How will Hyperion continue the momentum?

[BigBenAussie]

They will release an X500 shortly after the X1000 with lower specs for the rest of us.

But first, what is the X1000?

Answer: The X1000 is made by Varisys and has the following processor:

PA Semi’s new PWRficient™ family of CPUs delivers ground breaking performance and efficiency: performance per Watt is up to 300% better than existing embedded computing solutions. The first PWRficient™ device – the PA6T-1682M – combines a pair of 2GHz 64-bit Power Architecture™ cores with AltiVec™ compatible VMX instructions, 2MB of on-chip L2 cache, dual DDR2-1067 memory controllers, and 24 high speed SerDes serial I/Os capable of supporting PCI Express, SGMII and XAUI.

A PPC board with above spec was delivered in 2007 can be found here:
http://varisys.co.uk/vm31.html

Varisys are partners with XMOS, and even spoke at a recent XMOS conference.
http://www.xmoslinkers.org/conference-online-wf

News from their website March 18th, 2009

After two and half years of working together Varisys and XMOS have complete their 16th development/evaluation board.

[BigBenAussie]

Ok. So building on that we can see that the X1000 processor is a PWRficient PA6T-1682M from P.A. Semi.

I believe a lower cost single core chip will be available in an X500.
Assuming it is pin compatible, a single core PA6T-13xxM could lower the costs, and allow them to reach some sort of economy of scale with the motherboard.
Also assuming it is pin compatable, chips with up to 8 cores(right now) and 16 in future could fill an even higher specced line-up perhaps designated X2000 and beyond.
I would kinda think specifying the amount of cores by having the model number divisible by 500 would make more sense than it did in classic days.

I also assume Varisys are supplying this board for what they see as their key markets:
 Video and image processing
 Telecommunications
 Military and aerospace
 Radar and Sonar
 Speech recognition
 Industrial Control
 Vetronics

So, perhaps leveraging these markets could help lower costs.

[BigBenAussie]

@AndyFC
I believe infrastructure components that you would expect, like networking, should be included in the operating system, and possibly very basic apps provided, but I contend that Hyperion shouldn't go further than that.

And now without further ado.... my obligatory treatise on spurring a viable Amiga software market.

I believe the eventual goal should be a viable and sustainable software market, which entices developers to take part for financial reward(even if only for sustenance), and with competition to drive the quality and features of that software to new heights. There is a danger that if you include software with the OS that is too good from the outset you let the air out of the balloon and potentially undercut developers and curtail software development.

A viable software market is important because it is clear that competition results in software of higher quality. There is a reason for the term "Commercial Quality" and that is what I believe we must strive for.

Ultimately, free applications are a false economy which only hurts the platform in the long run. It is often commented, that like anything, if software is given away for free, it is typically considered of little or no value. Regardless of whether it is true or not, of course, it is the general perception that unfortunately pervades in the general public. The mentality that everything must come for free is also anathema to a sustainable software market. It might sound aggressively capitalistic, but without people willing to pay for software, there will be little impetus to further software development, such is the case at the moment. Of course, piracy is an issue that arguably hurt the classic platform in the long run and we should be vigilant of the return of such a mentality.
 
I understand that the AmigaOS is on life support, and that to some extent, free software is a necessity right now. However, love of the platform can only get you so far, if it wont put food on the table. As the userbase (hopefully) grows and applications mature a software landscape overrun with free software will hurt progress. I am supportive of bounties and ports that enhance software development for every developer or create integral operating system features for every user, but contend that actual applications and games should be paid for. Bounty amounts are but charity and are not the main motivation for the task. However, free early beta downloads and shareware are most acceptable.

Microsoft is a master of providing just enough in their solutions to allow software developers to flourish. They provide all manner of tools but they are careful to never plug up all the holes in their tech.... and they do that on purpose. I can't tell you how many Microsoft technologies I have come across that provide little more than scaffolding. They know that it is in their best interests to nourish third party software and third party service providers to increase their technology's market share. The feedback loop of vested interests promote that technology further.

Hyperion should take aspects of Microsoft's approach with Windows, as AmigaOS is also commercial proprietary OS, and the most successful one at that. AmigaOS is not Linux, which pretty much throws everything in bar the kitchen stink(not a typo). Linux typically only succeeds when businesses pay for it in a service capacity(such as is the case with RedHat and Novell for instance), and I believe the services sector is most unlikely to be a viable form of revenue for the Amiga. There is no denying that the approach I am advocating has helped Apple's iPhone software market, while in contrast, their Mac software market with, what appears to me, to be the alternative approach of providing so much software up front with the OS, is often criticised for the sparsity of its software market. When you provide iLife with each new Mac I believe you get significantly fewer companies bothering to raise the bar in those areas and I contend that this could be the case with any application software bundled with AmigaOS.

Imagine how much great software the Amiga platform would have if developers could make a living from it. United in cause, all Amiga users and developers have the power to make this a reality.

[BigBenAussie]

Hell yea. I know if I suddenly out of the blue had a business opportunity with some major mega corp and a meeting with top brass, I'd tell 'em to shove it! Absolutely! It'd be the most logical decision ever.

Are you managing Amiga Inc?

[BigBenAussie]

@BigBenAussie

Your research seems fairly sound, but the only part that doesnt fit is that Hyperion claim that the board was built to thier specification, which the boards youve proposed being an X1000 wasnt. There's also "Hyperion AmigaOne" printed on the motherboard.

The board I see on the Varisys website, only demonstrates that Varisys has built something with similar specs before, that in fact was on sale as far back as 2007. A board containing the XMOS chip, which I contend must be the basis for the X1000, was referred to in the March 18th 2009 news item. The news item states that "After two and half years of working together Varisys and XMOS have completed their 16th development/evaluation board." This would imply that either, A-Eon/Hyperion contracted them for the design work as as far back as late 2006, or they have pretty much.........um........found a lucky convergence of ideas.

It would be nice if it is the PA6T they are using in the X1000, seeing as they have experience with it, and obviously had supply of them, but given Varisys has experience in all manner of PPC chips, it could just as likely be a lower specced chip from another vendor, which would be a shame. The argument against PA6T is that it's difficult to get nowadays, but we don't know what deal Varisys had with P.A.Semi nor if they hoarded supply. Let's face it, Amiga sales could well be a drop in the ocean in terms of supply.

[BigBenAussie]

@EvilGuy
Dude, I practically got my head chopped off by the Friendens for saying something like that in 2004. Welcome to 2010.

I still see where you're coming from. There have been Amiga window managers for Linux, but do you see them being prominantly used? Anubis is said to be sort of an in between, that is somewhat more palatable a solution, which is using the Linux kernal, but building the user experience up to be more closely Amiga-like. This gets you hardware support on par with Linux, but hopefully, without the bloat.

Look, probably the hardest part of AmigaOS4 is done, and it is just a matter of building it up, so there is no point for them going back to the scenario you mention.
Getting to x86 is still easier from where they are now than what you suggest. You have to remember that it would be very hard to drag the userbase over with what you suggest.

As for the future, the era of the OS will probably only run another decade, and then user experiences will be downloadable to a non-descript OS on any hardware over the internet practically instantaneously, and configured to your preferences. You could have an AmigaOS preference, with access to the same computing resources as any other. With the trend moving to internet based user experiences, and data rates increasing exponentially I see this as an inevitability. You may have low access or bandwidth right now, but do you really think it won't improve by the end of the decade.