FPGA based A500 accelerator

[thgill]

Name recommendation: Sledgehammer

It will be very easy to come up with logos and images with that and it makes it pretty clear what you think it’s performance is like.

[Mathesar]

Originally Posted by Lemaru

I'm guessing the core will be opensource but the hardware won't be?

Please let everything for Amiga be open source from now on...
No closed hardware anymore!

[Lemaru]

Originally Posted by Mathesar

Please let everything for Amiga be open source from now on...
No closed hardware anymore!

Totally agree. There are some extremely talented people out there who can produce some amazing stuff based on opensource hardware. Unfortunately people like to keep it closed and stick a rediculous price on stuff though

[kipper2k]

Originally Posted by Lemaru

Totally agree. There are some extremely talented people out there who can produce some amazing stuff based on opensource hardware. Unfortunately people like to keep it closed and stick a rediculous price on stuff though

This is a recurring discussion about whether people who spend a lot of time creating HW for the Amiga and other devices are willing to forego compensation for all the development costs. The cost nowadays for complicated HW is not cheap and developers need rewards for keeping the various platforms alive. No one expects manufacturers of phones and other devices to be open source so why should this be. Please dont assume there will be a ridiculous price on this accelerator, it will be feature loaded and the price will be much lower than other accelerators that are on the market.

I do not know what Mike's plans are as regards open source but ultimately it is his decision

[jbilander]

Yes please open source this. I'm curious to learn more about the glue logic with the Amiga learning from people who know how to do this the right way. Ofc people who put a lot of time and talent in this kind of work should be rewarded, in that case a PP Donate button or a Patreon account can help out with putting food on the table and recovering dev and HW costs. Kickstarter is also an option. Anyway, whatever you decide it is good with more options out there. In the end it is up to people to decide whether it is worth buying or not.

[Niklas]

Originally Posted by kipper2k

The cost nowadays for complicated HW is not cheap

On the contrary, the cost of making advanced hardware designs has never been lower due to advances in FPGAs, the low cost of having PCBs made in China, etc.

Originally Posted by kipper2k

and developers need rewards for keeping the various platforms alive.

The feeling of contributing to and being a part of a community is enough of a reward for many, assuming you have a day job that pays your bills.

The real problem to me with closed source is that buying an accessory to my Amiga is not an end in itself, it's a starting point. I want to be able to play with the thing in any manner I like, typically by hacking it, learning how it works, and improving it. Not releasing the design as open source diminishes the value of the product significantly, to the point that I would not consider buying things for my Amiga that are not open source.

Also, not releasing the design as open source is short sighted; eventually someone will make the same thing as open source, and then that thing can be built upon and improved further. As seen by the several similar designs that have been mentioned in this thread, this is neither the first nor the last FPGA based accelerator.

[kipper2k]

the board will contain BGA components with very fine pitch. This will make the board require professional assembly. Soldering them is very difficult without the proper tools and trouble shooting in the event of a failure would be a nightmare.

Originally Posted by Lemaru

Totally agree. There are some extremely talented people out there who can produce some amazing stuff based on opensource hardware. Unfortunately people like to keep it closed and stick a rediculous price on stuff though

There are a lot of smart cookies out there that could probably use Github and probably find enough info to put together something as regards an accelerator and other things. But as i mentioned it has got to the point where it cannot be built by the masses and someone needs to invest capital, and take the risks of having these boards professionally assembled and then of course sell them on. when you buy electronic components nowadays you get massive discounts when buying for large batches which will reduce the cost. If you try to buy for yourself then the cost and the risk would be all on you so some things i believe are best left to the board houses

[Mathesar]

Originally Posted by kipper2k

the board will contain BGA components with very fine pitch. This will make the board require professional assembly. Soldering them is very difficult without the proper tools and trouble shooting in the event of a failure would be a nightmare.

Well... I happen to have acces to a vaporphase oven at work ..

But seriously, FPGA boards all follow the same recipe: Power supply, FPGA, SDRAM, flash. And, in the case of Amiga FPGA boards: levelshifters. That's it. Nothing magic there but note that I started with the power supply Now, if someone wants to copy/build it, they will build it. I see plenty of A3660's ("professionally built!") on Amibay, as well as TF530's and A500 RAM expansions. And there is a market for them as well, plenty of people who can't built it themselves or just don't want the hassle.

However, why can't it be both? Open source and available as a ready built project? Just look at the SCSI2SD V4 or the MNT VA2000, completely open source but also available as ready built products. The MNT people even sell their scan doubler / monitor switch as a DIY kit! How wonderful is that!

I personally like to built stuff and making your own Amiga expansions is very satisfying. I also like the fact that I will always be able able to repair it because it is open source. After all, most of us just want to keep our Amiga's alive for the next few decades so this is an important point.

But it is ultimately Mike's choice what he wants to do with this very cool project

[kipper2k]

i can't give any info as to some of the characteristics yet, once we show you some pictures then you can gauge the buildability of the board. I am biting my tongue again, but i cant tell you any more but i really think you will be very happy. And of course the pricing will be very fair

[kipper2k]

one thing also is the board is so complicated that you will need this amount of solder...
https://i.ebayimg.com/images/g/oycAA...6f/s-l1600.jpg

[robinsonb5]

Originally Posted by kipper2k

one thing also is the board is so complicated that you will need this amount of solder...
https://i.ebayimg.com/images/g/oycAA...6f/s-l1600.jpg

Are you building it in the steampunk style?!

[kipper2k]

Originally Posted by robinsonb5

Are you building it in the steampunk style?!

A little light hearted humour to lighten the mood

[TroyWilkins]

Originally Posted by thgill

Name recommendation: Sledgehammer

It will be very easy to come up with logos and images with that and it makes it pretty clear what you think it’s performance is like.

Sorry, but this one has already been used too: https://en.wikipedia.org/wiki/AMD_K8

The AMD K8 Hammer, also code-named SledgeHammer, is a computer processor microarchitecture designed by AMD as the successor to the AMD K7 Athlon microarchitecture. The K8 was the first implementation of the AMD64 64-bit extension to the x86 instruction set architecture.

[thgill]

Originally Posted by TroyWilkins

Sorry, but this one has already been used too: https://en.wikipedia.org/wiki/AMD_K8

I'm not sure that's really relevant. It's not like someone is going to think this is an AMD cpu for their Amiga.


Besides, that was a codename, not a production/retail name.

[TroyWilkins]

Originally Posted by thgill

I'm not sure that's really relevant. It's not like someone is going to think this is an AMD cpu for their Amiga.


Besides, that was a codename, not a production/retail name.

Yeah, I agree, but that was the first thing I thought of when I read Sledgehammer, well, that and the song, haha.

[ Show youtube player ]

Anyway, I'm really interested in this project, it has a lot of potential, and I look forward to future updates.

[Mathesar]

Originally Posted by kipper2k

one thing also is the board is so complicated that you will need this amount of solder...
https://i.ebayimg.com/images/g/oycAA...6f/s-l1600.jpg

I guess I will need a new soldering iron then

[AJCopland]

Just call it "Bob500" or something the name hardly matters so long as people can Google for it.

It'd be nice to have it open source but only so others can make A1200, etc versions more easily. Starting from a known good example is always easier than starting from scratch.

Either way this is an idea who's time has come and I'm glad there's an FPGA based design coming which is more open than the Vampire.

[kipper2k]

Work is ongoing on improving the numbers for the board trying to squeeze out every available dhrystone, so far it is Very encouraging

[AJCopland]

Kipper2k you guys could ship it right now, as-is, and it'd sell like hot cakes

[kipper2k]

Originally Posted by AJCopland

Kipper2k you guys could ship it right now, as-is, and it'd sell like hot cakes

I think it will be popular, in the last couple of months we have upped the specs identifying some things that people will want. there will not be a lot of unused space on the board, and of course the board is multilayer

if you check Mikes blog you will see a reply to a question in there referring to whether it will work in other models, here is his reply..."Don’t see why not (size aside, and this will shrink dramatically). The picture showing the DE0 is just for dev and the final size will be nowhere near that size

The De0 board will soon reach its limit for dev testing as components will be needed that are not included on the DE0 board making it unusable for the final product