Costs are roughly proportional to area but if more transistors can be placed into the same area, then the cost per transistor is less and that is what primarily drives investment into new process generations even at the expense of performance.
Either way, the target audience for expensive flash drives is mainly Mac users, and current Mac laptops don't have any legacy USB ports—only USB-C. So building such an expensive product and giving it only a legacy USB port is a really, really bad idea, and has been for at least the last year or so.
But now Kingston can sell you an adapter and tell you that you are using it wrong. I wonder if Apple will sue them.
If one 5 GHz radio can receive without degradation while the other 5 GHz radio is transmitting, then I might consider it "tri-band". Otherwise it is just Linksys marketing lies which is nothing new for Linksys. Why would I ever buy anything from them again?
I don't think desktop computing is dead. It's just turning into a niche market. There will continue to be people who need the maximum amount of computing power available, along with multiple displays and input devices, and those people will continue to buy desktops.
Especially when the fragile unrepairable laptops have chickelet keyboards and glossy screens.
The impression I'm getting in recent years is that we're transitioning towards a computing world where individual consumers primarily want portables, or alternately, "all in one" or super small form-factor desktops which just use mobile motherboards and CPUs anyway.
The high-end "power users" who tell you they still need a desktop machine for the work they do are best served by a "workstation" class system, vs. a regular desktop PC. The primary differentiation between a "desktop" and a "workstation"? Seems to be the inclusion of a Xeon class processor, originally intended to go into servers. Secondarily, workstations tend to offer the highly costly video cards optimized for use with CAD/CAM and other graphics design packages.
I have been using "workstation" class processors on my desktops since the K6-3 where "workstation" means large amounts of DRAM, ECC support (I curse Intel's marketing segmentation), and lots of fast local storage.
I think what has changed is that previously consumers needed a desktop class processor for routine tasks and passive entertainment and now a mobile processor and often mobile form factor is sufficient. On the other hand if you want to get serious work done, you still need at least a desktop form factor and often a workstation class processor.
Big bucks and still 4 lousy cores. Huge amount of R&D went into 10% overall performance increase compared to Skylake (or really anything semi-recent). I want more damn cores, and drop the useless GPU that is wasting silicon area. 6-8 kickass cores should be the norm these days, but intel wants a massive premium for that.
The huge amount of R&D, and it was not that huge compared to the gains, went into designing a core which could be used on servers, workstations, and desktop products. If you want a server or workstation CPU, then market segmentation means that speed cost money. How fast do you want to go?
There is no incentive to innovate, so there is no innovation. Desktop CPUs will remain in a holding pattern until something happens to force their hand.
Intel has innovated but in ways other than increasing single threaded scalar performance like increasing vector performance, lowering cost, and lowering power which is a requirement for increased transistor densities. If there was an easy way to continue increasing single thread scalar performance, then it would have been applied to server CPUs and in some cases it has like diamond composite heat spreaders.
This was the entire gist of my post. If new offerings are only a marginal improvement over what I have now, and I'm likely not to notice much of a change in performance, why should I upgrade? And this is a self-perpetuating problem. The more times they release lackluster improvements, the more times we opt not to upgrade, they lose more profit and decide against developing better cuz better isn't selling.
Upgrade when you can double or quadruple the amount of RAM for the same price.
I would like to see the CPU RAM bottleneck get as much attention as the CPU itself has. Unless we get that addressed I really don't see faster chips doing much good.
That will be a tough row to hoe unless DRAM is replaced with something else. Increasing integration helps with bandwidth and parallelism but helps little with sense amplifier latency. That leaves increasing memory parallelism and caching as the only options and we have been doing that.
Some old high performance computer designs replaced DRAM with SRAM but at considerably increased cost and lower density.
Thing is, even for people who actually use a computer to do work - how many tasks are really CPU-limited anymore? Obviously there are some niches where a faster CPU will improve the efficiency of their workflow... but that's a small percentage of an ever-declining overall percentage.
Transcoding, simulations, and my email client doing searches are CPU limited for me but the biggest time wasters are poorly programmed applications like Firefox.
If i remember correctly, Essen superconductor is an ittrium barium copper oxide one.
Do to the difficult mechanical properties of high temperature superconductors, they are enclosed in metal which is usually (always?) silver. And while the super conducting element itself is super conducting, the cable or wire as a whole is not. Professor Eagar referred to this as "cold silver" and said the improvement was 20% in this video. Considering the cost compared to aluminum, superconducting power transmission is only going to be used where space and weight constraints are critical.
In Portugal, the excess energy from solar (during day) or wind (usually during night) is used to pump water upstream back to hydroelectric dams reservoirs, back to potential energy, to be used later, when there is no enough wind or solar energy. Of course, like any solution, it have its limits and all countries know that they have to have several energy sources in parallel, so when one is weak, the remaining ones should fill the gap... Even coal, if for some reason the a boiler stop working, you have to fetch energy from other places
Well then, we can just build more dams and water reservoirs just as soon as the Green stop protesting against them and the nuclear power plants.
Impractical fantasy you say? There is also a 1 km liquid nitrogen cooled superconducting installation in Essen that has been working just fine as a part of Essen power grid for several years already. This installation has replaced a 100kV AC powerline. No helium was needed and not that much liquid nitrogen either thanks to a good insulation. It just works.The reason for that installation was a different one, though - there was no room left in the underground channels for additional power lines and that superconducting cable transfers 5 times as much power as a normal copper cable with the same diameter.
Considering that the resistance of copper wire drops to about 1/7th at that temperature and most of the composition of the superconductive wire is silver metal, that is not much of a deal. It is really a liquid nitrogen cooled silver wire with some impurities.
The major breakthrough came in the 90s with the advent of the cell phone. Suddenly there was a huge market incentive for investing into battery research to maximize power density. That is why we have 200+ mile range on electric cars now, rather than the 30-70 mile range they could reach back then.
What? Power density in batteries has not been a problem in electric vehicles for a long time; it mattered in power tools though. Energy density has been and energy for price is the largest issue and that has not been solved.
Yep, natural gas and nuclear can provide power when solar isn't providing enough at the moment, for whatever reason. That's a great mix. The cheapest, cleanest energy when it's available, reliable energy that's still clean and reasonably cheap when the more preferred energy isn't sufficient at the moment.
This works well for natural gas because the consumable is a large part of the cost but not for a nuclear power plant where you want to operate at as large a capacity factor as possible. One of the deliberate ploys the greens use is to push for mandatory use of solar and wind power because it will lower the capacity factor of nuclear power making it less economical.
Of course if people want to pay for both the energy delivered *and* when it is provided, then night time power provided by nuclear and hydroelectric plants will be a lot more economical considering that solar power will not be available at any cost. I wonder how much overcapacity in wind it would require to provide reliable power at night.
Challenger was launched right before Reagan's SOTA address, which was a massive scheduling fuckup right there. Obviously in those circumstances there would be high-level pressure not to delay the launch, and that was confirmed during the post-crash investigation.
Intel says transistors produced in this way will be cheaper than those that came before, continuing the decades-long trend at the heart of Moore's Law -- and contradicting widespread talk that transistor-production costs have already sunk as low as they will go.
Err, what now? I thought smaller transistors were desirable for performance reasons. Has the marginal per-transistor cost been what's holding us back all these years?
I was under the impression that the costs for microprocessor fabrication had to do with their design and then building the foundry. The per-unit cost (and thus per-transistor cost) is utterly negligible, right?
This is a salient point because it implies that in decades to come we're eventually going to see a steep drop-off in prices for not just CPUs, but also RAM and flash memory once enough patents expire and enough high-output fabs come online, which promises to be a utterly world changing solution-in-search-of-a-problem. (Specifically, I predict this will be the point at which AI really takes off.)
Moore's law has *always* been primarily of economic importance. Decreasing the cost per transistor is what makes later fabrication node economically feasible.
In technology, 14-20 years is effectively 100 years. Technology is old news in 5 years and almost useless in 10 years. Since we're talking about CPU companies, let me know how competitive a 14 year old CPU is. Patents are great for innovative breakthroughs. They are bad for evolutionary next steps. Instead of making lots of quick steps and evolving technology quickly, create artificial gaps between each step and slow things down.
But that doesn't change the mechanism in finances, and the time it would require to recoup one's costs. The tech may be old, but if $X billion has been spent in trying to develop Y, it's not gonna be recouped in 5 years just b'cos Y is obsolete in 5 years. So they'd either have to hike prices of Y, which would then make it more difficult to sell, and longer to recoup $X B or they can charge the patent costs and split those costs upfront
In the case of high performance logic semiconductors, achieving the same performance on a 10 year old process costs 3 times more than if you had developed a new process following Moore's Law so there is a great incentive to pay for research and development; if you stop, then you will not be competitive and this will remain true as long as the cost per transistor continues to decrease.
The important criteria is the cost per transistor which must decrease to make the next fabrication node economical. This comes even at the cost of reduced performance.
There was never a road-map or intention to eventually bring it to the desktop as a i386 replacement.
That is not what Intel's actions and marketing indicated. They said Itanium would replace the last generation of x86 which was represented by the Pentium 4 and Itanium included hardware support for executing x86 code.
I've been building my own computers for nearly 20 years, and I don't understand why anyone buys Intel over AMD.
What's the point of paying double/triple the price? Better performance?
Intel's compilers and libraries only take advantage of features found in Intel processors. AMD is great if you want your software to ignore the various instruction set extensions. So the extra cost for Intel CPUs is actually a licensing cost for their compilers and libraries to pay for Intel's programmers to go out of their way to cripple performance on AMD; that does not come cheap.
Slashdot Mirror
Costs are roughly proportional to area but if more transistors can be placed into the same area, then the cost per transistor is less and that is what primarily drives investment into new process generations even at the expense of performance.
Intel's William Holt gave a recent lecture on the subject - Moore’s Law: A Path Forward.
Either way, the target audience for expensive flash drives is mainly Mac users, and current Mac laptops don't have any legacy USB ports—only USB-C. So building such an expensive product and giving it only a legacy USB port is a really, really bad idea, and has been for at least the last year or so.
But now Kingston can sell you an adapter and tell you that you are using it wrong. I wonder if Apple will sue them.
You're gonna be fucked when your house becomes sentient.
That was his house posting using his account.
If one 5 GHz radio can receive without degradation while the other 5 GHz radio is transmitting, then I might consider it "tri-band". Otherwise it is just Linksys marketing lies which is nothing new for Linksys. Why would I ever buy anything from them again?
I don't think desktop computing is dead. It's just turning into a niche market. There will continue to be people who need the maximum amount of computing power available, along with multiple displays and input devices, and those people will continue to buy desktops.
Especially when the fragile unrepairable laptops have chickelet keyboards and glossy screens.
The impression I'm getting in recent years is that we're transitioning towards a computing world where individual consumers primarily want portables, or alternately, "all in one" or super small form-factor desktops which just use mobile motherboards and CPUs anyway.
The high-end "power users" who tell you they still need a desktop machine for the work they do are best served by a "workstation" class system, vs. a regular desktop PC. The primary differentiation between a "desktop" and a "workstation"? Seems to be the inclusion of a Xeon class processor, originally intended to go into servers. Secondarily, workstations tend to offer the highly costly video cards optimized for use with CAD/CAM and other graphics design packages.
I have been using "workstation" class processors on my desktops since the K6-3 where "workstation" means large amounts of DRAM, ECC support (I curse Intel's marketing segmentation), and lots of fast local storage.
I think what has changed is that previously consumers needed a desktop class processor for routine tasks and passive entertainment and now a mobile processor and often mobile form factor is sufficient. On the other hand if you want to get serious work done, you still need at least a desktop form factor and often a workstation class processor.
Big bucks and still 4 lousy cores. Huge amount of R&D went into 10% overall performance increase compared to Skylake (or really anything semi-recent). I want more damn cores, and drop the useless GPU that is wasting silicon area. 6-8 kickass cores should be the norm these days, but intel wants a massive premium for that.
The huge amount of R&D, and it was not that huge compared to the gains, went into designing a core which could be used on servers, workstations, and desktop products. If you want a server or workstation CPU, then market segmentation means that speed cost money. How fast do you want to go?
There is no incentive to innovate, so there is no innovation. Desktop CPUs will remain in a holding pattern until something happens to force their hand.
Intel has innovated but in ways other than increasing single threaded scalar performance like increasing vector performance, lowering cost, and lowering power which is a requirement for increased transistor densities. If there was an easy way to continue increasing single thread scalar performance, then it would have been applied to server CPUs and in some cases it has like diamond composite heat spreaders.
This was the entire gist of my post. If new offerings are only a marginal improvement over what I have now, and I'm likely not to notice much of a change in performance, why should I upgrade? And this is a self-perpetuating problem. The more times they release lackluster improvements, the more times we opt not to upgrade, they lose more profit and decide against developing better cuz better isn't selling.
Upgrade when you can double or quadruple the amount of RAM for the same price.
I would like to see the CPU RAM bottleneck get as much attention as the CPU itself has. Unless we get that addressed I really don't see faster chips doing much good.
That will be a tough row to hoe unless DRAM is replaced with something else. Increasing integration helps with bandwidth and parallelism but helps little with sense amplifier latency. That leaves increasing memory parallelism and caching as the only options and we have been doing that.
Some old high performance computer designs replaced DRAM with SRAM but at considerably increased cost and lower density.
Thing is, even for people who actually use a computer to do work - how many tasks are really CPU-limited anymore? Obviously there are some niches where a faster CPU will improve the efficiency of their workflow... but that's a small percentage of an ever-declining overall percentage.
Transcoding, simulations, and my email client doing searches are CPU limited for me but the biggest time wasters are poorly programmed applications like Firefox.
If i remember correctly, Essen superconductor is an ittrium barium copper oxide one.
Do to the difficult mechanical properties of high temperature superconductors, they are enclosed in metal which is usually (always?) silver. And while the super conducting element itself is super conducting, the cable or wire as a whole is not. Professor Eagar referred to this as "cold silver" and said the improvement was 20% in this video. Considering the cost compared to aluminum, superconducting power transmission is only going to be used where space and weight constraints are critical.
In Portugal, the excess energy from solar (during day) or wind (usually during night) is used to pump water upstream back to hydroelectric dams reservoirs, back to potential energy, to be used later, when there is no enough wind or solar energy. Of course, like any solution, it have its limits and all countries know that they have to have several energy sources in parallel, so when one is weak, the remaining ones should fill the gap... Even coal, if for some reason the a boiler stop working, you have to fetch energy from other places
Well then, we can just build more dams and water reservoirs just as soon as the Green stop protesting against them and the nuclear power plants.
Impractical fantasy you say? There is also a 1 km liquid nitrogen cooled superconducting installation in Essen that has been working just fine as a part of Essen power grid for several years already. This installation has replaced a 100kV AC powerline. No helium was needed and not that much liquid nitrogen either thanks to a good insulation. It just works.The reason for that installation was a different one, though - there was no room left in the underground channels for additional power lines and that superconducting cable transfers 5 times as much power as a normal copper cable with the same diameter.
Considering that the resistance of copper wire drops to about 1/7th at that temperature and most of the composition of the superconductive wire is silver metal, that is not much of a deal. It is really a liquid nitrogen cooled silver wire with some impurities.
... and use deep wells/deep heat pipes to extract the heat converting with high efficiency
Well, there's your problem.
The major breakthrough came in the 90s with the advent of the cell phone. Suddenly there was a huge market incentive for investing into battery research to maximize power density. That is why we have 200+ mile range on electric cars now, rather than the 30-70 mile range they could reach back then.
What? Power density in batteries has not been a problem in electric vehicles for a long time; it mattered in power tools though. Energy density has been and energy for price is the largest issue and that has not been solved.
Yep, natural gas and nuclear can provide power when solar isn't providing enough at the moment, for whatever reason. That's a great mix. The cheapest, cleanest energy when it's available, reliable energy that's still clean and reasonably cheap when the more preferred energy isn't sufficient at the moment.
This works well for natural gas because the consumable is a large part of the cost but not for a nuclear power plant where you want to operate at as large a capacity factor as possible. One of the deliberate ploys the greens use is to push for mandatory use of solar and wind power because it will lower the capacity factor of nuclear power making it less economical.
Of course if people want to pay for both the energy delivered *and* when it is provided, then night time power provided by nuclear and hydroelectric plants will be a lot more economical considering that solar power will not be available at any cost. I wonder how much overcapacity in wind it would require to provide reliable power at night.
Challenger was launched right before Reagan's SOTA address, which was a massive scheduling fuckup right there. Obviously in those circumstances there would be high-level pressure not to delay the launch, and that was confirmed during the post-crash investigation.
What happened even has historical president: "Will no one rid me of this troublesome priest?"
Intel's William Holt gave a recent lecture on the subject - Moore’s Law: A Path Forward.
Intel says transistors produced in this way will be cheaper than those that came before, continuing the decades-long trend at the heart of Moore's Law -- and contradicting widespread talk that transistor-production costs have already sunk as low as they will go.
Err, what now? I thought smaller transistors were desirable for performance reasons. Has the marginal per-transistor cost been what's holding us back all these years?
I was under the impression that the costs for microprocessor fabrication had to do with their design and then building the foundry. The per-unit cost (and thus per-transistor cost) is utterly negligible, right?
This is a salient point because it implies that in decades to come we're eventually going to see a steep drop-off in prices for not just CPUs, but also RAM and flash memory once enough patents expire and enough high-output fabs come online, which promises to be a utterly world changing solution-in-search-of-a-problem. (Specifically, I predict this will be the point at which AI really takes off.)
Moore's law has *always* been primarily of economic importance. Decreasing the cost per transistor is what makes later fabrication node economically feasible.
In technology, 14-20 years is effectively 100 years. Technology is old news in 5 years and almost useless in 10 years. Since we're talking about CPU companies, let me know how competitive a 14 year old CPU is. Patents are great for innovative breakthroughs. They are bad for evolutionary next steps. Instead of making lots of quick steps and evolving technology quickly, create artificial gaps between each step and slow things down.
But that doesn't change the mechanism in finances, and the time it would require to recoup one's costs. The tech may be old, but if $X billion has been spent in trying to develop Y, it's not gonna be recouped in 5 years just b'cos Y is obsolete in 5 years. So they'd either have to hike prices of Y, which would then make it more difficult to sell, and longer to recoup $X B or they can charge the patent costs and split those costs upfront
In the case of high performance logic semiconductors, achieving the same performance on a 10 year old process costs 3 times more than if you had developed a new process following Moore's Law so there is a great incentive to pay for research and development; if you stop, then you will not be competitive and this will remain true as long as the cost per transistor continues to decrease.
The important criteria is the cost per transistor which must decrease to make the next fabrication node economical. This comes even at the cost of reduced performance.
2016 Plenary Session 1 - Moore’s Law: A Path Forward
There was never a road-map or intention to eventually bring it to the desktop as a i386 replacement.
That is not what Intel's actions and marketing indicated. They said Itanium would replace the last generation of x86 which was represented by the Pentium 4 and Itanium included hardware support for executing x86 code.
I've been building my own computers for nearly 20 years, and I don't understand why anyone buys Intel over AMD.
What's the point of paying double/triple the price? Better performance?
Intel's compilers and libraries only take advantage of features found in Intel processors. AMD is great if you want your software to ignore the various instruction set extensions. So the extra cost for Intel CPUs is actually a licensing cost for their compilers and libraries to pay for Intel's programmers to go out of their way to cripple performance on AMD; that does not come cheap.
In the original, Han was the only one to shoot.