Active Topics
-
Maemo repository (3)
to Community by torsen - 2 hrs, 44 mins ago -
Camera phone competition March 2023 "Endless" (29)
to General by Maemish - 3 hrs, 2 mins ago -
Error after installing Leste on sdcard (11)
to Maemo 7 / Leste by Maemish - 1 day ago -
TLS 1.2 for N9 is ready (27)
to MeeGo / Harmattan by smatkovi - 1 day, 5 hrs ago -
Introducing ubiboot N9 (multiboot OS loader) (1,393)
to Alternatives by smatkovi - 2 days, 6 hrs ago - more...
|
2010-04-05
, 11:49
|
|
Posts: 36 |
Thanked: 8 times |
Joined on Apr 2010
|
#892
|
Originally Posted by javispedro
Neither source gives any argument why 125 MhZ may be unstable. Only "sayings" ("I've heard it may be unstable").
125Mhz comes from https://bugs.maemo.org/show_bug.cgi?id=7116 (as you can see, testing firmwares had scaling_min_freq set to 125Mhz). There has also been comments about that on IRC
Originally Posted by javispedro
This seems to be a little vague, too.
720Mhz comes from the 3530 datasheet, which is public. It also indicates it should be run at the same core voltage as 600Mhz, which may or might not apply to the 3430...
Actually, seems that only applies to newer 3530s.
Only real argument ist some slide from the posting some postings above this one... thin, really thin.
So I would summarize: No facts against higher clockspeeds - and no facts in favour of higher clockspeeds
 |
|
2010-04-05
, 11:52
|
|
Posts: 2,355 |
Thanked: 5,249 times |
Joined on Jan 2009
@ Barcelona
|
#893
|
Originally Posted by Matan
For the 3530 (which I even mentioned in my post!), yes.
Unfortunately for the fear mongers, the expected lifetimes are public. They are even reported in this thread. You even thanked that post: http://talk.maemo.org/showpost.php?p...2&postcount=91
|
|
2010-04-05
, 11:53
|
|
Posts: 14 |
Thanked: 4 times |
Joined on Mar 2010
|
#894
|
Just oc'd from 800mhz to 900mhz. 900mhz performs way better multitasking than the 750mhz oc. See the video.
http://www.youtube.com/watch?v=bdvhHya3A6c
http://www.youtube.com/watch?v=bdvhHya3A6c
 |
|
2010-04-05
, 12:02
|
|
Posts: 844 |
Thanked: 521 times |
Joined on Jan 2009
@ UK southampton
|
#895
|
had mine on 900/125 for 12 hours now all gravy so far.
One thing though in conky it never seems to drop below 500mhz even with 2 percent usage.
One thing though in conky it never seems to drop below 500mhz even with 2 percent usage.
 |
|
2010-04-05
, 12:02
|
|
Posts: 1,559 |
Thanked: 1,786 times |
Joined on Oct 2009
@ Boston
|
#896
|
Originally Posted by Matan
Way to post a sensible, objective response.
Unfortunately for the fear mongers, the expected lifetimes are public. They are even reported in this thread. You even thanked that post: http://talk.maemo.org/showpost.php?p...2&postcount=91
I'll summarize (again) the data: the CPU is supposed to work for 100000 hours if not more than 23000 hours of those are at the highest voltage level (the one used for 600MHz in the standard kernel and for higher speeds in the overclocked kernels). In other words:
If you stress the CPU for 6 hours every day, it is supposed to hold on for 10 years. Even if the CPU is locked always on 600MHz it should work for 5 years.
Please note that TI talk about voltages, not frequencies.
In my kernel (which already works on my device with no problems for a few months), I use OPP3 for 600MHz and OPP4 for above that, and it seems stable, so I even lower risks then with Nokia's default configuration.
It's not "fear mongering" to pass on warnings given by Nokia engineers to enthusiasts given free phones. They know the design limitations of the phone and and chipset combination better than any of us, after all.I appreciate your information, but the presentation was unneccessarily subjective (to the point of factual misstatements).
__________________
Unofficial PR1.3/Meego 1.1 FAQ
Unofficial PR1.3/Meego 1.1 FAQ
***
Classic example of arbitrary Nokia decision making. Couldn't just fallback to the no brainer of tagging with lat/lon if network isn't accessible, could you Nokia?
|
|
2010-04-05
, 12:03
|
|
Posts: 36 |
Thanked: 11 times |
Joined on Nov 2009
|
#897
|
"With the ability to scale in speed from 600MHz to greater than 1GHz, the Cortex-A8 processor can meet the requirements for power-optimized mobile devices"
Source: http://www.arm.com/products/processo.../cortex-a8.php
It was said hundret times that Cortex A8 was made with great scalability of speed in mind. Long live OCed N900's
Source: http://www.arm.com/products/processo.../cortex-a8.php
It was said hundret times that Cortex A8 was made with great scalability of speed in mind. Long live OCed N900's
|
|
2010-04-05
, 12:09
|
|
Posts: 42 |
Thanked: 7 times |
Joined on Nov 2009
|
#898
|
Ok, stock 600mhz phone cpu life is 10 years.
1000mhz phone cpu life is 3-4 years? But very better multi-tasking and fast response.
After 1-2 years coming dual-core cpu, better ram and very better display.
1000mhz phone cpu life is 3-4 years? But very better multi-tasking and fast response.
After 1-2 years coming dual-core cpu, better ram and very better display.
|
|
2010-04-05
, 12:09
|
|
Posts: 2,355 |
Thanked: 5,249 times |
Joined on Jan 2009
@ Barcelona
|
#899
|
Originally Posted by miwalter
Of course not. What argument could I give you here? The only thing I can give you is that on previous firmwares, the minimum frequency was 125Mhz, and then it was upped to 250Mhz. You can guess that was done because:
Neither source gives any argument why 125 MhZ may be unstable. Only "sayings" ("I've heard it may be unstable").
a) It was unstable at 125Mhz
b) Power savings were negligible
c) Nokia is evil*
d) Nokia is incompetent**
e) All of the above
Choose your own option.
Originally Posted by miwalter
So a kernel hacker telling you that is a "thin argument"?
This seems to be a little vague, too.
You might say "so I know that the 3530 has its lifetime reduced by a half when working at OPP5, thus I'll consider that a similar clause applies to the 3430 and thus I'll do it". Fine. But at this point, the 3430 datasheet might say that either OPP5 is completely fine and does not reduce lifetime, or might say that OPP5 reduces lifetime to 1/18: we don't know.
Of course, does it matter to you? That's the issue here. You don't care, you want higher clock speed, you will replace the device in 6months either way (which btw is the average shelf time for $600 phones), ...
*If you think "Nokia is evil", consider that they did increase the maximum operating frequency of the N800 in a firmware upgrade when they found it safe to do. God, is it so hard to understand that Nokia is composed of people that also want to reach the device limits?
**If you think this is the case, we might even get Nokia to make it happen in the official kernels. Feel free to file a enhancement request!
Last edited by javispedro; 2010-04-05 at 12:15.
| The Following User Says Thank You to javispedro For This Useful Post: | ||
|
|
2010-04-05
, 12:12
|
|
Posts: 2,355 |
Thanked: 5,249 times |
Joined on Jan 2009
@ Barcelona
|
#900
|
Originally Posted by bartekxyz
The Cortex-A8 is not a physical thing, but a design (aka "a piece of garbled paper"). A design has a maximum theoretical frequency of operation. An implementation has a real one.
"With the ability to scale in speed from 600MHz to greater than 1GHz, the Cortex-A8 processor can meet the requirements for power-optimized mobile devices"
Source: http://www.arm.com/products/processo.../cortex-a8.php
I'd like to see one 3430 clocked at more than 720Mhz by a commercial manufacturer. Please find it for me!
 |
«
Previous Thread
|
Next Thread
»
| Thread Tools | |
|
All times are GMT. The time now is 20:29.
If same chipset is runnig by default at 800mhz when implemented by other manufacturers, Nokia must buy the same too...
What lifetime do you need? I spent first half year with a dummy device. With overclocking and PR1.2 (if I live enought to see the update) I will use n900 for no more of two years... and then I will overclock to 1050 and when burning thow it to the sea like Vikings.
The only thing is that now I have a nice phone, if it only lives two years, enough for me.