Is Amiga NG underpowered?

[danbeaver]

I thought we were discussing how under powered the PPC CPU and its architecture were...   Not the price.

[TheBilgeRat]

I thought we were discussing how under powered the PPC CPU and its architecture were...   Not the price.

Why do you hate capitalism?

[ElPolloDiabl]

Would there be enough Power8 chips to scavenge to make upgrade cards? You only need a around 2000.

[agami]

I thought we were discussing how under powered the PPC CPU and its architecture were...   Not the price.

Your attempt at witty subterfuge is poorly veiled. It's not about PPC it's about Amiga NG systems. And they are underpowered by many measuring standards. But if you prefer a better term for it then call them 'underwhelming'.

ARM is more "powerful" than the Amiga NG PPC CPUs in terms of price/performance ratio. And there's a lot more buzz at this end of the SKUs.

If people don't know it already they should; The Power 7, 7+, and now 8 are some of the most powerful socketed CPU ICs ever to come off a wafer production line. The only other things in this area of sheer performance per watt are the Intel Xeon Phi and the Nvidia Tesla cores, though the latter two have more specialist architectures and applications.

But as it has already been stated, these things have high price tags attached to them, and none of the Amiga NG OSs could take advantage of them anyway.

[ChuckT]

There's going to be a mandatory ROM lockout of non-approved (hobby) OSs on all new Windows 10 systems.  NG Amigas have no chance of becoming mainstream.   I'm going back to the classics using FPGA technology.  Classics have a better chance by sheer virtue of having more software.

Maybe it will cause more homebrew computers which would be a good thing.  I am glad they are forcing all of you.

[danbeaver]

Why do you hate capitalism?

WOW! Were we discussing capitalism?  Oh, that's right, we weren't.

Will let's update that "list" a little more.

[KimmoK]

I thought we were discussing how under powered the PPC CPU and its architecture were...   Not the price.

Page 13 should tell that PPC is not underpowered, at least not for every use...
http://www.freescale.com/files/training_pdf/FTF/2012/americas/WBNR_FTF12_NET_F0027.pdf
;-)

Some examples of blender speed:
23s blender 2,58 G5 2.5Ghz?? QUADcore (OSX)
27s Blender 2.73a AMD A4-3400 APU 2.7 Ghz (2 cores in use) WinXp pro 32bit
112s blender 2.62 PA6T 1.8Ghz Dualcore (LinuxMint)
113s blender 2.62 G4 1.8Ghz DualCPU !

(IIRC, world fastest blender render tests are done in 3s or so, at those speeds one surely must have also very fast HW overall (SSD and perhaps some GL-GPU help even?).)

It would be cool to see how fast blender renders on T4240 development board... could it reach 50% of fastest single CPU i7 systems?
(I have so far not seen any render tests done by e5500 or e6500 core devices.)

more.... vs price thinking:
Locally I found available:
-PlaystationTV e50
-RPi2+case+PSU+etc e90
-tablet e70...xxxx
-modern games console e300...500
-x64 Win8 laptop e199...1xxx
-x64 Win8 Desktop PC e199...5xxx
-x64 Win8 gaming pc e700...5xxx
-Apple OSX computer e1000...4xxx
-old, working G5 mac e100

(-my AOS4.1 system cost me ~1000 including everything, it has RPi caliber CPU but better I/O)

some components:
-intel CPU e80...3XXX (local)
-AMD cpu e60...2XX (local)
-t1014 e40?
-T1042 e80
-T208X e150

.... in HW wise, IMO we mainly lack A1200 kind of NEW low end AmigaLike at game console price range or cheaper, so more people could try it to see if it is any FUN or usable

more...:
-With tablet caliber CPU (2...4 cores @ 1,4...1,8Ghz) and 2D GFX we should be able to do similar productivity stuff than is possible with mainstream tablet HW (we need the SW)
-when RadeonHD GPU is added, we should be able to do similar 3D stuff as tablets and low end games consoles & cheapest low end x86 laptops & dekstops (we need the SW)
-For some high performance needs e6500 core based SoCs should be ok (but not super good and not for every need, especially if maximized single core throughput is essential). (but we need the SW)

We need more powerfull CPUs when we have SW that is slow on the HW that we have (/can afford).

So, IMO, in this order
#1 we need more SW
#2 we need more affordable low end HW (with useable performance, IMO: tablet caliber of performance)
#3 we need multicores taken in use
...

[whabang]

PPC is dead from a consumer perspective. While there's nothing stopping it from being a hobbyist platform, it's expensive as hell and there's not a lot of hardware to choose from.

X86 is well-supported and cheap, though there's a zillion different hardware configs. Getting it to run on everyone's PC's can be tricky because of this.

ARM is cheap as hell, and there's plenty of real cheap hardware available (the omnipresent Pi and boards like it). The hardware is a bit more uniform, but performance is lower than X86. Still, it's cheap as hell.

I'd prefer a Pi port, but that's subjective.

[guest11527]

While there's nothing stopping it from being a hobbyist platform, it's expensive as hell and there's not a lot of hardware to choose from.

Well, not quite so: There's something very important that stops it from a Hobbyist platform, and you said it: The price!

The real problem is here the way how some Amiga vendors try to address the market: They are basically following a (failed) 1980's strategy of a locked-down special hardware and a closed-source vendor-specific Os on top.

While this strategy worked for some time (shorter for CBM, longer for MS) it becomes more and more obsolete. MS has the market penetration to reach their goals and force vendors into their model, but Amiga has not, so it really makes no sense to follow this idea anymore.

All the "cheapo" hardware products hence operate in an open-source type of market to attract hobbyists, and it seems that this strategy is somewhat more successful.

[wawrzon]

Well, not quite so: There's something very important that stops it from a Hobbyist platform, and you said it: The price!

The real problem is here the way how some Amiga vendors try to address the market: They are basically following a (failed) 1980's strategy of a locked-down special hardware and a closed-source vendor-specific Os on top.

While this strategy worked for some time (shorter for CBM, longer for MS) it becomes more and more obsolete. MS has the market penetration to reach their goals and force vendors into their model, but Amiga has not, so it really makes no sense to follow this idea anymore.

All the "cheapo" hardware products hence operate in an open-source type of market to attract hobbyists, and it seems that this strategy is somewhat more successful.

strange to read it from you. are you starting to sympathise with aros?

[kolla]

If you want something substantially different, there is Linux. Or Windows. Or IOS. Runs on modern hardware, pretty powerful, avoids a lot of the design problems of AmigaOs. NG is just incompatible enough to 3.x to be uninteresting, but not incompatible enough to be modern.

Exactly, that it is the problem I have with "NG", and since 3D games are totally unimpressing to me, Warp3D etc are of little interest. And you should be aware by now that I do use Linux and various BSD (including OSX and iOS) already. Windows just doesn't have the tools I need daily, but it's ok as an occational service in the cloud or in temporary vm.

[guest11527]

strange to read it from you. are you starting to sympathise with aros?

I'm neither for nor against it, I just cannot contribute as a developer without putting it into danger. I'm just giving you my comments about development models. And yes, the open source development model also has a couple of serious drawbacks. Make your pick...

As a meta-comment: You should probably understand that I don't place me on a specific side. I'm not a particular fan of Windows, but I understand why most people use it as "least troublesome Os". I'm not a particular fan of the open source development model, even though I use such products a lot. And contribute...

I've done both. Open source, closed source... It depends on the situation what makes sense for you and your clients. Closed source Amiga might make also some sense. Open source might make some (different) sense depending on what you want.

All I'm saying is that I don't understand the Os 4 closed source model or the Morphos model if that suits you better - maybe it is reasonable for some people, but I personally cannot make much sense of it, so I don't care. Yet, if somebody from Hyperion would come and actually make me an offer to develop something for them, *and actually pay money for that*, would i regret? No, why? It's a job. I don't make this a fundamental decision for or against a specific system. It's the special client-gold coin-developer relationship I feel. (-: No money, no service, quite simple. I'm not *intrinsically* motivated for any NG system to offer anything for free.

[wawrzon]

I'm neither for nor against it, I just cannot contribute as a developer without putting it into danger.

you should have realized by now, that i understand your concerns and am not trying to force you towards aros. i just observed that you have rather a critical view on open source, and was a bit amazed that you seemed to shift your sympathies a bit in your above post.

I'm not a particular fan of Windows, but I understand why most people use it as "least troublesome Os". I'm not a particular fan of the open source development model, even though I use such products a lot.

me too. and also here i understand your concerns. open source software is in most cases inferior and shaped after its closed source counterpart. closed model is more efficient to achieve particular goals if there is good administration behind it. its similar to how it behaves in politics or other areas.

All I'm saying is that I don't understand the Os 4 closed source model or the Morphos model if that suits you better - maybe it is reasonable for some people, but I personally cannot make much sense of it, so I don't care.

i cant say i dont understand it. in the meantime i tend to think i can understand these motivations. still, i generally dont see much sense in closed source software development in such a limited community, especially taking the point of view of users, also as potential contributors, into account.

No money, no service, quite simple. I'm not *intrinsically* motivated for any NG system to offer anything for free.

even if i cant contribute what you theoretically could, i feel different about it. i would not offer my service to a company for free, but i have no problem to offer help to a free solution that benefits everybody.

[guest11527]

even if i cant contribute what you theoretically could, i feel different about it. i would not offer my service to a company for free, but i have no problem to offer help to a free solution that benefits everybody.

As said, I do contribute to some projects *if I'm intrinsically motivated*. That depends on the project, and it happened in the past. Os 4 or Morphos is not such a project. That's what I'm saying.

[wawrzon]

As said, I do contribute to some projects *if I'm intrinsically motivated*. That depends on the project, and it happened in the past. Os 4 or Morphos is not such a project. That's what I'm saying.

right, im aware about your free contributions of course.

[danbeaver]

IF opinions were noses, then everyone would sneeze in your face as if only they had that the right to do so.