Re: Really pocketable 15C clone Message #21 Posted by Mike Morrow on 19 Dec 2011, 2:09 p.m., in response to message #1 by Massimo Gnerucci (Italy)
Judging from statements at the DM-15CC website:
1. "based on the original microcode of the HP-15C" and "runs on a battery saving LPC1114 ARM processor emulating the NUT processor"
Is there no danger that HP will assert ownership of the firmware?
2. "uses one CR2032 battery"
That's a much better approach than paralleling two cells as the HP models do. That's just bad engineering by HP!
3. "runs up to 60 times faster than the original calculator"
This would be about half the speed of the new HP 15C-LE. That's good, especially if it means that that energy consumption from the single CR2032 is half or less than that of the 15C-LE.
The thing that's somewhat disappointing about this exciting product is its use of emulation of the old firmware rather than using all new native LPC1114 firmware coding.
A comparison of benchmark run times on the emulated HP 15C-LE versus the non-emulated HP 30b shows that identical tasks take about eight times longer on the HP 15C-LE compared to the HP 30b, even though both use the same SAM7L SoC and clock frequency. That shows that a task performed on the 15C-LE will use eight times the energy of the same task on the 30b.
Such gross inefficiency means that emulation is a really really bad idea for programmable calculators operating from a very small battery. In terms of reasonable battery life, emulation works for the HP 12C+ because user programming on the 12C+ is typically negligible. It remains to be seen if it will be practical in the 15C-LE, which is much more likely to be employed for user programming.
Regardless, the DM-15CC is irresistible! Here's a very interesting and more detailed demo video of the DM-15CC prototype. Fantastic work.
Edited: 19 Dec 2011, 3:09 p.m.
|