The second major foreign interview magazine, included a thorough "investigation" on the Italian manufacturer of Amiga NG equipment, ACube Systems Srl. Once again, we did not hesitate and we believe that we asked all the "difficult" questions, the ones that every Amiga NG owener and user would like to know their answer. The interview was released on the 4th issue and of course translated into Greek. However, we decided to publish the full version in English as a act of good will to the community.
Η δεύτερη μεγάλη ξενόγλωσση συνέντευξη του περιοδικού, περιελάμβανε την διεξοδική "ανάκριση" του Ιταλού κατασκευαστή Amiga NG, ACube Systems Srl. Κατά την πάγια τακτική μας υποβάλαμε όλες τις δύσκολες ερωτήσεις, αυτές που κάθε χρήστης και κάτοχος Amiga NG θα ήθελε να γνωρίζει τις απαντήσεις τους. Η συνέντευξη κυκλοφόρησε στο 4ο τεύχος μεταφρασμένη στα ελληνικά. Την δημοσιεύουμε αυτούσια στην αγγλική για ολόκληρη την κοινότητα.
Hello and welcome to RetroPlanet, our new paper magazine! Hope you enjoy the following interview, though we are prone on asking all the difficult questions that come to our minds and don’t hold back. Well, you can confirm this with our good friend Mr. Trevor Dickinson who started our interview series! Anyway, let us start:
R.P.: ACube was formed by the synergy of three Italian firms. Care to explain us with more detailed, the past of each company separately and its affiliation with the Amiga market?
A.C.: Enrico Vidale of Virtual Works and Max Tretene of Soft3 were Amiga fans already. They were the biggest Amiga resellers in Italy. Virtual Works doing Amiga and PC support, Soft3 doing software developments.
R.P.: Do the companies still exist as separate entities, following individual commercial projects as well? Do these projects have anything to do with the Amiga?
A.C.: Virtual Works is still working on Amiga and PC markets, doing sales and support. Soft3 developed software for AmigaOS like the Ringhio notification system, Hyperclock, XNet-RSS and did the porting of some games. Alternative Holding Group doesn't exist anymore and its founder is not involved in Amiga projects anymore.
R.P.: Cube was born in January 2007. How long the negotiations did took before you finally formed the synergy?
A.C.: It all started with the idea behind the Sam440ep, in early 2006. Later when the situation of Eyetech and the Amiga market was clear, it took only a few months to create the company and start the adventure.
R.P.: What was the initial and/or decisive factor that pushed you to form this type of synergy?
A.C.: Passion :-) And also the fact that Eyetech leaving the Amiga market, we really had to do something to bring new hardware to run AmigaOS.
R.P.: Project Codename Samantha 440 was your first commercial project. First question to begin with: Why name it Samantha?
A.C.: The prototype board sold by AMCC showing their 440ep SoC was called the Yosemite board. This name led us to the name of the funny character on some cartoons, Yosemite Sam. Then, as the Classic Amiga chipset had girl names (Denise, Paula...) it was more logical to use Samantha over Sam :-)
R.P.: How long it took you from paper design to actual commercial release of the board?
A.C.: Approximately, 1 year. Doing a first hardware project in less time was not possible. In such project there is always a lot to do (many tests, new revisions to build...) before the product can be ready for a release.
R.P.: Was it based on any AMCC test boards that you made amendments or you designed it from scratch?
A.C.: The first evaluation were made with the Yosemite prototype board under Linux. As this board is short on components, we had to design a new board with the components we needed.
R.P.: How many alpha and beta phases it took before verifying that the board is actually ready for mass production and how many boards you had to make in the process? Which were the difficulties that you faced?
A.C.: We built 2 series of 5 prototypes for the Sam440ep board. About difficulties it seems we had all of them. There is always something than can go wrong like shortage of components between the project design time and the time when the board is ready for mass production.
|Sam440 Flex edition|
R.P.: Hardware wise, why did you put a 1GB RAM barrier on the 440? Was this the maximum amount that the CPU could handle?
A.C.: The maximum capacity of this SystemOnChip is 1GB of RAM. For the rest we try to use 100% of the possibilities of our boards.
R.P.: Sam440 boasts an FPGA Lattice on board. Regretfully we have failed to see any valuable usage until now. Why did this happen?
A.C.: Well, we really thought Amigans would have come with some ideas to use this extension. We added this feature that is not present on PC motherboards hoping that someone would have taken the opportunity to use it.
R.P.: Around 4 years ago, I was reading opinions throughout many forums that suggested to take advantage of the FPGA same way Minimig did. From a technical point of stand, was/is this even possible? Did the FPGA had the required power to begin emulate an OCS/ECS system to begin with? What other useful ways of exploiting the FPGA would you suggest to the users?
A.C.: It could be done with a bigger FPGA from the same family. The idea we had in mind was a special UAE version where FPGA could have been used to do small parts of the emulation: floppy control, joystick and mouse ports...etc. And all the rest would have been emulated by UAE itself.
R.P.: How long it took you to port AmigaOS4 in Sam440? Did the port was done by ACube alone, with the aid of Hyperion, the other way around or by licensing programmers from Hyperion directly?
A.C.: Well it's a long time ago :-) The kernel and the HAL was done by the Frieden brothers and Hyperion. Drivers were done by ACube and OS4 developers so it was done all together :-) ACube paid developers for the development of some drivers.
R.P.: How many people were involved into AmigaOS porting, which were the primary obstacles they faced and how did they overcome them?
A.C.: Really I don't remember who were involved exactly. The biggest part were the ATI driver for the M9 and the development of Uboot but there was no particular problems. It was just the first time we did it so it took some time.
R.P.: Why did AmigaOS4 remain at beta stage for the Sam440, for such a long period of time and why the same happens once again for the 460?
A.C.: It happened with the X1000 too and it will happen again for every hardware because it can't be done differently. When a new hardware is done, AmigaOS is ported to it and as soon as it works nicely, the product can be released. Building a hardware costs so much that as soon as possible, the hardware must be released to get some return on investment. Without doing it like this, new hardware would not be possible. Also Amigans really want to have that hardware as fast as possible so we can't make them wait too long ;-)
R.P.: Should we consider that the life cycle of added features (new drivers, optimisizations) for the 440 ended? Also, what other improvements should be awaiting the following months on the 460 front?
A.C.: Even if the Sam440 reached end of life, we'll offer support as much as we can and as long as we can. As an example, if we have to do an update, we'll do it. About the Sam460, we are planning something but it's too early to talk about it.
R.P.: We haven’t heard any news for a long time on the MorphOS 460 port. Why’s that? Is MorphOS team experiencing any specific difficulty in accomplishing the task?
A.C.: The MorphOS team had to do a new release of MOS for the Mac platform so their Sam port had to be delayed a bit. After the release of MOS 3.4 they did a great job to have something to show for the Pianeta Amiga in December 2013. Did they find the port difficult? You'll have to ask them directly.
R.P.: Samantha series were created solely as an AmigaOS4 project or you had requests for specific features found on the boards from embedded clients? If yes, which features were?
A.C.: We had requests for an FPGA chip and extension vor very specific applications on which I cannot comment.
R.P.: In which exactly sectors of the embedded market do you believe that Sam series could have a chance, taking account the price - performance ratio of your systems.
A.C.: Industrial automation. Public information display (multimedia kiosks) like Ferrule Media.
R.P.: Have you sold any worth mentioning amount of boards outside the Amiga market? Either to hobby / computer enthusiasts (mostly Linux community) and/or embedded clients? A-EON has responded that their sole market is the Amiga market.
|Beta version of MorphOS running on a Sam460 EX|
A.C.: We sell boards also outside the Amiga market even if Amigans are the most of our customers.
R.P.: Many had criticized you regarding the initial pricing of your machines. What’s your respond?
A.C.: The price is very small compared to the low volume of a production batch. We should have sold them at a higher price to get our investment back but we wanted to keep the price as low as possible for all Amigans. We also decided to build them entirely in Italy instead of China to better control their quality and also to give work to European companies.
R.P.: Mr. Dickinson spoke us thoroughly regarding the NRE (Non - Recurring Engineering Costs). In which percentage the NRE costs affect the retail price of your systems?
A.C.: Because of the small price NRE costs have a pretty high percentage on the final cost.
R.P.: Did the 460 had cut down NRE costs in contract to the 440? If yes/no why’s that?
A.C.: Sam460 and Sam440 are very different but NRE costs are approximately equal. The 460 was not more complex or simple than the 440 so costs stayed the same.
R.P.: You (and AEON) have managed to succeed, where others failed. And I am speaking regarding the actual release of a product instead of hot air or promised vapor projects that ranged from late 90s to late 00s (i.e from boxer to MCC and from Panda to Natami. Why so many vapor projects in our platform? Because we must be holding a record or something :D
A.C.: It happens because these people have the Amiga passion and they want to do wonderful things. When they start a project they understand that it is extremely difficult to build a motherboard and it costs A LOT! :) Doing hardware on the current Amiga market is just impossible.
A.C.: Apple were leaving the PPC MacMini behind and it could be a good idea to purchase some of them and offer them on the Amiga market. Later Hyperion/Amiga Inc didn't want to follow us on this and it ended there. Too bad because it was the right time to do it and MorphOS had a good success on this platform so it was a nice thing to do.
|Minimig housed on a crystal case|
R.P.: Was Hyperion in knowledge of the port?
A.C.: Really I don't remember. It was so long time ago..
R.P.: Based on your experience, what Moana ISO lacks (i.e. kernel modules) and how much work would it need (in general) to make AmigaOS4 run properly on the Mac Mini PPC? I remember reading about 1 year work full time on/off work for the MorphOS in order to make it happen.
A.C.: Well, there was still a lot to do but as it's dead, now there's no reason to think about it.
R.P.: Looking back, from a marketing point of view, did you find worthwhile renaming/reselling Sam460 as complete systems under the brand name of “AmigaOne”? Would you do the same for the 440 if you had the chance?
A.C.: We wanted to use the AmigaOne name for the Sam440Flex but we decided to keep it for the Sam460 only which was ready to be released at that time.
R.P.: Licensing the “AmigaOne” brand name was done by Hyperion or Amiga Inc.?
A.C.: Hyperion is the owner of the AmigaOne name. We never had any contact with A.Inc.
R.P.: If tomorrow you were given the “OK” from Hyperion for backing you up on the porting sector, would you go for an AmigaOS4.x laptop solution? If yes, around which parts (i.e. type/speed of CPU, amount/type of RAM, onboard GFX, etc) would you built it? Which retail price would you consider feasible?
A.C.: Too many ifs :-) We do want to do a laptop but it is so difficult to do. And of course it can't be at the price of current PCs unfortunately.
|The infamous MOANA screenshot|
R.P.: We have already witnessed commercializing a community project (Minimig). Does any other current hobby product attract you?
A.C.: We are always interested to build new products for Amigans. If we find some nice products to build we'll do it. We have no new projects right now but we are looking all around the web to find one.
R.P.: Since you are mostly a manufacturing company, what made you went into openin an e-shop as well?
A.C.: As we were already selling to Italian Amigans, opening a shop was a logical step to be able to sell worldwide. And all resellers have an online shop. There's nothing strange here :-)
R.P.: Can you mention a small hint regarding your upcoming Secret Projects? How do you predict the NG Amiga market developing the future 1- 2 years?
A.C.: We can't say anything about projects in development, sorry. For the future we need absolutely new features in AmigaOS to make it continue growing: full 3D support of Radeon cards, printing system. Maybe it's about time to replace the old Picasso96 sub-system. With all this it would be a lot easier to move forward. Will we have this one day? I don't know but I hope.
ACube Systems Srl
© RETRO PLANET