cpsro

About

Username
cpsro
Joined
Visits
931
Last Active
Roles
member
Points
3,702
Badges
1
Posts
3,286

Reactions

2.5KLike34Dislike253Informative

  • 2019 iPhone sticking with Face ID, but Android moving to in-screen fingerprint sensors

    cpsrocpsro
    We'll see if the screen sensor is secure enough to be sanctioned by the EMV consortium (Eurocard/MasterCard/VISA) for monetary transactions. The new druid phones might retain the old backside sensor for monetary transactions and use the screen sensor for convenience, just like they've had facial recognition for convenience but couldn't use it for monetary transactions because the druid implementations aren't secure enough.

    ronnwatto_cobra

  • Apple's 2018 iPhone may charge faster because of new material in Qi charging coil

    cpsrocpsro
    Thermodynamics are complex, but given all other factors equal, increasing the wattage of the coil also increases the heat generated in the circuit roughly exponentially with that linear increase in wattage...
    I believe you mean the heat generated increases with the square of the wattage, not exponentially with the wattage, assuming the voltage remains constant. This comes from Ohm's Law, where E = IR (voltage = current * resistance).  Also P = EI (power or wattage = voltage * current).

    Power lost as heat due to resistance, R, of the coil is P = I^2R.  If the voltage is doubled, the power (EI) can be kept constant by halving the current. The resistive power loss as heat (I^2R) is then one-fourth.  (This is the basis for efficient, long-distance transmission of electrical power using high-voltage power lines.) Alternatively, resistive power loss can be reduced linearly with decreasing resistance of the coil, such as by switching to a more conductive material.
    watto_cobra

  • Turkey's president calls for boycott of Apple products

    cpsrocpsro
    Putin welcomes Erdogan's switch to Android--it makes it so much easier to stay in touch. And any move away from the Allies is a good move for Putin.
    baconstangwilliamlondonjony0

  • Hands on with Apple's 15-inch 2018 MacBook Pro with i7 processor

    cpsrocpsro
    cpsro said:
    cpsro said:

    Geekbench 4 returned 4,884 for single core and 22,179 for the multi-core. [...] The two-year-old model earned a 20,908 while the 2018 base model 15-inch earned a 20,179.
    Which is it then:  22,179 or 20,179? My own testing at an Apple store repeatedly got 5,040 single-core and 22,500 multi-core on a base model 15".
    Geekbench browser shows the top-of-the-line 2016 2.9 GHz i7 model scores 4,305 and 14,137. That's very different from the 20K+ figure you're reporting.

    For the top-of-the-line 2017 3.1 GHz i7 model, AI strangely reports 4,360 single-core and 11,979 multi-core, which are far from what Geekbench browser shows (4,627 and 15,549). In my own testing of the top-of-the-line 2017 model, I got 4822 single-core and 15,652 multi-core.

    IMHO there are far too many inconsistencies and errors in this article to have passed editing, particularly when people are looking so closely now at performance and throttling.
    All of these numbers are well within the margin of error. The benchmarks apply to that particular run on that particular machine. Because the numbers don't match yours, doesn't mean that they're wrong. That's the nature of benchmarking.

    If every machine got the same reading every time, then there wouldn't be a range of benchmarks listed in the Geekbench database.
    Not true about your margin of error claim. You quote 22,179 in one paragraph and then 20,179 shortly thereafter for the same model. That looks like one of those figures is a typo... but which one?
    You quote 11,797 for the top-of-the-line 2017 15-inch MBP, which is ridiculously low.

    Some people know how to obtain meaningful (i.e., accurate) benchmarks and adhere to the necessary procedures to obtain them (when they care). That's why benchmarks (such as Geekbench) I run and report are consistently very near the top of the reported range, not just involving Macs or in the present situation but for all platforms and going back decades.
    "Very near the top of the reported range" doesn't mean accurate, it just means very near the top of the reported range. Try not to break your arm patting yourself on the back. 

    I'll talk to Andrew about it.
    How about you stick to areas you know something about? Benchmarks I've run in situations where I cared enough to take time to do them right are very near the top of reported ranges... and reproducible.  And fix the ridiculous inconsistencies in the article.
    The wide-ranging variability seen in Geekbench results are often because of inaccurate platform descriptions and because many people run the app without remedying factors that lower the scores.
    williamlondon

  • Tested: Thermal conditions in the 2018 i9 MacBook Pro dramatically hampering performance

    cpsrocpsro
    FWIW: I repeatedly got GB4 single-core scores of 504X and multi-core scores of 224XX on a 2.2 GHz model on display in an Apple Store. If GB4 scaled linearly with clock speed, one would predict a multicore score of 29600 for the Core i9 model. This isn't a completely unreasonable expectation because the i9 has 12 MB L3 cache vs. only 9 MB in the 6-core i7 models. Due to throttling and the fact that main memory is the same speed across all models, the best i9-8950HK benchmarks don't even come close to 29600. Even PCs with faster 2666 memory aren't close to 29600.
    d_2Avieshek