100 Hand Painted Special Edition Pandoras

[torpor]

This is yet another typical PR disaster from the master of such things, in my opinion ..

 

[Monk]

I still don't get it, I ordered a 600Mhz Pandora. If I get anything better, its a bonus, but not something I paid for. I feel I have no reason to be upset for OPT deciding to make some money off of better quality boards; boards I didn't pay anything extra for. If I want an 800Mhz board then I will be able to order one from the special edition Pandoras.

Also, I have not seen anything in writing that we will be able to overclock (or overvolt, etc). In fact, I specifically remember craig saying we won't be allowed to overclock and demanding order numbers in the process.

So why do people feel so ripped off? You didn't pay for a board that will run at 800Mhz and part of the warranty states we cannot overclock. The + on the website was put there before this was clarified. Also, are the terms of the warranty set in stone? Isn't OPT allowed to change their warranty at any point anyway, regardless if anything has shipped?

Some people ordered back when OpenPandora were loudly proclaiming that test boards were running at 800MHz or more, that 900MHz ought to be quite possible. Forum members generally ran with what they were being told on here and some even got excited enough to accidentally raise the number to 1GHz IIRC. Wow - think of this, a 1GHz rocket in your pocket!

So while YOUR expectation may have only been for 600MHz, other people who'd been shown 800MHz and told more was possible paid over a year before that overclocking fiasco thread. Even so it's unclear whether Craig was being serious or facetious in that thread, which makes it a fairly dangerous thread to quote in seriousness.

Since forum goers were getting excited about 800MHz plus, despite having ordered 600MHz plus (that plus is important!), some die-hard killjoys (such as, IIRC, WizardStan and myself) have been known to throw in the odd "you can't COUNT on it doing over 600MHz. Yes, they promise 600MHz PLUS but you can't COUNT on 720, or 800 - honest, they don't guarantee it!. That does not stop people treating OpenPandora's OpenBragging about the speed as an incentive to buy. Yes, it goes to AT LEAST 600MHz. You didn't expect the extra speed tha tOpenPandora showed us years ago. That's good for you. Some other people, who have seen the numbers quoted on the boards time and again really do expect to get what they've been told/sold. Reducing their chance of getting it IS like cherry picking the winning lottery tickets and selling only "no win" tickets. Or rather, it would be if OpenPandora ever seriously considered doing that.

People feeling let down result in sad, unhappy posts. People berating them for wanting what they've been told they are buying on the basis that "I don't want/need/expect that" make for sad and angry THREADS. That doesn't FEEL productive, to me, personally.

 

[shadow.8]

I still don't get it, I ordered a 600Mhz Pandora. If I get anything better, its a bonus, but not something I paid for. I feel I have no reason to be upset for OPT deciding to make some money off of better quality boards; boards I didn't pay anything extra for. If I want an 800Mhz board then I will be able to order one from the special edition Pandoras.

Also, I have not seen anything in writing that we will be able to overclock (or overvolt, etc). In fact, I specifically remember craig saying we won't be allowed to overclock and demanding order numbers in the process.

So why do people feel so ripped off? You didn't pay for a board that will run at 800Mhz and part of the warranty states we cannot overclock. The + on the website was put there before this was clarified. Also, are the terms of the warranty set in stone? Isn't OPT allowed to change their warranty at any point anyway, regardless if anything has shipped?

Some people ordered back when OpenPandora were loudly proclaiming that test boards were running at 800MHz or more, that 900MHz ought to be quite possible. Forum members generally ran with what they were being told on here and some even got excited enough to accidentally raise the number to 1GHz IIRC. Wow - think of this, a 1GHz rocket in your pocket!

So while YOUR expectation may have only been for 600MHz, other people who'd been shown 800MHz and told more was possible paid over a year before that overclocking fiasco thread. Even so it's unclear whether Craig was being serious or facetious in that thread, which makes it a fairly dangerous thread to quote in seriousness.

Since forum goers were getting excited about 800MHz plus, despite having ordered 600MHz plus (that plus is important!), some die-hard killjoys (such as, IIRC, WizardStan and myself) have been known to throw in the odd "you can't COUNT on it doing over 600MHz. Yes, they promise 600MHz PLUS but you can't COUNT on 720, or 800 - honest, they don't guarantee it!. That does not stop people treating OpenPandora's OpenBragging about the speed as an incentive to buy. Yes, it goes to AT LEAST 600MHz. You didn't expect the extra speed tha tOpenPandora showed us years ago. That's good for you. Some other people, who have seen the numbers quoted on the boards time and again really do expect to get what they've been told/sold. Reducing their chance of getting it IS like cherry picking the winning lottery tickets and selling only "no win" tickets. Or rather, it would be if OpenPandora ever seriously considered doing that.

People feeling let down result in sad, unhappy posts. People berating them for wanting what they've been told they are buying on the basis that "I don't want/need/expect that" make for sad and angry THREADS. That doesn't FEEL productive, to me, personally.

I want to clarify I wasn't trying to be offensive or insulting. Rereading my post I can see I probably could have wording things better so I apologize if I came off as attacking someone for their point of view.

I ordered in October 2008 and have been following the project (albeit casually sometimes) the whole time. I recall Craig saying it could easily handle Doom 3 with a proper port and I remember hearing about the various possibilities of overclocking. To be perfectly honest I was always skeptical and didn't really expect that to be a real-world result. I was seriously asking why people were so upset and hadn't considered this point of view so I do appreciate your reply. Thanks

So how are they deciding which boards to use?

EDIT: fixed to the correct year, 2008 not 2009

 

[Usoppe]

As I ve understood, Craig wants a special edition Pandora because
a ) its cool if you see in 10 years a special edition Pandora selling on Ebay.
b ) big companies offer special editions too.
c ) get some money in for build together in UK
All the points are cool and ok.

My opinion is that instead cherry picking boards which can go to 800 MHz, offer special editions with airbrush which customers can influence by own thoughts.
For example, I would say to your airbrusher friend who will do this job that I want a Final Fantasy VII style Pandora. He paints it on the Pandora and I am happy to have a custom design. Thats very special!

Cherry picking some boards, too, for the special edition could be counter productive to the goals of the Pandora and its sells, because
a ) .. we "just" have an 600MHz OMAP build in every Pandora. Devolpers must live with it. If the program doesnt perform well they need to optimize it - if theres no way, offer frameskipping, etc. - and shouldnt bet on overclocking capabilities. Fact is that we do know nothing about overclocking capabilities of the OMAPs and the second batch could be less overclockable than this batch. So, it would be counterproductive to show 100 units of 800MHz Pandoras on the market and let people hope that all Pandoras can come near to that "speed". Even developers could be influenced by "wrong" thoughts like, "Its ok, I just need to overclock it a bit and my software will run well".
b ) ... I feel a bit "I want a 800MHz OMAP Pandora. I need it" - like. Thats not good. Marketing Pandoras like that could be counterproductive. 600 MHz is that what we all get. If youre lucky its a bit overclockable. But as a) said, there shouldnt be a need to overclock it.
c ) ... even big companies doesnt offer consoles in special edition with higher clocks. They "just" let it paint, put some stuff with it like more HDD capacity (focus on peripheral equipment), whatever.

So in my eyes a solution could be:
- Offer a Pandora which will be painted by an airbrusher in a customer wished design. So customers have the opportunity to influence the airbrush paint by own thoughts - and not to choose between just 10 ready-to-go airbrush images for example. Let customers influence the design like, "I want a Final Fantasy VII/Little Big Adventure style with Cloud/Twinsen on it", whatever. The airbrusher paints it on the Pandora and voila: Customers get in that way a REAL special Pandora because it just exists one time in that design - and you kick all big companies in the .. because they dont offer something like that!
- Put some extra stuff in it: For example 2x SanDisk Extreme III 8GB/16GB SDHC cards.
- Some special designed carrybags.
- And a kiss on the display with rouge that you cant wash out !

A special edition Pandora! And I bet that it I will be sold out very soon when offered although its standard clocked.

 

[Monk]

I want to clarify I wasn't trying to be offensive or insulting. Rereading my post I can see I probably could have wording things better so I apologize if I came off as attacking someone for their point of view.

I ordered in October 2008 and have been following the project (albeit casually sometimes) the whole time. I recall Craig saying it could easily handle Doom 3 with a proper port and I remember hearing about the various possibilities of overclocking. To be perfectly honest I was always skeptical and didn't really expect that to be a real-world result. I was seriously asking why people were so upset and hadn't considered this point of view so I do appreciate your reply. Thanks

So how are they deciding which boards to use?

EDIT: fixed to the correct year, 2008 not 2009

Huge thanks for an adult reply (they aren't impossible to find on here just... appreciated). Apolopies right back at you if I came across as an ass. It's too easy to see one side or the other seemingly repeated and rush out something ill phrased


Some people believe everything that OpenPandora say about the Pandora, which is an exercise in making room for contradictions. I prefer the multiple-universe explanation myself... but yes, even now you'll see people suggesting that most if not all Pandoras will go to 800MHz mostly via a combination of faith and hope. Powerful allies, faith and hope. That so many will run to or beyond 800MHz is even used as one of the arguments to not get upset about the possiblity of cherry-picking the Special Edition units.

How will they be picked? Sadly, we don't know yet - and may never do. OpenPandora aren't being very open about that, as yet. This is probably rattling those who have lost some of that "faith and hope" I mentioned earlier

 

[Pleng]

I recall Craig saying it could easily handle Doom 3 with a proper port

To be fair to Craig I don't think he ever said it could easily handle Doom 3. I think he said Doom 3 was probably about the limit of the hardware.

 

[juniorm33]

Also the earlier revision of the boards seemed to clock higher than the last for whatever reason. I believe it may have been Pickle that made that observation but I could be wrong. And even that observation could've been incorrect....

But if that is the case, the sort of overclocks that we were originally seeing is probably the exception, where they used to be the norm.

That's one of the reason's I've been so curious as to what could've caused the difference, although it's academic.

 

[Monk]

To be fair to Craig

Oi! Don't go changing the rules of the game NOW...

Also the earlier revision of the boards seemed to clock higher than the last for whatever reason. I believe it may have been Pickle that made that observation but I could be wrong. And even that observation could've been incorrect....

But if that is the case, the sort of overclocks that we were originally seeing is probably the exception, where they used to be the norm.

That's one of the reason's I've been so curious as to what could've caused the difference, although it's academic.

Where would we be without academia though? I've mused that either the PCB design got more cmplex/fussy/prone to issues which could foul overclocking as the revisions went on (remember when each new revision of board seemed to be a different colour? THOSE were the days!) OR that the overclocking has reduced for the same reason that the Wifi needs a patch resistor adding to the board - the difference between expensive prototype boards and cheap mass-production boards

 

[juniorm33]

Where would we be without academia though? I've mused that either the PCB design got more cmplex/fussy/prone to issues which could foul overclocking as the revisions went on (remember when each new revision of board seemed to be a different colour? THOSE were the days!) OR that the overclocking has reduced for the same reason that the Wifi needs a patch resistor adding to the board - the difference between expensive prototype boards and cheap mass-production boards

The WIFI/taces was my thought also, but someone told me it wouldn't make a difference.
Although I have to admit I wasn't convinced it couldn't be the reason, but they probably know more about that stuff than me so I except it.

I just find it interesting, not in a OPT did something bad/wrong way though.
Wonder if it progressively got worse or, just the last board revision.
Which colour was the best :lol:

 

[Monk]

There's an EU Salt Mountain created purely from the "pinch of salt" I take with everything I'm told on the net... like, I've been told "this will not cause a delay" so often that it make my whole face twitch now when someone says that.

I'd be interested in how and why later boards have a lower maximum speed than older ones too - whether it's a design issue or a fabrication issue etc. For my part I think the red ones probably went the fastest, just because they were red. Like a good sportscar Pigmentation has a definite effect on speed...

 

[Wootson]

I recall Craig saying it could easily handle Doom 3 with a proper port

To be fair to Craig I don't think he ever said it could easily handle Doom 3. I think he said Doom 3 was probably about the limit of the hardware.

It have seen Doom 3 running on a Pentium 3 (don't know how many Mhz precisely, but it was one of the best P3 CPUs ever made) with a 3DFX Voodoo 3 (3000 series if I remember correctly). I can't say that it will be EASY for the Pandora to run it, but it is certainly possible. If some very skilled programmers are put onto this project, it can run at 30FPS for sure. But not on the highest graphical settings though.

And most likely no programmer will ever use the hardware video capabilities of the Pandora to the max. Just like that never happened with the GPH handhelds and the Dingoo.

 

[MWeston]

I'd be interested in how and why later boards have a lower maximum speed than older ones too - whether it's a design issue or a fabrication issue etc. For my part I think the red ones probably went the fastest, just because they were red. Like a good sportscar Pigmentation has a definite effect on speed...

Actually, it's the other way around. Rev1's were the slowest technically. It had nothing to do with the PCB, but the newer batch of chips that went on rev2 and beyond.

Rev1 did 625MHz at OPP3 before losing stability (well crashed at 650 for sure). They did 900MHz at OPP5. Rev2 and later do on average 750MHz at OPP3 but we've never tested OPP5 because my old kernel panics at the UART init due to some changes somewhere and we have never had the voltage scaling in the kernels since. I'm not going to speculate on what OPP5 will do on the newer chips but my fingers are *unofficially* crossed.

 

[Monk]

I'd be interested in how and why later boards have a lower maximum speed than older ones too - whether it's a design issue or a fabrication issue etc. For my part I think the red ones probably went the fastest, just because they were red. Like a good sportscar Pigmentation has a definite effect on speed...

Actually, it's the other way around. Rev1's were the slowest technically. It had nothing to do with the PCB, but the newer batch of chips that went on rev2 and beyond.

Rev1 did 625MHz at OPP3 before losing stability (well crashed at 650 for sure). They did 900MHz at OPP5. Rev2 and later do on average 750MHz at OPP3 but we've never tested OPP5 because my old kernel panics at the UART init due to some changes somewhere and we have never had the voltage scaling in the kernels since. I'm not going to speculate on what OPP5 will do on the newer chips but my fingers are *unofficially* crossed.

Oooh, intersting - some actual data. Thanks MWeston.

So you're saying... vague half-recollected memories of spculation made on The Internet are, on this occasion, not entirely accurate? I'm shocked.


However, the idea that the Pandoras have been getting faster rather than slower is nice. I haven't looked at the boards in ages - can you tell me what "chips" refers to in that paragraph? The SOC, I'm hoping?

MANY thanks for that info - it's the kind of thing that this thread has wanted for a while, some actual data rather than just guesswork

 

[juniorm33]

Actually, it's the other way around. Rev1's were the slowest technically. It had nothing to do with the PCB, but the newer batch of chips that went on rev2 and beyond.

Rev1 did 625MHz at OPP3 before losing stability (well crashed at 650 for sure). They did 900MHz at OPP5. Rev2 and later do on average 750MHz at OPP3 but we've never tested OPP5 because my old kernel panics at the UART init due to some changes somewhere and we have never had the voltage scaling in the kernels since. I'm not going to speculate on what OPP5 will do on the newer chips but my fingers are *unofficially* crossed.

Wow thanks for the input. This has answered a few questions that I had. (And where I started to convince myself otherwise)
Voltage scaling isn't currently working in this kernal.
All recent speed reports are based on OPP3. (e.g. 600Mhz is at OPP3)
When we had those (really) high figures before they where running at OPP5 and therefore not really equivalent to what we get now.

Once again thanks for the info, I think this has helped clarify a lot. :lol:

 

[Patrick R Ludvigsen]

http://focus.ti.com/lit/ds/symlink/omap3530.pdf

There's that one document that talks about this, on page 123. I'm not quite sure what the implications of it are, it might be that switching between voltage levels also causes considerable degradation... I don't know much about this at all, just remembered seeing it on the forums before

 

[Drak]

@Drak: I still had you on ignore from years ago when you were annoying.

 

[may88]

I'm not going to speculate on what OPP5 will do on the newer chips but my fingers are *unofficially* crossed.

Stop typing and get on that plane.


edit: @Na-Noo - OPP5 is overclocking. Not for the likes of us. What's your Order No.?

 

[WizardStan]

edit: @Na-Noo - OPP5 is overclocking. Not for the likes of us. What's your Order No.?

OPP5 is officially supported by TI and in fact "required" (according to the spec sheet) for 600Mhz.

 

[shadow.8]

Huge thanks for an adult reply (they aren't impossible to find on here just... appreciated). Apolopies right back at you if I came across as an ass. It's too easy to see one side or the other seemingly repeated and rush out something ill phrased


Some people believe everything that OpenPandora say about the Pandora, which is an exercise in making room for contradictions. I prefer the multiple-universe explanation myself... but yes, even now you'll see people suggesting that most if not all Pandoras will go to 800MHz mostly via a combination of faith and hope. Powerful allies, faith and hope. That so many will run to or beyond 800MHz is even used as one of the arguments to not get upset about the possiblity of cherry-picking the Special Edition units.

How will they be picked? Sadly, we don't know yet - and may never do. OpenPandora aren't being very open about that, as yet. This is probably rattling those who have lost some of that "faith and hope" I mentioned earlier

No offense taken

It have seen Doom 3 running on a Pentium 3 (don't know how many Mhz precisely, but it was one of the best P3 CPUs ever made) with a 3DFX Voodoo 3 (3000 series if I remember correctly). I can't say that it will be EASY for the Pandora to run it, but it is certainly possible. If some very skilled programmers are put onto this project, it can run at 30FPS for sure. But not on the highest graphical settings though.

And most likely no programmer will ever use the hardware video capabilities of the Pandora to the max. Just like that never happened with the GPH handhelds and the Dingoo.

Is this what you mean? http://www.3dfxzone.it/enboard/topic.asp?TOPIC_ID=1462

I'm still fairly skeptical about that. Comparing an x86 proc with an ARM proc is the whole apples and bananas thing. And its not just the processor, the RAM is different, the speed of ram is different, the graphics chip is different, etc. I would love to be proved wrong but there just seem to be too many variables for me to be believe it is very possible.

Actually, it's the other way around. Rev1's were the slowest technically. It had nothing to do with the PCB, but the newer batch of chips that went on rev2 and beyond.

Rev1 did 625MHz at OPP3 before losing stability (well crashed at 650 for sure). They did 900MHz at OPP5. Rev2 and later do on average 750MHz at OPP3 but we've never tested OPP5 because my old kernel panics at the UART init due to some changes somewhere and we have never had the voltage scaling in the kernels since. I'm not going to speculate on what OPP5 will do on the newer chips but my fingers are *unofficially* crossed.

Thanks for the post, its great to hear some real data

 

[Grench]

Actually, it's the other way around. Rev1's were the slowest technically. It had nothing to do with the PCB, but the newer batch of chips that went on rev2 and beyond.

Rev1 did 625MHz at OPP3 before losing stability (well crashed at 650 for sure). They did 900MHz at OPP5. Rev2 and later do on average 750MHz at OPP3 but we've never tested OPP5 because my old kernel panics at the UART init due to some changes somewhere and we have never had the voltage scaling in the kernels since. I'm not going to speculate on what OPP5 will do on the newer chips but my fingers are *unofficially* crossed.

Thank you for the relative numbers. It helps to put perspective around things.

If the average Rev2 and later Pandora consoles can do 750MHz...

Making some assumptions, 750 is likely the mode as well. In consumer electronics devices things are usually more similar than different, so likely half or more of the consoles would perform at or very close to this point?

So, the distribution might be something like:
Proportion Max MHz
10% 600
15% 720
50% 750
20% 800
5% >800

The distribution is certainly -NOT- exactly like that above. The idea of this is to illustrate that performance curves have tails - and it sounds like the 800MHz+ units are going to be in the minority.

What I'm trying to outline though, is from MWeston's comments above that 750 is the 'average' on OPP3 that units that can do 800 on OPP3 are likely a small slice of the overall set.

Given the noises from customers above, it might be a good idea to do the 2nd batch as a speed-binned and differentially priced set. Classic market segmentation. The objective of any good market segmentation being convincing the customers who can spend more on your product do so.

It would look something like:
600MHz = 15% discount
720MHz = 5% discount
750MHz = 0% discount
800MHz = 15% upcharge
>800MHz = 30% upcharge

Of course this would add a level of complexity to the order system that I'm not sure you would want to support.

The above is nothing but my humble opinion, speculation and random thoughts. Any semblance to worthwhile ideas or notions should be considered over a good nights sleep.