BLITTER RE-CREATION THOUGHTS

[exxos]

Hopefully all that glitters won't be on fire again

[PaulJ]

Hopefully all that glitters won't be on fire again

Just keep in mind all that glitters is not gold!!

[rubber_jonnie]

That is awesome work guys, it really is.

Of course knowing that I like to build and test stuff, you have a willing volunteer if you need one

[Icky]

All that Glitters is golden

We have a booting STFM with an FPGA Blitter Board (Sparkalaphobia).

Now although it boots, we are getting test results its performing at 58% of a stock Blitter. We need to do some more work to get it up to normal speed.

However that said we are now truly on the road towards the ST Remake FPGA Chipset era.

[rpineau]

[dhedberg]

Wow! That's quite an achievement! Congratulations!

[Cyprian]

awesome!

[Maximilian]

NIce!!!

[Icky]

Posting some GB6 benchmarks to show where we are currently. We think that the Blitter is under performing due to how the state machine works in the core and we are looking into this. We did a test of the Blitter at 16MHz to see how it was behaving but it created screen corruption. Managed to get the results from GB6 in that mode by running all the tests then turning the blitter off and running the ALT RAM test to clean the screen corruption when it refreshes.

I learnt something new that by clicking on the individual test text in GB6 runs just that test.

First image is of the FPGA Blitter running at 8MHz against a stock STFM without a Blitter

IMG_4256.jpg (148.44 KiB) Viewed 5096 times

Second image is of the FPGA Blitter running at 8MHz against a stock STFM with a Blitter

IMG_4255.jpg (142.75 KiB) Viewed 5096 times

Third image is of the FPGA Blitter running at 16MHz against a stock STFM with a Blitter

IMG_4254.jpg (153.36 KiB) Viewed 5096 times

[PhilC]

Looking good nonetheless. The hardest bit is to get it running in the first place. The tweaking will take time but I'm willing to bet you'll be able to see improvements almost with every step.