By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - Gaming Discussion - Mobile devices will be more powerful than PlayStation 4, Xbox One in 2017, ARM forecasts

elektranine said:
curl-6 said:

Actually, high end mobile chips beat PS3/360 in single precision too. Tegra K1, for example, peaks at 326 Gigaflops single precision, and that's a mobile chip from 2014.

And that's before we even start factoring in advantages outside of Gigaflops, like the far more modern graphical feature sets that modern mobile chips have compared to the ancient DX9-era stuff in PS3/360.

You do know that the Tegra K1 is used in zero smartphones. There is a reason smartphone manufacturers stay away from all of Nvidia's mobile stuff lately. They can see through all of Nvidia's marketing hogwash. Nvidia badly wants into the mobile market but Qualcomm is king. Basically the only devices running Nvidia's mobile chipsets is Nvidia along with a token tablet here or there from other companies. There is no reaching 326 gigaflops in the real world that is with active cooling and an active power supply. And even Nvidia's own tablets have this really weird short battery life even shorter than laptops (2-3 hours for gaming). 

Look at this thing it is huge. Its the size of a small motherboard. Its basically a scaled down laptop gpu with a quadcore arm cpu thrown in. Nobody is using this.

 

And you bring up DirectX. I don't see why, OpenGL is king when it comes to graphics. OpenGL is used in the vast majority of graphical applications and its supported by more hardware than DX ever was. Just because MS tries to force people to upgrade their OS by not updating the features of their API doesn't mean OpenGL has to do the same. OpenGL and their legacy support is really top notch and feature wise there is only a small subset of what you cant do on the PS3 versus with modern hardware.

Tegra K1 is used in the nvidia shield tablet, that is a mobile device, and that was in 2014.

 

JRPGfan said:
curl-6 said:

There have been smartphones faster than the PS3 for a while now.

AFAIK no there hasnt been. Not in ability to run things, and not theoretically either.

Theres some tablet chips that are theoretically faster than the PS3, but those use around 10-15watts (not the 2-3w of a smartphone).


Marketing departments have claimed Smart phones have console quality image effects ect, but its all just hot air.

Nvidia love to make headlines like that, why? because they want to sell their chips.

 

So far there isnt a smartphone chip (running at 2-3watts), that can beat the PS3 in terms of GPU performance.

Sony's PS3 benchmarks are also essentially "hot air", as in they are inflated theoretical maximums that can only be attained in a specific benchmark test, not in an actual game.



Around the Network

"Sony's PS3 benchmarks are also essentially "hot air", as in they are inflated theoretical maximums that can only be attained in a specific benchmark test, not in an actual game."

The same is true for ARM though.
They cant show us a game, that runs better on a smart phone than a PS3 yet.


If smart phones are soon faster than PS4s, show me 1 game, thats on a PS3, and also runs on a phone, that runs at higher res, more details, higher FPS, than the PS3 version. (also would need to prove its not due to really bad programming on the PS3s side).



curl-6 said:

Tegra K1 is used in the nvidia shield tablet, that is a mobile device, and that was in 2014.

And that's just too bad since that will limit the chips maximum TDP ... 

The ENTIRE system needs to run at sub 5 watts so the chip hogging all the power isn't good enough ... 



JRPGfan said:
"Sony's PS3 benchmarks are also essentially "hot air", as in they are inflated theoretical maximums that can only be attained in a specific benchmark test, not in an actual game."

The same is true for ARM though.
They cant show us a game, that runs better on a smart phone than a PS3 yet.

Ultimately, tech demos and specialized benchmarks will always show a level of performance that actual games cannot attain, as the random nature of a game means you cannot tailor it as specifically. You can't predict when a player's input or a chance convergence of events will create a scenario where a certain element of the hardware is bottlenecked, for instance.

All I'm saying is, it's a bit unfair to call out mobile for using overly flattering benchmarks that don't reflect real world performance (which they do) when consoles do the exact same thing.



curl-6 said:
JRPGfan said:
"Sony's PS3 benchmarks are also essentially "hot air", as in they are inflated theoretical maximums that can only be attained in a specific benchmark test, not in an actual game."

The same is true for ARM though.
They cant show us a game, that runs better on a smart phone than a PS3 yet.

Ultimately, tech demos and specialized benchmarks will always show a level of performance that actual games cannot attain, as the random nature of a game means you cannot tailor it as specifically. You can't predict when a player's input or a chance convergence of events will create a scenario where a certain element of the hardware is bottlenecked, for instance.

All I'm saying is, it's a bit unfair to call out mobile for using overly flattering benchmarks that don't reflect real world performance (which they do) when consoles do the exact same thing.

I ll take actual proof then :p

1 PS3 game, that also runs on Android, where its maxing out the GPU on the PS3, but runs higher Image quality / resolution / fps, on the Phone.

 

If you dislike us just useing theoretical numbers.

I still think they are good guide lines, and it shows that Phones arnt yet more powerfull than the playstation 3.

Saying next year, they will have smart phones more powerfull than the Playstation 4, it is a strait up lie.



Around the Network
JRPGfan said:
curl-6 said:

Ultimately, tech demos and specialized benchmarks will always show a level of performance that actual games cannot attain, as the random nature of a game means you cannot tailor it as specifically. You can't predict when a player's input or a chance convergence of events will create a scenario where a certain element of the hardware is bottlenecked, for instance.

All I'm saying is, it's a bit unfair to call out mobile for using overly flattering benchmarks that don't reflect real world performance (which they do) when consoles do the exact same thing.

I ll take actual proof then :p

1 PS3 game, that also runs on Android, where its maxing out the GPU on the PS3, but runs higher Image quality / resolution / fps, on the Phone.

If you dislike us just useing theoretical numbers.

I still think they are good guide lines, and it shows that Phones arnt yet more powerfull than the playstation 3.

Saying next year, they will have smart phones more powerfull than the Playstation 4, it is a strait up lie.

Trine 2 and Doom 3 BFG Edition run at 1080p on Tegra X1 versus 720p (and in former, even lower) on PS3 and Xbxo 360.



curl-6 said:
JRPGfan said:

I ll take actual proof then :p

1 PS3 game, that also runs on Android, where its maxing out the GPU on the PS3, but runs higher Image quality / resolution / fps, on the Phone.

If you dislike us just useing theoretical numbers.

I still think they are good guide lines, and it shows that Phones arnt yet more powerfull than the playstation 3.

Saying next year, they will have smart phones more powerfull than the Playstation 4, it is a strait up lie.

Trine 2 and Doom 3 BFG Edition run at 1080p on Tegra X1 versus 720p (and in former, even lower) on PS3 and Xbxo 360.

Alright Actually found compairsions for both.

The Tegra X1 actually does come out ahead for the Nvidia Shield TV X1.

But thats a TV console, not a mobile device.

And its at 20watts or so, when gameing, not the typical 2-3watts most smart phones have to be under.

 

Basically Smart phones would need to be 10 times as effecient as the Nvidia Shield TV X1, and it would need to be about 10 times more powerfull than that to match a Playstation 4.

You see where Im going with this?

 

Basically before Smart phones can match a playstation 4.

They need to get 100x performance/watt of what the Tegra X1 can do.

 

Do you think next year, ARM pulls that off?



JRPGfan said:
curl-6 said:

Trine 2 and Doom 3 BFG Edition run at 1080p on Tegra X1 versus 720p (and in former, even lower) on PS3 and Xbxo 360.

Alright Actually found compairsions for both.

The Tegra X1 actually does come out ahead for the Nvidia Shield TV X1.

But thats a TV console, not a mobile device.

And its at 20watts or so, when gameing, not the typical 2-3watts most smart phones have to be under.

Google's Pixel C tablet uses a Tegra X1, that's a mobile device.



curl-6 said:
JRPGfan said:

Alright Actually found compairsions for both.

The Tegra X1 actually does come out ahead for the Nvidia Shield TV X1.

But thats a TV console, not a mobile device.

And its at 20watts or so, when gameing, not the typical 2-3watts most smart phones have to be under.

Google's Pixel C tablet uses a Tegra X1, that's a mobile device.

 

"At first glance it looks as though the computing heart of the Pixel C is identical to that of Nvidia Shield TV console, but there is a subtle difference. The Tegra X1 processor in the Pixel C has 'only' four ARM Cortex A57 cores, all of which can be clocked at a maximum of 1.9 GHz."

 

Everything inside the Pixel C, runs much slower than the Nvidia Shield TV X1.

It might not be as fast as the playstation 3 anymore.

 

Still theres a big jump from being about equal to a Playstation 3, to being more power than a Playstation 4.



JRPGfan said:
curl-6 said:

Google's Pixel C tablet uses a Tegra X1, that's a mobile device.

"At first glance it looks as though the computing heart of the Pixel C is identical to that of Nvidia Shield TV console, but there is a subtle difference. The Tegra X1 processor in the Pixel C has 'only' four ARM Cortex A57 cores, all of which can be clocked at a maximum of 1.9 GHz."

Everything inside the Pixel C, runs much slower than the Nvidia Shield TV X1.

It might not be as fast as the playstation 3 anymore.

Still theres a big jump from being about equal to a Playstation 3, to being more power than a Playstation 4.

I was never claiming they're more powerful than PS4, only that they have passed PS3, which technically they have.

The vanilla Tegra X1 in the Android TV has 512 Gigaflops of single precision floating point power in its GPU. That's much more than PS3, (enough to run some 720p or sub-HD PS3 games at 1080p) so there is plenty of room to trim without dropping below last gen consoles standards.

Also, apparently the Pixel C's GPU retains the 256 core-count of the Android TV version: http://www.androidcentral.com/pixel-c-specs

Pixel C also offers 3GB of RAM; 6 times as much as PS3 or 360.