Right. Like I even know when the fans kick in on my beeefy ass work laptop that the CPU usage hitting 100% is only in certain instances, same for GPU. so I just don't understand this concern that it won't give the exact power when needed in 90% of. Scenarios.
That's an excellent point. This is probably a dumb question but is there a reason why some developers don't add frame rate caps to pause screens and the like? My ps4 pro sounds the worst on menu screens and map screens because I think the frame rate is uncapped. Is there something preventing them from capping the frame rate on those screens?
Yeah they can lower the clock that's why they can reach higher peak clocks, but that's not what I'm arguing about.
Yeah they can lower the clock that's why they can reach higher peak clocks, but that's not what I'm arguing about.
Regarding power consumption, yeah it decreases exponentially, but that also means it increases exponentially from the base 1.7-1.8ghz the gpu was first designed to run on.
As I posted before, this is a 5700XT, RDNA1, at 2.1ghz consumption:
I'm assuming close to 100% load because it's running Far Cry @1440p in 80-100fps. But look at the power drawn.
RDNA2 will bring these values down, sure, but we are way past what's reasonable on the console. Which is why I'm saying we don't actually have nearly enough data to make the claim of how Ps5 clocks will vary, because we don't know any of the variables (base and peak wattages, etc)
Yeah they can lower the clock that's why they can reach higher peak clocks, but that's not what I'm arguing about.
Regarding power consumption, yeah it decreases exponentially, but that also means it increases exponentially from the base 1.7-1.8ghz the gpu was first designed to run on.
As I posted before, this is a 5700XT, RDNA1, at 2.1ghz consumption:
I'm assuming close to 100% load because it's running Far Cry @1440p in 80-100fps. But look at the power drawn.
RDNA2 will bring these values down, sure, but we are way past what's reasonable on the console. Which is why I'm saying we don't actually have nearly enough data to make the claim of how Ps5 clocks will vary, because we don't know any of the variables (base and peak wattages, etc)
It's not just that. The entire paradigm is basically designed so that the worst case heating issue is essentially the same as the best case. Constant power instead of constant frequency. This way fan speed is a constant.
Simply has to do with the types of instructions needed.
Yup and that's more than likely exactly what we should expect. The super interesting thing in ALL of this. Is what this translates to future Sony consoles. I don't know how prevalent a solution like this is power draw vs board temps. But this could mean for crazy high GPU MHz in the future with less CUs that would allow for them to add more things to chips.
Oh hell yeah they would've which is pretty crazy to think
Wheels, joysticks, the Novint Falcon, VR controllers, laparoscopy training rigs...there's lots of things with haptics. They're not directly comparable, of course, and some of these devices are very expensive to build in total. But the point isn't the total cost, just the incremental cost to add FFB, as a percentage of the overall BOM.
Of course, but obviously Epic have cracked that. The demo was running on an actual PS5. So I'm not sure where your concern comes from about fitting so much geometry into memory. They certainly can, even if we don't know precisely how that's achieved.
You don't know this for sure, because you're basing that statement on RDNA1. Improved power consumption is a bullet point feature for its successor, and this is very liable to change the exact shape of the curve.
No, it's up 72W. And again, this is RDNA1. We don't know what RDNA2 will look like.
People saying PS5 design being more efficient aren't necessarily speaking in power draw terms. They may be referring to faster caches, rasterization, etc.
Sony did tell us how their design works. The thing you're missing is that the PS5 approach is not just letting clocks be variable, like uncapping a framerate. That would indeed have no effect on the lowest dips in frequency. But they've also changed the trigger for throttling from temperature to silicon activity. And that actually changes how much power can be supplied to the chip without issues. This is because the patterns of GPU power needs aren't straightforward.
Stellar post.Sony did tell us how their design works. The thing you're missing is that the PS5 approach is not just letting clocks be variable, like uncapping a framerate. That would indeed have no effect on the lowest dips in frequency. But they've also changed the trigger for throttling from temperature to silicon activity. And that actually changes how much power can be supplied to the chip without issues. This is because the patterns of GPU power needs isn't straightforward.
Here's a depiction of the change. (This is not real data, just for illustrative purposes of the general principle.) The blue line represents power draw over time, for profiled game code. The solid orange line represents the minimum power supply that would need to be used for this profile. Indeed, actual power draw needs to stay well below that. Power supplies function best when actually drawing ~80% of their rating. And when designing a console the architects, working solely from current code, will build in a buffer zone to accommodate ever more demanding scenarios projected for years down the line.
You'd think the tallest peaks, highlighted in yellow, would be when the craziest visuals are happening onscreen in the game: many characters, destruction, smoke, lights, etc. But in fact, that's often not the case. Such impressive scenes are so complicated, the calculations necessary to render them bump into each other and stall briefly. Every transistor on the GPU may need to be in action, but some have to wait on work, so they don't all flip with every tick of the clock. So those scenes, highlighted in pink, don't contain the greatest spikes. (Though note that their sustained need is indeed higher.)
Instead, the yellow peaks are when there's work that's complex enough to spread over the whole chip, but just simple enough that it can flow smoothly without tripping over itself. Unbounded framerates can skyrocket, or background processes cycle over and over without meaningful effect. The useful work could be done with a lot less energy, but because clockspeed is fixed, the scenes blitz as fast as possible, spiking power draw.
Sony's approach is to sense for these abnormal spikes in activity, when utilization explodes, and preemptively reduce clockspeed. As mentioned, even at the lower speed, these blitz events are still capable of doing the necessary work. The user sees no quality loss. But now behind the scenes, these events are no longer overworking the GPU for no visible advantage.
But now we have lots of new headroom between our highest spikes and the power supply buffer zone. How can we easily use that? Simply by raising the clockspeed until the highest peaks are back at the limit. Since total power draw is a function of number of transistors flipped, times how fast they're flipping, the power drawn rises across the board. But now, the non-peak parts of your code have more oomph. There's literally more computing power to throw at the useful work. You can increase visible quality for the user in all the non-blitz scenes, which is the vast majority of the game.
Look what that's done. The heaviest, most impressive scenarios are now closer to the ceiling, meaning these most crucial events are leaving less untapped resources. The variability of power draw has gone down, meaning it's easier to predictively design a cooling solution that remains quiet more often. You're probably even able to reduce the future proofing buffer zone, and raise speed even more (though I haven't shown that here). Whatever unexpected spikes do occur, they won't endanger power stability (and fear of them won't push the efficiency of all work down in the design phase, only reduce the spikes themselves). All this without any need to change the power supply, GPU silicon, or spend time optimizing the game code.
Keep in mind that these pictures are for clarity, and specifics about how exactly much extra power is made available, how often and far clockspeed may dip, etc. aren't derivable from them. But I think the general idea comes through strongly. It shows why, though PS5's GPU couldn't be set to 2GHz with fixed clocks, that doesn't necessarily mean it must still fall below 2 GHz sometimes. Sony's approach changes the power profile's shape, making different goals achievable.
I'll stop with this (slowly) animated version of the above.
Great post. One of the best explanations I have read on the subject.Sony did tell us how their design works. The thing you're missing is that the PS5 approach is not just letting clocks be variable, like uncapping a framerate. That would indeed have no effect on the lowest dips in frequency. But they've also changed the trigger for throttling from temperature to silicon activity. And that actually changes how much power can be supplied to the chip without issues. This is because the patterns of GPU power needs isn't straightforward.
Here's a depiction of the change. (This is not real data, just for illustrative purposes of the general principle.) The blue line represents power draw over time, for profiled game code. The solid orange line represents the minimum power supply that would need to be used for this profile. Indeed, actual power draw needs to stay well below that. Power supplies function best when actually drawing ~80% of their rating. And when designing a console the architects, working solely from current code, will build in a buffer zone to accommodate ever more demanding scenarios projected for years down the line.
You'd think the tallest peaks, highlighted in yellow, would be when the craziest visuals are happening onscreen in the game: many characters, destruction, smoke, lights, etc. But in fact, that's often not the case. Such impressive scenes are so complicated, the calculations necessary to render them bump into each other and stall briefly. Every transistor on the GPU may need to be in action, but some have to wait on work, so they don't all flip with every tick of the clock. So those scenes, highlighted in pink, don't contain the greatest spikes. (Though note that their sustained need is indeed higher.)
Instead, the yellow peaks are when there's work that's complex enough to spread over the whole chip, but just simple enough that it can flow smoothly without tripping over itself. Unbounded framerates can skyrocket, or background processes cycle over and over without meaningful effect. The useful work could be done with a lot less energy, but because clockspeed is fixed, the scenes blitz as fast as possible, spiking power draw.
Sony's approach is to sense for these abnormal spikes in activity, when utilization explodes, and preemptively reduce clockspeed. As mentioned, even at the lower speed, these blitz events are still capable of doing the necessary work. The user sees no quality loss. But now behind the scenes, these events are no longer overworking the GPU for no visible advantage.
But now we have lots of new headroom between our highest spikes and the power supply buffer zone. How can we easily use that? Simply by raising the clockspeed until the highest peaks are back at the limit. Since total power draw is a function of number of transistors flipped, times how fast they're flipping, the power drawn rises across the board. But now, the non-peak parts of your code have more oomph. There's literally more computing power to throw at the useful work. You can increase visible quality for the user in all the non-blitz scenes, which is the vast majority of the game.
Look what that's done. The heaviest, most impressive scenarios are now closer to the ceiling, meaning these most crucial events are leaving less untapped resources. The variability of power draw has gone down, meaning it's easier to predictively design a cooling solution that remains quiet more often. You're probably even able to reduce the future proofing buffer zone, and raise speed even more (though I haven't shown that here). Whatever unexpected spikes do occur, they won't endanger power stability (and fear of them won't push the efficiency of all work down in the design phase, only reduce the spikes themselves). All this without any need to change the power supply, GPU silicon, or spend time optimizing the game code.
Keep in mind that these pictures are for clarity, and specifics about how exactly much extra power is made available, how often and far clockspeed may dip, etc. aren't derivable from them. But I think the general idea comes through strongly. It shows why, though PS5's GPU couldn't be set to 2GHz with fixed clocks, that doesn't necessarily mean it must still fall below 2 GHz sometimes. Sony's approach changes the power profile's shape, making different goals achievable.
I'll stop with this (slowly) animated version of the above.
Sony did tell us how their design works. The thing you're missing is that the PS5 approach is not just letting clocks be variable, like uncapping a framerate. That would indeed have no effect on the lowest dips in frequency. But they've also changed the trigger for throttling from temperature to silicon activity. And that actually changes how much power can be supplied to the chip without issues. This is because the patterns of GPU power needs isn't straightforward.
Here's a depiction of the change. (This is not real data, just for illustrative purposes of the general principle.) The blue line represents power draw over time, for profiled game code. The solid orange line represents the minimum power supply that would need to be used for this profile. Indeed, actual power draw needs to stay well below that. Power supplies function best when actually drawing ~80% of their rating. And when designing a console the architects, working solely from current code, will build in a buffer zone to accommodate ever more demanding scenarios projected for years down the line.
You'd think the tallest peaks, highlighted in yellow, would be when the craziest visuals are happening onscreen in the game: many characters, destruction, smoke, lights, etc. But in fact, that's often not the case. Such impressive scenes are so complicated, the calculations necessary to render them bump into each other and stall briefly. Every transistor on the GPU may need to be in action, but some have to wait on work, so they don't all flip with every tick of the clock. So those scenes, highlighted in pink, don't contain the greatest spikes. (Though note that their sustained need is indeed higher.)
Instead, the yellow peaks are when there's work that's complex enough to spread over the whole chip, but just simple enough that it can flow smoothly without tripping over itself. Unbounded framerates can skyrocket, or background processes cycle over and over without meaningful effect. The useful work could be done with a lot less energy, but because clockspeed is fixed, the scenes blitz as fast as possible, spiking power draw.
Sony's approach is to sense for these abnormal spikes in activity, when utilization explodes, and preemptively reduce clockspeed. As mentioned, even at the lower speed, these blitz events are still capable of doing the necessary work. The user sees no quality loss. But now behind the scenes, these events are no longer overworking the GPU for no visible advantage.
But now we have lots of new headroom between our highest spikes and the power supply buffer zone. How can we easily use that? Simply by raising the clockspeed until the highest peaks are back at the limit. Since total power draw is a function of number of transistors flipped, times how fast they're flipping, the power drawn rises across the board. But now, the non-peak parts of your code have more oomph. There's literally more computing power to throw at the useful work. You can increase visible quality for the user in all the non-blitz scenes, which is the vast majority of the game.
Look what that's done. The heaviest, most impressive scenarios are now closer to the ceiling, meaning these most crucial events are leaving less untapped resources. The variability of power draw has gone down, meaning it's easier to predictively design a cooling solution that remains quiet more often. You're probably even able to reduce the future proofing buffer zone, and raise speed even more (though I haven't shown that here). Whatever unexpected spikes do occur, they won't endanger power stability (and fear of them won't push the efficiency of all work down in the design phase, only reduce the spikes themselves). All this without any need to change the power supply, GPU silicon, or spend time optimizing the game code.
Keep in mind that these pictures are for clarity, and specifics about how exactly much extra power is made available, how often and far clockspeed may dip, etc. aren't derivable from them. But I think the general idea comes through strongly. It shows why, though PS5's GPU couldn't be set to 2GHz with fixed clocks, that doesn't necessarily mean it must still fall below 2 GHz sometimes. Sony's approach changes the power profile's shape, making different goals achievable.
I'll stop with this (slowly) animated version of the above.
Amazing explanation.Sony did tell us how their design works. The thing you're missing is that the PS5 approach is not just letting clocks be variable, like uncapping a framerate. That would indeed have no effect on the lowest dips in frequency. But they've also changed the trigger for throttling from temperature to silicon activity. And that actually changes how much power can be supplied to the chip without issues. This is because the patterns of GPU power needs isn't straightforward.
Here's a depiction of the change. (This is not real data, just for illustrative purposes of the general principle.) The blue line represents power draw over time, for profiled game code. The solid orange line represents the minimum power supply that would need to be used for this profile. Indeed, actual power draw needs to stay well below that. Power supplies function best when actually drawing ~80% of their rating. And when designing a console the architects, working solely from current code, will build in a buffer zone to accommodate ever more demanding scenarios projected for years down the line.
You'd think the tallest peaks, highlighted in yellow, would be when the craziest visuals are happening onscreen in the game: many characters, destruction, smoke, lights, etc. But in fact, that's often not the case. Such impressive scenes are so complicated, the calculations necessary to render them bump into each other and stall briefly. Every transistor on the GPU may need to be in action, but some have to wait on work, so they don't all flip with every tick of the clock. So those scenes, highlighted in pink, don't contain the greatest spikes. (Though note that their sustained need is indeed higher.)
Instead, the yellow peaks are when there's work that's complex enough to spread over the whole chip, but just simple enough that it can flow smoothly without tripping over itself. Unbounded framerates can skyrocket, or background processes cycle over and over without meaningful effect. The useful work could be done with a lot less energy, but because clockspeed is fixed, the scenes blitz as fast as possible, spiking power draw.
Sony's approach is to sense for these abnormal spikes in activity, when utilization explodes, and preemptively reduce clockspeed. As mentioned, even at the lower speed, these blitz events are still capable of doing the necessary work. The user sees no quality loss. But now behind the scenes, these events are no longer overworking the GPU for no visible advantage.
But now we have lots of new headroom between our highest spikes and the power supply buffer zone. How can we easily use that? Simply by raising the clockspeed until the highest peaks are back at the limit. Since total power draw is a function of number of transistors flipped, times how fast they're flipping, the power drawn rises across the board. But now, the non-peak parts of your code have more oomph. There's literally more computing power to throw at the useful work. You can increase visible quality for the user in all the non-blitz scenes, which is the vast majority of the game.
Look what that's done. The heaviest, most impressive scenarios are now closer to the ceiling, meaning these most crucial events are leaving less untapped resources. The variability of power draw has gone down, meaning it's easier to predictively design a cooling solution that remains quiet more often. You're probably even able to reduce the future proofing buffer zone, and raise speed even more (though I haven't shown that here). Whatever unexpected spikes do occur, they won't endanger power stability (and fear of them won't push the efficiency of all work down in the design phase, only reduce the spikes themselves). All this without any need to change the power supply, GPU silicon, or spend time optimizing the game code.
Keep in mind that these pictures are for clarity, and specifics about how exactly much extra power is made available, how often and far clockspeed may dip, etc. aren't derivable from them. But I think the general idea comes through strongly. It shows why, though PS5's GPU couldn't be set to 2GHz with fixed clocks, that doesn't necessarily mean it must still fall below 2 GHz sometimes. Sony's approach changes the power profile's shape, making different goals achievable.
I'll stop with this (slowly) animated version of the above.
This current fly on the ass of this thread pulled the same nonsense back at the start of this gen too. It's same shit, it's boring.
Sony did tell us how their design works. The thing you're missing is that the PS5 approach is not just letting clocks be variable, like uncapping a framerate. That would indeed have no effect on the lowest dips in frequency. But they've also changed the trigger for throttling from temperature to silicon activity. And that actually changes how much power can be supplied to the chip without issues. This is because the patterns of GPU power needs aren't straightforward.
Here's a depiction of the change. (This is not real data, just for illustrative purposes of the general principle.) The blue line represents power draw over time, for profiled game code. The solid orange line represents the minimum power supply that would need to be used for this profile. Indeed, actual power draw must stay well below the rated capacity. Power supplies function best when actually drawing ~80% of their rating. And when designing a console the architects, working solely from current code, will build in a buffer zone to accommodate ever more demanding scenarios projected for years down the line.
You'd think the tallest peaks, highlighted in yellow, would be when the craziest visuals are happening onscreen in the game: many characters, destruction, smoke, lights, etc. But in fact, that's often not the case. Such impressive scenes are so complicated, the calculations necessary to render them bump into each other and stall briefly. Every transistor on the GPU may need to be in action, but some have to wait on work, so they don't all flip with every tick of the clock. So those scenes, highlighted in pink, don't contain the greatest spikes. (Though note that their sustained need is indeed higher.)
Instead, the yellow peaks are when there's work that's complex enough to spread over the whole chip, but just simple enough that it can flow smoothly without tripping over itself. Unbounded framerates can skyrocket, or background processes cycle over and over without meaningful effect. The useful work could be done with a lot less energy, but because clockspeed is fixed, the scenes blitz as fast as possible, spiking power draw.
Sony's approach is to sense for these abnormal spikes in activity, when utilization explodes, and preemptively reduce clockspeed. As mentioned, even at the lower speed, these blitz events are still capable of doing the necessary work. The user sees no quality loss. But now behind the scenes, the events are no longer overworking the GPU for no visible advantage.
But now we have lots of new headroom between our highest spikes and the power supply buffer zone. How can we easily use that? Simply by raising the clockspeed until the highest peaks are back at the limit. Since total power draw is a function of number of transistors flipped, times how fast they're flipping, the power drawn rises across the board. But now, the non-peak parts of your code have more oomph. There's literally more computing power to throw at the useful work. You can increase visible quality for the user in all the non-blitz scenes, which is the vast majority of the game.
Look what that's done. The heaviest, most impressive scenarios are now closer to the ceiling, meaning these most crucial events are leaving fewer resources untapped. The variability of power draw has gone down, meaning it's easier to predictively design a cooling solution that remains quiet more often. You're probably even able to reduce the future proofing buffer zone, and raise speed even more (though I haven't shown that here). Whatever unexpected spikes do occur, they won't endanger power stability (and fear of them won't push the efficiency of all work down in the design phase, only reduce the spikes themselves). All this without any need to change the power supply, GPU silicon, or spend time optimizing the game code.
Keep in mind that these pictures are for clarity, and specifics about exactly how much extra power is made available, how often and far clockspeed may dip, etc. aren't derivable from them. But I think the general idea comes through strongly. It shows why, though PS5's GPU couldn't be set to 2GHz with fixed clocks, that doesn't necessarily mean it must still fall below 2 GHz sometimes. Sony's approach changes the power profile's shape, making different goals achievable.
I'll end with this (slowly) animated version of the above.
It's not based on power draw at all. Re-read or re-watch carefully what cerny told us. If it was based on power draw alone, there would be variations between PS5 because of silicon lottery.
Sony did tell us how their design works.
<snip>
Sony did tell us how their design works. The thing you're missing is that the PS5 approach is not just letting clocks be variable, like uncapping a framerate. That would indeed have no effect on the lowest dips in frequency. But they've also changed the trigger for throttling from temperature to silicon activity. And that actually changes how much power can be supplied to the chip without issues. This is because the patterns of GPU power needs aren't straightforward.
Here's a depiction of the change. (This is not real data, just for illustrative purposes of the general principle.) The blue line represents power draw over time, for profiled game code. The solid orange line represents the minimum power supply that would need to be used for this profile. Indeed, actual power draw must stay well below the rated capacity. Power supplies function best when actually drawing ~80% of their rating. And when designing a console the architects, working solely from current code, will build in a buffer zone to accommodate ever more demanding scenarios projected for years down the line.
You'd think the tallest peaks, highlighted in yellow, would be when the craziest visuals are happening onscreen in the game: many characters, destruction, smoke, lights, etc. But in fact, that's often not the case. Such impressive scenes are so complicated, the calculations necessary to render them bump into each other and stall briefly. Every transistor on the GPU may need to be in action, but some have to wait on work, so they don't all flip with every tick of the clock. So those scenes, highlighted in pink, don't contain the greatest spikes. (Though note that their sustained need is indeed higher.)
Instead, the yellow peaks are when there's work that's complex enough to spread over the whole chip, but just simple enough that it can flow smoothly without tripping over itself. Unbounded framerates can skyrocket, or background processes cycle over and over without meaningful effect. The useful work could be done with a lot less energy, but because clockspeed is fixed, the scenes blitz as fast as possible, spiking power draw.
Sony's approach is to sense for these abnormal spikes in activity, when utilization explodes, and preemptively reduce clockspeed. As mentioned, even at the lower speed, these blitz events are still capable of doing the necessary work. The user sees no quality loss. But now behind the scenes, the events are no longer overworking the GPU for no visible advantage.
But now we have lots of new headroom between our highest spikes and the power supply buffer zone. How can we easily use that? Simply by raising the clockspeed until the highest peaks are back at the limit. Since total power draw is a function of number of transistors flipped, times how fast they're flipping, the power drawn rises across the board. But now, the non-peak parts of your code have more oomph. There's literally more computing power to throw at the useful work. You can increase visible quality for the user in all the non-blitz scenes, which is the vast majority of the game.
Look what that's done. The heaviest, most impressive scenarios are now closer to the ceiling, meaning these most crucial events are leaving fewer resources untapped. The variability of power draw has gone down, meaning it's easier to predictively design a cooling solution that remains quiet more often. You're probably even able to reduce the future proofing buffer zone, and raise speed even more (though I haven't shown that here). Whatever unexpected spikes do occur, they won't endanger power stability (and fear of them won't push the efficiency of all work down in the design phase, only reduce the spikes themselves). All this without any need to change the power supply, GPU silicon, or spend time optimizing the game code.
Keep in mind that these pictures are for clarity, and specifics about exactly how much extra power is made available, how often and far clockspeed may dip, etc. aren't derivable from them. But I think the general idea comes through strongly. It shows why, though PS5's GPU couldn't be set to 2GHz with fixed clocks, that doesn't necessarily mean it must still fall below 2 GHz sometimes. Sony's approach changes the power profile's shape, making different goals achievable.
I'll end with this (slowly) animated version of the above.
