How About The 720Mhz Omap For The Second Batch?

[paddy]

I hope OP don't do some mad thing like actuly upgrade the Pandora cpu or other main functions
within months of the first batch ,this would be the biggest kick in the teeth all us first batchers
could ever receive ,never mind the two year wait.

I don't know where all this Pandora 2 stuff comes from but if OP were to do a new Pandora
within the first 2 years of the Pandora's life span then i would personaly kick my Pandora
out the window followed by a youtube video of the death.

hehe - joke btw ,everyone is out to make money at the end of the day
so good luck to them but hurry and get my Pandora to my door.

 

[nihilon]

A better way would to make Batch 2 cheaper. Imagine having paid $330 and wait for one or two years, and then see people buying the Pandora a few months later for a lower price!

Uh, I seriously hope you realize this will happen over time right? The Pandora will in all likelihood be cheaper in 1 year than today. It happens with pretty all tech equipment out there. Over time the components will get cheaper as newer tech is developed pushing down the price of parts used in the Pandora.

So the price will most certainly drop over time.... it is the price for buying bleeding edge electronics. Folks that wait 6-12 months almost always get the same equipment for less. Of course that won't stop folks at this forum from bitching and complaining when that happens. Sigh.

Not true. Craig has already confirmed that the second batch will be MORE expensive due to exchange rates and the loss that they are taking on the first batch. (Of course, assembly was supposed to start a week ago, so this can easily change)

I clearly said the price will be less in a year and I stand by that statement.

As well, the price variation you are talking about is about currency fluctuation not material costs. To remove currency fluctuation from a product you either hedge the currency OR you sell the product in the currency that comprises most of the cost of the product. I would guess that the US$ is the currency used to purchase most of the components in the Pandora. Given that assumption, the price of the Pandora in US$ will decrease over time.

I even suspect that the price of the 2nd batch in US$ will be less than the cost of the first batch in US$.

If the subsequent batches are more than $330 then it'll be a no-go for me unless there are alot more apps and development going on then to justify the cost. It's been a bumpy two years I know, but the tech is already aged and anything a notch more than $330 is a premium price to me. I'll let those with money out of their ears adopt 1st and see what's available then or even if a netbook might work as well for me.

How much more expensive are they talking about anyways?

 

[mali]

^ ED's price will most likely stay at 299 Euros.

 

[Sphinxter]

Why should not every generation improve on the previous. Early adoption is always at a price, second batchers deserve a faster chip and some more memory for having to wait so much longer.

 

[Tom`]

Why should not every generation improve on the previous. Early adoption is always at a price, second batchers deserve a faster chip and some more memory for having to wait so much longer.

http://en.wikipedia.org/wiki/Osborne_effect

 

[Exophase]

720MHz OMAP3 has nothing to do with how TI configures the chip, and it's still a different version sold than the 600MHz one. TI is likely doing some kind of binning to show that the 720MHz are capable of 720MHz, and not performing this on the 600MHz parts. They're probably confident that a high enough percentage of 600MHz parts will reach that rating w/o having to verify it, but don't have that same confidence with the 720MHz one. So the price difference probably accounts for the verification they perform. That isn't to say the 600MHz chips failed a 720MHz verification, but if any did they'd probably end up being sold as 600MHz.

By the way, this verification is going to be per-lot, they can't actually check the individual processors they're selling you.

 

[VRAndy]

Why should not every generation improve on the previous.

Because at some point OpenPandora will want to/is entitled to sit back and make a profit. They (presumably) don't want to go through this development hell for every (rather small) batch of product they turn out.

 

[Sphinxter]

Why should not every generation improve on the previous. Early adoption is always at a price, second batchers deserve a faster chip and some more memory for having to wait so much longer.

http://en.wikipedia..../Osborne_effect

Will it make me want to eat a bat?

 

[puppydee]

I'm all for change but here's a thought, why don't we just get the first batch out of the way first. You lot are trying to run before you can walk. Oh and btw, it's NEVER gonna happen, just thought I'd mention it.

 

[WolfSpider]

I would feel like Ive been ripped off if they added more speed to the second batch, especially since we've probably been waiting longer for the Pandora than most people that will order the second batch.

 

[Ollonk]

Why should not every generation improve on the previous. Early adoption is always at a price, second batchers deserve a faster chip and some more memory for having to wait so much longer.

http://en.wikipedia..../Osborne_effect

Will it make me want to eat a bat?

No, but you might lose all oratorical skill other than singing.

 

[WizardStan]

I would feel like Ive been ripped off if they added more speed to the second batch, especially since we've probably been waiting longer for the Pandora than most people that will order the second batch.

My only exception being if the 600 were no longer physically available, but that's not going to happen either.
Although I certainly wouldn't be one to complain if the second batch really was cheaper than the first. That's generally how it goes. I openly mocked all those iPhone purchasers that complained when Apple dropped the price, it'd be pretty hypocritical for me to act differently here.
(meaning I will openly mock anyone who complains that a subsequent batch is cheaper than earlier batches)

 

[Ollonk]

It makes sense for the prices to drop, at least in the UK/EU. However, the Dollar is now much weaker than when the Pandora was announced, and $330 is infeasible. (At least until costs drop some more or exchange rates improve.)

Edit: That might have came out wrong. What I'm trying to say is that $330 now is a significant price drop from $330 two years ago because the Dollar is worth so much less. I don't know the ins and outs of the Euro and Pound, but I know that when I received my order, the GBP amount was equal to $375. When I paid a few hours later it was nearly $380.

 

[darkblu]

720MHz OMAP3 has nothing to do with how TI configures the chip, and it's still a different version sold than the 600MHz one. TI is likely doing some kind of binning to show that the 720MHz are capable of 720MHz, and not performing this on the 600MHz parts. They're probably confident that a high enough percentage of 600MHz parts will reach that rating w/o having to verify it, but don't have that same confidence with the 720MHz one. So the price difference probably accounts for the verification they perform. That isn't to say the 600MHz chips failed a 720MHz verification, but if any did they'd probably end up being sold as 600MHz.

By the way, this verification is going to be per-lot, they can't actually check the individual processors they're selling you.

Actually there's a per-die verification process on each wafer produced, though i'm not sure if that would normally include the speed-binning. But each and every processor ever sold has passed a sort of a QA test at the fab. Simply because there's not such thing as a ~100% yield in VLSI. At rates of ~90% (and that's already optimisitic) you'd have 1 faulty die out of 10 - that's too much of a problem not to verify. For really advanced stuff yields of %50 or less are not uncommon - you cannot afford not to check each die. The costs of any yields issues, testing included, are included in the price of the product. That's why a piece of silicon which would otherwise be the second most-abund solid compound on the planet, after frozen water, can cost tens of dollars (if not hundreds) a piece in bulk from the vendor - it's all from the cost of the producton process.

 

[verm71]

It's been discussed too many times. The idea is always the same. Make some upgrades, and soon, stuff will come out that only runs well on those upgraded systems. That pisses off earlier buyers. It's OK for the PC gaming industry, but not the console industry, where if you pick up a a game for it, you know it'll work, and work well.

How does hardware need to be uniform

Because it would create a class of people who can run better games than the other class of people.
Also, some of the first batch people are like "hey you can't go adding better hardware to the later ones it's not fair I bought it first" or something.

It's just too much trouble, too many polls and useless threads of us posting the same shit over and over again. OP isn't going to waste time thinking about it.

 

[diegor]

I would feel like Ive been ripped off if they added more speed to the second batch, especially since we've probably been waiting longer for the Pandora than most people that will order the second batch.

My only exception being if the 600 were no longer physically available, but that's not going to happen either.
Although I certainly wouldn't be one to complain if the second batch really was cheaper than the first. That's generally how it goes. I openly mocked all those iPhone purchasers that complained when Apple dropped the price, it'd be pretty hypocritical for me to act differently here.
(meaning I will openly mock anyone who complains that a subsequent batch is cheaper than earlier batches)

Sorry, but I think there is an error. The 600 and 720 are exactly the same chip.

TI at the begin rate the 3530 cpu at 600Mhz becouse they are not confident that 100% of the cpu are able to work fine at more of 600Mhz. After some time they find they work fine to 720Mhz (often 800Mhz), so now they say:

"We have told you that the max speed guaranteed was 600, ok now we say you that is 720".

But the chip is the same...

That what's was said on BeagleBoard newsgroup....

 

[may88]

It's been said that the 2nd batch will be the same spec as the first.
I HOPE this is true mainly because I want there to be a larger community of Pandoras with the same spec.
This will mean that we will not end up with software being available for "second batch only".

Beyond that I think it poor form to be worried about what someone else is getting or paying.
I've opened my wallet and given OP the money required for the spec published at the time. That's my deal. And I'll be more than happy when I have it in my little mitts.

 

[WizardStan]

Sorry, but I think there is an error. The 600 and 720 are exactly the same chip.

TI at the begin rate the 3530 cpu at 600Mhz becouse they are not confident that 100% of the cpu are able to work fine at more of 600Mhz. After some time they find they work fine to 720Mhz (often 800Mhz), so now they say:

"We have told you that the max speed guaranteed was 600, ok now we say you that is 720".

But the chip is the same...

That what's was said on BeagleBoard newsgroup....

Then why are they still selling two completely distinct processors, one rated at 600 the other at 720? They are apparently chips that come out of the same system, and in that respect they are the same processors, but that doesn't mean they will all run at 720Mhz. Either they are speed testing them to figure out which can or cannot run at 720Mhz, or when a customer requests 720Mhz, they perform tests then to ensure that the processor they send can meet expectations.
Either way, the point still stands: The Pandora has the 600Mhz rated processor, and it would be in very poor form for the devs to switch to the 720Mhz processor unless they had no choice.

 

[MWeston]

My answer from TI on this about the 600/720 chips was:

Wafers are probed for certain tests. Then they are packaged and fully tested. That is the point at which the binning is done. No SW is run on the devices. They are basically clocked and the timing is checked.

All this means is that these devices are the fastest possible and will work over the full temp range.

They test all the parts the same, it just means if the data is better in some areas, it goes into a different box.

So as I think myself or someone has mentioned in the past, overclocking a 720 (or even just running at 720) will run cooler than a chip sold as a 600 at the same speed. The speculation begins as to whether it means everyone can overclock the same as long as you stay in an temperature controlled environment like an office or house where 25C is attainable. Further speculation based on the Motorola Droid, Palm Pre and Nokia N900 overclock tests suggests that the silicon can gate the data at overclocked speeds but these devices run considerably warmer when pushed so it is unknown if the extra heat will eventually mean their death.

If they have 600MHz chips inside, then it is likely they are running hotter than if they are sold with 720 chips. I don't know how much hotter the difference is. It could be as insignificant as 1 degree celcius or it could be 20 degrees (that sounds a bit extreme!). There's probably more breathing room in the Pandora case than these phones so that might also benefit overclocking. Also, since the POP stack on the Pandora PCB is on the bottom side, during normal usage (not setting the unit upside down on your table) all 515 balls will pass the heat of the CPU up to the PCB and dissipate through the copper planes instead of getting trapped in the RAM chip and having to pass through it to the air. I have no data to suggest this will provide any significant benefit but it seems better than overheating the RAM chip!

 

[atomicthumbs]

 Also, since the POP stack on the Pandora PCB is on the bottom side, during normal usage (not setting the unit upside down on your table) all 515 balls will pass the heat of the CPU up to the PCB and dissipate through the copper planes instead of getting trapped in the RAM chip and having to pass through it to the air.  I have no data to suggest this will provide any significant benefit but it seems better than overheating the RAM chip!

This seems to me like a very clever solution for heat dissipation.