benchmark new smartphones: (Old Thread)

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

Terry McCracken
Posts: 16465
Joined: Wed Aug 01, 2007 4:16 am
Location: Canada

Re: benchmark new smartphones:

Post by Terry McCracken »

Terry McCracken
Solrac1970
Posts: 15
Joined: Sat Jul 27, 2013 4:53 pm

Re: benchmark new smartphones: The 1.0 GHz perspective

Post by Solrac1970 »

Thanks for posting the LG 2 score. So it seems that the Note 2 will be faster than the Note 3 on Stockfish even though in every other benchmark the Note 3 wins.

Doesn't 64 bit systems run something like 1.6 times faster for certain chess programs than 32 bit systems? Does anyone know if Stockfish will be taking advantage of this in IOS or Android when they get it?
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: benchmark new smartphones: The 1.0 GHz perspective

Post by JuLieN »

Solrac1970 wrote:Thanks for posting the LG 2 score. So it seems that the Note 2 will be faster than the Note 3 on Stockfish even though in every other benchmark the Note 3 wins.

Doesn't 64 bit systems run something like 1.6 times faster for certain chess programs than 32 bit systems? Does anyone know if Stockfish will be taking advantage of this in IOS or Android when they get it?
Hello Carlos. If you read this thread you'll see that we already discussed that iOS' version of SF already makes good use of 64 bits CPUs (currently, the iPhone 5S is the only smartphone with a 64 bits CPUs). And yes, it's nearly twice as fast as anything else, despite running on only two cores and at a lower frequency.
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
User avatar
mclane
Posts: 18748
Joined: Thu Mar 09, 2006 6:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub

Re: benchmark new smartphones: The 1.0 GHz perspective

Post by mclane »

the note2 is still a very good mobile for chess. all this qualcom shit that even cheats in benchmarks (it made an antutu of 35.6xx) is IMO a step backwards.
we will have to wait how good the Nexus5 is. and how the next android kitkat performs. i read it is faster.
What seems like a fairy tale today may be reality tomorrow.
Here we have a fairy tale of the day after tomorrow....
Jhoravi
Posts: 291
Joined: Wed May 08, 2013 6:49 am

Re: benchmark new smartphones: The 1.0 GHz perspective

Post by Jhoravi »

JuLieN wrote:Hello Carlos. If you read this thread you'll see that we already discussed that iOS' version of SF already makes good use of 64 bits CPUs (currently, the iPhone 5S is the only smartphone with a 64 bits CPUs). And yes, it's nearly twice as fast as anything else, despite running on only two cores and at a lower frequency.
Your post can be easily misinterpreted that the jump from 32 to 64bits is responsible for the near two times speed. I just want to clarify that it's not. Though the jump to 64bit gives some 25% to 35% nps boost for Stockfish, its actually the better CPU architecture of A7 that is responsible for most of the gains. Most obvious is its extra large 128kb total L1 cache.
h1a8
Posts: 508
Joined: Fri Jun 04, 2010 7:23 am

Re: benchmark new smartphones: The 1.0 GHz perspective

Post by h1a8 »

The crazy thing is that Samsung made the A7 chip for Apple. Yet they chose to go the cheapest route of the snapdragon. The next Samsung generation of phones is supposed to have 64bit.
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: benchmark new smartphones: The 1.0 GHz perspective

Post by JuLieN »

h1a8 wrote:The crazy thing is that Samsung made the chip. Yet they chose to go the cheapest route of the snapdragon. The next Samsung generation of phones is supposed to have 64bit.
No, Samsung didn't make the chip. Apple did. Samsung is just the producer. They only receive the masks and melt the chips. So they have no idea how it works. Except if they reverse-engineer it. Which is forbidden and would be obvious.
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
h1a8
Posts: 508
Joined: Fri Jun 04, 2010 7:23 am

Re: benchmark new smartphones: The 1.0 GHz perspective

Post by h1a8 »

JuLieN wrote:
h1a8 wrote:The crazy thing is that Samsung made the chip. Yet they chose to go the cheapest route of the snapdragon. The next Samsung generation of phones is supposed to have 64bit.
No, Samsung didn't make the chip. Apple did. Samsung is just the producer. They only receive the masks and melt the chips. So they have no idea how it works. Except if they reverse-engineer it. Which is forbidden and would be obvious.
Isn't reverse engineering legal as long as one doesn't implement it or disclose the source?
User avatar
JuLieN
Posts: 2949
Joined: Mon May 05, 2008 12:16 pm
Location: Bordeaux (France)
Full name: Julien Marcel

Re: benchmark new smartphones: The 1.0 GHz perspective

Post by JuLieN »

h1a8 wrote:
JuLieN wrote:
h1a8 wrote:The crazy thing is that Samsung made the chip. Yet they chose to go the cheapest route of the snapdragon. The next Samsung generation of phones is supposed to have 64bit.
No, Samsung didn't make the chip. Apple did. Samsung is just the producer. They only receive the masks and melt the chips. So they have no idea how it works. Except if they reverse-engineer it. Which is forbidden and would be obvious.
Isn't reverse engineering legal as long as one doesn't implement it or disclose the source?
It's illegal if you are a competitor.

Chips are protected with an ad-hoc protection for semi-conductors, that protects the "topography" of the chip. (Most of intellectual property is covered either with copyright or patent protections, so this one is an exception, just like new varieties of plants are also protected with an ad-hoc protection.)

Of course, it's very difficult to prove that your competitor reverse-engineered your design...

So why doesn't Apple get its own chip foundry, you might ask ? That way they would get a much safer protection for their chips. Well, it's not that simple : "printing" a chip is also a very complicated process that involves a lot of research with patented processes. Apple chose Samsung to print their chips because Samsung masters the sub-20 nanometer processes, for which they invented robust technologies only very few of their competitors can compete with (Intel, to name a few). TSMC is still not up to the challenge, and that's why Apple had to chose Samsung. Apple can't easily build it's own factory because all those processes are either patented or would require enormous amounts of time and money to develop. There is still another hypothesis : buying a chip producer. A few years ago, when asked what Apple would do with all their money, Steve Jobs answered they were saving it for a very big deal. Many people understood that Apple was probably planning to buy a chip manufacturer. Four years later they still haven't spent a dime of their fat money stash. But you can be sure they're not happy to rely on Samsung to print their chips.
"The only good bug is a dead bug." (Don Dailey)
[Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]
h1a8
Posts: 508
Joined: Fri Jun 04, 2010 7:23 am

Re: benchmark new smartphones: The 1.0 GHz perspective

Post by h1a8 »

JuLieN wrote:
h1a8 wrote:
JuLieN wrote:
h1a8 wrote:The crazy thing is that Samsung made the chip. Yet they chose to go the cheapest route of the snapdragon. The next Samsung generation of phones is supposed to have 64bit.
No, Samsung didn't make the chip. Apple did. Samsung is just the producer. They only receive the masks and melt the chips. So they have no idea how it works. Except if they reverse-engineer it. Which is forbidden and would be obvious.
Isn't reverse engineering legal as long as one doesn't implement it or disclose the source?
It's illegal if you are a competitor.

Chips are protected with an ad-hoc protection for semi-conductors, that protects the "topography" of the chip. (Most of intellectual property is covered either with copyright or patent protections, so this one is an exception, just like new varieties of plants are also protected with an ad-hoc protection.)

Of course, it's very difficult to prove that your competitor reverse-engineered your design...

So why doesn't Apple get its own chip foundry, you might ask ? That way they would get a much safer protection for their chips. Well, it's not that simple : "printing" a chip is also a very complicated process that involves a lot of research with patented processes. Apple chose Samsung to print their chips because Samsung masters the sub-20 nanometer processes, for which they invented robust technologies only very few of their competitors can compete with (Intel, to name a few). TSMC is still not up to the challenge, and that's why Apple had to chose Samsung. Apple can't easily build it's own factory because all those processes are either patented or would require enormous amounts of time and money to develop. There is still another hypothesis : buying a chip producer. A few years ago, when asked what Apple would do with all their money, Steve Jobs answered they were saving it for a very big deal. Many people understood that Apple was probably planning to buy a chip manufacturer. Four years later they still haven't spent a dime of their fat money stash. But you can be sure they're not happy to rely on Samsung to print their chips.
Thanks, but I don't see where it says it's illegal if you are a competitor. I'm referring to reverse engineering only. Not using the technology in any way or form. What law says a competitor can't reverse engineer even if they don't reproduce or import the technology?