The NatAmi project, which aims to recreate a classic Amiga, progress a little bit every day. Many (billions :-)) Amigans have express interest in this machine that looks promising. At the same time, the team around this project see the emergence of new heads, such as Guillaume Michalakakos, which will handle the music part.
Hello Guillaume. Could you introduce yourself to our readers?
Hello, I was born June 18, 1984 in Paris, within a family of musicians. I currently live in Taverny in the Val d'Oise. I work in a concert hall : I am technician and manager in sound and light. Next to that, I make my own little home studio, as a musician (of course, just taking a quick look to at family, I could not escape: harpsichord, "Ondes Martenot", guitar, graduation from classical music theory study, Jazz School,...), I always dreamed of producing my own music, to make my own album from begining to end. It required a lot of work but now it's finally possible, and at the right time to join the NatAmi Team as a musician. ^^
I took the opportunity to start producing for other bands. I discovered the Amiga in 1991 or 1992 when my parents bought me an Amiga 600 : one of my greatest gifts. At this time, I was using it to play games, it was only much later, by searching on Google about the Amiga by nostalgia, I came across the impact the Amiga had. I then discovered that the Amiga was not just a game console, but a real small single computer, responsive and productive. I quickly bought an Amiga 1200: my passion for the Amiga was born.
Your nickname is "Einherjar". I'm curious, that a Nordic-sounding and I would say warrior, right? Why choose such a nickname? Within you is the strength and the will of many warriors of Odin?
In fact, in Scandinavian/Germanic mythology, Einherjars are the warriors who died heroically in battle, and chosen by the Valkyries to be taken to Valhalla, where they will train to fight all day for finally feasting alongside Odin. This in anticipation of Ragnarok, the burn of the powers (a sort of Apocalypse), where in a huge battle, all the Einherjar will perish, including Odin. This will be followed by a renewal...
The choice of this nickname is actually very simple. I have another passion: the ancient history and medieval re-enactments. And (guess what) I'm in Viking re-enactment. I have two swords, two axes, inclusding one to throw, two shields, a helm, a chainmail, a scramasax, and I will not talking about clothing, jewelry, and other drinking horns, and on gastronomy. The medieval European fight is a very interesting martial art, more sophisticated than it seems, and very effective. You just have to be very careful to protect yourselves fine and not go too hard on a beginner or someone in difficulty. Even if blunt, this is weapons, not toys. I really like this environment, it is a return to nature, crafts, we meet many interesting people and passionate and we discovered many talented people. Einherjar is a nickname I have since a lot of time within this environment. I finally keep it.
Like many Amigans, I had a lot of Amiga, and many different expansion cards that go well. But now I have kept only my good old Amiga 1200, with Blizzard 1260-50 MHz, 64MB of memory and a 16GB Compact Flash card as a hard drive.
What do you do with it?
I experiment with the system, I play some demos, I play a few games and I create sounds. I do this sometimes to create sound samples on my Amiga, and to improve quality on my G5, then use these sounds in my compositions. Thanks to Aminet, I also found a way to convert any sound recorded MIDI data directly, and from there directly into musical score. It is very useful, and sometimes it makes me save a lot of time following what I want to do.
What are your thinking about the modern Amiga?
I think there are many good ideas and lots of good will within the Amiga community in the broadest sense. There are very interesting people who do a lot with not much, and this will to push more to creation than consumption of services attracts me greatly. Unfortunately, I also find it particularly regrettable that many members / groups spend considerable time and energy to denigrate the other systems. In short, it is drawn in the legs during this time, it does not produce and, ultimately, all the "camps" come out the loser. I had already noted that to my entry in the Amiga world, but now that I invested myself much more, it hits me particularly. We can present almost anything without people of another "side", often uninformed, what is more, just do anything to denigrate, misinform, brief attempt to destroy what others are striving to create. This is true for all forms of the Amiga : Amiga Classic, NatAmi, AmigaOS 4, MorphOS and AROS. There is no "evil" or "nice camp". Although I agree on the fact that we have our preferences or that we really do not like some systems for various reasons, as I do not agree with having to scuttle each other ! This helps to weaken the community for alternative computer users, which is already limited. I really despise those who act dramatically so.
I think it is still really amazing that despite the small market that we represent, there are still new solutions, upgrades, and new computers coming.
What are your impressions on the evolution of the Amiga NG systems (AROS, MorphOS, AmigaOS 4.x, etc.) ?
For now, I have never tested AROS. I know it's changing a lot right now, and it's getting very interesting. I think that it has a bright future, and because it's free it should have a long life.
I had tested MorphOS on my Mac mini G4, unfortunately, because I needed money, I had to sell it. This is an excellent system, apart from some (rare) problems of stability, and takes advantage of computer’s performance much better than Mac OS X (and even better than Linux). For me this is the leading Amiga NG system. But I have to ask myself some questions about the future hardware support. G4 Macs are starting to seem old, there are many machines that have problems over time, and having a faster frequency processor and better graphic card does not mean a faster memory bandwidth, which is the downside of the G4. To have had these two machines, you will have a very little memory bandwitch gain between a Power Mac G4 MDD with a Dual Card Sonnet 1.8GHz G4 and a Mac mini. It would be nice if there was new hardware for this system.
I once tested AmigaOS 4, and I must say that I didn't really like it. I think this system isn't very mature. I have seen a lot of crashes, but I found a few programs that are really interesting… I plan to try it again later, when it becomes more mature. I also have some concerns about the X1000 : when will AmigaOS 4 really be able to exploit the characteristics of this machine? How many programs will take advantage of it? How many machines will be sold with the announced price range? Hopefully, there is a new SAM. :-)
Now that we know you better, we'll take a fascinating subject which is being debated today: the NatAmi. This project that many of us eagerly awaiting. Recent events (launch video of the machine) shows that the project is making progress. What is the history of the NatAmi Project?
Thomas Hirsch, the creator, wanted to revive the Amiga long ago. He worked for many years rebuilding the OCS and AGA chipsets.
Gunnar von Boehn and Peter Kaltstein were working on their daily jobs in the same department of a hardware development lab as Thomas. They teamed up with Thomas to help him to get the NatAmi out, and were very motivated by seeing the 1st prototype in action at an Amiga-meeting in 2008. Soon after, more people joined the team.
The original prototype was designed like an A4000 using 50MHz memory. But starting in 2008, Thomas has been redesigning the Amiga DMA engines to make full use of modern memory. This means that memory bandwidth is greatly enhanced.
Late January 2010, I had the honour of being the first musician to join the team. My main role is to compose music for games, video presentations and more. I work at creating music that have a lot of diversity in styles but also of high quality using both electro sound recordings of real professional musicians. In short, the Amiga had excellent sound possibilities in her time, at least I want the music on NatAmi to be the highest quality possible. I have a lot of musical projects for NatAmi (and it's good, I just start playing around with my equipment at home) and I have many pieces in reserve, waiting to be recorded/mixed. For the moment everything is going well for me as team members like my songs, what inspires me to continue.
Of course I also offer my help to other things, like helping to finish the French translation for Total Chaos.
How did the choice of processor come about? What is involved in its development?
We wanted the fastest 68k possible. And we evaluated several options.
Use the 68060. The current 68060 cards work fine. And later 68060s run fine at 100MHz. When combined with fast memory, this gives the best 68060 CPU card performance. But Freescale (formerly Motorola semiconductor) does not develop the 68k line any more, and only produces slower models today. So the supply of 68060 chips is quite limited and there is no where to go for better performance.
Use Coldfire? Coldfire is a successor to the 68k line optimized for the embedded market. Many instructions were dropped from the 68k, meaning that there was a lot of slow emulation going on for native 68k code. Out of the Coldfire generations (V1-V5), V1-V3 were too slow, and V4 was only about on par with the 68030/040. The V5 is much faster, but they are not openly sold by Freescale. The other problem is that Coldfire is also not being developed.
Use an existing 68k softcore. We evaluated four, TG68, Suska 68k, and two others. We saw that today's FPGAs allow a perfect implementation of a 68k CPU. But none of these cores were optimized for performance. They were made like the old 68000.
From these evaluations, we decided that we could design our softcore using modern CPU principles. The advantages are: that it can be fast, at least as fast as a 68060; it makes us fully independent from outside suppliers or market fluctuations; and perhaps in the future with the proper budget, we can “bake” some real ASIC chips which would have a much higher clockrate.
Will the NatAmi be supplied with AmigaOS and/or a Kickstart?
Yes, we have a stock of OS licenses. We want to ship “complete” systems only.
What is the distribution of tasks within the team?
Thomas is the "creator of the Natamiverse". He did the chipset, IDE, video in, PCI, the mainboard, and the 68060 card.
The 68050 is done mostly by Jens and Gunnar. Andi, Keith, Jens, and Gunnar wrote 68k testcases for it.
Tami, the texture mapper, is still far in the future. The initial design was by Gunnar, Andrew Copland, and Jens. Andrew also develops the OpenGL library for it.
There are other parts being talked about, but no real details can be said yet. The team has grown big in short time, and it helps the project with misc contributions.
Does Team NatAmi need some help? If so, with what?
Software development! There is a need for new drivers and libraries to use NatAmi's functionality. There is a need for new games, including coders, graphics artists, sound/music artists, and so on. Having more people would help make the work go faster.
Do you have perhaps a remote idea of the availability of the machine?
We had to experience that estimations can be very difficult for a part time project, so even when we finally are very close to a first production, please understand we will not give out another date estimation.
The initial date of availability was announced for summer 2008. What happened?
The original NatAmi was based on the same memory architecture as the A1200/A4000. This means that the memory was all direct access. It severely crippled performance.
Thomas had the pcb plan ready for a non-DDR design, including the 68060 onboard. This design would have been very expensive to build, and not as fast and flexible as the LX pcb based on DDR2 and a 68060-card.
Since we aimed for the NatAmi to be much more, the design was completely rewritten to take advantage of modern pipelined burst memory. As you see, the delay in NatAmi release is driven by very positive improvements.
NatAmi has an enhanced architecture compared to classic Amigas. This offers new opportunities for software development. Are there plans for software that uses the new power?
The new memory architecture greatly improves performance for all old software. It may not be visible for well tuned software, but other software should be very apparent. For example, Workbench has never been this fast and smooth before.
We are also working on games that show the new power. Deluxe-194x shows off the NatAmi blitter, which is about 100x the performance of the AGA blitter.
We will also publish sources of games as coding examples for others to use.
Is there a support line envisioned to help independent developers work with the NatAmi features (N050, SuperAGA, etc)?
Yes. Proper manuals are forthcoming for both the N050 and SAGA.
The N050 can be used out of the box as any 68k. But it also has many improvements, such as new instructions and addressing modes. The N050 should be the easiest 68k CPU to code for so far. It is more orthogonal than any previous 68k.
Will the NatAmi be shipped with a custom case? Will the customer have the possibility to buy only the motherboard?
We plan to ship complete systems to resellers. We believe that complete, pre-installed, plug and play, 100% ready and working systems are the most user friendly way to go. Mainboards alone should also be available, to be able to use a custom case.
Do you have a scoop for the Obligement' readers ?
Team Chaos is cooking up a top-secret enhanced version of Total Chaos AGA for the NatAmi.
Do you have any final thoughts?
The Amiga was fun. Now the fun continues with NatAmi.