According to a PostDoc's paper, TSMC is much better than SAMSUNG 6s' battery efficiency

Discussion in 'iPhone' started by Prince134, Oct 14, 2015.

  1. Prince134 macrumors 6502

    Joined:
    Aug 17, 2010
    #1
  2. cableguy84 macrumors 6502a

    cableguy84

    Joined:
    Sep 7, 2015
    Location:
    Essex
    #2
    alot of evidence is now saying the tsmc performs better, regardless of anomalies like the toms hardware report
     
  3. Shazamadoo macrumors member

    Shazamadoo

    Joined:
    Sep 26, 2015
    #3
    All that article does is state he ran geek bench multiple times on both phones.

    Then creates a c++ algorithm for no apparent reason, maybe to make his statement seem relevant I suppose...
     
  4. ssls6 macrumors 6502a

    ssls6

    Joined:
    Feb 7, 2013
    #4
    That paper is ridiculous. You cannot use statistics to generate meaningful information from flawed tests. Any test can favor one subject more than another but NOT represent the real mission profile (actual use). This is especially true with accelerated testing......Borrowed for many car manufacturers "I put this engine on a dyno for the equivalent of 150k miles thus in 30 days I can simulate 12 years of ownership"....BS you can. All you can really say is it didn't break down in 30 days of dyno time.

    The real fact is this....if you bought a phone just to run geek bench until the battery is depleted, then the data is relevant to you. If you bought a phone to use as a phone, then it probably isn't.
     
  5. CraigGB macrumors regular

    Joined:
    May 30, 2010
    #5
    he didn't even run the test, he is using the data already gathered by that benchmark provided to him by John Poole, all that is a summary of what we already knew. there's nothing new here at all but now we have another topic. we already knew there is a massive difference between the two chips in the geekbench battery test benchmark. :p
     
  6. terraphantm macrumors 68040

    Joined:
    Jun 27, 2009
    Location:
    Pennsylvania
    #6
    Well it should shut everyone who kept saying we can't use the results since it's just "n=1"
     
  7. TechnicallyTee macrumors 65816

    TechnicallyTee

    Joined:
    May 14, 2013
    Location:
    Atlanta
  8. soupcan macrumors 6502a

    soupcan

    Joined:
    Nov 21, 2014
    Location:
    Netherlands
    #9
    Sigh... Another one of these "Chipgate" threads stating a source that uses Geekbench when we all know Geekbench puts the A9 at max load on all the cores which has a 99.9% probability of not happening in real life.

    He can call himself professor doctor master genius wizard all he wants, using a benchmark app to define real-world performance is just dumb and not representative of real-world performance.
    /thread.
     
  9. MaciMac100 macrumors 6502

    Joined:
    Oct 26, 2014
    #10
    Wrong. Check your facts.
     
  10. sunking101 macrumors 603

    sunking101

    Joined:
    Sep 19, 2013
    #11
    I have solved this with my own test. I emailed the Samsung CEO and asked him 'who makes the best chips?'. He replied that Samsung do.
    I then emailed the CEO of TSMC and asked him who makes the best chips. He replied that his company do. So it's a dead heat.

    *Unless you include the fish shop in the village where I used to live. For me they make the best chips. Period.
     
  11. penajmz macrumors 68030

    penajmz

    Joined:
    Sep 11, 2008
    Location:
    New York City
    #12
    Seriously?! Another one of these....? SERIOUSLY??
     
  12. CraigGB macrumors regular

    Joined:
    May 30, 2010
    #13
    seems legit, chipgate is over, you've solved it, both chips are old news apparently, now how do I get one of those fish shop chips in my phone? :D
     
  13. CNeufeld macrumors 6502a

    Joined:
    Nov 25, 2009
    Location:
    Edmonton, AB
    #14
    Did you read the page? All he shows is that the TSMC chips did better at the GeekBench test than the Samsung chips did. He didn't state anything about the "battery efficiency" (whatever that is). He even says that "for the most common tasks like browsing the web and even light graphics usage, the screen has the highest battery consumption, and thus differences in battery lifetime between models with TSMC and Samsung chip will be small". Doesn't even touch on how radio usage (cell, WiFi, or Bluetooth) will also chew up battery power the same for both chips, negating any CPU differences.

    C
     
  14. MaciMac100 macrumors 6502

    Joined:
    Oct 26, 2014
    #15
    Samsung should have replied: TSMC

    After all, they stole their FinFET technology from TSMC...

    Official: Samsung stole trade secrets from TSMC

    According to a report published Wednesday by Taiwanese trade publication DigiTimes, the court has determined that Liang Mong-song, a former senior director of research and development at TSMC, revealed TSMC’s trade secrets and patents related to its advanced FinFET process technology to Samsung Electronics.

    The report makes no mention of Apple, but the connection couldn’t be clearer: Samsung might have been able to leverage the stolen secrets to win orders for Apple’s next-generation ‘A9’ processor. Prior reports have posited that both Samsung and TSMC got to build Apple’s A9 chips on the advanced 14-nanometer FinFET process technology which uses entirely new three-dimensional transistors.

    http://www.idownloadblog.com/2015/08/26/samsung-stole-trade-secrets-from-tsmc/


     
  15. sunking101 macrumors 603

    sunking101

    Joined:
    Sep 19, 2013
    #16
    I thought TSMC's A9 chip was 16Nm?
     
  16. MaciMac100 macrumors 6502

    Joined:
    Oct 26, 2014
    #17
    TSMC uses second generation 16nm FinFET+ (FF+) on Apple A9.
     
  17. Prince134 thread starter macrumors 6502

    Joined:
    Aug 17, 2010
    #18
    because those are not related to his analysis. And we know those factors will mask the result, and being out of focus..

    You still dont believe there is difference?
     
  18. CNeufeld macrumors 6502a

    Joined:
    Nov 25, 2009
    Location:
    Edmonton, AB
    #19
    I believe that the statement Apple sent out (within 3% of each other), the test by Tom's Hardware and Ars Technica, as well as the article by Anandtech are likely closer to the truth for most users in a day-to-day use case than a single artificial benchmark.

    The sky is not falling.

    C
     
  19. Prince134 thread starter macrumors 6502

    Joined:
    Aug 17, 2010
    #20
    I am tired with your Another day to day, regular person, normal usage statement. You are Apple fan or not? They once told you "THINK DIFFERENT". But you only want to think normal. Why cant you show your respect to the true analysis?
     
  20. CNeufeld macrumors 6502a

    Joined:
    Nov 25, 2009
    Location:
    Edmonton, AB
    #21
    Why can't you show your respect to logic and reason?

    C
     
  21. soupcan macrumors 6502a

    soupcan

    Joined:
    Nov 21, 2014
    Location:
    Netherlands
    #22
    Oh, you're one of those guys. Apple fan or not, this "true analysis" is NOT what happens in real life which is what matters in the first place. Unless you're that guy that runs benchmarks all the time instead of, I don't know, USING THE PHONE LIKE EVERY SINLGE OTHER PERSON IN THE WORLD and then gets surprised when the as-you-called-them "regular person" comes up and says you're wrong. In this case, you're wrong. The TRUE analysis are the dozens of real-world tests of both the A9s which have proven that there is a minuscule difference in performance and battery life between the 16 nm FinFET+ TSMC A9 chip and the 14 nm FinFET Samsung A9, which is the case with all electronic devices since none of them are perfect and never will be. You can believe your synthetic benchmarks all day every day, and I'm not stopping you from doing so, but quit slapping everybody in the face because their opinion isn't suiting you. If you're not satisfied with the iPhone 6s and 6s+ because they have different chips in them, get an iPhone 6 or 5S or an Android device for that matter. Just stop spreading ******** around here.

    Mods, I recon you close this thread before everything goes ******* which will happen fairly shortly.
     
  22. Prince134 thread starter macrumors 6502

    Joined:
    Aug 17, 2010
    #23
    I do that's way I said.
     
  23. mbaran macrumors regular

    Joined:
    Jun 10, 2008
    #24
    You can't argue with Prince134, he's a troll extraordinaire.
     
  24. soupcan macrumors 6502a

    soupcan

    Joined:
    Nov 21, 2014
    Location:
    Netherlands
    #25
    I noticed.
     

Share This Page