dnetc benchmarks

[amigadave]

@kas1e

Because the X1000 was created to run OS4 and to OS4 users it is the fastest machine the can get, that is the real substance of it all. I can't run OS4 on a mac so I don't give a damn about the PPC Mac I have in the garage gathering dust. If I could run on it and it was faster than the X1000 then obviously I'd prefer using a faster machine. But here we are proposing benchmarks which have nothing to do with OS4 and I run OS4. I hope it's clear.

The only relevance I can see in these tests is that they raise the old question "why did Hyperion abandon porting OS4 to the G4 MacMini (and then other G4 Mac hardware like MorphOS did) and provide OS4 users with cheap, easily found systems to run OS4 on?"  It can't be because MorphOS was already going down that path, as after the decision to not support the G4 MacMini was announced, support for the Peg2 was released.

I am happy that I have purchased my X1000 and it will arrive next week, but there are many OS4 users (and probably potential OS4 users too) that cannot afford an X1000 and might not be able to afford a SAM that could easily afford a G4 MacMini, or other G4 Mac hardware, if only Hyperion would port OS4 to such hardware.

I know that the soon to be released Netbook will fill the low price OS4 problem with hardware, but it will probably be as slow, or slower than the current SAM hardware and can't compete with the power of the used G4 Mac's.

So, Hyperion, please reconsider porting OS4 to G4 Mac systems.  It can only help grow your user base and increase your sales of OS4.  More users means the possibility of more developers too.  Please port OS4 to G4 Mac hardware for people that can't afford to buy new OS4 hardware but want the speed shown in the benchmarks being presented in these many threads.

[Iggy]

David, Daddy,
I wholly agree with you both.
I spent months poring over MPC8641, ULi 1541 and ATI SB600 data sheets.
I even started to enter initial information into some design software I have.

All the while I knew that any system based on what I was working on would be expensive to produce, difficult to sell and support, and have a very limited market.

And even after I'd given up on the idea, I found it tempting.

So I really DO understand Treavor.

And I admire Aeon's desicision on who to source the design to.

Varisys designers pointed me to the Qorlq line when I was still investigating the 86XX processors. They are some very knowledgeable people and I, for one, never doubted that the X1000 would get produced.

But I think we all have to admit this is a vanity project, not a sensible business venture (and I still envy them).

[zylesea]

at the end of the day who the hell cares?

All those ppl who want to have the fastest Amiga ever. While I am not a big fan of OS4 myself even I would have considered an X1000 if it were really fast and giving me a big computing power benefit over what  have already. Well, it doesn't. I think that's highly interesting and valuable information. Saves me $3000 at the end of the day.

[Akiko]

Has anyone did any benchmarks yet with Linux running on both cores?

It would be interesting to see the possible future Potential with SMP support for OS4x.

[zylesea]

I know that the soon to be released Netbook will fill the low price OS4 problem with hardware.

Do you really believe that netbook will come soon? I have my concerns on this. No public demonstration, nothing except the talk at Amiwest. And confronted with the question whether it will be a THTF/MTC 5121 based design or not only silence. It will. And hence it will come with all the 5121 weaknesses.
SSolie not even sure about the screen size is telling how serious that whole thing is. It is just a placeholder for "oh yes, we have a laptop, too", telling the crowd between the lines not to buy an "old crappy Powerbook" to run MorphOS on it.

[amigadave]

I remember the days, weeks, months, where you were discussing the possibility of producing a new PPC motherboard design Jim.  I also remember that the cost of such a project always was a big factor preventing it from happening or going forward.

Trevor just said (not a direct quote, but I am guessing he said something like this) "to hell with how much it is going to cost, we need a new, more powerful Amiga(One)".  Then he made it happen, and without him being a hardware engineer he had to depend on others to choose which parts to use.

Edit: @zylesea, "Soon" is a relative term and in Amigaland can mean a lot of different things.  If it arrives this year, it will be soon, compared to usual timelines from some people for projects.  If it arrives at all, it will be a welcome addition (by many OS4 users) and the first portable PPC AmigaOne system.  I also suspect that the announcement of the OS4 Netbook from Hyperion was not much more than a tactic from Hyperion to delay users from getting a G4 PowerBook and trying MorphOS, but I don't blame them or have any ill will toward them if that is the true reason for the timing of their announcement, as long as they deliver a product to the Amiga community to back up the announcement.  Nothing worse than companies that disappoint Amiga users by announcing something and then not delivering on their announcement.  It has hurt the community too many times in the past.  I may not like some things about Hyperion, but at least they have delivered most things they have announced, even if they have been late.  I think Hyperion's timing for announcements of products or features has been terrible and they should have learned by now to NOT announce something until it is closer to being finished, but that is just my preference and some people would rather know what direction Hyperion is going in, no matter how long it takes them to get there, instead of waiting for information just a few weeks before it is in the hands of the buyers.

[zylesea]

Has anyone did any benchmarks yet with Linux running on both cores?

It would be interesting to see the possible future Potential with SMP support for OS4x.

Well for a perfectly symmetrically balanced system you mag gain a speed increase of the factor lim 2. In evryday usage it depends quite a lot. Take that lame test. alme itself is nt multithreaded, but a nice frontend can just launch several instances of lame. On a system more or less idle execpt doing the lame encoding this could really yield to a speed increase of nearly the factor of 2. Thing is, you need an SMP system to benefit from the 2 cores in a balanced way, or an AMP system with quite some apps written to use the 2nd core if available.
SMP is out of question for OS4 as long as a high backward compability will not get discarded, AMP is rather easy possible, but requires Applications to support that (also possible, but they don't com efrom above like the Manna). I *highly doubt* AmigaOS 4.x will benefit seriously from the 2nd core before 2014.

[BuzzBrain]

So, Hyperion, please reconsider porting OS4 to G4 Mac systems.  It can only help grow your user base and increase your sales of OS4.  More users means the possibility of more developers too.  Please port OS4 to G4 Mac hardware for people that can't afford to buy new OS4 hardware but want the speed shown in the benchmarks being presented in these many threads.

This will never happen,  reason?
Well if hyperion ports OS4.x to mac G4 systems they will undermine the sale of X1000 and sam460. Why use a lot of money on a SAM or x1000, when they can buy a mac systms for less then nothing and get the same or better speed?

Only a minor group will use that amount on money just to get "new system" or the fpga chip on x1000.



Well, this is my opinion.

[Tripitaka]

And, btw, you're not plugging a Radeon HD4650 into an old Mac (even if you could, there's no BIOS support).

A very good point. Graphics cards are doing a lot of work these days and support for them is a  very relevant point when considering these systems against each other.

Come to think of it, let's see a benchmark using FPS framerates with the best graphics card each system supports.

[amigadave]

This will never happen,  reason?
Well if hyperion ports OS4.x to mac G4 systems they will undermine the sale of X1000 and sam460. Why use a lot of money on a SAM or x1000, when they can buy a mac systms for less then nothing and get the same or better speed?

Only a minor group will use that amount on money just to get "new system" or the fpga chip on x1000.

Well, this is my opinion.

First, it is the SAM440ep, 440flex and 460ex that have a fpga chip on their motherboards, not the X1000, which has an XCORE chip which is different than a fpga.

Unless Hyperion is getting a kick-back of some of the money made for each X1000, or SAM sale, besides the OEM license fee for OS4.x being included in the sale, which I don't think is the case, why would they care if cheaper alternatives for OS4.x hardware were available?  They ported OS4.x to the Pegasos2, which can be found for sale used cheaper than any SAM system and the Pegasos2 is faster than any of the SAM's (in most ways, if not all tests, IIRC).

I don't think I have ever read any reasons from Hyperion that make any sense for why they did not, and continue to not provide a port of OS4.x to any of the G4 Mac models.  Does anyone outside Hyperion know the answer for certain?  I would really like to know.  It does not make any sense for the MorphOS Dev. Team to port MorphOS to the X1000 or any of the SAM hardware, but there are a lot of Amiga users that would like to be able to run both OS4.x and MorphOS (and even AROS PPC) on the same computer and would buy a license for OS4, if only Hyperion would port OS4 to one of the better G4 Mac models.  Hyperion has no obligation to ACube, or A-Eon to protect their sales by not doing a port of OS4.x to any G4 Mac models.  Porting to a G4 Mac would probably have a negative impact on SAM sales, but probably not as much as you would think, as some buyers still just want a new computer opposed to used gear.  The X1000 is unique and would be negatively affected even less than the SAM's by a port of OS4 to any G4 Mac hardware.  With the limited supply and high cost of the PA6T CPU's, it is not known how many X1000's Trevor and AmigaKit plan on producing, so by the time it would take Hyperion to do a port of OS4 to any G4 Mac model, it is possible that A-Eon might be done producing X1000's by then and have moved on to a different design.  I wish for nothing but success for Trevor and Matthew Leaman of AmigaKit in their sales of X1000's and any future designs, but for the health and growth of OS4, I see a port of OS4 to G4 Mac computers as a good thing.  I apologize to Trevor and Matthew if my views are harmful to any present of future plans for hardware that they have.  As a matter of fact, Hyperion's own plans for the PPC Netbook for OS4.x at a price between $300 to $500 is just as big a threat, if not more so, than a port to used G4 Mac hardware.  Many OS4 users have already stated that they do not want used hardware and prefer to spend more on new hardware with a warranty, but many more of them would be interested in a G4 Mac port to obtain cheap AND fast OS4 hardware.

It is hard to tell just how many new OS4 licenses would be sold if Hyperion ported OS4 to one or more G4 Mac model, but the price of entry into the world of using OS4 has definitely been a limiting factor for many potential OS4 users in the past, and will continue to be that way until the day the $300 to $500 Netbook is released.  Even then, a $0 to $300 G4 Mac will still be cheaper and much faster option that will appeal to many potential OS4 buyers, with an added benefit (for some that see it as a benefit) is the ability to run both OS4 and MorphOS on the same computer and not need two different systems if you are interested in both PPC NG Amiga systems.

Not porting OS4 to any G4 Mac hardware has always puzzled me.  Since I own several G4 Mac's, I would like to see Hyperion port OS4 to one or more of them in the future (but not until they are finished optimizing OS4.2 for the X1000 and other already supported AmigaOne systems and completing work on porting to their Netbook (even though I have no plans to purchase a Netbook, or any other currently supported OS4 system).

Some people have an aversion to any Apple products, but the number of current and former Amiga users who like and use Apple products far outnumbers the few that have Apple phobia.

My interest is growing the Amiga community as large as possible, as fast as possible.  Including trying to win back as many developers as possible.  We need tens of thousands, if not hundreds of thousands of users again, if that will ever be possible?  How we get there is not my primary concern.  I just want more users and more developers and I include OS4.x, AROS, MorphOS and AmigaOS3.x on original Amigas and FPGA devices as well in my definition of what the Amiga community is.

Time to go back to lurking mode.

[Tripitaka]

Understanding Hyperion is not so hard. Just consider that they produce OS4 for the G4, Well that is by definition a niche, the market will never expand beyond the amount of G4s still in working order. A new machine has an infinite amount of market expansion available by comparison.

Now you can argue this any way you damn well like but if your trying to get people to part with capital to support you, that's as much as they needed to read before choosing which way to go. Whatever your national equivalent of Dragons Den* is should have taught you that.  Of course, once down the new machine route you can't compete with yourself by producing OS4 for the G4 later either.


*http://en.wikipedia.org/wiki/Dragons'_Den#Versions

[Iggy]

Thank God this thread has gotten back to a more civil discourse (instead of the "mine's better than yours" rut).
Each NG OS and its hardware offers specific advantages relative to one another.
The fact that there are this many choices is a testament to how much interest remains in the Amiga platform.

[Piru]

A very good point. Graphics cards are doing a lot of work these days and support for them is a  very relevant point when considering these systems against each other.

Come to think of it, let's see a benchmark using FPS framerates with the best graphics card each system supports.

Well, nothing new to see here, MorphOS still crushes AmigaOS4 in this department. In fact, the MPlayer video decoding benchmark deliberately excluded displaying the decoded video, as this would have seriously crippled the X1000 result (the Radeon HD driver doesn't yet support overlay).

[fishy_fiz]

Wow, my 4 year old budget amithlon box, running an emulated 68k cpu gets better rc5-72 results than A1X1000. Didnt expect that. I'll have to try the other benchmarks floating around and see how they compare.

Slight sidebar, but I was surprised to see that '060 versions not only work on the emulated '040 cpu core of amithlon, but also works faster than '040 versions.

Might have to try to compile lame and dnetc for amithlon x86 as well.

[amigadave]

Understanding Hyperion is not so hard. Just consider that they produce OS4 for the G4, Well that is by definition a niche, the market will never expand beyond the amount of G4s still in working order. A new machine has an infinite amount of market expansion available by comparison.

The amount of available used G4 Mac computers that are still in "working order" is in the millions, and far more by orders of magnitude than will ever be produced by ACube, or A-Eon, or anyone else, since producing PPC desktop motherboards from anyone else is about non-existent these days.  Also, the niche of Amiga OS4 users has great potential to grow several times larger than it currently is today, under the right circumstances.  Will it happen?  I don't know, but there certainly were hundreds of thousands of Amiga users in the not too distant past that are potential new users for the niche OS4 community of today and tomorrow.  Why not do everything possible to try to get more of them to return?

Now you can argue this any way you damn well like but if your trying to get people to part with capital to support you, that's as much as they needed to read before choosing which way to go. Whatever your national equivalent of Dragons Den* is should have taught you that.

Huh?  If you want people to spend money to return to using any kind of Amiga experience then the rules of giving them the most for their money still apply.  There are some that want only new computers, there are some that hate Apple and will never buy any of it, and then there are the rest that buying a used G4 Mac makes perfect sense to run OS4 on and if you want to have as many OS4 users as possible, you should try to satisfy all of the three groups and specially the largest of the three groups, which is probably the group that wants the best performance per dollar/pound/euro and would be interested in buying OS4.x for a G4 Mac computer.

Of course, once down the new machine route you can't compete with yourself by producing OS4 for the G4 later either.

Not true, even Hyperion has ported to used G4 computer AFTER they had already ported OS4 to the SAM440ep.  IIRC, the port to the Peg2 came after the port to the SAM440ep.  There is no valid reason to not port OS4.x to used G4 Mac computers that I can see.  It will only result in more sales for Hyperion and more OS4 users in total.

It could hurt ACube SAM sales and might hurt A-Eon's sales of the X1000 (but I doubt it because by the time Hyperion could finish a port of OS4 to any G4 Mac model, I doubt A-Eon will still be producing X1000's due to the cost and difficulty in obtaining PA6T CPU's, but I could be wrong).

What is more important to the OS4 users, having more fast and cheap hardware to tempt new users and developers with to join the party, or supporting ACube & A-Eon by only having the option to use OS4 on more expensive custom hardware?

I want to support OS4 hardware companies too, but not at the expense of limiting the community in it's attempt to gain more users and developers.

Maybe after a port to used G4 Mac models the community as a whole could concentrate more on getting more developers and software and worry less about what the future of OS4 hardware is for a while (but still keep exploring all different possibilities for future hardware to move to when all existing hardware starts to fail.  I plan on using my X1000 for the next 20 years:) )

Edit:  After further thought, my request for Hyperion to port OS4 to G4 Mac hardware will surely hurt A-Eon's future chances for any other hardware designs.  I really like Trevor and want him to succeed, but I still think the best thing for the OS4 users and developers would be to provide a port of OS4 to one or more G4 Mac models.  So, who does Hyperion support, the users, or the hardware designers/suppliers?  Tough choice.

Since I have already purchased my X1000, it is easy for me to suggest that the porting to G4 Macs is the better choice, now that I have all that I need for the next many years.  Others might resent this idea as they want their own X1000, or some later new design, and so do not want Hyperion to port OS4 to any G4 Mac's if that means that future OS4 custom hardware might then get canceled.

[Rodomoc]

Wow, my 4 year old budget amithlon box, running an emulated 68k cpu gets better rc5-72 results than A1X1000. Didnt expect that. I'll have to try the other benchmarks floating around and see how they compare.

I was wondering when this one was going to come up :laugh1:. It is possible to run more modern x86 boards now with the new kernel and its drivers. Not a perfect Amiga solution but Amithlon is still a damn interesting concept even after all these years. I need to get that 3Ghz Xeon cranked up in this kernel4 capable Asus board I have. At any rate, it would be interesting to see some similar performance comparisons (if possible) using Amithlon.

As for the published performance data in this thread, it is clear what will happen on my end. And that is registering a nice G4 laptop when 3.0 comes out. Power, portability, stability. I'm in on that deal. I see complaints about old Mac garbage, etc... Well that old garbage is still good, cheap, and running well against other PPC by the looks of it. The selection of Mac hardware by MOS developers was obviously a shrewd one. The resulting amount of supported hardware should speak for itself. And it will only get better. If Amiga developers were only 1/2 as shrewd, they would port over to Mac as soon as possible.

Regarding the new X1000 board, well....'best of luck to ya'.