Reality Check - Would you really pre-order on a KS?

Would you give $500 to a Pandora 2 Kickstarter with an 18 month timeline?


  • Total voters
    77

@ Grench, suggest you change the poll question to : Would you give $500 to a Pandora 2 Kickstarter from OP GmbH with an 18 month timeline?

If you did my answer would be a big yes.

Sadly Craig has shown that he hasn't got a clue (or maybe just can't be bothered) about how to run a successful kickstarter. The utterly lamentable lack of communication / effort to engage / enthuse the backers for the iCP2 is really rather sad.
 
Well, there are a few things I can reply to.

If I do a Pandora successor preorder using Kickstarter, you can be sure I won't start it before I got a fully working prototype incl. case, etc.

Software might not be polished, but I need to be sure that the hardware doesn't have serious bugs left.

I want to have it ready so I can just say "GO!" when Kickstarter is successfull.

I wouldn't expect many delays from then on, as production would happen at GC (and we know they are reliable) and I'd also do a case production in Europe.

Additionally, GC could order the parts and start production right away when Kickstarter is successful, no need to wait until the money is on my account (they do trust me, you know)

Right now, it's still Pandora 1 time though :)

Let's concentrate selling more and delivering the remaining units, we need something to finance the production as a successor first, right? :)
That sounds like a great plan - but where will the $$$,$$$ come from to finance the prototype & hardware?
Don't forget that we got 350K EUR from investors.

That money was used to finance the Pandora 1, but with the business I made, that money will be back once 3800 Pandora 1 has been delivered (to both old and new orders).

We're at about 2700 right now, so there is money coming back into the pot.
We should consider having some kind of drive to sell a bunch of Pandoras... I'm thinking we could wait until the Craig queue is completed and push like hell on every geek site known to man.
 
Last edited by a moderator:
If craig had absolutely noting to do with the project I would back it. If I got even a hint he was involved then no, never.
 
I would  do it .period .
 
Last edited:
500$ is quite a number. If it was running a very power-efficient X86 like the upcoming tablet generation, if that is possible to cramp in a similar-sized case, yes. Less Emulation, some really awesome games becoming pocketable. Playing aquanox on the go, hell yeah.

If the Pandora 2 is the same but faster it doesn't make sense for me personally. What I want to play on an ARM device is possible already with the existing CC model. Would I play Q3 engine games if they ran perfectly smooth on the new model? Quite possibly and I'd enjoy it. But there would still be so many awesome stuff that I couldn't use on it, all the closed source games, so the step up wouldn't really be big for me. So, no buy, not even for two thirds of the money.

IMO the next Pandora should adress as wide a market as possible. X86 can give you that. Just look at how much $$ people have spent over the last years on half-baked windows UMPCs and were put off by the usability, heat and battery life.. the technology is in reach. All of this is of course only viable if the battery life wouldn't differ much from the ARM options. And being able to run windows XP or 7 doesn't mean it wouldn't be a native Linux device with all the bells and whistles and freedom associated.

/me fits Armor Thermic Hardener II to his pod and warps away at 6AU/s.
 
Last edited by a moderator:
500$ is quite a number. If it was running a very power-efficient X86 like the upcoming tablet generation, if that is possible to cramp in a similar-sized case, yes. Less Emulation, some really awesome games becoming pocketable. Playing aquanox on the go, hell yeah.


If the Pandora 2 is the same but faster it doesn't make sense for me personally. What I want to play on an ARM device is possible already with the existing CC model. Would I play Q3 engine games if they ran perfectly smooth on the new model? Quite possibly and I'd enjoy it. But there would still be so many awesome stuff that I couldn't use on it, all the closed source games, so the step up wouldn't really be big for me. So, no buy, not even for two thirds of the money.


IMO the next Pandora should adress as wide a market as possible. X86 can give you that. Just look at how much $$ people have spent over the last years on half-baked windows UMPCs and were put off by the usability, heat and battery life.. the technology is in reach. All of this is of course only viable if the battery life wouldn't differ much from the ARM options. And being able to run windows XP or 7 doesn't mean it wouldn't be a native Linux device with all the bells and whistles and freedom associated.


/me fits Armor Thermic Hardener II to his pod and warps away at 6AU/s.
B) I am not the only one anymore...
 
OOOOOooo I like the idea of x86. (or even betrer would be x64 if that's possible). And I would think that Windows 8 would make a better OS for a UMPC than 7 or XP. And since there are plenty of x86 Linux distros and even AndroX86 then the thought of having Android, Debian, and Windows all on one device that fits in my pocket becomes a reality that I would absolutely love!!! But sadly the power needed to run x86 chips is too great and it would lead to a horrid battery life.
 
IMO the next Pandora should adress as wide a market as possible. X86 can give you that. Just look at how much $$ people have spent over the last years on half-baked windows UMPCs and were put off by the usability, heat and battery life.. the technology is in reach.
Did you just use overheating and short battery life as an argument in favour of x86 in a handheld device?

:blink:
 
IMO the next Pandora should adress as wide a market as possible. X86 can give you that. Just look at how much $$ people have spent over the last years on half-baked windows UMPCs and were put off by the usability, heat and battery life.. the technology is in reach.
Did you just use overheating and short battery life as an argument in favour of x86 in a handheld device?
:blink:
I said "but sadly"


I realize that an x86 Pandora just win't happen.
 
Yes, I basically did that. Because people bought those half-assed devices for loads of money, just because they ran the software they were used to!

I should have said "The technology to put away with those caveats is in reach" though.
 
How much would it cost to put a 64 bit x86 CPU, a small nuclear reactor and a hydro-cooling system inside of our current case?  Can't be that much...

Okay, scratch that idea...but these things are getting smaller and more powerful all of the time.

Here's one recent example---> http://www.techhive.com/article/2026540/the-gizmo-board-is-a-tiny-amd-powered-x86-microcontroller.html

Less than 10 watts of power draw.  Still a lot, I know.  Maybe in a year or two we'll see better developments - maybe Mr. Weston could work his magic and draw up something amazing...and I kind of like ParadisoShlee's idea of using pixie dust...   :rolleyes:
 
People cannot see their own BIAS but.... thanks to advancements in the mobile space. ARM SOCs are going to get much more powerful with hopefully less power usage and less heat.

blah blah blah. Things designed for small, light and mobile will win out over trying to jam something that is new into the side of a mobile device.
 
^ You're probably right.  And with companies like Samsung and Qualcomm (etc) constantly pushing the envelope in the mobile phone arenas with their faster processors, we as Pandora owners/fans will come out the winners.  I was mostly being a bit cheeky there in my last post...
 
While I think Intel is exaggerating their marketing, their Silvermont x86 SoCs are still likely to be tangibly more efficient and powerful than anything else on the market at the time of release. Especially anything Cortex-A15 based. Intel has a lot of advantages in this space:

1) Having a better manufacturing process than anyone else, one that is now heavily optimized for low power. TSMC's 20nm won't be available until a few months after the first Silvermont devices at the earliest, and even then it won't have FinFET so it'll still be behind in performance.

2) Tons of experience with CPU uarch design that few others in the industry have, and a lot of that talent is now horded at Apple where nothing but Apple devices can benefit

3) Tons of experience with CPU physical design that few others in the industry have, same issue with them vs Apple who bought out Intrinsity

4) Tight integration between CPU design and fab that no one else has - the only other SoC vendor with their own fab are currently fitting standard ARM cores to it.

5) Lots of experience with important uncore stuff like fast memory controllers, others in the industry are getting better at this but there's still a gap

They already did very well with the traditional Atom cores despite barely changing the uarch in forever, and that was never a very good design fundamentally.. This was without a process advantage. Now that they've caught up in process and design they're quite formidable. The disadvantage to using x86 is still there but it's small compared to the advantages they have and the cost gets smaller the more complex the cores get (they still lose somewhat to being restricted to 2-way decoders but this hasn't really held AMD back with Jaguar).

This doesn't apply to GPUs though, there you're stuck with whatever Intel puts in it and I don't think even Baytrail's Gen 7 derived GPU is going to be anything close to class leading. There's also the possibility that Intel still doesn't make its phones and tablet chips 64-bit capable, despite it being part of the uarch.

I think we're going to see some players who make tons of revenue in smartphones start paying a lot more money to try to catch up, especially Samsung who has fabs and is rumored to be doing a custom core. Who knows how things will play out in the next few years. And it's not like Intel is going to gobble up the market just because the have a CPU with the best perf/W, phone sales depend on far more than just that. But the idea that x86 automatically means much worse power efficiency is obsolete, bordering on archaic. And it's not like this is what I want, I have more investment in the ARM ISA than most of you :p But you can't deny reality. The writing was on the wall for this one for a long time.

People cannot see their own BIAS but.... thanks to advancements in the mobile space. ARM SOCs are going to get much more powerful with hopefully less power usage and less heat.
Actually I think it'll be the opposite, phones are pushing the limits of power consumption now and uarchs have already plucked a lot of the efficiency tricks they can, and manufacturing processes are improving more slowly for foundries (ie not Intel). We saw a big increase in CPU and GPU power because phones have gotten larger/bigger batteries and much better power management so it's only used when needed. Once the power wall is hit you start seeing something like what you see in Intel desktop CPUs today, much more incremental improvement generation to generation, but a big perf/W lead over the next best competitor.
 
Last edited by a moderator:
What if all the money being spent on optimising x86 for lower power consumption were spent of ARM designs?
 
What if all the money being spent on optimising x86 for lower power consumption were spent of ARM designs?
But it's not like all of Intel's advantages come from spending money specifically on low power x86. Their manufacturing advantage is amortized by all the money they make on desktops, laptops, and servers.

Right now no one in the industry is as integrated at making ARM CPUs as Intel is at making x86 CPUs. That goes beyond a willingness to spend money. Even if you dropped tens of billions of dollars on fabs they'd still take a few years to catch up and it's hard to out-bid decades of experience; part of that is influenced by corporate culture and not just the ability to poach the best workers.

If Intel were making a low power ARM CPU instead of an x86 one it could be slightly better at best. Only it wouldn't have the advantage of better code generation that x86 still enjoys in some areas. It'd have been great if Intel stuck with ARM cores and influenced improved code quality on ARM, so that we could be using the same code everywhere, but they're committed to pushing x86 wherever they can.
 
Back
Top