Ok, to put this into perspective: iPad4 is 3 times as slow as the Playstation3. (77GFLOPS versus 230GFLOPS in single precision floating point operations). If we look at power use the difference is incredible: the iPad is 35 times as efficient (latest generation Playstation 3, 70Watt, versus 4 Watt for the iPad 4, of which half is used for the display).
By the way, looking at some other posts: I do think that the numbers show the real difference, that's because the games (that potentially use all GPU power) use OpenGL to achieve that. And it's reasonal to assume that the implementation of OpenGL on a platform that has it's sole purpose in gaming (and exist for some time) is comparable to that of Apple.
Ok, to put this into perspective: iPad4 is 3 times as slow as the Playstation3. (77GFLOPS versus 230GFLOPS in single precision floating point operations).
If we look at power use the difference is incredible: the iPad is 35 times as efficient (latest generation Playstation 3, 70Watt, versus 4 Watt for the iPad 4, of which half is used for the display).
J.
Is GFLOPS a good way to measure gaming performance?
...
Also now we know why the iPad 3 was discontinued it was not balance, iPad 2, mini and now iPad 4 have the hardware that can drive its display...
I was saying the exact same thing, even before the iPad3 was released.
It should have had at least 4 times the GPU performance of the iPad2 because it had 4 times the pixels.
If I remember correctly, the iPad 3 was in fact, only three times as fast, and relatively slower than the iPad2.
The A6X corrects this.
Ok, to put this into perspective: iPad4 is 3 times as slow as the Playstation3. (77GFLOPS versus 230GFLOPS in single precision floating point operations).
If we look at power use the difference is incredible: the iPad is 35 times as efficient (latest generation Playstation 3, 70Watt, versus 4 Watt for the iPad 4, of which half is used for the display).
By the way, looking at some other posts: I do think that the numbers show the real difference, that's because the games (that potentially use all GPU power) use OpenGL to achieve that.
And it's reasonal to assume that the implementation of OpenGL on a platform that has it's sole purpose in gaming (and exist for some time) is comparable to that of Apple.
J.
But the Xbox 360 also has 77gflops and the performance is the same as the PS3...
Ok, to put this into perspective: iPad4 is 3 times as slow as the Playstation3. (77GFLOPS versus 230GFLOPS in single precision floating point operations).
If we look at power use the difference is incredible: the iPad is 35 times as efficient (latest generation Playstation 3, 70Watt, versus 4 Watt for the iPad 4, of which half is used for the display).
By the way, looking at some other posts: I do think that the numbers show the real difference, that's because the games (that potentially use all GPU power) use OpenGL to achieve that.
And it's reasonal to assume that the implementation of OpenGL on a platform that has it's sole purpose in gaming (and exist for some time) is comparable to that of Apple.
J.
But the Xbox 360 also has 77gflops and the performance is the same as the PS3...
So in CPU i believe they are in the same ball park but what about GPU? is the 554 MP4 in the same ball park as Xbox 360-PS3?
But the Xbox 360 also has 77gflops and the performance is the same as the PS3...
So in CPU i believe they are in the same ball park but what about GPU? is the 554 MP4 in the same ball park as Xbox 360-PS3?
Maybe you confuse some things.
I was talking about GPU performance but the PS3 has no GPU, at least not by 'definition'.
The PS3 has 7 cores (cells) that act together as a GPU. You could say that it's a GPU formed by software and several cell cores.
So to answer your question: the PowerVR 554 (GPU) of the iPad 4 has 77GFLOPS (single precision) and is 3 times as slow and 12 times as power efficient as the Cell processor cores (organized by software as a GPU) of the PS3.
This excludes the performance of the dual core Arm processor (1.4GHz) of the iPad4, so technically the iPad is faster than 77GFLOPS. I don't know the number of GFLOPS for the dual core Arm, but compared to GPU performance this isn't much.
The point is that this performance isn't used for games, because the game software layer OpenGL uses the GPU only.
I was talking about GPU performance but the PS3 has no GPU, at least not by 'definition'.
The PS3 has 7 cores (cells) that act together as a GPU. You could say that it's a GPU formed by software and several cell cores.
So to answer your question: the PowerVR 554 (GPU) of the iPad 4 has 77GFLOPS (single precision) and is 3 times as slow and 12 times as power efficient as the Cell processor cores (organized by software as a GPU) of the PS3.
This excludes the performance of the dual core Arm processor (1.4GHz) of the iPad4, so technically the iPad is faster than 77GFLOPS. I don't know the number of GFLOPS for the dual core Arm, but compared to GPU performance this isn't much.
The point is that this performance isn't used for games, because the game software layer OpenGL uses the GPU only.
J.
Actually you are wrong first the PS3 does have a GPU
Xbox GPU; ATI Xenos
PS3 GPU: NVIDIA RSX "Reality Synthesizer"
iPad 4; 554 MP4
Second the gflops is the CPU;
Xbox; 77gflops
PS3; 230gflops
iPad 4; 78gflops
Also I'm pretty sure that both CPU and GPU are used in OpenGL...
Actually you are wrong first the PS3 does have a GPU
Xbox GPU; ATI Xenos
PS3 GPU: NVIDIA RSX "Reality Synthesizer"
iPad 4; 554 MP4
Second the gflops is the CPU;
Xbox; 77gflops
PS3; 230gflops
iPad 4; 78gflops
Also I'm pretty sure that both CPU and GPU are used in OpenGL...
I didn't know that the PS3 had a GPU, thanks for the information.
I didn't talk about the Xbox (but I did know that it had a separate GPU).
The GFLOP number I was referring to was from Anadtech and specifies the PowerVR 554 GPU (it's 77GFLOPS) of the iPad4.
Your right that the GFLOPS of the PS3 I mentioned was about the CPU (Cell processor), that's because I mistook it for a software GPU.
OpenGL runs on a CPU thread (although even this thread might run on the GPU), but the point is that all calculation intensive tasks are offloaded to the GPU.
So everything that has to do with performance is on the GPU side, that's why the GPU GFLOPS count and the CPU is largely irrelevant.
I'am not sure about the PS3 implementation because it seems that the GPU is only 10GFLOPS or so (from a quick search on the internet) and it seems strange to ignore 230GFLOPS from the Cell processors.
It seems to me that the OpenGL driver of the PS3 uses the CPU as one or more GPUs and that the total of 230 plus 10 (?) GFLOP of the discrete GPU is the actual performance of the PS3.
Maybe someone knows exactly how this is implemented?
I didn't know that the PS3 had a GPU, thanks for the information.
I didn't talk about the Xbox (but I did know that it had a separate GPU).
The GFLOP number I was referring to was from Anadtech and specifies the PowerVR 554 GPU (it's 77GFLOPS) of the iPad4.
Your right that the GFLOPS of the PS3 I mentioned was about the CPU (Cell processor), that's because I mistook it for a software GPU.
OpenGL runs on a CPU thread (although even this thread might run on the GPU), but the point is that all calculation intensive tasks are offloaded to the GPU.
So everything that has to do with performance is on the GPU side, that's why the GPU GFLOPS count and the CPU is largely irrelevant.
I'am not sure about the PS3 implementation because it seems that the GPU is only 10GFLOPS or so (from a quick search on the internet) and it seems strange to ignore 230GFLOPS from the Cell processors.
It seems to me that the OpenGL driver of the PS3 uses the CPU as one or more GPUs and that the total of 230 plus 10 (?) GFLOP of the discrete GPU is the actual performance of the PS3.
Maybe someone knows exactly how this is implemented?
J.
that's the problem consoles gflops numbers are of CPU and ipad 4 gflops is GPU. Anyone knows the gflops of the A6X CPU? or the gflops of the GPU's opf Xbox 360 & PS3?
Comments
If we look at power use the difference is incredible: the iPad is 35 times as efficient (latest generation Playstation 3, 70Watt, versus 4 Watt for the iPad 4, of which half is used for the display).
By the way, looking at some other posts: I do think that the numbers show the real difference, that's because the games (that potentially use all GPU power) use OpenGL to achieve that.
And it's reasonal to assume that the implementation of OpenGL on a platform that has it's sole purpose in gaming (and exist for some time) is comparable to that of Apple.
J.
Is GFLOPS a good way to measure gaming performance?
Yes, it's directly related. See my previous comment (I added some lines to it).
J.
I was saying the exact same thing, even before the iPad3 was released.
It should have had at least 4 times the GPU performance of the iPad2 because it had 4 times the pixels.
If I remember correctly, the iPad 3 was in fact, only three times as fast, and relatively slower than the iPad2.
The A6X corrects this.
J.
Quote:
Originally Posted by jnjnjn
Ok, to put this into perspective: iPad4 is 3 times as slow as the Playstation3. (77GFLOPS versus 230GFLOPS in single precision floating point operations).
If we look at power use the difference is incredible: the iPad is 35 times as efficient (latest generation Playstation 3, 70Watt, versus 4 Watt for the iPad 4, of which half is used for the display).
By the way, looking at some other posts: I do think that the numbers show the real difference, that's because the games (that potentially use all GPU power) use OpenGL to achieve that.
And it's reasonal to assume that the implementation of OpenGL on a platform that has it's sole purpose in gaming (and exist for some time) is comparable to that of Apple.
J.
But the Xbox 360 also has 77gflops and the performance is the same as the PS3...
Quote:
Originally Posted by jnjnjn
Ok, to put this into perspective: iPad4 is 3 times as slow as the Playstation3. (77GFLOPS versus 230GFLOPS in single precision floating point operations).
If we look at power use the difference is incredible: the iPad is 35 times as efficient (latest generation Playstation 3, 70Watt, versus 4 Watt for the iPad 4, of which half is used for the display).
By the way, looking at some other posts: I do think that the numbers show the real difference, that's because the games (that potentially use all GPU power) use OpenGL to achieve that.
And it's reasonal to assume that the implementation of OpenGL on a platform that has it's sole purpose in gaming (and exist for some time) is comparable to that of Apple.
J.
But the Xbox 360 also has 77gflops and the performance is the same as the PS3...
So in CPU i believe they are in the same ball park but what about GPU? is the 554 MP4 in the same ball park as Xbox 360-PS3?
Quote:
Originally Posted by ShAdOwXPR
But the Xbox 360 also has 77gflops and the performance is the same as the PS3...
So in CPU i believe they are in the same ball park but what about GPU? is the 554 MP4 in the same ball park as Xbox 360-PS3?
Maybe you confuse some things.
I was talking about GPU performance but the PS3 has no GPU, at least not by 'definition'.
The PS3 has 7 cores (cells) that act together as a GPU. You could say that it's a GPU formed by software and several cell cores.
So to answer your question: the PowerVR 554 (GPU) of the iPad 4 has 77GFLOPS (single precision) and is 3 times as slow and 12 times as power efficient as the Cell processor cores (organized by software as a GPU) of the PS3.
This excludes the performance of the dual core Arm processor (1.4GHz) of the iPad4, so technically the iPad is faster than 77GFLOPS. I don't know the number of GFLOPS for the dual core Arm, but compared to GPU performance this isn't much.
The point is that this performance isn't used for games, because the game software layer OpenGL uses the GPU only.
J.
Quote:
Originally Posted by jnjnjn
Maybe you confuse some things.
I was talking about GPU performance but the PS3 has no GPU, at least not by 'definition'.
The PS3 has 7 cores (cells) that act together as a GPU. You could say that it's a GPU formed by software and several cell cores.
So to answer your question: the PowerVR 554 (GPU) of the iPad 4 has 77GFLOPS (single precision) and is 3 times as slow and 12 times as power efficient as the Cell processor cores (organized by software as a GPU) of the PS3.
This excludes the performance of the dual core Arm processor (1.4GHz) of the iPad4, so technically the iPad is faster than 77GFLOPS. I don't know the number of GFLOPS for the dual core Arm, but compared to GPU performance this isn't much.
The point is that this performance isn't used for games, because the game software layer OpenGL uses the GPU only.
J.
Actually you are wrong first the PS3 does have a GPU
Xbox GPU; ATI Xenos
PS3 GPU: NVIDIA RSX "Reality Synthesizer"
iPad 4; 554 MP4
Second the gflops is the CPU;
Xbox; 77gflops
PS3; 230gflops
iPad 4; 78gflops
Also I'm pretty sure that both CPU and GPU are used in OpenGL...
Quote:
Originally Posted by ShAdOwXPR
Actually you are wrong first the PS3 does have a GPU
Xbox GPU; ATI Xenos
PS3 GPU: NVIDIA RSX "Reality Synthesizer"
iPad 4; 554 MP4
Second the gflops is the CPU;
Xbox; 77gflops
PS3; 230gflops
iPad 4; 78gflops
Also I'm pretty sure that both CPU and GPU are used in OpenGL...
I didn't know that the PS3 had a GPU, thanks for the information.
I didn't talk about the Xbox (but I did know that it had a separate GPU).
The GFLOP number I was referring to was from Anadtech and specifies the PowerVR 554 GPU (it's 77GFLOPS) of the iPad4.
Your right that the GFLOPS of the PS3 I mentioned was about the CPU (Cell processor), that's because I mistook it for a software GPU.
OpenGL runs on a CPU thread (although even this thread might run on the GPU), but the point is that all calculation intensive tasks are offloaded to the GPU.
So everything that has to do with performance is on the GPU side, that's why the GPU GFLOPS count and the CPU is largely irrelevant.
I'am not sure about the PS3 implementation because it seems that the GPU is only 10GFLOPS or so (from a quick search on the internet) and it seems strange to ignore 230GFLOPS from the Cell processors.
It seems to me that the OpenGL driver of the PS3 uses the CPU as one or more GPUs and that the total of 230 plus 10 (?) GFLOP of the discrete GPU is the actual performance of the PS3.
Maybe someone knows exactly how this is implemented?
J.
Quote:
Originally Posted by jnjnjn
I didn't know that the PS3 had a GPU, thanks for the information.
I didn't talk about the Xbox (but I did know that it had a separate GPU).
The GFLOP number I was referring to was from Anadtech and specifies the PowerVR 554 GPU (it's 77GFLOPS) of the iPad4.
Your right that the GFLOPS of the PS3 I mentioned was about the CPU (Cell processor), that's because I mistook it for a software GPU.
OpenGL runs on a CPU thread (although even this thread might run on the GPU), but the point is that all calculation intensive tasks are offloaded to the GPU.
So everything that has to do with performance is on the GPU side, that's why the GPU GFLOPS count and the CPU is largely irrelevant.
I'am not sure about the PS3 implementation because it seems that the GPU is only 10GFLOPS or so (from a quick search on the internet) and it seems strange to ignore 230GFLOPS from the Cell processors.
It seems to me that the OpenGL driver of the PS3 uses the CPU as one or more GPUs and that the total of 230 plus 10 (?) GFLOP of the discrete GPU is the actual performance of the PS3.
Maybe someone knows exactly how this is implemented?
J.
that's the problem consoles gflops numbers are of CPU and ipad 4 gflops is GPU. Anyone knows the gflops of the A6X CPU? or the gflops of the GPU's opf Xbox 360 & PS3?
Found a few consoles gflops cpu/gpu numbers;
Xbox | CPU: 1.5 GFLOPS | GPU: 5.8 GFLOPS | Combined: 7.3 GFLOPS
Xbox360 | CPU: 77-115 GFLOPS | GPU: ???-240 GFLOPS | Combined: 355 GFLOPS
Dreamcast | CPU: 1.4 GFLOPS | GPU: 0.1 GFLOPS | Combined: 1.5 GFLOPS
Wii | CPU: 60 GFLOPS | GPU: 1 GFLOPS | Combined: 61 GFLOPS
PS2 | CPU: 6 GFLOPS | GPU: 0 GFLOPS | Combined: 6 GFLOPS
Wii U| CPU: 50-260 GFLOPS | GPU: ???-600 GFLOPS | Combined: 61 GFLOPS
iPad 4 | CPU: ??? GFLOPS | GPU: 78-??? GFLOPS | Combined: ??? GFLOPS
both xbox 360 and Wii U real gflops is the first one not the theoretical, apparently the real and theoretical are very different...