3D Mark - which tool to choose

[KubGeek]

Hey guys,
I'm technical writter and I review technological products in Czech Republic. I have gone through a lot forums and webpages and still I'm not sure about which 3D Mark benchmark is good to use to compare graphic performance of Android/Apple (iPadOS/iOS) units. Wild Life seems to be hardly prefering Apple products, Sling Shot seems to show way more aligned scores between Android and iOS. Could you please tell me, which of these tests is more accurate for comparsion please?

Thnak you for your time and your GR8 job in LTT 
Love you guys!

Best regards from Czech Republic,
KubGeek

[SkyHound0202]

Short answer: Unlimited tests in 3DMark and Offscreen testes in GFXBench.

 

Long answer:

Spoiler

It's hard to compare graphic performance though synthetic software. The main problem is that Apple uses their proprietary Metal API on iOS (OpenGL ES has been deprecated in iOS 13) while Android used OpenGL ES and/or Vulkan. So they are not the same from the ground up since the API overhead and driver optimization for hardware is different on both platform.

 

This is further complicated by other variables like software edition and hardware design for a device using the said processor (example: same A15 in iPhone 13 & 14, iOS updates, benchmark software). What's more, even the same processor in the same device performs differently due to other factors like background activities, battery health, screen brightness, temperature and so on.

 

In a word, results from cross-platform synthetic benchmark software like 3DMark and GFXBench only tells how well a device can run the test, not necessarily a full-on comparison between devices or brands.

 

If you want to do a proper comparison, you can look up the result database of the said software and analyze them yourself. Otherwise you will need a lot of data on various devices through repeated tests using unlimited/offscreen style test (after clearing as much background apps as possible) since they are render at fixed resolution and is more comparable between runs/devices and more representative of the graphic power of the chip.

 

[KubGeek]

On 12/18/2022 at 5:10 PM, SkyHound0202 said:

Short answer: Unlimited tests in 3DMark and Offscreen testes in GFXBench.

 

Long answer:

  Reveal hidden contents

It's hard to compare graphic performance though synthetic software. The main problem is that Apple uses their proprietary Metal API on iOS (OpenGL ES has been deprecated in iOS 13) while Android used OpenGL ES and/or Vulkan. So they are not the same from the ground up since the API overhead and driver optimization for hardware is different on both platform.

 

This is further complicated by other variables like software edition and hardware design for a device using the said processor (example: same A15 in iPhone 13 & 14, iOS updates, benchmark software). What's more, even the same processor in the same device performs differently due to other factors like background activities, battery health, screen brightness, temperature and so on.

 

In a word, results from cross-platform synthetic benchmark software like 3DMark and GFXBench only tells how well a device can run the test, not necessarily a full-on comparison between devices or brands.

 

If you want to do a proper comparison, you can look up the result database of the said software and analyze them yourself. Otherwise you will need a lot of data on various devices through repeated tests using unlimited/offscreen style test (after clearing as much background apps as possible) since they are render at fixed resolution and is more comparable between runs/devices and more representative of the graphic power of the chip.

 

Thank you man, I thought it will be like that