Pandora Lifetime

[craigix]

OMG ! Is there a Pandora with newer TI's OMAP chip, at 1ghz or more, I'll buy one directly !!!

Clearly Craig and ED should get a dozen or so of the new and uber expensive chips and have them made into quasi-prototype 1.2Ghz Pandora XZRS models and sell them to the highest bidder on Ebay. Not enough of them to make a push/switch in code base. Uses a different board/chip population so it won't affect pre-orders. Should sell for stupid amounts of money.

Is it as simple as putting a different chip in the queue for the pick&place machine? If so - go for it.

After batch 1&2 are fulfilled, selling new units with faster CPUs - I don't see a problem with that if:
1. Replacement/upgrade boards are made available to us first batchers at a reasonable (with profit) price.
2. We get some core credit at the store for returning the old boards to you for use in warranty replacements.

There won't be many people with the hardware needed to replace this chip on their own.

First things first though, get those first batch orders filled please. We're rapidly approaching two years.

It's not that simple and requires new RAM and software too.

We might be making some as devkits for that chipset to sell to companies who want to get their hands that chipset early. There would be no support from us for them and they will be *very* expensive due to the custom made nature of them.

 

[conso]

We might be making some as devkits for that chipset to sell to companies who want to get their hands that chipset early.

lol

 

[maplesugarlover]

We might be making some as devkits for that chipset to sell to companies who want to get their hands that chipset early.

Do you mean something like the ti blaze?

 

[craigix]

We might be making some as devkits for that chipset to sell to companies who want to get their hands that chipset early.

Do you mean something like the ti blaze?

Well we have been asked by several developers if such a thing could be done. This way they don't have to make their own boards to try the new chip. So it's something we are considering if they put their money on the table

 

[ari64]

DM3730 by any chance? One thing I like about that chip is that it doubles L1 cache size. I'm very curious to see how that would improve scores on typical applications (with all other variables kept equal), including ones hurt by the memory hierarchy of the system. My guess is that it doesn't help nearly as much as doubling L2 would, since the L2 performance on Cortex-A8 is very high. The higher clock, especially for the SGX, is nice too; you can get those on OMAP3630 but I doubt that'll be as readily available in lower quantity.

I'm curious how much the larger L1 improves things too. OMAP3530 does actually spend quite a lot of time in L2 wait. I've also seen some issues with the write buffers getting full, which I assume is blocking on L1->L2 writeback.