@tlosm

Quote:

tlosm wrote: @Shiryu oh my gosh, trools think sam is a router

The Sam would make a terrible router. But embedded stuff is what the CPU's used in SAM's was good for/was meant for. Certainly not desktop usage.


Quote:

x 1000 was crappy,

The idea behind it and the design was crappy (PA6T, a dead-end CPU already at the design stage, the Xorro/Xena that not even AeonKit knows what it's for, etc), its price was crappy, its performance was crappy (especially considering its price). Proclaiming that this was the future for OS4 made many lose hope, and as such, the X1000 is probably the single biggest cause of people leaving the OS4 platform in modern years. That's kind of crappy too.


Quote:

e5500 is slower than a g5

On a clock-by-clock comparison, the X5000 has a slight edge when it comes to integer performance. But then again, so has the G4 as well, which isn't exactly anything new, and pretty unremarkable actually. But when it comes to FPU and memory performance, the situation is different, the G5 has the edge over X5000 there. And on top of that, the X5000 completely lacks a SIMD unit. These three latter things has a real impact on practically all multimedia applications for example. And on top of it all, when not doing a clock-by-clock comparison anymore, but real speed of the machines, the G5 wins simply because of the higher clock frequency alone.

But looking at it through "2017 desktop level glasses", I'd say the X5000 is about on par with all the other PPC stuff from more than a decade ago. Meaning: Who on earth cares what is slightly faster or slower in one or another area of measurement, when the situation in total is so completely off the charts behind? The rest of the world moved not generations, but several dimensions ahead since these performance levels of the G5/X5000 were relevant in desktop computing.

@TRIPOS

Just read your ann.lu posts about AmigaOne written more than decade ago. Some people never change...

@fishy_fis

Quote:

fishy_fis wrote: I can't imagine Apollo specific software will be very prominent. There's simply too few owners, ergo developers of the hardware. Compared to "real" 68k CPU, or emulation users its a niche user base. Can't say I agree with the idea that Vampire users are more likely to spend money and/or more active than non-vampire users either. The vast, vast majority of recent 68k software has been developed on either real 680x0 based systems or UAE. ACA accelerators always sell out quickly, Indivision sold like hotcakes (vampire wasn't even on the radar back then),etc,etc. As an Amithlon user it'd be disappointing, and ironic, if software developed with more resources in mind didn't work because someone targeted it at the comparatively slow vampire cards.

The retro crowd is divided in two segments, just like in automobile retro crowd (very similar). There are the purists/preservation enthusiasts that wants as much as possible to be original and untouched. They collect, restore, preserve, recap, stores, trades, and uses the gear with a great sense of nostalgia. When using it, they enjoy real floppies (the waiting, the clicks and hums is part of the experience) with games/apps made for the standard H/W specifications (and limitations) they have. They are perhaps not even using a CPU turbo card at all, but if they do, it's probably a Phase 5 or similar (from the same era as the computer) in mint condition. Then there are the pimpers, modders etc that uses the original gear as a base for further modifications. Putting A1200's in custom tower cases, floppy emulators, bolting on all kinds of expansions and modifications, building their own custom Franken-miggys.

The former category isn't likely very interested in stuff like Vampire. The latter category however would obviously love to have a 200MHz "68060".

But when you start introducing new, incompatible ISA's, features never present in Amiga, etc, then you are effectively moving out of the retro scope altogether and into the "NG" realms. But if the PPC and the implementation of OS4/MorphOS fell short of NG enthusiasts dreams (which it did), then an incompatible 68k with special S/W base will fall even shorter. It will be completely unusable for all the modern stuff NG enthusiasts wants, every single thing.

I don't see any evolution in Workbench, no development beyond Workbench 3.X. It isn't feasible. New developments are in MorphOS, OS4 and AROS. The number of people prepared to buy and use a "Classic-NG Franken-miggy" will probably be counted in tens. Maybe a hundred or two. Special S/W? Nah, don't think so. Maybe a couple of demos. Which could be cool. But it will never become a platform. If you think about why people are using NG, and why other people are using Amiga 68k, then you should see how "Classic-NG Franken-miggy" will find itself stuck in the middle, not being either one of them.

@TRIPOS

I am not sure if you are right...

there are a lot of comments from hardcore amigans who will not use Vampires in big boxes but for A500 or A600 (or A1200) where extensions are rare or not existing. I do not see why people cannot do both.

Regarding OS you seem to forget that AROS exists for 68k. I hope that finally people will be able to use both 3.X and AROS.

And regarding software really specific software will be rare, in my view such specific hardware features should be used on OS level, not in 3rd party software.

@TRIPOS

Quote:

TRIPOS wrote: @fishy_fis The number of people prepared to buy and use a "Classic-NG Franken-miggy" will probably be counted in tens. Maybe a hundred or two.

Have you been hiding under a rock somewhere? The number of users interested in the Vampire accelerators is already many hundreds, and probably already over 1,000. It is possibly the single most sought after Amiga hardware project in the last 20 years. I would not be surprised if 2,000 to 3,000 Vampire boards are sold over the next 2 years, if they can produce them quickly enough. When a version for the A1200 is finally released, that version might be the most popular, and sell in even higher numbers.

I think that you will be surprised when the number of Vampire users exceeds both the number of AmigaOne w/OS4, and PPC Mac's w/MorphOS, 2 to 3 years from now.

@BoingBear

http://orders.apollo-accelerators.com/

1844 people wanting to be vampirized !

And then you have the 100s of people that already got Vampires (like myself).

@TRIPOS

Quote:

The number of people prepared to buy and use a "Classic-NG Franken-miggy" will probably be counted in tens. Maybe a hundred or two. Special S/W? Nah, don't think so. Maybe a couple of demos. Which could be cool. But it will never become a platform. If you think about why people are using NG, and why other people are using Amiga 68k, then you should see how "Classic-NG Franken-miggy" will find itself stuck in the middle, not being either one of them.

Wow if there ever was a classic troll comment this is it.

You are seriously uninformed if you believe that above statement. Maybe you should check out the project in more detail. There are some seriously intelligent people doing some quite remarkable things with a low-end FPGA.

Is this an Intel-killer? No way. Is it a low-end PPC accelerator killer? Absolutely. As a bonus you get some very fast memory, an RTG system and HDMI out. What's not to love?

This product is causing many Amigans that gave up on their machines and didn't go PPC to "come out of the closet". This can only be good for the entire Amiga community.

@Overflow

soon there will be more Vampire users than MorphOS user and in two years I guess the vampire market (together with standalone device) will exceed the whole NG market, not counting UAE user.

Some of the community never fail to amaze me with quite negative comments about Vampire becoming a platform...blah blah blah. Simple fact is that Vampire pushes our Amiga's further and as long as we have better hardware who cares.

We have been a retro hobby platform for years (too many years) and however you wish to describe it the Amiga is in it's strongest marketable position for many years and that is in part to the Vampire. Just roll with it. Buy whatever hardware or software suits your needs but don't try to look 2/3 years down the line at what may or may not be - just embrace what you have and enjoy it - thats the Amiga spirit.

Many years ago natami was announced as fast and cheap alternative to ppc.

68k faster than 200 MHz G4 for under 100 E will be something super cool.

After many years of lies there is vampire, for 280 E plus another 100 E for C= hardware you get 68060 performance.

Five time worse price performance ratio than ppc?

This crap vampire is really disappointing.

@ppcamiga1

they always talked about PPC is future and now after many years my PC runs several times faster at a fraction of the costs, they even sell embedded hardware for desktop. The crap PPC is disappointing

(sorry at the PPC fans, I could not resist at our hero)

@OlafS25

Agreed the price to performance ratio isn't a very useful discussion. But it's not like they can do much else at this point. The embedded PPC parts from NXP are quite performant and use very little power. It's either that or cram a full fat POWER CPU on an AmigaOne which would run circles around any household PC, but which would be way to costly and which would have an OS that can, as of yet, not use it's massively parallel nature.

I'd say the future of AOS4 would still be with PPC, however that will play out in the future, hard to say. But I can see why the Vampire has so much appeal and if they really succeed in making a very performant m68k platform, then by all means I wish them a lot of success.

@ppcamiga1

Noooo, not you again... uffff...

Do you sell PPC-Stuff or what?

Please stop writing nonsense.

_________________
Sometimes we give people a lot of credit just because they’re writing nice sentences even if it isn’t adding up to much.

@ppcamiga1

We don't need a useless 68k vs PPC war.

Some of you lowlife trolls never stop to amaze me... what's inside your skull? Dust? Garbage? Quantum singularity? I bet everything but a working brain.

Douche.

@ppcamiga1

Quote:

ppcamiga1 wrote: Many years ago natami was announced as fast and cheap alternative to ppc. 68k faster than 200 MHz G4 for under 100 E will be something super cool. After many years of lies there is vampire, for 280 E plus another 100 E for C= hardware you get 68060 performance. Five time worse price performance ratio than ppc? This crap vampire is really disappointing.

The Apollo cores are a continuation of the NatAmi. The SuperAGA core will need some simplification to be useful in a Vampire but the CPU core is the same core that originated as the N68050, progressed into the N68070 design, and now is the AC68080. And in case you haven't heard, the Gold2 core is much faster than the Gold1 core and its AMMX vector unit is proving to be as useful in real-world applications as the integer operations of AltiVec.

@WolfpackN64

I did not want to start a fight against PPC fans here. I know that some people like it even if I do not share that. But talking about making sense and mentioning PPC in 2016 as the light in the darkness is simply stupid, PPC today makes as much sense as 68k in todays world... zero. I do not know much about POWER CPUs from IBM but IBM still uses them for servers so they make sense there. But nobody here is interested to create a server farm so I do not think they make any sense to us. Freescale obviously is dropping PPC now so in a couple of years there will be no PPC CPUs left that are in production. The future of 4.X will stay PPC I assume but not because PPC has a future but Hyperion has no resources to port it.

@OlafS25

Freescale was already bought up by NXP, which will now be bought up by Qualcomm, the future for the old Freescale Power embedded line is indeed looking grim.

But I'm still positive better alternatives might emerge from the OpenPOWER initiative, which has been gathering steam in the past years. I wouldn't be surprised to see lower powered POWER chips emerging again since they're near ubiquitous in military applications.

And yes, Hyperion has no real resources to port the OS, but that's what you get with a not really that portable system. That's the nature of the classical OS' and it's kinda what makes an Amiga an Amiga. If it would run on everything under the sun, it wouldn't be computer family anymore.

@Samurai_Crow

Quote:

Samurai_Crow wrote: The Apollo cores are a continuation of the NatAmi. The SuperAGA core will need some simplification to be useful in a Vampire but the CPU core is the same core that originated as the N68050, progressed into the N68070 design, and now is the AC68080. And in case you haven't heard, the Gold2 core is much faster than the Gold1 core and its AMMX vector unit is proving to be as useful in real-world applications as the integer operations of AltiVec.

According to Gunnar it is a new core design? The pipeline design and workings that have "leaked" (for some reason even things like instruction encoding is kept secret for now) indicate that it isn't an improved "68050" rather than a new design.

Comparing MMX to Altivec is funny BTW, the differences in computing power is significant.

@megol

The pipeline design has improved much since the days of the NatAmi. Also, AMMX has more registers than Intel's MMX: 24 dedicated vector registers and direct access to the data registers for a maximum of 32. As the Gold 2 core becomes available with documentation, the capabilities of the Vampire board will become apparent to non-members soon.

Duplicate post.

Last edited by Samurai_Crow on 03-Jan-2017 at 08:15 PM.