posted by hmurchison
I think we hit 3Ghz before hitting the wall. IBM is going to have to utilize some of the more expensive processes to eek out the performance(Ghz) and then scale horizontally. I'm not worried I'd take a quad 2.8Ghz any day.
I agree. I think it'll probably hit 3.2GHz at some point, but probably not by MWSF. I would rather see Apple make motherboard modifications, and open the graphics system to the outside rather than a speed increase right now. Nobody is getting much faster, so in the mean time try, and address some of the other areas that prevent Mac sales to the greater public.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Originally posted by onlooker
I agree. I think it'll probably hit 3.2GHz at some point, but probably not by MWSF. I would rather see Apple make motherboard modifications, and open the graphics system to the outside rather than a speed increase right now. Nobody is getting much faster, so in the mean time try, and address some of the other areas that prevent Mac sales to the greater public.
Opening the "graphics system" to the outside isn't going to happen. Apple uses Open Firmware, and there's no way in hell they're going to implement an ATX style BIOS. . . ever. This isn't a bad thing. The only reason why there aren't a ton of graphics cards out there for mac is because Apple has to supply the firmware for them, and the mac isn't a gaming platform, so the manufacturers have no incentive to make there own mac-compatible firmware.
I'm not interested in gaming on the Mac nor are most of the people that I know that would like more graphics options from Apple. Highend 3D cards is what we are after. Not the kind intended for use with games.
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Originally posted by AirSluf
Notice the article doesn't say Intel will never produce higher Hz chips ever, just not higher Hz P4's because the multi-core version will be out before they could get proper ROI for a 4GHz single core P4.
I'm not holding out alot of hope that Intel can bring to market a fast dual core chip either. The problem with the P4 is thermal and it won't get that much better with a dual core chip.
Sure they could lower the clock rate to manage thermal issues but that won't satisfy everyone. I certainly wouln't complain about a dual core P4 running at 2.6GHz or so, but that doesn't mean everybody can make good use of slower cores. SMP support in Linux is really good now, it would be nice to have low cost hardware to take advantage of that software.
And that for now they are working on a wide number of options to increase performance--just as IBM has been saying and designing for over a year now. Those are huge differences in the business details from the main thesis of that painfully expired thread.
There are still options floating about for improving 90nm and smaller processes. So I will not be throwing in the towel on faster processors any time soon. I would not be surprised at all to see Intel though come out with a faster processor that does away with alot of the baggage that P4 carries around. Intel may very well build dual processor chips on Dothan type cores, this could happen much faster that trying to squeze 2 P4's on a die. Dothan would allow them the physical space along with the possibility of scaling frequency in the future. That is my take anyways Dothan could be the basis of Intels dual core efforts.
Dave
Originally posted by wizard69
Sure they could lower the clock rate to manage thermal issues but that won't satisfy everyone. I certainly wouln't complain about a dual core P4 running at 2.6GHz or so, but that doesn't mean everybody can make good use of slower cores. SMP support in Linux is really good now, it would be nice to have low cost hardware to take advantage of that software.
They'd have to break the megahertz myth.. which they started in the first place.
"bleh, this xeon 3.6ghz is like way faster than this dual core 2.8Ghz 64bit IBM wuzamajibbit... hah! it even costs more!.. morons.. What's this fancy 'dual core' thing anyway, I bet it means the processor is painted purple or something."
Originally posted by AirSluf
Gaming card revenue drives the R&D budget for supposed high end Pro cards, there's your tie-in. Not to mention that most of those high end pro cards are just non-crippled versions of the same hardware the gamers get. Then the requisite QA and Cert process is done--I'll grant the manufacturers need to be paid for that, but they could go about the whole vertical lineup much more effectively and reduce those outlandish prices by spreading the QA/Cert over a larger number of shipped cards--which don't get crippled--even though they are just Gamer boxes they are shipped in..
That is total speculation from point 1. And crippled is not anywhere near an accurate way to describe the distinct differences in Pro 3D, and gaming card drivers.
Originally posted by AirSluf
While there was a kernel of truth deeply buried in what Nr9 said, the rest of the histrionics around his prediction separate it enough from this that it bears too little resemblance to give him much if any credit.
Notice the article doesn't say Intel will never produce higher Hz chips ever, just not higher Hz P4's because the multi-core version will be out before they could get proper ROI for a 4GHz single core P4. And that for now they are working on a wide number of options to increase performance--just as IBM has been saying and designing for over a year now. Those are huge differences in the business details from the main thesis of that painfully expired thread.
whatever you say man..
you will see..
Now I guess they should have banned me rather than just shut off posting priviledges, because kickaha and Amorph definitely aren't going to like being called to task when they thought they had it all ignored *cough* *cough* I mean under control. Just a couple o' tools.
Don't worry, as soon as my work resetting my posts is done I'll disappear forever.
Nr9's multi-core PPC 440 predictions is also looking to be true, though probably not for Apple machines. The BlueGene/L (dual-core 440) will be the heart, well 65k hearts, of IBM's upcoming 360 TFLOP supercomputer.
Also remember that there are no free lunches. Multi-core is nice, but it also has diminishing returns with increasing cores. They will become more and more inefficient due to network traffic (communication between cores) as more cores are added. And only certain applicatioins will benifit fully from it. The 700 MHz BlueGene/L supercomputer won't run many applications much faster than an iMac G5.
So there is pressure for companies to both increase single threaded performance (clock rate) and multi-threaded performance for a long while yet.
There will come a point when cores catch up with the hertz, and the two will increase slowly side by side, but for the next few years it seems that cores will be favourable to speed.
I think there will come a time in about 10 years, when personal computer will become so powerful that there just won't be a need for the rapid development we've seen so far.
If you have an 8 core chip running at 10ghz with a 128 pipe graphics card, that can render 200fps of exceptional photorealistic graphics on a 21"screen with 300 dpi resolution, with 128 channels of sound, realword physics and particles systems, 32 GB of memory, then I don't think the demand will be there to push the development at anywhere near the speed we have seen.
Some people will want more of course, but they'll be buying the 32 way Power8's - that will cost about the same as a kitted out Powermac 2.5 today.
http://www.geek.com/news/geeknews/20...1018027458.htm
Im sure there were also articles on The Register and Arstechnica yesterday, but they seem to have been pulled.
Clock speed is going backwards! and the fatest single chip you can expect from intel for some time is a 3.8GHZ 2MB cache.
http://techreport.com/onearticle.x/7417
That extra 500MHz comes from the fact that IBM got that much out of the 970FX. Now AMd is actually in the postion of supplying the low power processor so one has to wonder how much they will be willing to crank up the current design.
As to IBM and our supplier of 970's I do wonder when the low power variants will be on line. With the rumors of the iBook updates coming soon there is a remote possibility that the new low power 970FX is ready to hit the street. What would be even better is if this low power 970 was the GX variant with the larger cache. Larger cache is now a popular way to help control power demands in portable chips.
Dave
Originally posted by MarcUK
OTOH AMD's 90nm process seems to be working
http://techreport.com/onearticle.x/7417
Eventually, a viability wall will be hit and the quantum phenomena (which will be responsible for the eventual downfall of current technologies) will prove to be our saving grace. Processing technology built on quantum states will make current computing technology look like vacuum tubes. http://www.sciencedaily.com/releases...1007082711.htm