ARM is RISC origins so it almost always gets throughput of one instruction per clock. The difference of Cortex vs ARM11 is true SIMD instructions with lower latencies between and better power efficiency vs just pipelining in the ARM11.
All in all, not much difference between them if you run them on a crappy OS like Windows Mobile as it only uses mostly ARMv4 instructions without any activated extension support for NEON, VFP, and such. Android has the same issue actually, and most (except possibly Symbian), do not have optimized GPU drivers, but just random blobs that only pass the conformance tests.
Of course, using any SoC processor that's not the Qualcomm MSM72xx is always a HUGE win in speed. The MSM72xx has SINGLE-HANDEDLY held back the "mobile" platform for the last 3 generations due to its poor ARM11 performance vs its XScale and Samsung cousins, as well as tight-fisted, unreleased documentation for tuning and accelerating due to Qualcomm's money-grubbing licensing policies. Qualcomm touch-screen lag, slow insensitive GPS, and htcclassaction.org are good examples of the infamy Qualcomm single-handedly created.
Notice the mobile marketplace is suddenly perking up now that the MSM72xx platform is now on its way to the graveyard? The OMAP3 is definitely a platform to be on.
As flashy as the Qualcomm Snapdragon is, it's still the one-off platform ride to hell as it was for the MSM72xx.. Just don't ride that bxxxch.
Any lag you experience is an artifact of the Qualcomm cpu. This has only been a problem for even Windows Mobile when HTC switched to the Q chip starting with the Kaiser (AT&T Tilt) phone.
It's been evident ever since, so it's no surprise that Android suffers from it also especially with HTC's scrolling hacks all over it.
The lag goes away once you switch back to a XScale cpu, or anything else that isn't a Qualcomm cpu.
They bite off more than they can chew, deny it through canned CS responses, and then try to drag it out until it's sorta fixed, or people loose interest.
It's bettable that they wrote all the Hero code without realizing they'll fall under GPL. It's probably also all statically linked together and they didn't bother separating it properly, which is why they can't "release" any of the code "right now".
It's only fair to get a court order to get the code immediately released before they do any more damage to hide their violations.
Another reason nobody wants to seriously develop for Windows Mobile is because its the off-platform for graphics that has only be highly successful in holding the platform back. Thank you Don Couch.
Most graphics developers doing cross-platform have settled on OpenGL as the lowest-cost and highest portability API to use. But, Microsoft continues to push DirectDraw and Direct3D where 90% of their Window Mobile ODMs don't even bother hardware-accelerating by not shipping drivers like HTC (HTCClassAction.org), or ship inferior 3D in-house hardware with crappy or expensive drivers (Samsung/Qualcomm), and no FPU support in the OS even when the shipped CPU has one.
So, combined with the fact Windows Mobile is basically Windows 98 graphics-wise, WinCE5 IS Win98 (32MB max ram per process for instance), you have an ancient platform that doesn't support any advanced hardware, and cannot do so no matter how much you beat it.
Then MS goes on the anemic.NETCF push that is not only performance-sapping as writing your game apps in HTML on WebOS, but avoids the much needed rearchitecture for a high-performance modern GUI. Even Qt4 is more functional with more capability than.NETCF. I am beginning to believe at least a major portion of the Windows Mobile's software being based from India as a reason for all this lazy, low performance.
WM6.5 is Copeland -- WM7 will be Rhapsody.
These are serious fundamental, architectural problems that shipping a new OS and/or skin is NOT going to solve especially after ramping up development 5 years late.
Balmer, DUMP DirectDraw/Direct3D/.NETCF/WinCE5 and move forward from your current utter fail!
Problem is Blizzard isn't doing that anymore. They don't want their programmers pulling Duke Nukem 3Ds anymore, so instead they just do Agile development and ship it as soon as possible. Then, follow-up with a "horde" of patches.
He could've spent a fraction of that $38K to reduce his power bills to something closer to $100 per month. Better insulation, more energy efficient PSs in his computers, not leaving electronics on 24/7, change out to CF bulbs, and so on -- seriously!
He's going to hit Jevon's Paradox and end up not actually saving anything, IMO.
I'm more in support of community solar (have the HOA do it, or in a park, etc) which is more beneficial and lower cost. Mass solar.
Seconded. Babylon5 attacked and discussed the very subjects TFA says has never done before, besides TOS.
The difference is there was an overall story-arc and the guy won a writer's award (Nova?) for it.
TV == short memories.
rewrite Qt in C++ templates?
on
Qt Becomes LGPL
·
· Score: 1
Would Qt be better written now using C++ templates and the latest template meta-programming instead of the meta-language Qt is currently using?
We're at the point where most main-stream platform compilers can handle template code properly and efficiently, so Qt's non forward thinking objections are beside the point now.
Hermann Minkowski had proposed in 1908 that light momenta is proportional to a material's refractive index then the following year, another German theorist, Max Abraham proposed the opposite -- momentum is inversely proportional to a material's refractive index.
"Given this result, Minkowski has been declared the new winner and light momenta is directly proportional to the material it is travelling through."
And yet Abraham is declared the winner in TFA and Abraham's equation says it's inversely proportional.
It's nice it's using the Omap3430, the successor to the one used in the (king of GLBenchmark) Nokia N95. Not only that, it's a next-gen ARM Cortex (verison 7) architecture that can scale from 600MHz to 1GHz dynamically.
Doesn't hurt that it should have a more advanced 3D GPU from PowerVR that powered the Nokia N95, and the iPhone.
Good thing Palm didn't get stupid and go with a super-sluggish and driver-less Qualcomm SoC, but locking with Sprint seems rather silly. The iPhone sure didn't launch with CDMA to establish itself as a device for world consumption.
While they revel in open-sourcing drivers for Linux, they still take away on the other hand any form of support for their Imageon processors.
Imageon GPUs are implemented in many phone and pda based embedded devices, and supposedly offer hardware 3D acceleration for such devices.
However, AMD/ATI has completely refused to give out any non-NDA documentation to let 3rd-party developers write apps against their ATI Handheld Interface (AHI), or to support the specific ATI extensions in the available OpenGL ES drivers.
Apparently, Oregon implemented this long ago and has been trying to convince the other surrounding States that it's a good idea.
What Oregon continuously doesn't get is road wear is a matter of physics where the heavier the vehicle, the more road damage -- not the miles driven. This means SUVs and the tractor trailers.
Instead of taxing vehicles for becoming more efficient, why not tax vehicles for becoming more massive? You know where this is going, but that's the point.
Taxes have tended to encourage people to do the right thing, and this tax instead digs its own grave.
It's kinda like building hybrid and small vehicles. The better strategy is to take the leap of faith and invest in what you know is right.
Holding back is the SUVs/more-lanes-on-highways strategy that is nothing more than burying your head in the sand, and will never give you a good position for the eventual expected transitions.
And, rail will pay for itself as toll roads currently have, you charge riders for premium schedules, and for increased worker productivity.
It's already been quantified how much productivity is lost every minute spent in traffic on the road instead of in the office doing work. It's not just the exact time lost in traffic, but the missed opportunities, morale drain, and fatigue involved.
For many in the USA, sitting in traffic is the only way of life. Anybody trying to change it would be a crackpot or a terrorist, while none of the Execs care since they got their corporate jets.
How are you going to store the hydrogen and use it efficiently?
Nature has already over millions of years developed the most efficient hydrogen fuel-cell batteries, it's called Petroleum, and it was powered by the solar energy, and distillated by geothermal energies.
We need to stop bailing out Americans. They need to hit bankruptcy (Depression) so that they can shed their ways that requires all these wasteful energy needs and "wants".
Hydrogen is a ENERGY CARRIER (battery), not a Power Source Hydrogen is a ENERGY CARRIER (battery), not a Power Source Hydrogen is a ENERGY CARRIER (battery), not a Power Source
Currently, the best energy carrier we have using hydrogen is Petroleum. The most energy efficient way we can extract hydrogen from water takes more energy in extraction than we recover burning it in fuel-cells. You cannot name solar/wind power because of unreliability and unavailability.
Unless you're fusing hydrogen so you're getting much more energy out of the process to be worth the cost and time, it's just a carrot dream with no real payoff.
The problems of storage, lack of energy density (CtoH bonds hold more energy than HtoH), unstable energy extraction in relation to environmental temperatures, etc will keep fuel-cells a fringe concept, and never a production reality.
The irony of this truth is the SUV lobby deflected this hammering about: o the hybrid premium (myth) o hybrids only last 150K miles (myth) o CNR's hybrids destroy more of the environment over its lifetime than a DHummer (myth) o hybrids don't deserve the ($3000) tax credit/deduction
Slashdot Mirror
ARM is RISC origins so it almost always gets throughput of one instruction per clock. The difference of Cortex vs ARM11 is true SIMD instructions with lower latencies between and better power efficiency vs just pipelining in the ARM11.
All in all, not much difference between them if you run them on a crappy OS like Windows Mobile as it only uses mostly ARMv4 instructions without any activated extension support for NEON, VFP, and such. Android has the same issue actually, and most (except possibly Symbian), do not have optimized GPU drivers, but just random blobs that only pass the conformance tests.
Of course, using any SoC processor that's not the Qualcomm MSM72xx is always a HUGE win in speed. The MSM72xx has SINGLE-HANDEDLY held back the "mobile" platform for the last 3 generations due to its poor ARM11 performance vs its XScale and Samsung cousins, as well as tight-fisted, unreleased documentation for tuning and accelerating due to Qualcomm's money-grubbing licensing policies. Qualcomm touch-screen lag, slow insensitive GPS, and htcclassaction.org are good examples of the infamy Qualcomm single-handedly created.
Notice the mobile marketplace is suddenly perking up now that the MSM72xx platform is now on its way to the graveyard? The OMAP3 is definitely a platform to be on.
As flashy as the Qualcomm Snapdragon is, it's still the one-off platform ride to hell as it was for the MSM72xx .. Just don't ride that bxxxch.
Any lag you experience is an artifact of the Qualcomm cpu. This has only been a problem for even Windows Mobile when HTC switched to the Q chip starting with the Kaiser (AT&T Tilt) phone.
It's been evident ever since, so it's no surprise that Android suffers from it also especially with HTC's scrolling hacks all over it.
The lag goes away once you switch back to a XScale cpu, or anything else that isn't a Qualcomm cpu.
This is just so typical of HTC when the http://www.htcclassaction.org/ was going on ...
They bite off more than they can chew, deny it through canned CS responses, and then try to drag it out until it's sorta fixed, or people loose interest.
It's bettable that they wrote all the Hero code without realizing they'll fall under GPL. It's probably also all statically linked together and they didn't bother separating it properly, which is why they can't "release" any of the code "right now".
It's only fair to get a court order to get the code immediately released before they do any more damage to hide their violations.
Or he could be a genius and trolling so extreme to prevent it from happening. Now if we only had this kind of troll to be against public healthcare ...
Another reason nobody wants to seriously develop for Windows Mobile is because its the off-platform for graphics that has only be highly successful in holding the platform back. Thank you Don Couch.
Most graphics developers doing cross-platform have settled on OpenGL as the lowest-cost and highest portability API to use. But, Microsoft continues to push DirectDraw and Direct3D where 90% of their Window Mobile ODMs don't even bother hardware-accelerating by not shipping drivers like HTC (HTCClassAction.org), or ship inferior 3D in-house hardware with crappy or expensive drivers (Samsung/Qualcomm), and no FPU support in the OS even when the shipped CPU has one.
So, combined with the fact Windows Mobile is basically Windows 98 graphics-wise, WinCE5 IS Win98 (32MB max ram per process for instance), you have an ancient platform that doesn't support any advanced hardware, and cannot do so no matter how much you beat it.
Then MS goes on the anemic .NETCF push that is not only performance-sapping as writing your game apps in HTML on WebOS, but avoids the much needed rearchitecture for a high-performance modern GUI. Even Qt4 is more functional with more capability than .NETCF. I am beginning to believe at least a major portion of the Windows Mobile's software being based from India as a reason for all this lazy, low performance.
WM6.5 is Copeland -- WM7 will be Rhapsody.
These are serious fundamental, architectural problems that shipping a new OS and/or skin is NOT going to solve especially after ramping up development 5 years late.
Balmer, DUMP DirectDraw/Direct3D/.NETCF/WinCE5 and move forward from your current utter fail!
Yup, and now retracting by giving developers more native access, as they should've done from the beginning. Take a hint Palm!
Problem is Blizzard isn't doing that anymore. They don't want their programmers pulling Duke Nukem 3Ds anymore, so instead they just do Agile development and ship it as soon as possible. Then, follow-up with a "horde" of patches.
Sounds like you need to run NoScript and Adblock at least to stabilize your FF experience.
Well, would be same test as bringing a guy into Fry's Electronics/Yodabashi/Bic Camera too.
How about the stock market? As bad as the recent recession has been, it'll still be about ~10% ROI in 10 YEARS.
He could've spent a fraction of that $38K to reduce his power bills to something closer to $100 per month. Better insulation, more energy efficient PSs in his computers, not leaving electronics on 24/7, change out to CF bulbs, and so on -- seriously!
He's going to hit Jevon's Paradox and end up not actually saving anything, IMO.
I'm more in support of community solar (have the HOA do it, or in a park, etc) which is more beneficial and lower cost. Mass solar.
And Visual Basic inbred itself into .NET!! Oh what horrors.
But, maybe the better solution is the Street Fighter II strategy: easy to get into, difficult to master.
Seconded. Babylon5 attacked and discussed the very subjects TFA says has never done before, besides TOS.
The difference is there was an overall story-arc and the guy won a writer's award (Nova?) for it.
TV == short memories.
Would Qt be better written now using C++ templates and the latest template meta-programming instead of the meta-language Qt is currently using?
We're at the point where most main-stream platform compilers can handle template code properly and efficiently, so Qt's non forward thinking objections are beside the point now.
Considering how light-weight WTL is, is it worth porting WTL apps to Qt? (versus going from heavyweight MFC to heavyweight Qt).
Make up your mind!!!
Hermann Minkowski had proposed in 1908 that light momenta is proportional to a material's refractive index then the following year, another German theorist, Max Abraham proposed the opposite -- momentum is inversely proportional to a material's refractive index.
"Given this result, Minkowski has been declared the new winner and light momenta is directly proportional to the material it is travelling through."
And yet Abraham is declared the winner in TFA and Abraham's equation says it's inversely proportional.
It's nice it's using the Omap3430, the successor to the one used in the (king of GLBenchmark) Nokia N95. Not only that, it's a next-gen ARM Cortex (verison 7) architecture that can scale from 600MHz to 1GHz dynamically.
Doesn't hurt that it should have a more advanced 3D GPU from PowerVR that powered the Nokia N95, and the iPhone.
Good thing Palm didn't get stupid and go with a super-sluggish and driver-less Qualcomm SoC, but locking with Sprint seems rather silly. The iPhone sure didn't launch with CDMA to establish itself as a device for world consumption.
While they revel in open-sourcing drivers for Linux, they still take away on the other hand any form of support for their Imageon processors.
Imageon GPUs are implemented in many phone and pda based embedded devices, and supposedly offer hardware 3D acceleration for such devices.
However, AMD/ATI has completely refused to give out any non-NDA documentation to let 3rd-party developers write apps against their ATI Handheld Interface (AHI), or to support the specific ATI extensions in the available OpenGL ES drivers.
3rd-party developers can't even create alternative 3D drivers to get past Qualcomm's current inept implementation
AMD/ATI is not as nice as you think they are.
Apparently, Oregon implemented this long ago and has been trying to convince the other surrounding States that it's a good idea.
What Oregon continuously doesn't get is road wear is a matter of physics where the heavier the vehicle, the more road damage -- not the miles driven. This means SUVs and the tractor trailers.
Instead of taxing vehicles for becoming more efficient, why not tax vehicles for becoming more massive? You know where this is going, but that's the point.
Taxes have tended to encourage people to do the right thing, and this tax instead digs its own grave.
It's kinda like building hybrid and small vehicles. The better strategy is to take the leap of faith and invest in what you know is right.
Holding back is the SUVs/more-lanes-on-highways strategy that is nothing more than burying your head in the sand, and will never give you a good position for the eventual expected transitions.
And, rail will pay for itself as toll roads currently have, you charge riders for premium schedules, and for increased worker productivity.
It's already been quantified how much productivity is lost every minute spent in traffic on the road instead of in the office doing work. It's not just the exact time lost in traffic, but the missed opportunities, morale drain, and fatigue involved.
You're one of the lucky few, educated ones.
For many in the USA, sitting in traffic is the only way of life. Anybody trying to change it would be a crackpot or a terrorist, while none of the Execs care since they got their corporate jets.
How are you going to store the hydrogen and use it efficiently?
Nature has already over millions of years developed the most efficient hydrogen fuel-cell batteries, it's called Petroleum, and it was powered by the solar energy, and distillated by geothermal energies.
We need to stop bailing out Americans. They need to hit bankruptcy (Depression) so that they can shed their ways that requires all these wasteful energy needs and "wants".
Okay, say it with me:
Hydrogen is a ENERGY CARRIER (battery), not a Power Source
Hydrogen is a ENERGY CARRIER (battery), not a Power Source
Hydrogen is a ENERGY CARRIER (battery), not a Power Source
Currently, the best energy carrier we have using hydrogen is Petroleum. The most energy efficient way we can extract hydrogen from water takes more energy in extraction than we recover burning it in fuel-cells. You cannot name solar/wind power because of unreliability and unavailability.
Unless you're fusing hydrogen so you're getting much more energy out of the process to be worth the cost and time, it's just a carrot dream with no real payoff.
The problems of storage, lack of energy density (CtoH bonds hold more energy than HtoH), unstable energy extraction in relation to environmental temperatures, etc will keep fuel-cells a fringe concept, and never a production reality.
The irony of this truth is the SUV lobby deflected this hammering about:
o the hybrid premium (myth)
o hybrids only last 150K miles (myth)
o CNR's hybrids destroy more of the environment over its lifetime than a DHummer (myth)
o hybrids don't deserve the ($3000) tax credit/deduction
Only if you like paying for something that gets you raped either by gate pirates, lottery fraud, or something else more sinister than real life.