By Sam on AWN.

Quote:

The RAM speed values are indeed impressive, but the MIPS value is way below expectation (should be 3960MIPS/core). The VRAM values are also not really overwhelming. Better drivers may change that.
Anyway, the MIPS value is really disappointing.
Is the benchmark program worging correctly?

[ Editiert durch Zylesea 31.01.2012 - 23:55 ]

And we have a thread on the performance on a machine that we know won't be used for MorphOS, why?

Just for fun?

pampers,
Quote:

But we're not here for the fun, are we?

pampers,
Quote:

For the lulz is an expectable acceptable answer.

[ Edited by stephen_robinson 01.02.2012 - 15:23 ]

Oh no. You guys ain't gonna start a red vs blue pissing match are you? It's immature.

PS I think we're gonna win because Morphos is cool and AOS4.x sucks

Quote:,

Since when has there not been discussions about potential PPC based hardware? This isn't a religion we are talking about where we don't want to offend people, it is a piece of computer hardware.

If the performance truly blew the Power Mac G5 out of the water and it turned out to be a true monster, i suspect the MOS team would at least consider it. This does not seem to be the case.

[ Edited by redrumloa 01.02.2012 - 07:50 ]

Impressive RAM-Speed, indeed.

If the performance truly blew the Power Mac G5 out of the water and it Quote:

Well expensive and limited hardware is not a good option no matter the performance. Why spend so much in X1000 when you could get a PowerMac G4 or G5 and run MorphOS in high performance?

>Well expensive and limited hardware is not a good option no matter the performance. Why spend so much in X1000 when you could get a PowerMac G4 or G5 and run MorphOS in high performance?

Uh, because the G5 isn't supported?

Maybe Morphosteam could support it because at least someone tries to keep the PPC operating system alive with new hardware ? It's not that these aren't getting sold. And I also see more crossplatform development these days.

But I can understand the decision not to support ... who knows what he plan is for MorphOS, if any.

I would love to be able to run MorphOS and OS4 on my X1000 some day, but I am pretty sure it will never happen and I am okay with that result.

Even if I could raise enough money to donate a free X1000 to 4 or 5 members of the MorphOS Dev. Team, I doubt they would agree to port MorphOS to it, or if they did, they would not release it for sale to anyone else. Something about supporting hardware that has OS4 bundled with it seems to make such a port impossible, and I think we all have read some of the reasons for that attitude at one time or another in the past.

So, no MorphOS for X1000 ever and I can live with that. Many MorphOS users are still interested in the X1000 hardware just out of curiosity, not because they would ever buy one. Not many MorphOS users are as crazy as I am. Impossible to explain the feelings that led me to purchase it, as it is definitely out of my price range and I could have used the money elsewhere very easily. At my age money just isn't as important any more though, so I am quite happy with my purchase of the First Contact X1000 computer. I never bought a new Commodore Amiga when they were being produced. Closest I came to that was getting a NOS CD32 from China, the last New NTSC A1200 that AmigaKit had in stock that they were selling (they had one more for their own use for testing third party hardware on), and buying a brand new, unopened CDTV from Carl Sassenrath at the 2010 AmiWest Show. Even though this is an AmigaOne, not a Commodore Amiga, it still "feels good" to be buying one of the first publicly available NG computer systems designed solely to run OS4.x. I don't recommend it for everyone, and doubt that hardly anyone on this forum will understand, or agree with my decision, or copy my actions, but that is also okay with me. I don't expect anyone here to agree with me.

I still remain a huge supporter of MorphOS and believe that it is still superior to OS4.x, but I have hopes that OS4 will continue to improve in the coming months and years ahead. I also hope that MorphOS continues to improve and move forward, because if OS4.x should ever catch or surpass MorphOS in capabilities, speed and features, I am not sure MorphOS could still survive. (probably not a popular thing to write on this forum, but since it is not likely to happen, I don't think anyone here need worry about it.)

Just for fun I ran ragemem via os4emu on my Powermac 3.4 with MaxPower G4 7448 @2GHz installed. Here are the results:

RAGEMEM v0.37 - compiled 11/06/2010

CPU: Unknown 0.0 @ 1999 Mhz
Caches Sizes: L1: 16 KB - L2: 1024 KB - L3: none
Cache Line: 0

---> CPU <---
MAX MIPS: 5982

---> L1 <---
READ32: 7606 MB/Sec
READ64: 15209 MB/Sec
WRITE32: 6518 MB/Sec
WRITE64: 5072 MB/Sec

---> L2 <---
READ32: 2166 MB/Sec
READ64: 2560 MB/Sec
WRITE32: 2445 MB/Sec
WRITE64: 3348 MB/Sec

---> RAM <---
READ32: 276 MB/Sec
READ64: 276 MB/Sec
WRITE32: 804 MB/Sec
WRITE64: 810 MB/Sec
WRITE: 752 MB/Sec (Tricky)

---> VIDEO BUS <---
READ: 32 MB/Sec
WRITE: 180 MB/Sec

Side by side comparison's are easier to read.

Mac numbers--------------------------X1000 numbers
RAGEMEM v0.37 - compiled 11/06/2010 RAGEMEM v0.37 - compiled 11/06/2010

CPU: Unknown 0.0 @ 1999 Mhz ------ CPU: P.A. Semi PWRficient PA6T-1682M B1 @ 1800 Mhz
Caches Sizes: L1: 16 KB - L2: 1024 KB - L3: none -- Caches Sizes: L1: 64 KB - L2: 2048 KB - L3: none
Cache Line: 0 --------------------- Cache Line: 64

---> CPU <---
MAX MIPS: 5982 ------------------- MAX MIPS: 3084

---> L1 <---
READ32: 7606 MB/Sec ----------- READ32: 6851 MB/Sec
READ64: 15209 MB/Sec ---------- READ64: 13682 MB/Sec
WRITE32: 6518 MB/Sec ---------- WRITE32: 6851 MB/Sec
WRITE64: 5072 MB/Sec ---------- WRITE64: 13681 MB/Sec

---> L2 <---
READ32: 2166 MB/Sec ----------- READ32: 3276 MB/Sec
READ64: 2560 MB/Sec ----------- READ64: 4784 MB/Sec
WRITE32: 2445 MB/Sec ---------- WRITE32: 2531 MB/Sec
WRITE64: 3348 MB/Sec ---------- WRITE64: 4090 MB/Sec

---> RAM <---
READ32: 276 MB/Sec ------------ READ32: 2857 MB/Sec
READ64: 276 MB/Sec ------------ READ64: 4000 MB/Sec
WRITE32: 804 MB/Sec ----------- WRITE32: 2732 MB/Sec
WRITE64: 810 MB/Sec ----------- WRITE64: 3383 MB/Sec
WRITE: 752 MB/Sec (Tricky) ---- WRITE: 352 MB/Sec (Tricky)

---> VIDEO BUS <---
READ: 32 MB/Sec --------------- READ: 15 MB/Sec
WRITE: 180 MB/Sec ------------- WRITE: 161 MB/Sec

(formatting might be messed up above)

[ Edited by amigadave 01.02.2012 - 11:39 ]

@ Ruud, amigadave

Interesting comparison there...

interesting indeed...

Interesting values. The RAM speed of the X1000 remains pretty awsome, but I really wonder what they did with teh MIPS value. Also from paper numbers your 7448 2GHz should deliver 4600MIPS (2.3MIPS/MHz) only, I wonder where the "additional" 1382 MIPS come from. From paper numbers the PA6T sould deliver 2.2 MIPS/MHz/core, i.e. the 1.8GHz version should deliver 3960MIPS..

The lame test of http://forum.amigaone.pl/topic65.html shows numbers that are much closer to my expectation. X1000 about twice the speed of a Mac mini. Waiting for more numbers...

Quote:

...at about 10x - 40x the cost?

Well, speed increasement is not a linear function of monetary cost. Nevertheless, I think X1000 is waaaay overpriced and has its weaknesses. But one must stay fair: Ram speed *is* impressive. And while G5 support is not there yet/on hold for MorphOS, the X1000 is probably the fastest ppc Amiga yet. But the numbers presented so far make me feel the X1000 will have a pretty hard time once G5 support will come.But since the two more weeks things isn't much of my cup of tea, but rather am all for comparing what's actually here and now, one must note that G5 support isn't there whiel the X1000 is there now (okay, more than 1.5 years later than announced, but, well at last it came).

[ Editiert durch Zylesea 02.02.2012 - 00:50 ]

amigadave,
Quote:

Dave,

I think we should do as you have done, personal sacrifice to help the MOS community. We have a very special OS with lots of wonderful features. 3.0 will breath lots of fresh air into the scene for us, and we might all be surprised how many new users we pick up. I have lots of ex Amiga now Mac guys that will buy MOS for their PB's for the fun of using an Amiga like system on their HW. There is a lot of Amiga users that went to Mac in the mid to late 90's that are a real target market we still can bring back. The biggest problem as I see it, is just finding the time to push the OS and finding old Amiga buddies that still have spare time to enjoy MOS.

Can someone explain in layman's terms the RAM write (tricky)?

It looks like the X1000 RAM read and write speeds are almost 10 times faster until you get to the write (tricky) figure where it is a little more than half of what the 2GHz G4 shows.

> the MIPS value is way below expectation (should be 3960MIPS/core).
> [...] the MIPS value is really disappointing. Is the benchmark
> program worging correctly?

Yes, it is. You're just confusing MIPS (PA6T in this benchmark: 1.7 per MHz and core) with DMIPS (PA6T: 2.2 per MHz and core).

> I really wonder what they did with teh MIPS value. Also from paper
> numbers your 7448 2GHz should deliver 4600MIPS (2.3MIPS/MHz) only,
> I wonder where the "additional" 1382 MIPS come from. From paper
> numbers the PA6T sould deliver 2.2 MIPS/MHz/core, i.e. the 1.8GHz
> version should deliver 3960MIPS..

See previous posting.

I disagree about possibility to port MOS to the x1000. It would require time and resources that the MOS Team could better spend to work on MOS release for the PowerMac G5 instead.
MOS Team is not an enormous team, we all know this, so every hardware added to the MorphOS compatibility list take them about a couple of years (we're still waiting for Powerbook support..) so it would be very frustrating to see them waste a couple of years of development to support an overpriced hardware that almost nobody here would buy! It has no sense. I would prefer them to spend their development time to support hardware that we all can buy (read: PowerMac G5s..)