Right before thanksgiving I started a forum topic Chimera Queues Brainstorming. The concensus there was that the LPC2103's limited RAM and the slow SRAM speed was going to constrain the system too much. This caught me a little off-guard and has caused me to change priorities. I didn't want to significantly increase the cost of the cart to add extra hardware that was not needed by the Supercharger subsystem. With the 2103, I think we could say we hit that sweet-spot. But by moving to the 2368, and the subsequent shift from a 2 to a 4-layer board, Chimera is no longer a Supercharger replacement with as many extras that don't cost a lot to add. The extras are becoming the focus, and will invariably raise the retail cost of the unit.

Delicon and I would both like to see people use these features, and so we have to listen closely to requests. Saving money is not going to do any good if developers do not embrace it because it appears hamstrung (as ironic as it may seem given the usual task of coding for the 2600). I still think that a lot of what Chimera could do even with just the 2103 is not being fully understood or appreciated. Hopefully when my version 9.2 board comes back I will be able to illustrate some of these things. Nevertheless, the 2368 should provide plenty of headroom for the kinds of bells and whistles mentioned in the brainstorming thread.

So the news tonight is that version 10's circuit layout is complete.

As you can see, the CPLD at the bottom is surrounded on all 4 sides by connections. If you look closely at the ARM, you'll see that there are also scant few pins that aren't already connected to something. The wider headers take up a little extra "emergency space". There is now a power socket on the right. That is meant to power dongles. (The 5v coming from the cart port is only enough juice for the cart and to provide 3.3v (from the ARM) for lower-power dongles like controller-port and RS232.). behind the power connector to the top are the JTAG pins. Delicon doesn't think they will be useful. You never know.

I'm crossing my fingers that this board checks out when it gets built. I'm really looking forward to closing the book on the hardware and getting more involved in the firmware. The ARM provides almost too much flexibility in that respect. We've theorized a lot about how the timing is going to work out and what this is going to enable for games. We just have to try it out and see.