Are you Big Geekbench Believers?

[anon(10210323)]

Is Geekbench is a valid and recognized way to rate the speed and power of smartphones? I've heard of it, and I know some respected sites (like GSMArena) use Geekbench to compare smartphones in their reviews. If Geekbench is legit, then our lil' inexpensive iPhone SE's are in lofty company, and, if I'm understanding it correctly, they rank as the 6th most powerful smartphone currently on the market - which means probably in history. And it also means that during the summer months of 2016 it was number two, only behind the Huawei P9.

On Geekbench 4, the SE is beaten only by the iPhone 7 and 7 Plus; the Galaxy S7 and S7 Edge; and the Huawei P9.

My personal SE scores 4433 in the multi-core test, whereas the SE is listed as scoring 4057 at the Geekbench site.

So, can I use this to boast to my geeky friends? Especially considering the price, Apple put out a really special phone in the SE. Or am I putting too much weight on these scores?

Link: [redacted]

 

[Just_Me_D]

Some people care about stuff like that, but I don't. I use real world experiences to judge the speed of my devices.

 

[Ledsteplin]

You can brag if you like. That's a good score. But scores will vary from phone to phone for several reasons. Average scores for the SE will be different from those of the 7 Plus. The older and more used a phone is will score lower. But it's fun to brag. Go for it and enjoy your phone!

 

[anon(10210323)]

Some people care about stuff like that, but I don't. I real world experiences to judge the speed of my devices.

I like both - benchmarks and real world experience. When using an SE it's obvious that it is fast and has a lot of power, but it is still interesting to see how it compares with its "peers".

 

[Just_Me_D]

I like both - benchmarks and real world experience. When using an SE it's obvious that it is fast and has a lot of power, but it is still interesting to see how it compares with its "peers".

Understood...

 

[jmr1015]

Is Geekbench is a valid and recognized way to rate the speed and power of smartphones? I've heard of it, and I know some respected sites (like GSMArena) use Geekbench to compare smartphones in their reviews. If Geekbench is legit, then our lil' inexpensive iPhone SE's are in lofty company, and, if I'm understanding it correctly, they rank as the 6th most powerful smartphone currently on the market - which means probably in history. And it also means that during the summer months of 2016 it was number two, only behind the Huawei P9.

On Geekbench 4, the SE is beaten only by the iPhone 7 and 7 Plus; the Galaxy S7 and S7 Edge; and the Huawei P9.

My personal SE scores 4433 in the multi-core test, whereas the SE is listed as scoring 4057 at the Geekbench site.

So, can I use this to boast to my geeky friends? Especially considering the price, Apple put out a really special phone in the SE. Or am I putting too much weight on these scores?

Link: [redacted]

It shouldn't be all that surprising. That little inexpensive SE is running the same basic chip as the iPhone 6S, but on a much smaller screen with less pixels to push. It should benchmark at least as well as a 6S or 6S Plus, but performing slightly better wouldn't surprise me either.

In my opinion the scores themselves are a very valid basis to compare different phones. "Real world" performance is very subjective and equates to feeling a "seat of the pants" difference in cars, where as a chassis dynamometer doesn't lie. Put two different cars on the same dyno on the same day, and you'll get a valid basis to compare their power output. Likewise, run the same benchmark on two different phones at the same time, and the difference in score is a pretty solid indicator of which phone has more processing power.

How efficiently the phone/OS uses that power is a different story, and that is where I think you'll get you're "real world" usage differences of opinion.

 

[Matty]

@RefrigeratorRaider,

I think Geek-bench definitely has a place and it allows us to accurately determine the performance of a device. Which allows us o easily compare devices and give definite answers to which will perform better. Some OEM over-clock their GPU & CPU to to obtain a higher score lol.

OnePlus:

n order to give users a better user experience in resource intensive apps and games, especially graphically intensive ones, we implemented certain mechanisms to trigger the processor to run more aggressively. (When Geek-bench Is Running

Geekbench is a lot like speed test in cars going from 0 - 60 or 0 - 100 (depending on where you live). While the numbers are great to know and it gives you a nice idea on performance when trying to compare it to other cars, it doesn't really affect most people since they not going to push their cars that hard, ever

To me, i prefer to watch a video loop test and see how long the battery lasts.

 

[anon(10210323)]

@Matty -- Thanks for the thoughtful response. I guess my "angle" in bringing up Geekbench is the monetary cost of the iPhone SE in relation to its performance. The SE is not considered a flagship by the smartphone press, and their reasoning is "OK" - it's small, it has a 1.2 megapixel selfie camera, and the screen is not up to Flagship standards, although it is very, very nice. But although it is not considered a flagship, it still very easily outperforms many smartphones that are considered "flagships" -- like the LG V10 and V20, the new Motorolas, etc. etc.

Geekbench confirms the iPhone SE's performance is amongst the "greats". Also, it is, as @jmr1015 mentions, an iPhone 6S in disguise - the same Twister (A9), the same great camera, 2gb memory, etc. so the Geekbench ratings shouldn't be a surprise. What I like is the fact that Apple launched this device at only half the cost of "flagships" it clearly outperforms. Add to that the support and superior level of customer service provided by Apple.

I was just wondering if members of the iMore community use Geekbench ratings as seriously as some of the smartphone press does to "rate" devices. It appears not lol -- Apple iPhone users are more secure than others

 

[mogelijk]

@RefrigeratorRaider,

I think Geek-bench definitely has a place and it allows us to accurately determine the performance of a device. Which allows us o easily compare devices and give definite answers to which will perform better. Some OEM over-clock their GPU & CPU to to obtain a higher score lol.

OnePlus:


Geekbench is a lot like speed test in cars going from 0 - 60 or 0 - 100 (depending on where you live). While the numbers are great to know and it gives you a nice idea on performance when trying to compare it to other cars, it doesn't really affect most people since they not going to push their cars that hard, ever

To me, i prefer to watch a video loop test and see how long the battery lasts.

The issue with this is that, unlike 0-60, this isn't a test that works perfectly between manufacturers. While Geekbench, and other like it (such as AnTuTu), try to compensate for the differences in chips -- it is still not a perfect test. The other issue in your analogy is that 0-60 is just one attribute of an engine, it doesn't take into account the hauling power or top speed -- whereas Geekbench attempts to bring various attributes together into a single score (or two scores, with single and multi-core). But the fact remains, if it was a "perfect" test, you wouldn't have different tests (again, like AnTuTu) giving different scores -- based on how they value the various processor attributes and how they account for the processor architectures. There is also the problem of them running "on top of" the various OS versions -- particularly when the score for a single device can vary, depending on what else is running at the time and what else the OS might be doing.

 

[mogelijk]

@Matty -- Thanks for the thoughtful response. I guess my "angle" in bringing up Geekbench is the monetary cost of the iPhone SE in relation to its performance. The SE is not considered a flagship by the smartphone press, and their reasoning is "OK" - it's small, it has a 1.2 megapixel selfie camera, and the screen is not up to Flagship standards, although it is very, very nice. But although it is not considered a flagship, it still very easily outperforms many smartphones that are considered "flagships" -- like the LG V10 and V20, the new Motorolas, etc. etc.

Geekbench confirms the iPhone SE's performance is amongst the "greats". Also, it is, as @jmr1015 mentions, an iPhone 6S in disguise - the same Twister (A9), the same great camera, 2gb memory, etc. so the Geekbench ratings shouldn't be a surprise. What I like is the fact that Apple launched this device at only half the cost of "flagships" it clearly outperforms. Add to that the support and superior level of customer service provided by Apple.

I was just wondering if members of the iMore community use Geekbench ratings as seriously as some of the smartphone press does to "rate" devices. It appears not lol -- Apple iPhone users are more secure than others

One issue I have is the way you are looking at "flagship." Flagship largely means it is overall a companies best phone (or phones, such as the 7 and 7 Plus). This does not necessarily mean the "fastest processor," though it typically is the fastest. It also includes things like the radio, the camera, the display, etc. Also, if this were a true "processor test", then the screen (and current resolution) wouldn't change the score (such as how Geekbench rates the 6s and the SE differently).

The original Moto X is an example of a phone that was a flagship but didn't have a great processor -- instead Motorola matched two processors -- one that gave decent performance and another that had strong low power performance (prior to a CPU that did both great), to give the phone "always on" features while keeping the price of the phone down.

So, on one hand, the SE is not a "flagship," not because of the processor but because of the weaker cameras, lack of force touch, 1st gen Touch ID, etc. Additionally, trying to tout the Geekbench score is going to cause others to claim (as some here have pointed out) that the great score is in large part because of the "lackluster" screen, particularly as opposed to Android devices that have much larger displays and pixel densities.

I don't mean to take anything away from the SE, it is a great phone; just that Apple intentionally did not make it a flag-ship, more of a high midrange phone with a great processor.

 

[Matty]

There is also the problem of them running "on top of" the various OS versions -- particularly when the score for a single device can vary, depending on what else is running at the time and what else the OS might be doing.

Agreed. This was something i noticed and found quite funny. I ran geek-bench and got xx score, i then put my phone on charge and got a higher score. So it seems like the amount of power also plays apart. So i guess it boils down to, it gives you a round about performance rating which is close to what your actual 'top' performance is.