| Author |
Samsung Galaxy S4 announced |
etaab
Joined: Jan 23, 2004
Posts: > 500
From: UK - South Yorkshire
PM |
Of course im serious. That's exactly how benchmarking has worked in the computer technology and micro electronics world for the past twenty years. And guess what ? modern mobile phones are simply micro computers which use technology as powerful as a desktop pc, if not more powerful in some respects.
Ive never heard of any benchmarking where you'd expect to use the same settings in daily life if you turned them up to its absolute maximum. Check out what its meant by benchmarking, by checking out this simple explanation on Wikipedia..
"Computer manufacturers are known to configure their systems to give unrealistically high performance on benchmark tests that are not replicated in real usage"
Check me out on Instagram ! search for etaab ! |
|
|
reeflotz
Joined: Jun 15, 2010
Posts: > 500
From: Philippines
PM |
^^ So what is the purpose of benchmarking? just to cheat scores? that is what I understand from your post, also please don't link to wikipedia because anything that is written there can easily be edited.
As for your example in computer benchmarking that is completely wrong imo  . You do know the fact that computers are set to the highest stable clock speed when they want to get the most out of it during benchmarks, meaning you can use the computer in those clock speeds even for long periods of time because it is stable and that you will get the full potential of its computing power without damaging it. Any unstable clock speed is not listed in computer benchmarks because clearly it is useless.
Perfect example are gamers who overclock their computers to max so they can get 30 or 60 fps and minimize frame drops when gaming, and yes those benchmarks reflect real life usage when gaming because you need at least 30 fps to be able to play decently.
|
Wishmaster89
Joined: Jun 02, 2006
Posts: 333
From: Warsaw, Poland
PM |
^
Agree with you completely. Benchmarking for the sake of highest scores is nothing more than sports, you're just doing it to get the highest possible score even if it means cooling your device with liquid hydrogen 
In this case it is exactly this kind of benchmarking, to see what kind of performance this platform is capable of but it can't sustain it and use it in day to day scenario.
^^
To etaab and Tsepz_GP
So you're fine with the fact that when you're looking at the scores you know that it's not the performance you're going to get when using the device? At this point I'll repeat myself, I'm not saying that S4 is slow and laggy, that's not the point.
About your Bugatti Veyron example.
You just prove that it's nothing more than a show off, impressive show off but it's completely different situation.
It can't go as fast for longer periods of time because it can't but because there are limitations to physics(tires can't stand the generated heat). The engine itself could probably sustain the speed for longer periods of time. In the case of S4, it couldn't sustain higher clocks due to the generated heat and battery life.
It's nice to know that you can go as fast but it creates unfair comparisons with devices that are using their max performance in benchmarks as well as in day to day use. Remember that there are many people that buy devices based on how it fares in benchmarks.
One thing I have to agree with you, turbo mode.
All recent intel CPU's have turbo mode and I think that if possible it should be used in ARM but in this case it can't be considered turbo mode if it only launches in selected benchmarks. If it would launch when you're running some very demanding game then I would agree that samsung was clever to include this kind of turbo mode but this is nothing more than as the name implies benchmark booster.
Maybe HTC should raise use similar benchmark boosters and raise CPU clock to 2Ghz and GPU to 600mhz. This way they would be faster than S4, in benchmarks I wonder what would you say if HTC was the one using this kind of 'special code'.
After all, those companies that artificially improved their scores due to special drivers or simplified rendering in PC market were bashed by community for cheating, so why shouldn't we do the same in this situation?
And if you believe that samsung will just change the clocks for standard usage then you're very naive. They'll rather remove benchmark booster at all or hide it better this time |
randomuser
Joined: Sep 13, 2011
Posts: > 500
PM |
With such replies, there is NO doubt that etaab and Tsepz_GP are the BEST fanboys on Esato or probably the whole of the interwebs.  And they say they are NOT samsung fanboys 
@wishmaster89 and reeflotz
Imagine the snide remarks they would have made had it been Sony and not Samsung caught cheating 
[ This Message was edited by: randomuser on 2013-07-31 13:27 ] |
daviep
Joined: Apr 04, 2005
Posts: > 500
From: bonnie scotland
PM |
samsung denies cheating in bencmark scores
galaxy note8 |
randomuser
Joined: Sep 13, 2011
Posts: > 500
PM |
Cmon, who are they trying to fool here ?
But then what else can be expected from Samsung  A lame statement.
Anandtech already provided evidence that the GPU doesn't hit 533mhz for ANY app except certain benchmarking apps
[ This Message was edited by: randomuser on 2013-07-31 13:45 ] |
Wishmaster89
Joined: Jun 02, 2006
Posts: 333
From: Warsaw, Poland
PM |
Their reply indicates totally different situation than what we are dealing with. So it's nothing more than samsung trying to tell people that white is black and black is white hopping that people will believe it, even though their eyes see it a little bit different.
Hopefully no other company is doing something like this or else it would mean that there is no point in checking benchmark scores before deciding what device to buy. |
Tsepz_GP
Joined: Dec 27, 2006
Posts: > 500
From: Johannesburg, South Africa
PM |
Wishmaster89
This being Android, I know already that not every app is even capable of taking full advantage of the hardware, to me benchmarks give an idea of just how far the device can be pushed.
randomuser
Lol, so because we don't agree with what you and your fellow sony fans think, we are fanboys?
Phone: iPhone 15 Pro Max Black Ti 512GB
Tablet: iPad Pro 11” 2020 Space Gray 256GB
Watch: Series 3 Nike Edition Space Gray
Droid: Huawei Mate 40 Pro 256GB |
reeflotz
Joined: Jun 15, 2010
Posts: > 500
From: Philippines
PM |
Lots of problems with samsung statement:
First they said it is listed to allow 533mhz. It turns out you don't get 533mhz, what you get is 480mhz. That would be misrepresenting specifications. You only get 533mhz on benchmark apps, go figure. 
Then to cover up and make people believe that their S4 runs at indeed 533mhz, they make it run at that clock speed on benchmarks to produce scores which does not reflect when you use it on apps that you run on the phone which is where it is supposed to be used for.
They said "the Galaxy S4 has been designed to allow a maximum GPU frequency of 533MHz". Then they stated that "However, the maximum GPU frequency is lowered to 480MHz for certain gaming apps that may cause an overload". So in short it was NOT designed to run at that clock speed because it can cause an overload.
Then we have people that before this was found out made statements like "the S4 blows the competition out of the water in terms of performance" basing their statements on the not so true benchmark scores.
Finally when it was found out, we now hear these people making excuses saying "but this is how benchmarks work anyways" "the purpose of benchmarks is to cheat scores as high as possible"
if "benchmarks give an idea of just how far the device can be pushed" then it should give scores relative to its true actual performance, not scores which are enhanced in order to delude the person using it.
/facepalm
[ This Message was edited by: reeflotz on 2013-07-31 15:15 ] |
rikken
Joined: Feb 23, 2007
Posts: > 500
From: Norway
PM |
For the great majority of users this is a non issue. Life goes on, it is after all just a phone. Samsung has delivered a great product and that is what matters
Who makes a purchase based on benchmarks anyway ?
http://mostly-tech.com/2012/0[....]ecret-about-mobile-benchmarks/
[ This Message was edited by: rikken on 2013-07-31 16:44 ] |
Wishmaster89
Joined: Jun 02, 2006
Posts: 333
From: Warsaw, Poland
PM |
On 2013-07-31 15:41:23, Tsepz_GP wrote:
Wishmaster89
This being Android, I know already that not every app is even capable of taking full advantage of the hardware, to me benchmarks give an idea of just how far the device can be pushed.
What's the point of this knowledge if you know that no application will be able to use this power
That's the whole point of what Samsung did.
Considering those differences I would say that no matter if you buy HTC One or Xperia Z or any s600 based device in day to day use you'll get similar performance in any application you choose to use
And please, don't call me a sony fanboy. I'm using sony device just because after 4 years with Samsung devices I decided that it's time to buy something that works and believe me, xperia z really works even though it's not perfect. |
Tsepz_GP
Joined: Dec 27, 2006
Posts: > 500
From: Johannesburg, South Africa
PM |
oh dear reeflotz...
They said "the Galaxy S4 has been designed to allow a maximum GPU frequency of 533MHz" . Then they stated that "However, the maximum GPU frequency is lowered to 480MHz for certain gaming apps that may cause an overload" . So in short it was NOT designed to run at that clock speed because it can cause an overload.
Considering that TouchWiz apps can use it at 533mHz, its listed clock, I see no issue here, Samsung are simply trying to control power consumption.
"the S4 blows the competition out of the water in terms of performance"
It does even if at 480mHz as shown by Anandtech's 'non boosted' benchmark scores.
"but this is how benchmarks work anyways" "the purpose of benchmarks is to cheat scores as high as possible"
Who exactly did you quote the bold part from or are you adding your own rubbish to someone's statement to make it swing your way?
if "benchmarks give an idea of just how far the device can be pushed" then it should give scores relative to its true actual performance, not scores which are enhanced in order to delude the person using it.
Your opinion.
Wishmaster89
As an enthusiast, it's just a matter of knowing how powerful the HW is. Also as Samsung mentioned, TouchWiz apps like S Browser, Camera, Gallery etc... Use it at full clock, so it's not entirely limited to 480mHz.
Phone: iPhone 15 Pro Max Black Ti 512GB
Tablet: iPad Pro 11” 2020 Space Gray 256GB
Watch: Series 3 Nike Edition Space Gray
Droid: Huawei Mate 40 Pro 256GB |
Wishmaster89
Joined: Jun 02, 2006
Posts: 333
From: Warsaw, Poland
PM |
Have you by any chance read updated post on Anandtech?
They checked that the only touchwiz application that pushes the GPU clock up to 532mhz is camera and only for a second or two due to the thermal limitation. This just proves that such high GPU clock wasn't destined for day to day use and only for benchmarks (read: for show off). It only proves that Samsung is doing what was highly criticized in PC market, fix benchmark results.
If you don't agree with what we're saying just browse the web and you'll see that most web portals are criticizing Samsung for what they have done and if not criticizing than raising doubts about this whole situation. All in all it's not about senseless attacking of Samsung, it's about some fair competition and principles.
But I have to agree on one thing with reeflotz.
Benchmarking should lead to determining highest clocks allowing stable day to day performance to get those extra fps in your favourite game, that's how I've done with my desktop PC 5 years ago and I'm running it OC by 30% stable and with much improved performance since then
As I said before, benchmarking for the sake of highest scores alone that won't be usable in any day to day scenario is nothing more than a sport. But I agree, it's nice to see quad core CPU's running at 6Ghz cooled with liquid nitrogen and maybe we should start doing similar things with our smartphones, but not this way and such scores shouldn't be compared to devices running stock clocks.
It reminds me of how Anand benchmarked early nexus 4 in freezer to counter very aggressive thermal limitation
What's funny is that he actually got 10-20% higher scores than when run in normal condition  |
etaab
Joined: Jan 23, 2004
Posts: > 500
From: UK - South Yorkshire
PM |
Sadly, as always people like randomabuser always prove that they're simply here to slag off anything that isn't Sony and purposefully to slag off the market leader. As soon as anything negative crops up against Samsung, they flood the forum waving their tiny little flags of short lived victory. But its always a shallow one at that. The childishness that is allowed on Esato continues. So the usual suspects like randomabuser can use it all they want, but soon it'll blow over and they will still be fans of the inferior products.
To me its totally irrelevant what reasoning such people bring to this thread, even GSM Arena agree with the fact companies clock their devices higher for benchmarking, except they also think its something new - they've probably never been into PC modding and don't understand that overclocking for benchmarks is actually perfectly normal. Its called Bench marketing.
"We know desktop GPU manufacturers do this stuff all the time with their graphics cards to boost the benchmarking scores. But until now we haven’t see such tweaking among the embedded mobile graphics and that’s why we consider it unfair"
No big deal to me. Sorry Sony fans, no win for you again.
[ This Message was edited by: etaab on 2013-07-31 22:24 ] Check me out on Instagram ! search for etaab ! |
Tsepz_GP
Joined: Dec 27, 2006
Posts: > 500
From: Johannesburg, South Africa
PM |
The amount of fuss they've made out of this is incredible haven't seen anything like it anywhere else, you'd swear they all had an S4. Should've seen them when there was a rumour that GS4 sales weren't good, of course that was soon shutdown when Samsung announced 20million sold.
As a GS4 I9500 user, I'm not fussed, it runs every HD game I throw at it beautifully, takes incredible pics and vids, UI is smooth and fast and when I run a benchmark the HW shows me what it's truly capable of which is all I want to know, the fact that it does all the above at 480mHz is even better as it consumes less power, even taking HD720P@120FPS slo-mo vids with no issue. The fact that it does use the GPU at 533mHz in Camera even for a moment is fine, my guess is that it sits at 266mHz for Browser and Gallery due to them simply not requiring the full 533mHz. Using CPU-Z and System Panel I hardly ever even see the CPU passing 1.2GHz, I once saw it at 1.6GHz when playing GTA, but that was about it, and that was with the frame cap set to off.
Phone: iPhone 15 Pro Max Black Ti 512GB
Tablet: iPad Pro 11” 2020 Space Gray 256GB
Watch: Series 3 Nike Edition Space Gray
Droid: Huawei Mate 40 Pro 256GB |
|
|
Posted: 2013-07-30 23:09
Hazel J20i [Bricked], Sony Xperia Go 
