There has been stormy weather this weekend here in the UK, the perfect weather for another post of “This Week in PiStorm”…
New Firmware Tools
We have had several users confused about the output of the tool used to flash the firmware for PiStorm, especially as to whether it succeeded or failed. Also several users aren’t even sure which tool to run as they do not know which CPLD is inside their Amiga.
This week I decided to solve both those problems, in “wip-crap” there have been several changes. First of all the “nprog” shell scripts now hide away their output in a log file, the scripts just return a basic success or failure message. Secondly the tools will not run if the emulator is running, giving an appropriate error message.
The really useful thing comes in the guise of a new script called
flash_experimental.sh if you want the experimental firmware). This automatically detects which CPLD you are using and runs the appropriate tool automatically. It also has some checks in place for common issues, such as not having the flash tool installed.
To use this tool you simply need to
cd into your PiStorm directory and run
./flash.sh. You don’t need to worry about using
sudo, this is taken care of in the scripts when needed. The scripts also come with the exec bits already set, so no need for
This is an example where I had the PiStorm emulator running via systemd before I started:
Amiga 600 PiStorm Progress
Leigh Russ has created a new prototype for an Amiga 600 PiStorm which appears to work. It features a method to gain extra power from the floppy drive power connector if needed (he has some splitters for this) along with a protection circuit when using this method.
It should be noted that whilst this works, it currently isn’t easy to get it working and it isn’t stable with current firmware. Leigh will not be releasing this into the wild until firmware exists that makes it more stable.
There appears to be a lot of chatter about the chip speed of the PiStorm. Whilst the experimental firmware does improve things, and Sysinfo 4.4 shows a speed of > 1.0, it turns out for certain operations the chip speed is running at around 70% the speed of a regular Amiga, for others it is running at > 100%. The situation is a little complicated due to how the Amiga works. The developers, as always, are working on improvements to this. But the really large improvements are not expected until the proto 4 firmware has been released.
A leader of another Amiga community has also made statements about what is and isn’t possible with PiStorm. Unfortunately their theoretical statements were not based on facts as they are today, and certainly not on facts as they will be.
PiStorm is in constant development and there are plans and ideas which could accelerate chip speed in many cases well beyond 100%.
One thing I have learned over many years of developing database software is benchmarks lie. They tell you an ideal picture of whatever the developer of the benchmark wants to tell you. They don’t give you any real-world view of how software will really react.
Some games run fast right now, some games run a little slower, but the compatibility gets better every week.
There have again been reports of attacks by other communities against our community. Even people getting threatened with bans from other communities for mentioning “PiStorm”. As I’ve mentioned in a previous TWiP post, the PiStorm leadership do not consider PiStorm competition to anything else out there. There is room for all Amiga accelerators and each have use cases that are different and each have overlaps.
I ask that people please be civil in other communities even if other communities may not always grant us the same kindness. We are doing cool stuff, they are doing cool stuff and we are happy that so many different approaches to acceleration keeps retro machines such as the Amiga alive.