@m0lebrain

FPGA's are too slow.. for 1.1Ghz.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@m0lebrain

Quote:

Wow...that's lots of money! Can't the whole thing be plopped into and FPGA? That would be cool.

That would be one way of making the Sam460 appear to be inexpensive hardware.

@Thread

Yay! PowerPC power!!!! Though somehow less exciting than a 080 FPGA 68k core/PiAmiga or a mini console Arm based product! Who knew?

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@amigang

Quote:

amigang wrote: I mean Acube might of just had some existing stock and thought it was easier to make a few of these than the A1222 Plus

It’s a new form factor motherboard so it’s kind of a new development. If it was about getting rid of a small stash of unused components, then it would have been easier to just throw them in the trash. The Sam460 was never free of issues and it’s so embarrassing out of date now anyway so it would have been more morally right to just move on. Come on, €1000+ for a system based on this is shameful.

Acube is AFAIK involved in building the open PPC Laptop based on the T2080 CPU. A MicroATX motherboard based on that (at half this price) would actually have been kind of interesting. They should have done that instead. A CPU with e6500 cores and Altivec at 1.8GHz, 2MB L2 and much better connectivity. That would still be completely out of date of course, but it would at least make those 1.5 decade old G4 Macs sweat a little. Sam460 is more in stone age G3 territory performance wise. And at this price!

m0lebrain Quote:

Wow...that's lots of money! Can't the whole thing be plopped into and FPGA? That would be cool.

Rob Quote:

That would be one way of making the Sam460 appear to be inexpensive hardware.

A FPGA with the same capabilities as the PPC460EX SoC ASIC would likely have roughly 1/10 the CPU performance and may cost 10x as much to produce.

TRIPOS Quote:

It’s a new form factor motherboard so it’s kind of a new development. If it was about getting rid of a small stash of unused components, then it would have been easier to just throw them in the trash. The Sam460 was never free of issues and it’s so embarrassing out of date now anyway so it would have been more morally right to just move on. Come on, €1000+ for a system based on this is shameful. Acube is AFAIK involved in building the open PPC Laptop based on the T2080 CPU. A MicroATX motherboard based on that (at half this price) would actually have been kind of interesting. They should have done that instead. A CPU with e6500 cores and Altivec at 1.8GHz, 2MB L2 and much better connectivity. That would still be completely out of date of course, but it would at least make those 1.5 decade old G4 Macs sweat a little. Sam460 is more in stone age G3 territory performance wise. And at this price!

There is quite a difference in technology between the original G3 and PPC 460.

1997 PowerPC 740 & 750 (G3) 260nm process
2006 PPC460EX SoC ASIC 90nm process
2012 Espresso (G3 in Nintendo's Wii U) 45nm process

The PPC460EX has an on chip L2 cache and supports PCIe while the original G3 had external L2 and PCIe wasn't introduced until 2003. The PPC460EX is based on a lower power PPC440 core embedded design while the G3 is a more balanced (for performance, power and area) design. The PPC460EX SoC has nice features and is a professionally designed embedded SoC developed by IBM although there are some cheaper cutdown embedded design choices like reusing the SATA and PCIe x1 SerDes lanes so they can't be used at the same time. The PPC460EX SoC doesn't have a builtin GPU so expensive PCIe SerDes lanes are required for graphics while limiting the total number of SerDes lanes keeps costs down. The most modern G3 based Espresso SoC for the Nintendo's Wii U has a AMD GPU but the
Radeon R600/R700 chipset is too old for Warp3D Nova support (PS3 uses Nvidia chipset) so harvesting from dead consoles is less attractive. That's pretty much it unless having a custom PPC SoC made with a GPU but I don't think it would sell well enough to mass produce. The rev6 68060s using a 420nm process and only running at about 100MHz are selling for several hundred dollars by themselves and are fast movers. Vampires are selling with 68k cores at less than 100MHz and aren't particularly cheap anymore. The original Amiga chipset used a 5000nm process and old original Amiga prices have been on the rise even as they are failing from old age, not unlike the users. The 68k AmigaOS 3 is outselling AmigaOS 4. Stay the course with PPC though and keep ignoring that iceberg ahead.

No RAM oboard, no GFX, no IDE controller, slow CPU. High price. Rather old design. And its another prepaid thing which you have to beliefe in and you dont really know how the experience of the acutual product will be *if* it really appears. It will be also limited in numbers. So if anyone buys and gets one: have fun.

Last edited by LarsB on 17-Feb-2022 at 07:49 AM.

@matthey

Quote:

A FPGA with the same capabilities as the PPC460EX SoC ASIC would likely have roughly 1/10 the CPU performance and may cost 10x as much to produce.

I don't know the exact speeds and prices but that was essentially my point.

@AmigaOldskooler

New SAM460 version? Nice!

Price is somewhat high, but that is not unusual in the world of Amiga. I will probably wait until the end of this year (Tabor, I´m still waiting for you! ) and then decide, which computer to purchase. Although my current NG machine is in some aspects better (CPU), nearly 20 years old hardware can die any day and I´m now too used to a non-emulated NG speed to return back.

@matthey

I find it a little ridiculous to compare performance between two extremely obsolete CPU’s. But if we are to talk G3 level CPU’s then a 750GX should come out as a winner at the same speed in an OS4 context performance wise, thanks to its on-die 1 MB L2 cache (the 460EX has only 256k L2). It’s popular to talk about improved general performance from SATA and PCIe, but in an OS4 context that shouldn’t even be noticeable at all since the amounts of data transfered in general usage is so small that it’s hardly even measurable compared to, say, Windows or MacOS X that constantly maps around gigabytes of virtual memory on the HDD. Besides, the standards used are *also* very old and obsolete. The main point in even having a PCIe gfx card at all, would be to be able to offload video playback to the GPU because the CPU’s are too weak to handle it. Forget about advanced 3D games. Nothing like that exists on OS4 and never will, and these CPU generations could never hope to saturate those newer generations GPU’s in 3D gaming the way it’s meant to be anyway.

The 750GX sold for the Amigaone Micro ran at 800MHz I believe (but can be overclocked) but the CPU itself existed in versions up to 1.1GHz. The 460EX used in Sam is also an 800MHz version that is overclocked if memory serves me right. And I recall discussions here on AW.net about the Sam460’s instability issues, where users reported how a combination of replaced heatsink with active fan, and downclocking it to the original 800MHz (or even underclocking it) improved the situation a lot.

And again, we are talking about a EUR €1000+ system here, with a 2000-2004 level performance in 2022.

Since they put in the efforts to do an ever so slight board redesign, and since they are involved in building the open PPC Laptop motherboard based on T2080, it would have been so much more interesting if they had set up a pre-payment scheme for a Micro-ATX version of *that* board instead, with a 1.8GHz CPU. It would have offered so much more. It could even have been very usable with Wayfarer and stuff like playing embedded youtube video on webpages etc. Not that OS4 will ever benefit from that, but I believe a MorphOS version would have been released for such a board. It would probably be noticeably faster and have other benefits over G4 Macs, so people would have bought it for MorphOS use exclusively (unlike X5000). I know at least one MorphOS developer having expressed some interest in the T2080 laptop project. But of course, MorphOS is always the black sheep, the dirty one, and it’s more “pure” to focus any HW development on dead-end OS4, since it’s teh reel!!1!

@TRIPOS

T2080 laptop is really interesting, I hope it will become something, I think it’s more interesting than the A1222, I’m sure A1222 can do some cool things, but be just too annoying having to recompile so many things to make it so.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

TRIPOS Quote:

I find it a little ridiculous to compare performance between two extremely obsolete CPU’s. ...

In the PPC Amiga time warp, the PPC460EX is closer to the A1222 than the original G3. NXP/Freescale ended active PPC development in 2013 and they were the last PPC developer.

1997 PowerPC 740 & 750 (G3) 260nm process
2006 PPC460EX SoC ASIC 90nm process
2008 QorIQ P1022 45nm process
2012 Espresso (G3 in Nintendo's Wii U) 45nm process

The PPC460EX and QorIQ P1022 are both cost reduced low power embedded designs while you want high performance cores but that does not make them obsolete. The QorIQ P1022 has better per core integer performance and is dual core but lacks the standard PPC FPU. The PPC460EX has nicer features even though the process difference makes it difficult to compete in performance. For low end computing and embedded markets, older processes are sometimes used to reduce cost although we are not seeing that here in the case of the PPC460EX (for example, the Raspberry Pi 3 SoC used a 40nm process). While an older process makes production cheaper initially, it may become more expensive as the process becomes obsolete if more production runs are desired. More likely though, the price depends on availability of existing chip stocks and how much the niche Amiga market will tolerate in markup of the final product.

TRIPOS Quote:

And again, we are talking about a EUR €1000+ system here, with a 2000-2004 level performance in 2022.

The performance/price is poor for sure, especially when comparing to low end hardware like Raspberry Pi products.

TRIPOS Quote:

Since they put in the efforts to do an ever so slight board redesign, and since they are involved in building the open PPC Laptop motherboard based on T2080, it would have been so much more interesting if they had set up a pre-payment scheme for a Micro-ATX version of *that* board instead, with a 1.8GHz CPU. It would have offered so much more. It could even have been very usable with Wayfarer and stuff like playing embedded youtube video on webpages etc. Not that OS4 will ever benefit from that, but I believe a MorphOS version would have been released for such a board. It would probably be noticeably faster and have other benefits over G4 Macs, so people would have bought it for MorphOS use exclusively (unlike X5000). I know at least one MorphOS developer having expressed some interest in the T2080 laptop project. But of course, MorphOS is always the black sheep, the dirty one, and it’s more “pure” to focus any HW development on dead-end OS4, since it’s teh reel!!1!

There was already the X1000 and X5000 with more performance and they were too expensive to sell well. The A1222 at least looks like an attempt to lower costs to expand the PPC AmigaOne user base. Something had to be sacrificed to make this possible which was a cheaper SoC, less expansion and smaller SBC board size. The value wouldn't have been horrible if e500mc cores were used with standard PPC FPU at perhaps half the cost of the current Sam460LE but without a standard FPU it's more like choose your performance handicap. It is difficult for PPC SoCs, lacking integrated GPUs, to compete with ARM SoCs which have integrated GPUs allowing for smaller and cheaper boards. They also have not only FPUs but SIMD units and multiple CPU cores. In some ways, the Vampire SA is further ahead as it is using a smaller board than either the Sam460LE or A1222 but it is handicapped because of using an affordable but cramped FPGA instead of ASIC.

Sam460LE
244mm × 244mm (microATX)
PCIe x4, x1

A1222
170mm × 170mm (Mini-ITX)
PCIe x4

Vampire SA
~142mm × 120mm

Raspberry Pi
85.6mm × 56.5mm

Raspberry Pi Zero (also $45 U.S. Amiga capable non-mass produced FleaFPGA Ohm)
65mm × 30mm

Raspberry Pi Pico
51mm × 21mm

The cost is roughly correlated with the board size (form factor) and all the smaller sizes have integrated GPUs of some kind except for the RPi Pico. The most competitive Amiga hardware maybe closer to the FleaFPGA Ohm using a 68k Amiga SoC ASIC with 10 times the performance of the Apollo core while cost reducing and mass producing the hardware.

Last edited by matthey on 18-Feb-2022 at 01:43 AM.

@NutsAboutAmiga

Yes I think the A1222 CPU is more trouble than it's worth. It's not like the 68K days were an FPU was optional. It's more like a ColdFire, as it is incompatible with standard CPU codes, but can run most of the common ones. So it needs software patching to work. I think this is the result of trying to keep it cheap. It's just too cheap and not effective in the long run. There wasn't any stocks of more standard PPC chips avail for a similar price I wonder.

But, it can also look a bit silly. At the end of day, all the CPU does is execute codes. An ARM executes similar codes being also a RISC chip. So it makes me wonder how it's so hard in this day and age to produce a chip that can execute PPC codes. Perhaps bolting on a PPC to ARM code converter would be easier and cheaper to produce than a decent PPC chip.

Or, call it a day after almost 30 years of PPC, as ten years is the limit for any CPU that has run it's course.

@matthey

Quote:

X1000 and X5000 with more performance and they were too expensive to sell well

actually i think the demand for the X1000 was quite strong, their was unfortunately some big issues, the development time it took to develop OS4 for X1000 was three years from X1000 first being giving the go ahead, a lot can happen in that time frame of CPU, and unlucky in the case of A-EON, Apple bought P.A. Semi and all projects/production effetely ended, as I understand PWRficient PA6T-1682M only got produce in small amount because of contract obligations more to the US military.

Because of the much smaller production run, cost per CPU was a lot higher than first thought and the amount available out strip demand so A-EON never was able to lower the cost of the nemo board and effetely all the great work and momentum they made with X1000 kinda had to be put on pause again for another 3 years for them to make the replacement the X5000.

I feel that was a big shame because i been recently looking back at the X1000 launch time and how much more buzz there was in the community for OS4 back then, I mean we also just had Timberwolf port as well as other big projects like Blender, AmiCygnix, QT, Hollywood, had updated ports, we got loads of games around then too.

I feel if they never had to stop producing the X1000 and maybe even managed to lower the price of that system, back in 2012, OS4 might of been in a much stronger position now. Plus if P.A. Semi where still in the PPC market they might of made better and cheaper PPC chips than the few companies that are left in the market now.

Who knows, we can only speculate, all I know the road OS4 market had to follow did not help its market. by the time the X5000 got out I think some interest and momentum had been lost and some promised project felt further away or unlikely to happen, (OS4.2, OpenOffice, Timberwolf update).

Last edited by amigang on 18-Feb-2022 at 03:33 PM.
Last edited by amigang on 18-Feb-2022 at 12:30 PM.

_________________
AmigaNG, YouTube, LeaveReality Studio

@Hypex



Quote:

Here lies the PowerPC architecture; it had a good run, adapted well to industrial and defence requirements and morphed into the Cell architecture and showed true potential through PA Semi's designs. All these gains are gone now and the Amiga community stands alone holding onto the embers of the embedded chips that once sat in industrial controllers and low energy appliances worldwide! This market is now Arm's and not even the open sourcing of the architecture by IBM can turn things around! "Goodbye PPC and thanks for all the fish!"

Last edited by BigD on 18-Feb-2022 at 12:21 PM.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@BigD

With 17-year-old G5 2.7 GHz being the fastest PPC processor in home computers. If nothing faster came since 2005, technology is dead for home computers. Sad but true.

@rzookol

And yet the IBM/Toshiba/Sony Cell and P.A. Semi's PA6T-1682M showed us a glimpse of alternate futures. I still marvel at the PS3 and the concept of a core CPU unit and multiple SPE units and its number crunching elegance. The PA6T-1682M was very efficient and no doubt inspired the 'Mx' Apple Silicon designs. Nothing out of the ashes of Motorola or IBM Power development have brought anything useful for the desktop in decades and when desktop development dried up so did the SoC designs!

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@BigD

PowerPC was used in printer, routers, in air craft Entrainment systems, blueray / dvd players, the military, it also used MRI’s and lots more that you won’t think of as PC, the low watt usage, high efficient, and reliability made it a nitche product.

PowerPC was never really focus for IBM, there interest was in Server market, the Power CPU. Freescale / Motroola as into embedded market, and Apple was in the Desktop market. They formed AIM alliance, that meant they were not competing they stayed out of etch others way, this Is what I expect drives the price up. I believe the CPUs were made in the USA, that drives price up as well. Compared to other chips made in Asia.

The ARM chip made into mobile market, and expanded quickly there, more and more multimedia was needed and like video, music, this made more and more useable in the embedded market, ARM was produced with less expensive license, In low-cost countries. With slave labor.

When IBM made PowerPC chips they designed the Power chip first, they made cut down version of it for desktop market, they did not design the chip to be the best desktop chip. That was a poor decision.

Simply put Apple was not IBM's priority, this why they switched to Intel.

The success of ARM is result of big investment from Google in Android for Smart phones, smart tv’s and Chromecast. Google is like Microsoft for Mobile phones.

Between ARM and PowerPC there is big cultural difference.

Last edited by NutsAboutAmiga on 18-Feb-2022 at 04:45 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@matthey

ARM was unless for desktop market until the M1 from Apple, that’s pretty recent event.

While Pi Pico, Pi Zero, offer nothing for the desktop market.
The CPU in PI 4, is a

Broadcom BCM2711 64-bit Quad-Core Cortex-A72 1.5 GHz SoC, its only slightly higher clocked then the AMCC460.

of couse we don't care about cores, not support by AmigaOS4.1, nor is ARM but, kind useless in this context anywhy.

https://en.wikipedia.org/wiki/ARM_Cortex-A72

As Desktop chip is pretty useless with not a lot of connectivity.

https://www.youtube.com/watch?v=x0wl3ZaOq_g

I go for M1 if I wonted desktop ARM system, BCM2711 is not design for it.

https://www.mobileshop.eu/apple/desktops/mac-mini-2020-256gb-8gb-ram-mgnr3/

so 2 years old mac cost about same as Sam460, only problem no PCIe slots


I guess ARM is perfect for Apple as they want there customer to buy new computer, instead of upgrading the hardware.

Last edited by NutsAboutAmiga on 18-Feb-2022 at 05:37 PM.
Last edited by NutsAboutAmiga on 18-Feb-2022 at 05:17 PM.
Last edited by NutsAboutAmiga on 18-Feb-2022 at 05:15 PM.
Last edited by NutsAboutAmiga on 18-Feb-2022 at 05:11 PM.
Last edited by NutsAboutAmiga on 18-Feb-2022 at 05:04 PM.
Last edited by NutsAboutAmiga on 18-Feb-2022 at 05:03 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@BigD

@Hypex

Yes it's sad. As stated, I liked the idea of PPC and still have a somewhat irrational love of the PS3's Cell probably due to it making 'The Last of Us' possible so late in the console's life! A genius design but hard to program for

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios