[PrinceAlbert]

Don't think it's marketing related as the UK page is still present.

[mich.bushi]

Gosh, maybe it is just STILL yet present?

Seriously, expecting from them agility and speed to market required to make it real to implement such last minute changes (given european launch coming close) is a little bit silly, especially with their track record of slowest mobile company in our galaxy and it's neighbourhood, don't you all think?

Stop dreaming, Pre 3 with high-end specs just ain't going to happen. It is still solid & very capable device, if you are not living in a benchmark world, I am sure 2.x will be simply flying on that thing. Which from end user's perspective is the only thing that matters, really (well apart from apps availability...)

Regards

[Targon]

Gosh, maybe it is just STILL yet present?

Seriously, expecting from them agility and speed to market required to make it real to implement such last minute changes (given european launch coming close) is a little bit silly, especially with their track record of slowest mobile company in our galaxy and it's neighbourhood, don't you all think?

Stop dreaming, Pre 3 with high-end specs just ain't going to happen. It is still solid & very capable device, if you are not living in a benchmark world, I am sure 2.x will be simply flying on that thing. Which from end user's perspective is the only thing that matters, really (well apart from apps availability...)

Regards

If the dual-core version of the processor has the same pin-outs as the single-core, then HP COULD have slipped the dual-core processor into the manufacturing process as easily as using a different source for capacitors would be for a motherboard company. A lot of the complexity in making a "faster version" would be if new comm boards or other additions were added. Basically, we have seen companies like AMD in the PC world make a CPU socket that could go from single to dual-core, and from dual-core to quad-core, and if it can be done with a socket, then surface mounting could work as well. I am not saying it WILL happen, but without any information to the contrary, we can still dream, can't we?

A bit of self delusion can go a long way to dealing with eternal delays.

[smooth3006]

Clearwire would have been bankrupt without a Sprint bailout, and the WiMax network rollout has been much slower than Sprint expected. There's no conspiracy as to why Sprint isn't interested in putting all their eggs into Clearwire anymore.

in case you haven't heard sprint will be going lte in the near future.

[mich.bushi]

If the dual-core version of the processor has the same pin-outs as the single-core, then HP COULD have slipped the dual-core processor into the manufacturing process as easily as using a different source for capacitors would be for a motherboard company.

true, technically they probably COULD, but for what good? Unlike capacitors, software would need to be seriously rewritten (most probably), to benefit at all from multicore architecture. And in any case, it would need to be thoroughly retested. Multithread/multiprocess programming is tricky, and a nightmare to test and debug. It adds order of magnitude in program complexity. It is not true, that it scales linearly, even for the best-scaling algorithms (so statements that 4 * 0.5GHz cores == 2 * 1GHz cores == equal power as 2GHz single core are completely false, and misunderstanding. You cannot do any math regarding clock speeds and number of cores, it doesn't work that way. For most tasks, multicores from the above "equation" will be (practically) less powerful than single-core one. For one reason - yes, their powers sum up to the "same" theoretical throughput, but splitting singular tasks (watching the video, downloading a file, etc.) into multi-thread ones, costs processor cycles. Combining these multiple results into one output (video frame, downloaded file), synchronizing them - costs processor cycles again. And sometimes, some cores finish faster than the others, and cannot process untill the last one finishes. And for some algorithms (that do not scale out), they will not benefit from multiple cores at all, and will be so much slower, as clock differences above shows - when task cannot be split into multiple parallel threads, so it will execute on one core only, other core(s) idling).

The only reason, why we are experiencing the multicore processing boom right now is that we cannot up the processor speed significantly, anymore - we have hit the ceiling of physics, at around 4-5GHz, and it is not practical/feasible, to try & increase processing power by upping the processor clock anymore. But they can increase output, while non-linear, by adding another core(s).

So, I frankly appreciate their decision to put faster single-core chip in Pre 3, to get it out the door quickly, and then worry about multi-cores, after they patch a gaping hole in their (non)available middlerange phone.

That's being said... Well, the keyword here is "to get it out the door quickly", which unfortunately is not going to happen... For that, damn you, HPalm!

Regards

[LCGuy]

mich.bushi-

If the Pre 3 launches with WebOS 3 on it, the OS WILL be dual core optimized (the TP, remember?).

[EvilKell]

in case you haven't heard sprint will be going lte in the near future.

My post was rationalizing why Sprint was abandoning Clearwire/WiMax...

[EvilKell]

true, technically they probably COULD, but for what good? Unlike capacitors, software would need to be seriously rewritten (most probably), to benefit at all from multicore architecture. And in any case, it would need to be thoroughly retested...

The Pre3 might be launching with webOS 2.2, but it's expected to get webOS 3.0 in the fall (along with the Veer and according to the product page, even the Pre 2). webOS 3.0 is obviously optimized for multi-core since the TouchPad already has multi-core technology.

Software re-writing isn't a big deal in this instance since webOS 2.x is being abandoned shortly after the Pre3's launch.

[mich.bushi]

Hmmm, these are very strong points... So,assuming they'll launch it with 3.0 onboard, they might as well go full in, and with upped specs...? I don't know, big assumptions, but seen in this light, maybe...

Still, that would be extremely awkward from marketing and pragmatic point of view (well now they would have introduced TWO absolutely substantial changes, instead of one), but technically, it makes a lot of sense (less fragmentation and coherent dev platform across the board is certainly a good thing...), and putting 2.x to rest sooner would allow them to focus entirely on 3.x

good points -lets wait n see, you are contagious ;-)

[dgdonovan]

HP must know the original specs announced for the Pre 3 are not competitive with either the latest high-end Android phones or the iPhone 5. We'll see what ships in Europe, but it needs a spec bump well before the iPhone 5 ships. Otherwise, HP doesn't have a complete family of handsets, just low and mid-range.

The webOS engineering team was focused on release of TouchPad. Other than the near-term OTA to address bugs, performance and incomplete features (3.0.?), they've probably pivoted resources to work on a consolidated version of webOS 3(.1.0) that will support TouchPad, Opal and the handsets.

[ToniCipriani]

HP must know the original specs announced for the Pre 3 are not competitive with either the latest high-end Android phones or the iPhone 5. We'll see what ships in Europe, but it needs a spec bump well before the iPhone 5 ships. Otherwise, HP doesn't have a complete family of handsets, just low and mid-range.

The webOS engineering team was focused on release of TouchPad. Other than the near-term OTA to address bugs, performance and incomplete features (3.0.?), they've probably pivoted resources to work on a consolidated version of webOS 3(.1.0) that will support TouchPad, Opal and the handsets.

I actually much prefer if they lower the price of the Pre 3 from high-end and just create another high-end phone. Many don't consider anything that's not a slab anywhere high-end.

And all the better for me, I get a cheaper Pre 3.

[milominderbinde]

I actually much prefer if they lower the price of the Pre 3 from high-end and just create another high-end phone. Many don't consider anything that's not a slab anywhere high-end.

And all the better for me, I get a cheaper Pre 3.

A family of products would be great. Put the Pre 3 at $0 and the Pre 4 at $199.

[Targon]

true, technically they probably COULD, but for what good? Unlike capacitors, software would need to be seriously rewritten (most probably), to benefit at all from multicore architecture. And in any case, it would need to be thoroughly retested. Multithread/multiprocess programming is tricky, and a nightmare to test and debug. It adds order of magnitude in program complexity. It is not true, that it scales linearly, even for the best-scaling algorithms (so statements that 4 * 0.5GHz cores == 2 * 1GHz cores == equal power as 2GHz single core are completely false, and misunderstanding. You cannot do any math regarding clock speeds and number of cores, it doesn't work that way. For most tasks, multicores from the above "equation" will be (practically) less powerful than single-core one. For one reason - yes, their powers sum up to the "same" theoretical throughput, but splitting singular tasks (watching the video, downloading a file, etc.) into multi-thread ones, costs processor cycles. Combining these multiple results into one output (video frame, downloaded file), synchronizing them - costs processor cycles again. And sometimes, some cores finish faster than the others, and cannot process untill the last one finishes. And for some algorithms (that do not scale out), they will not benefit from multiple cores at all, and will be so much slower, as clock differences above shows - when task cannot be split into multiple parallel threads, so it will execute on one core only, other core(s) idling).

The only reason, why we are experiencing the multicore processing boom right now is that we cannot up the processor speed significantly, anymore - we have hit the ceiling of physics, at around 4-5GHz, and it is not practical/feasible, to try & increase processing power by upping the processor clock anymore. But they can increase output, while non-linear, by adding another core(s).

So, I frankly appreciate their decision to put faster single-core chip in Pre 3, to get it out the door quickly, and then worry about multi-cores, after they patch a gaping hole in their (non)available middlerange phone.

That's being said... Well, the keyword here is "to get it out the door quickly", which unfortunately is not going to happen... For that, damn you, HPalm!

Regards

The OS is already capable of splitting applications up between the different cores. Remember, WebOS is built on top of a Linux kernel, and Linux already has support for multi-threading and multi-core processors. Individual apps may not be able to use more than one core at a time, but different apps CAN go to one processor core or the other even with WebOS 1.x.x.

This is really the thing that allows for multiple Android devices to be released, that the main OS can already handle it and it is really just about making a new phone with a different hardware configuration. Android is also built on top of a Linux kernel, so really, it's only HP that is holding back the release of 15 different devices at different speeds with different configurations of hardware.

[C-Note]

A family of products would be great. Put the Pre 3 at $0 and the Pre 4 at $199.

It's the future anyway. If Android starts feeling the heat from anyone (Apple, HP or Windows 7) I'd bet the farm that they will have sell their high end phones for about $200 and most everything else will be from 0$ to 50$ with lots of advertising and a exorbitant ETF.

C

[mich.bushi]

The OS is already capable of splitting applications up between the different cores. Remember, WebOS is built on top of a Linux kernel, and Linux already has support for multi-threading and multi-core processors.

That's an assumption. As I said before, KERNEL supporting multithreading is a requirement, but not a guarantee that webOS supports multithreading all the same. Remember, webOS apps do not run in the Kernel's environment, but the Browser's, sitting on top of that multi-threading kernel. So it is down to question, if that browser-like environment, sitting on top of kernel, does support multithreading all the same.
Which I don't know the answer to, but I strongly doubt is the case; certainly, it is not guaranteed "out of the box" for any application to be run in multithreaded way, it needs to be specifically implemented for. And I doubt they did it before, as up until TP, they only had single core HW, so that would make no sense to implement until they actually have multicore HW. So, stating that "it sure is multicore-enabled, because kernel supports multithreading", is a false statement.

Individual apps may not be able to use more than one core at a time, but different apps CAN go to one processor core or the other even with WebOS 1.x.x.

I guess you have checked it on your multicore HW running 1.x.x, right? . Not necessarily true, see above explanation. And individual apps will certainly not be able to use more than one core at a time, because (to my knowledge, someone please correct me), there is no multithreading API as such in webOS? Going back to system-level, that browser-like environment that sits on top of kernel, would need to be specifically designed for it, to delegate different "apps" (lets do not forget, these are CSS/javascript snippets, and NOT native code apps, having access to/running in kernel's environment directly) to different cores. Certainly doable, but far from "guarateed by multithreading-supporting kernel", and far from being trivial thing to implement properly/efficiently.

I think that the bottom line here is, as some earlier posters suggested; if Pre 3 ships with 3.x, there is a good chance they might have upped it to dual-core HW and make most of it, but if it ships with 2.x, there will be no gain in having multicore HW out of the box - but then certainly, it would future-proof it for the 3.x, shipping "in the coming months", huh huh, if we can believe in anything that HP is saying.

And it is not to say, that 3.x would run poorly on 1.4GHz single-core Snapdragon, it really shouldn't be slow by any means... If it was - boy, HPalm, get back to your drawing boards...

Regards,

[Targon]

That's an assumption. As I said before, KERNEL supporting multithreading is a requirement, but not a guarantee that webOS supports multithreading all the same. Remember, webOS apps do not run in the Kernel's environment, but the Browser's, sitting on top of that multi-threading kernel. So it is down to question, if that browser-like environment, sitting on top of kernel, does support multithreading all the same.
Which I don't know the answer to, but I strongly doubt is the case; certainly, it is not guaranteed "out of the box" for any application to be run in multithreaded way, it needs to be specifically implemented for. And I doubt they did it before, as up until TP, they only had single core HW, so that would make no sense to implement until they actually have multicore HW. So, stating that "it sure is multicore-enabled, because kernel supports multithreading", is a false statement.


I guess you have checked it on your multicore HW running 1.x.x, right? . Not necessarily true, see above explanation. And individual apps will certainly not be able to use more than one core at a time, because (to my knowledge, someone please correct me), there is no multithreading API as such in webOS? Going back to system-level, that browser-like environment that sits on top of kernel, would need to be specifically designed for it, to delegate different "apps" (lets do not forget, these are CSS/javascript snippets, and NOT native code apps, having access to/running in kernel's environment directly) to different cores. Certainly doable, but far from "guarateed by multithreading-supporting kernel", and far from being trivial thing to implement properly/efficiently.

I think that the bottom line here is, as some earlier posters suggested; if Pre 3 ships with 3.x, there is a good chance they might have upped it to dual-core HW and make most of it, but if it ships with 2.x, there will be no gain in having multicore HW out of the box - but then certainly, it would future-proof it for the 3.x, shipping "in the coming months", huh huh, if we can believe in anything that HP is saying.

And it is not to say, that 3.x would run poorly on 1.4GHz single-core Snapdragon, it really shouldn't be slow by any means... If it was - boy, HPalm, get back to your drawing boards...

Regards,

I'd verify it in a minute if I was still running WebOS 1.4.5, but I upgraded to 2.1 months ago(has it really been that long?!?) and all I can do is check it on WebOS 2.1, which I will do later if one of the WebOS Internals team members doesn't chime in. I've said that I didn't think that individual applications could use more than one core, but it's still a huge advantage to split the work load of different apps between available cores, so that's what I would expect to be happening under ALL versions of WebOS.

As far as slowness, it does seem like there is something in my WebOS 2.1 setup that causes my phone to pause whenever a SMS/MMS message is sent/received on my Pre Plus overclocked to 1.005GHz, I'm not sure why, but it is slightly annoying. This is why it annoys me that AT&T never signed off on WebOS 2.1, because I don't know how much of this is caused by the kludge needed to get the OS to work with AT&T, and how much is just a WebOS 2.1 issue.

[JayhawkOne]

As far as slowness, it does seem like there is something in my WebOS 2.1 setup that causes my phone to pause whenever a SMS/MMS message is sent/received on my Pre Plus overclocked to 1.005GHz, I'm not sure why, but it is slightly annoying. This is why it annoys me that AT&T never signed off on WebOS 2.1, because I don't know how much of this is caused by the kludge needed to get the OS to work with AT&T, and how much is just a WebOS 2.1 issue.

My overclocked Pre- will freeze up every time it receives a message (even the music app will skip sometimes) and is super slow while sending a message. Unless your problem is worse than that it may not be the 2.1 causing the problem.

/offtopic

[Targon]

My overclocked Pre- will freeze up every time it receives a message (even the music app will skip sometimes) and is super slow while sending a message. Unless your problem is worse than that it may not be the 2.1 causing the problem.

/offtopic

Well, F105 is stable except for that one problem, so I can live. AV8B has been pretty unstable for me using the stock voltages, but I don't feel like tinkering with the voltages to see what will be stable on my phone, which is why I switched.