Poster | Thread |
PhantomInterrogative
| |
A1222 and CPU multiplier Posted on 21-Jun-2017 18:06:55
| | [ #1 ] |
|
|
|
Cult Member |
Joined: 10-Sep-2004 Posts: 809
From: The Interrogative Lair | | |
|
| Taking into consideration that ACube's 440 and 460 series had variations with different CPU multipliers, what would keep A-Eon from selling A1222's at 1.2, 1.4, and 1.6 GHz configurations? Does the design of the P1022 allow for such speed increases without damaging the chip (I am assuming it would need a larger heat sink and fan)?
_________________ I sold my SAM460ex lite... waiting for money to buy a Raspberry Pi... or a Classic A1000 with Buffee... or an A1222... and OS4.3 FE update 11 |
|
Status: Offline |
|
|
NutsAboutAmiga
| |
Re: A1222 and CPU multiplier Posted on 21-Jun-2017 19:36:52
| | [ #2 ] |
|
|
|
Elite Member |
Joined: 9-Jun-2004 Posts: 12818
From: Norway | | |
|
| |
Status: Offline |
|
|
Rob
| |
Re: A1222 and CPU multiplier Posted on 21-Jun-2017 19:39:12
| | [ #3 ] |
|
|
|
Elite Member |
Joined: 20-Mar-2003 Posts: 6351
From: S.Wales | | |
|
| @PhantomInterrogative
I expect that there's headroom for overclocking the CPU cores but I think that A-EON are likely to stick with stock speed. |
|
Status: Offline |
|
|
PhantomInterrogative
| |
Re: A1222 and CPU multiplier Posted on 21-Jun-2017 19:50:19
| | [ #4 ] |
|
|
|
Cult Member |
Joined: 10-Sep-2004 Posts: 809
From: The Interrogative Lair | | |
|
| @NutsAboutAmiga
I wouldn't dare to overclock it as a user. I was wondering if the manufacturer, UltraVarisys, might be able to clock the P1022 at 1.4 or 1.6? I just checked the fact sheet from NXP, which stated that the P1022 could get up to 2 GHz. At that speed, I would assume the bus speed would be the maximum 1055 MHz. The A1222 bus is supposed to be 400 MHz. What would keep A-Eon/UltraVarisys from clocking it higher?
At one time there were hints that there was going to be a A12xx series. Perhaps they would ship with the higher bus speed? OK, I'll stop speculating and dreaming.
P1022 fact sheet Last edited by PhantomInterrogative on 21-Jun-2017 at 08:30 PM. Last edited by PhantomInterrogative on 21-Jun-2017 at 07:53 PM.
_________________ I sold my SAM460ex lite... waiting for money to buy a Raspberry Pi... or a Classic A1000 with Buffee... or an A1222... and OS4.3 FE update 11 |
|
Status: Offline |
|
|
Rob
| |
Re: A1222 and CPU multiplier Posted on 21-Jun-2017 20:17:33
| | [ #5 ] |
|
|
|
Elite Member |
Joined: 20-Mar-2003 Posts: 6351
From: S.Wales | | |
|
| @PhantomInterrogative
Quote:
just checked the fact sheet from NXP, which stated that the P1022 could get up to 2 GHz |
I haven't seen that fact sheet. Can you point me in the direction of it?
Quote:
At that speed, I would assume the bus speed would be the maximum 1055 MHz. |
I thought the memory controller supported up to 667Mhz memory. You don't have have to up the bus speed to overclock a CPU core. |
|
Status: Offline |
|
|
tonyw
| |
Re: A1222 and CPU multiplier Posted on 22-Jun-2017 1:36:07
| | [ #6 ] |
|
|
|
Elite Member |
Joined: 8-Mar-2003 Posts: 3240
From: Sydney (of course) | | |
|
| @PhantomInterrogative
I think you'll find that, in common with other CPUs, there are P1022s and P1022s. Not every P1022 model will be reliable at the higher clock rates. Also, the PCB tracks are much more critical at higher clock rates.
UltraVariSys may have been asked to design a "basic", low-cost board, using the current CPU, while a faster version, using a faster CPU, would cost considerably more to design and source components for.
_________________ cheers tony
Hyperion Support Forum: http://forum.hyperion-entertainment.biz/index.php |
|
Status: Offline |
|
|
dooz
| |
Re: A1222 and CPU multiplier Posted on 22-Jun-2017 9:58:58
| | [ #7 ] |
|
|
|
Member |
Joined: 17-Jul-2013 Posts: 48
From: Unknown | | |
|
| Last edited by dooz on 13-Jan-2018 at 05:32 PM. Last edited by dooz on 22-Jun-2017 at 10:01 AM.
|
|
Status: Offline |
|
|
PhantomInterrogative
| |
Re: A1222 and CPU multiplier Posted on 22-Jun-2017 11:48:19
| | [ #8 ] |
|
|
|
Cult Member |
Joined: 10-Sep-2004 Posts: 809
From: The Interrogative Lair | | |
|
| @tonyw
That's a pretty good explanation.
@dooz
I also like a quiet system. Perhaps A-Eon needs to source some Gelid or Noctua silent fans?
_________________ I sold my SAM460ex lite... waiting for money to buy a Raspberry Pi... or a Classic A1000 with Buffee... or an A1222... and OS4.3 FE update 11 |
|
Status: Offline |
|
|
TRIPOS
| |
Re: A1222 and CPU multiplier Posted on 22-Jun-2017 14:35:43
| | [ #9 ] |
|
|
|
Super Member |
Joined: 4-Apr-2014 Posts: 1205
From: Unknown | | |
|
| @tonyw
Quote:
tonyw wrote: @PhantomInterrogative
I think you'll find that, in common with other CPUs, there are P1022s and P1022s. Not every P1022 model will be reliable at the higher clock rates. Also, the PCB tracks are much more critical at higher clock rates. |
IMHO, these kind of embedded SoC's shouldn't be overclocked. The manufacturer rates their CPU's for a reason. If anything you may even want to underclock a notch. I'm pretty sure that the unpredictable stability issues that some users (not all) had with the Sam 460 was a result of the overclocking. Acube tried many fixes, like firmware updates, they tried disabling the L2 cache, they tried putting on a fan, but AFAIK nothing really cured it and then they simply stopped selling the product.
It's an extremely weak CPU anyway by any desktop standards, and no overclocking will ever change this fact in any noticeable way. So no, overclocking won't make it fast, it would only risk stability.
Put stability first!
|
|
Status: Offline |
|
|
dooz
| |
Re: A1222 and CPU multiplier Posted on 22-Jun-2017 14:43:51
| | [ #10 ] |
|
|
|
Member |
Joined: 17-Jul-2013 Posts: 48
From: Unknown | | |
|
| Last edited by dooz on 13-Jan-2018 at 05:36 PM.
|
|
Status: Offline |
|
|
PhantomInterrogative
| |
Re: A1222 and CPU multiplier Posted on 22-Jun-2017 14:57:49
| | [ #11 ] |
|
|
|
Cult Member |
Joined: 10-Sep-2004 Posts: 809
From: The Interrogative Lair | | |
|
| @TRIPOS
Quite so. Better speed usually means a better processor. I was just curious if the P1022/A1222 could achieve a little more speed at a 3.5x or 4x multiplier. If the design doesn't allow it, there are other options for speed (X5000, used X1000, any old PC, etc.).
When I had my AmigaOne XE, I bumped my 800 MHz 7450 to 866 MHz. Although a small bump in speed, it was noticeable. If the P1022 could achieve 1.4 GHz, I think users would notice the difference even if small. Edit: After checking a little more on the NXP site, I saw that the P1022 runs between 667 MHz and 1.2 GHz.
Mind you, as of the present, the A1222 has not been released, doesn't have a release version of AmigaOS, does not have SPE recompiled software, does not have benchmarks, etc. Thus, we don't know if the 1.2 GHz P1022 is more than enough power for the average NG user or not. Last edited by PhantomInterrogative on 22-Jun-2017 at 06:56 PM.
_________________ I sold my SAM460ex lite... waiting for money to buy a Raspberry Pi... or a Classic A1000 with Buffee... or an A1222... and OS4.3 FE update 11 |
|
Status: Offline |
|
|
Xenic
| |
Re: A1222 and CPU multiplier Posted on 22-Jun-2017 17:11:52
| | [ #12 ] |
|
|
|
Super Member |
Joined: 2-Feb-2004 Posts: 1246
From: Pennsylvania, USA | | |
|
| @TRIPOS Quote:
Agreed. I had one of the first 800MHZ SAM440 systems which was unstable and crashed frequently until ACube released a utility to change clock speeds. My SAM was stable when I set it to 733MHZ. Some other 800MHZ SAM440 owners had no stability problems so I guess I was just unlucky. I would say use a stable clock speed and possibly include a method of changing the clock speed which the user can set at his own risk._________________ X1000 with 2GB memory & OS4.1FE |
|
Status: Offline |
|
|
Hypex
| |
Re: A1222 and CPU multiplier Posted on 23-Jun-2017 5:57:06
| | [ #13 ] |
|
|
|
Elite Member |
Joined: 6-May-2007 Posts: 11215
From: Greensborough, Australia | | |
|
| |
Status: Offline |
|
|
dooz
| |
Re: A1222 and CPU multiplier Posted on 24-Jun-2017 15:05:45
| | [ #14 ] |
|
|
|
Member |
Joined: 17-Jul-2013 Posts: 48
From: Unknown | | |
|
|
Last edited by dooz on 13-Jan-2018 at 05:37 PM.
|
|
Status: Offline |
|
|
pavlor
| |
Re: A1222 and CPU multiplier Posted on 24-Jun-2017 15:36:52
| | [ #15 ] |
|
|
|
Elite Member |
Joined: 10-Jul-2005 Posts: 9588
From: Unknown | | |
|
| @dooz
Thanks again!
For comparison (all OGR-NG, KOGE 3.1 Scalar = NO AltiVec):
SAM460EX 1.0 GHz: 10,124,948 nodes/sec Mac mini G4 1.5 GHz: 14,881,017 nodes/sec |
|
Status: Offline |
|
|
Spectre660
| |
Re: A1222 and CPU multiplier Posted on 24-Jun-2017 16:01:21
| | [ #16 ] |
|
|
|
Elite Member |
Joined: 5-Jun-2005 Posts: 3918
From: Unknown | | |
|
| @dooz
root@Sam460ex:/home/julian# '/home/julian/Downloads/dnetc521-linux-ppc-uclibc/dnetc' -bench
distributed.net client for Linux Copyright 1997-2016, distributed.net RC5-72 PowerPC assembly by Malcolm Howell and Didier Levet Enhancements for 604e CPUs by Roberto Ragusa RC5-72 Altivec and OGR assembly by Didier Levet Please visit http://www.distributed.net/ for up-to-date contest information. Start the client with '-help' for a list of valid command line options.
dnetc v2.9112-521-CFR-16021317 for Linux (Linux 4.9.34-sam460ex-jm). Please provide the *entire* version descriptor when submitting bug reports. The distributed.net bug report pages are at http://bugs.distributed.net/
[Jun 24 14:55:41 UTC] Automatic processor type detection found a PowerPC 460EX/GT processor. [Jun 24 14:55:41 UTC] OGR-NG: using core #0 (KOGE 3.1 Scalar). [Jun 24 14:56:05 UTC] OGR-NG: Benchmark for core #0 (KOGE 3.1 Scalar) 0.00:00:20.09 [11,708,929 nodes/sec] [Jun 24 14:56:05 UTC] OGR-NG benchmark summary : Default core : #0 (KOGE 3.1 Scalar) 11,708,929 nodes/sec Fastest core : #0 (KOGE 3.1 Scalar) 11,708,929 nodes/sec [Jun 24 14:56:05 UTC] RC5-72: using core #0 (MH 2-pipe). [Jun 24 14:56:24 UTC] RC5-72: Benchmark for core #0 (MH 2-pipe) 0.00:00:16.74 [3,562,624 keys/sec] [Jun 24 14:56:24 UTC] RC5-72: using core #1 (KKS 2-pipe). [Jun 24 14:56:43 UTC] RC5-72: Benchmark for core #1 (KKS 2-pipe) 0.00:00:16.94 [3,812,679 keys/sec] [Jun 24 14:56:43 UTC] RC5-72: using core #2 (KKS 604e). [Jun 24 14:57:02 UTC] RC5-72: Benchmark for core #2 (KKS 604e) 0.00:00:16.38 [3,641,814 keys/sec] [Jun 24 14:57:02 UTC] RC5-72: using core #5 (MH 1-pipe). [Jun 24 14:57:21 UTC] RC5-72: Benchmark for core #5 (MH 1-pipe) 0.00:00:16.48 [3,435,065 keys/sec] [Jun 24 14:57:21 UTC] RC5-72: using core #6 (MH 1-pipe 604e). [Jun 24 14:57:39 UTC] RC5-72: Benchmark for core #6 (MH 1-pipe 604e) 0.00:00:16.43 [3,457,248 keys/sec] [Jun 24 14:57:39 UTC] RC5-72 benchmark summary : Default core : #1 (KKS 2-pipe) 3,812,679 keys/sec Fastest core : #1 (KKS 2-pipe) 3,812,679 keys/sec [Jun 24 14:57:39 UTC] Compare and share your rates in the speeds database at http://www.distributed.net/speed/ (benchmark rates are for a single processor core)
_________________ Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card |
|
Status: Offline |
|
|
dooz
| |
Re: A1222 and CPU multiplier Posted on 16-Jul-2017 18:44:57
| | [ #17 ] |
|
|
|
Member |
Joined: 17-Jul-2013 Posts: 48
From: Unknown | | |
|
| Last edited by dooz on 13-Jan-2018 at 05:38 PM. Last edited by dooz on 16-Jul-2017 at 06:47 PM.
|
|
Status: Offline |
|
|
Spectre660
| |
Re: A1222 and CPU multiplier Posted on 16-Jul-2017 20:34:06
| | [ #18 ] |
|
|
|
Elite Member |
Joined: 5-Jun-2005 Posts: 3918
From: Unknown | | |
|
| @dooz Sam460ex
lame '/home/julian/Documents/AKsack.wav' LAME 3.99.5 32bits (http://lame.sf.net) Using polyphase lowpass filter, transition band: 16538 Hz - 17071 Hz Encoding /home/julian/Documents/AKsack.wav to /home/julian/Documents/AKsack.mp3 Encoding as 44.1 kHz j-stereo MPEG-1 Layer III (11x) 128 kbps qval=3 Frame | CPU time/estim | REAL time/estim | play/CPU | ETA 10529/10529 (100%)| 0:41/ 0:41| 0:43/ 0:43| 6.6006x| 0:00 ------------------------------------------------------------------------------- kbps MS % long switch short % 128.0 100.0 74.4 13.4 12.2 Writing LAME Tag...done ReplayGain: +0.5dB
Last edited by Spectre660 on 16-Jul-2017 at 08:34 PM.
_________________ Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card |
|
Status: Offline |
|
|
dooz
| |
Re: A1222 and CPU multiplier Posted on 16-Jul-2017 21:55:39
| | [ #19 ] |
|
|
|
Member |
Joined: 17-Jul-2013 Posts: 48
From: Unknown | | |
|
| Last edited by dooz on 13-Jan-2018 at 05:35 PM.
|
|
Status: Offline |
|
|
jabirulo
| |
Re: A1222 and CPU multiplier Posted on 16-Jul-2017 22:12:18
| | [ #20 ] |
|
|
|
Regular Member |
Joined: 20-Jun-2004 Posts: 370
From: Donosti (GUIPUZCOA) | | |
|
| @dooz
SAM460ex under AOS4.1FE update1
#appdir:lame AKsack.wav LAME 3.99.5 32bits (http://lame.sf.net) Using polyphase lowpass filter, transition band: 16538 Hz - 17071 Hz Encoding AKsack.wav to AKsack.mp3 Encoding as 44.1 kHz j-stereo MPEG-1 Layer III (11x) 128 kbps qval=3 Frame | CPU time/estim | REAL time/estim | play/CPU | ETA 10529/10529 (100%)| 0:51/ 0:51| 0:52/ 0:52| 5.2903x| 0:00 ------------------------------------------------------------------------------- kbps MS % long switch short % 128.0 100.0 74.4 13.4 12.2 Writing LAME Tag...done ReplayGain: +0.5dB #list AKsack.mp3 4401109 ----rw-d Hoy 23:15:25 AKsack.wav 48686444 ----rwed 28-Dic-12 18:11:45 Clipboards Dir ----rwed Hoy 23:01:37 T Dir ----rwed Hoy 23:14:29 #
#SDH1:Comm/dnetc -bench
distributed.net client for AmigaOS Copyright 1997-2009, distributed.net RC5-72 PowerPC assembly by Malcolm Howell and Didier Levet Enhancements for 604e CPUs by Roberto Ragusa RC5-72 Altivec and OGR assembly by Didier Levet Please visit http://www.distributed.net/ for up-to-date contest information. Start the client with '-help' for a list of valid command line options.
dnetc v2.9105-511-CTR-09050816 for AmigaOS (OS 4.1, PowerPC). Please provide the *entire* version descriptor when submitting bug reports. The distributed.net bug report pages are at http://bugs.distributed.net/ Using email address (distributed.net ID) 'javierdlr@euskalnet.net'
[Jul 16 21:17:49 UTC] Automatic processor type detection did not recognize the processor (tag: "OS4:0x13") [Jul 16 21:17:49 UTC] OGR-NG: using core #0 (KOGE 3.1 Scalar). [Jul 16 21:18:09 UTC] OGR-NG: Benchmark for core #0 (KOGE 3.1 Scalar) 0.00:00:17.07 [10,554,334 nodes/sec] [Jul 16 21:18:09 UTC] OGR-NG benchmark summary : Default core : #-1 (undefined) Fastest core : #0 (KOGE 3.1 Scalar) [Jul 16 21:18:09 UTC] RC5-72: using core #0 (MH 2-pipe). [Jul 16 21:18:28 UTC] RC5-72: Benchmark for core #0 (MH 2-pipe) 0.00:00:16.28 [3,224,022 keys/sec] [Jul 16 21:18:28 UTC] RC5-72: using core #1 (KKS 2-pipe). [Jul 16 21:18:47 UTC] RC5-72: Benchmark for core #1 (KKS 2-pipe) 0.00:00:17.06 [3,498,518 keys/sec] [Jul 16 21:18:47 UTC] RC5-72: using core #2 (KKS 604e). [Jul 16 21:19:06 UTC] RC5-72: Benchmark for core #2 (KKS 604e) 0.00:00:16.35 [3,328,016 keys/sec] [Jul 16 21:19:06 UTC] RC5-72: using core #5 (MH 1-pipe). [Jul 16 21:19:25 UTC] RC5-72: Benchmark for core #5 (MH 1-pipe) 0.00:00:16.97 [3,113,986 keys/sec] [Jul 16 21:19:25 UTC] RC5-72: using core #6 (MH 1-pipe 604e). [Jul 16 21:19:45 UTC] RC5-72: Benchmark for core #6 (MH 1-pipe 604e) 0.00:00:17.02 [3,105,071 keys/sec] [Jul 16 21:19:45 UTC] RC5-72 benchmark summary : Default core : #-1 (undefined) Fastest core : #1 (KKS 2-pipe)
Last edited by jabirulo on 16-Jul-2017 at 10:15 PM.
|
|
Status: Offline |
|
|