Quote:

Interesting comment, but I'm not sure of your intent.
Do you know any better?

And yes Andreas, I knew that, but choose not to post it.
Frank deserved his confidentiality, until he chose to discuss the matter.

Quote:

I guess my comment wasn't clear. I agree we have the best developers in NG Amiga-like land and think they are far superior to any other group out there.

Also not meant as a dig at the various AROS guys, who are doing their own thing in a respectable manner. I just think on whole the MorphOS team are firing on all cylinders.

[ Edited by redrumloa 13.04.2015 - 11:56 ]

I kind of thought that was your intent.
Yeah, these guys are pretty talented.
AROS is interesting, but MorphOS is more stable and I just find it an overall better environment.

It will be interesting seeing the direct comparisons with OS4 as both OS' mature.
So far, the only system we have in common is the Pegasos and that is getting dated.

Quote:

I'm always interested in the comparisons, but OS4 users generally consider comparisons to be FUD only done by trolls.

> yes Andreas, I knew that, but choose not to post it. Frank deserved his
> confidentiality, until he chose to discuss the matter.

He gave up confidentiality when he brought his development machine to the show for everybody to see. And that he must have got it way before is pure common sense.

Quote:

Oh, absolutely. Trevor mentioned the idea of offering at least one to the developers a long time ago.
I'm just amazed that Frank (AND all the other developers) have the time and energy to pursue so many different projects.
While others may find the SAM460 more interesting (because of the much lwer price), I just don't see the point of buying a system with less power than a G4 Mac (except that they are available new and have a PCI-e expansion slot).
And it would not surprise me if the difficulties with the SAM port stem from issues with its processor.

In any case, I am thrilled with this development.
Its something many of us have been hoping for a few years.

This X5000 support is even more strange than the support for Sam 460. At least the Sam is "cheap" (but not really) and thus probably has 2-3x the userbase as the X1000 and X5000 can hope to accumulate *together* in their life spans. And still that won't be that many. IMHO it doesn't make sense.

But I see in bigfoot's post above that at least the Sam 460 resulted in "ironing out" some issues in Quark, which obviosly made it easier to port (given that it only took 2 days to get MorphOS booting on the X5000 thanks to this). And it possibly helps developing native drivers for new GFX cards. Both these things perhaps makes the X86 transition easier. Which IMHO would be the main benefit of this.

Because the X5000 is just as insane as the X1000, and just as with the X1000 it won't help OS4 in any way, and it will do even less for MorphOS. It takes a certain mind-set to buy €3,000+ EUR systems that is neither a chineese Power8 motherboard or at least a Mac Pro, and those who does it, does it only for OS4. For MorphOS there are far more suitable machines, and the X5000 will upon its launch come in second place compared to the fastest PPC Macs that MorphOS already supports, that's still available at a fraction of the cost.

@takemehomegrandma

P5020@2Ghz prices range from US$240 to US$320 for 1K pieces depending on the selected model.

Afair, it is way cheaper than the PA6T used in the X1000.

I hope the price tag for the X5000 will be saner than the X1000 one. If the X1000 board design is reused, it might also save a lot of buck from the R&D pov.

Quote:

As has already been stated elsewhere on these forums, the X5000 is very different than the X1000 is/was, so there was extremely little, if any, cost savings when developing the X5000 motherboard from what was learned on the X1000 design.

I can't remember the exact costs of the PA6T CPU's, but I think it was between $500 and $800 US dollars each, so if the CPU in the X5000 is only 50% to 70% less, it is not going to make the cost of the X5000 much cheaper than the X1000. Remember, Trevor already stated that he was not making any profit on the X1000, but expected to make a profit on each X5000 sold, to support continuing as a business.

I expect these facts to cause the cost of the X5000 to be nearly the same as the X1000, if not slightly higher.

@takemehomegrandma

I'm not sure if I can tell how much did I pay for X5000 mobo (so I won't) but it wasn't the price of X1000 so please don't spread fud in here.

> P5020@2Ghz prices range from US$240 to US$320 for 1K pieces depending on the
> selected model.

Current 2 GHz P5020 price range is from US$280 to US$340 at Freescale. US$240 to US$285 is the single-core P5010.

> If the X1000 board design is reused, it might also save a lot of buck from the R&D pov.

As WB_Coder said, completely different SoC with completely different pinout entails new board design from scratch.

Quote:

1. The price I mentioned was for a system, you spoke of a motherboard alone. Please don't compare apples and oranges.

2. Since the X5000 isn't publicly released yet, are you *really* sure that the price you got to pay for that *motherboard* is the same as what the final end-user price will be, and not some kind of discounted "betatester" price á la the X1000 "betatester" program? You actually claim that the final prices are set?

3. Here is from the horses mouth: "As for pricing. we are hoping that the new Cyrus Plus board with the P5020 will be slightly less expensive than the Nemo board, but no promises at this stage". This was in July 2013, almost two years ago, and the meter has been ticking ever since.

4. People here are discussing the cost of the CPU. What we really should discuss is the cost of R&D since the volume will be so small. They must have had quite high up-front development costs to that third party commercial design company they are hiring. They have actually designed several different prototype motherboards:

Cyrus Prototype 1



CyrusPlus Prototype 1



CyrusPlus Prototype 2



A loop of testing and new revisions after this is probably a given. Adding cost. Every time. This is what Trevor said about the X1000 development cost:

"All-in-all the NRE costs to produce the original prototypes, revision 2 modifications and new prototypes for hardware testing burnt almost $200K in pre-funding. That is the reality of the situation. If we could sell 200,000 units that would only add one dollar to the sales price. However, if we only sold 200 that that would add US$1000 to the sales price"

So how many do they realistically predict to sell to this ever-shrinking "market" that is probably already all-saturated by the X1000? The few hundred people actually prepared to pay that kind of money for an OS4 system with that backwards level of performance already did buy the X1000. And technically speaking the X5000 won't offer much that the X1000 didn't. Most of them wont go from €3000 to €6000 just to have two machines that are almost the same. AeonKit have to distribute their R&D costs on their sales predictions. And they can't be too optimistic here, they had better be realistic if they want to see their money back. This affects the price a great deal.


5. Trevor estimated the price of the X1000 to be "north of 1500 quid" at The Vintage Computing Festival 2010. Years later, when the product actually shipped for the first time and all R&D costs during the past years was added to the picture, it landed just below £2.000 AFAIK, making it surpass the $3.000 USD mark, excluding shipment and VAT (which is rather high for EU customers). In that interview linked to above from summer 2013 he said he *hoped* the price for the X5000 would be lower than X1000. But that was also years ago, and the product has been in R&D ever since and only now reaches a commercial state. The product could very well end up being "slightly less expensive than the Nemo board", but in my view it's not entirely impossible that it will be "slightly more expensive" either. For sure it won't be 25% or even 50% of the nemo board. Which is what could have made a difference.


So pampers, please don't throw groundless FUD accusations around you, especially when you are not contributing to the discussion in any meaningful way yourself. That's not very nice!

Quote:

It *isn't*! They have designed and produced several new board designs in their R&D process, from scratch, nothing re-used from the X1000. No economy of scale whatsoever.

> I can't remember the exact costs of the PA6T CPU's, but I think it was between $500 and
> $800 US dollars each

It was between 500 and 975 USD.

http://morph.zone/modules/newbb_plus/viewtopic.php?forum=3&topic_id=7183&start=894

With the merger of NXP and Freescale, the future development of Freescale's PPC products is uncertain.
I, for one, will be buying an X5000 in part because it will represent an interesting stage in the development of our platforms.

> They have designed and produced several new board designs in their R&D process,
> from scratch

Only Cyrus was designed from scratch. Cyrus+ (= Cyrus Plus = Cyrus rev2) is based on Cyrus, of course, even if the slot count and formfactor are different.

Quote:

I don't care if it's nice or not. I wish you could convert all that A-EON and Hyperion bashing posts into some articles about MorphOS and contribute this way to the community. Repeating the same stuff over and over again gives nothing. Sorry if it's not nice again but this is the way I see it.

Quote:

While high end PPC G5 Mac is hard to get, I hope it gets fully supported by MorphOS during its PPC era.

Mindset on X1000 purchase (on my personal example is)

a) Having no children or urgent mid term higher spendings (house, car etc.)
b) Wishing to have high end new Amiga(One), especially if you were either one of high end Classic owners, or you wanted it to be decades ago, but you couldn`t (East Europe)
c) Wishing to support the development. If there were no X1000 purchases, I doubt there would be x5000 and AEONs software purchases

Since I have tested X1000 under Linux and OS4, its fast enough, but no revelation. I wish for the future it was more G5 like in CPU performance (Altivec,SATA2 on fasted HDD or SSD, FPU and memory tests are fast) or it was a Power8 mobo. But it is what it is. Especially under AmigaOS where most of Nemo board can`t be used, and if compared in performance to SAM460ex where it offers equal performance gain to SAM460ex price. I found it to be very usable Linux system, which doesn`t make it that much of a non-usable toy.

I am saying all this, because X5000 is similar product. Its not in high end computing but in niche market of fastest available desktop PPC board. Whatever promotion and market visibility X5000 receives, there is a potential benefit for MorphOS.

One in all, while I dislike way Hyperion does things, AEON seems to do things nice way.

Quote:

Much of what you wrote I agree to, but it _may_ be different. Thing is, we don't know much real life power of the 50x0. While the PA6T is a pretty castrated chip of low performance this may not be the case for the 50x0. The ALU is fast (three instructions/cycle), the bus is fast, many modern peripherals are integrated to the chip. If Freescale did their job well it can be actually quite an interesting chip that probably performs _different_ than the 744x or 970xx do.
And yes, something like 3000 US$ is quite a lot of money, even only the half would be quite a very steep price, but depending on actual performance and state of my bank account even I can imagine scenarios where I could start thinking about buying one. Most of all: It's new and not a electric energy hog (like the G5s). My G4s are running fine still and except one harddrive I never had a hardware failure with any of my hardware ( C=, Phase5, bplan & Apple to name the ppc and 68k ones), but they are about 10 years old now. It's not granted they will run 10 more years. Sure, if there's a failure i can head over to ebay and get another machine for pocket money (would probably do so anyway), but that hardware is aging, too. Hence, new hardware has a certain appeal. Dunno though how much $$$ this appeal may be worth and I doubt it will do the math for me. But there are quite some ppl whose wallet opens quicker than mine. Hence, i wouldn't say it's a given that X5000 with MorphOS will be a "failure", but I also wouldn't bet money on a "success".

@vox

I understand that they are hard to find where you live, but it's pretty easy over here in Sweden. The other day I even found a G5 @ 2.7 GHz (without HDD) for 30€ (!). I'm bummed that I didn't buy it (even though I don't need it).

Quote:

Yes, they are cheap here too.
I purchased a 2.7 for about $50.

@Jim

Yeah! Sadly, the market is very uneven. Sometimes there are plenty of cheap G5:s, sometimes it's all dried up. I bought two of my G5:s during a dry spell and paid 300€ and 200€ for them. Overprices yes, but they where exactly what I wanted so I paid anyway

I wonder if you will ever see a used X5000 for 300€ or less

Quote:

If it auctions with no reserve, possible. Some years back I remember seeing a Sam440 w/OS4 on eBay go for ~$30(USD). I can't imagine the seller was all to thrilled. Haven't seen one sell with no reserve auction since (but haven't looked honestly).

@vox,

I thought you sold (traded for SAM460) your X1000 to some artist user who makes posters, or website designs?

> I thought you sold (traded for SAM460) your X1000 to some artist user
> who makes posters, or website designs?

Vox clarified later that this was no sale or trade but a case of mutual borrowing/lending.