Well, you could knock me down with a feather, if you thought remote programming was blisteringly fast.....
| File name | Time 1 | Time 2 | Time 3 | Time 4 | Time 5 | Ave Time (ms) |
| /blink.bit | 518 | 518 | 548 | 521 | 522 | 525 |
| /blink.bit.gz | 1140 | 1140 | 1142 | 1141 | 1141 | 1141 |
| /blinkc.bit | 86 | 84 | 87 | 83 | 86 | 85 |
| /blinkc.bit.gz | 189 | 189 | 189 | 189 | 189 | 189 |
| /sd/blink.bit | 1087 | 1087 | 1090 | 1086 | 1090 | 1088 |
| /sd/blink.bit.gz | 1148 | 1148 | 1150 | 1151 | 1151 | 1150 |
| /sd/blinkc.bit | 178 | 174 | 175 | 173 | 175 | 175 |
| /sd/blinkc.bit.gz | 193 | 196 | 193 | 193 | 196 | 194 |
The headline here has to be a blistering 85ms from the internal flash!!!
There is no surprise with compression methods, while some timings are nearer to each other than with remote programming, they are in line with what you'd expect.
As I understand it it might be possible to improve SD speeds (I might have the wrong end of the stick here), but even without an improvement I defy you to notice the difference between 175ms and 85ms by eye!
As the internal flash is limited it makes perfect sense to use --compress and gzip together, it being only a tiny bit slower than the fastest SD speed
Again that 85ms isn't a typo, fantastic! its easily conceivable you could have a massive library of bitstream files on your SD card, connect the board to your phone's AP and demonstrate your work anywhere....
I'm seriously thinking about finding a fast 640x480 LCD for this board.....
Not only is this board more than capable of retro emulation at home, it could very comfortably make a blisteringly capable portable game system.... (portable Amiga certainly, or maybe more!)