TheTechnician
Banned
I hope cerny will reveal at some Point wich one of the users here is his Account lol
Last edited:
If we add in the CPU components which at ~3.2GHz will likely consume about 50W then we get something like this:
.
Lol. The smaller the reference the larger the error. A slot is 12 cms, pretty easy to work it out from there. I think you do need to worry yourself with that.Someone has used the USB port on the ps5 dev kit to do measurements. U don’t need to worry urself with that
Actually 2080TI already runs it at 4K30 and almost 60fps at 1440pThe starwars RT demo was pretty pretty. When i watched TROS, there was a scene which reminds me of that demo.
Iirc it ran like 15fps on my 1080ti. The next 3080ti should run it comfortably at 4k30 at least.
RT is real deal.
I have a feeling Mark Sony will add as much RT hardware as possible to the V.
It actually looks pretty squarish to me. It's more likely to be as deep as it is wide. Besides that v cut out is still useful for air flow even.No way it's 3 controllers wide. Maybe 2 at best.
Use the disc slot to work out the width. A slot is 12 cms wide. So by my calculations the kit is 35.5cms wide.
10cm at it's highest and 5cms at its shortest. And about 24cms deep.
So if we allow for an average height of 8cms (due to the v cut out)
So volume is 35.5 x 8 x 24 = 6.816cm2
XSX is 31 x 16 x 16 =8,000cm2.
So not that big after all.
Actually 2080TI already runs it at 4K30 and almost 60fps at 1440p
Well PS4’s devkit was also huge. Yet PS4 base is considered by many to be disappointing.
![]()
I'm not interested in that argument, my post was about 2080TI running that in 4K30,There's nothing here that cannot be done without RT. I doubt that too much of a random memory access was used. Very small scene. The soldier that comes at us doesn't change the lightning on approach or after it disappears from the culled FOV.
Overall too much cheap reflections to make sure it stands out. Unless you want to play the whole game within glass and plastic corridors - infeasible.
At best it is two controllers deep, and controllers are 16cms. I say smaller than two, but if you are giving it the benefit of the doubt then say two.It actually looks pretty squarish to me. It's more likely to be as deep as it is wide. Besides that v cut out is still useful for air flow even.
By my calculations if it were a simple box it's be 36*36*10=12960cm3. The slice of pizza looks like it's 30cm deep, 12wide and 5 high. So it's volume is 0.5*12*30*5=900cm3
This nets a total volume of 12060cm3.
Still bigger than the XSX
I think the perspective is playing tricks on you. Especially the controller sitting on top. If you look at the second controller on the table you see what I mean, it looks way smaller than the first controllers and shows the perspective. 8 cm for the height is too small.At best it is two controllers deep, and controllers are 16cms. I say smaller than two, but if you are giving it the benefit of the doubt then say two.
So 35.5 x 8 x 32 = 9,088cm2.
I am being generous with the average of 8 for height as well.
Its just not as big as people are going on about, and as its a dev kit, size isnt something they care about, unlike retail. More plastic = more cost. More volume = greater shipping.
People clinging to the size of the dev kit as an indication of power are kidding themselves.
Actually 2080TI already runs it at 4K30 and almost 60fps at 1440p
Would pretty stupid of them to say "the most powerful console ever" if they don't know exactly what the competition is doing.
Oh ok, didn't know that since it wasn't specified.StarWars RT tech demo is using DLSS, so in reality it's running at 1440p
![]()
I wonder how much better 3080ti will run the same demo![]()
Actually 2080TI already runs it at 4K30 and almost 60fps at 1440p
Could you pinpoint me to that quote by Sony, couldn't find it anywhere on Google.Sony said it in February 2013: "Once again consistent with our heritage in gaming, delivering the most powerful platform ever". They knew back then, they know now.
You guys are delusional if you think both camps don't heard about specs this close to production.
Not saying the XSX won't be more powerful, just saying that PS5 might be on the small ball park and it could end up going either way.
Anyway if the difference is less than 1 TF I couldn't give two shits and will be going with the console with better exclusives. I probably won't be buying on day one either (didn't have that great of an experience with PS4 at day one) it will depend on games, OS maturity and price.
Could you pinpoint me to that quote by Sony, couldn't find it anywhere on Google.
It was during the PS4 reveal event (at around 8:30):
It was during the PS4 reveal event (at around 8:30):
Except that the 5700XT is 40 CU and 2150MHz is at its maximum power. 2000MHZ will use quite a bit less power because it's not going to be quite as close to it's voltage limit.
Besides, we already know how much power a 36 CU Navi consumes at ~2000MHz.
![]()
Flashing a Radeon RX 5700 to 5700 XT: Free 10% Performance Boost
In addition to our launch day reviews of the Radeon RX 5700 and 5700 XT, we followed up with extended comparisons, overclocking, and more. Most recently enthusiasts...www.techspot.com
People have flashed the 5700XT Bios to the base 5700. At around 2000MHz the 5700 draws 216W.
Let's not pretend that 216W is some thermally unmanageable monstrosity. It'll likely pull less power than a 56 CU GPU.
We can do a bit of armchair mathematics to prove it too.
36CU @ ~1700MHz = ~160W
36CU @ ~2000MHz = ~220W
56CU is 55% more CUs than 36CU.
If we assume power consumption scales linearly with larger chips then:
56CU @ 1700MHz = 160*1.55
= 248W
Thing is power doesn't necessarily scale linearly so it might actually be more than 250W for the 56CU part.
If we add in the CPU components which at ~3.2GHz will likely consume about 50W then we get something like this:
36 CU APU @ 2000MHz = ~ 270W
56 CU APU @ 1700MHZ = ~ 300W
Now Navi ships with pretty aggressive vcores so I think Sony and Microsoft might factory undervolt their chips to meet power consumption targets so end power consumption might be lower that we predict. But no matter which way you slice it the PS5 will draw less power than the XSX assuming the GitHub leaks are still representative of the final product.
And might I add. 250-300W total thermal output is not that big in the grand scheme of things.
AIB manufacturers can provide sufficiently quiet cooling for such TDPs without the explicit need for liquid cooling. And if the PS5 needs liquid, then the XSX sure as hell will need it too.
Actually yes, he didn't say "the most powerful console", he said "platform". Having listened to the bits beforehand makes it clear that he's speaking about ecosystem. This is why I couldn't find anyone quoting this as proof that PS4 will be more powerful.I think if you listen from 7 minutes on you'll see he was using platform in the ecosystem sense...not simply PS4 hardware. We could ask DIM for his call on this of course![]()
Guys, ignore that rumor. I uhh... "Shamelessly" created it on pastebin back in June to test how real some insiders were and how media outlets would report on it. BastienTech on Twitter put it on Twitter and it spread like wildfire on Reddit and elsewhere. There was an influx of pastebins back then too. Just rubbish everywhere.
Basically it proved to me you can't trust just anyone. Anyone takes any rumor and runs with it for the clicks, attention etc.
He's talking about two PS4 Pros duct-taped together LOL.
Duct taped as in 3d stacked?close, two 36 cu navis duct taped.
That of course makes much more sense. We are only speculating here, but in that case scenario, the two Navi10 36CUs chips would have been an earlier devkit, before the availability of the custom 72CUs definitive one.yeah, its basically the same as the hidden gpu in the power brick theory from last gen. now the roles have reversed.
there are simply cheaper ways to get to 14 tflops. 1.4 ghz for 40 cus is 80w. if 36 cus is 75w, you are looking at 150w for dual gpus. you could have one big gpu with 72 cus and it will be smaller and be more power efficient.
i remember this rumor from back then. this was when amd had just announced the rdna gpus and revealed the dual CU makeup of a shader engine. its just some guy from gaf or era making shit up.
3D stacking on new consoles? Watch from 12:31
I've already said it in this same thread, and nothing substantial and or concise enough, has come up yet that could point me to another direction _ i will bet a dollar with anyone that R600 is actually Mark Cerny, he was sent here to lower our expectations and Ninjad anyone that gets to close to the truth (askAre you telling me MS doesn't know yet @R600 does?![]()
I hope cerny will reveal at some Point wich one of the users here is his Account lol
The main reason that I have a hard time believing that dual GPU theory is that it sounds like trolling after MisterXMedia.
Also sounds incredibly expensive. Sony would be bleeding money selling that thing.
E3 2016 and e3 2018 they said most powerful console ever created for 1x and series x before Sony announce ps4pro and ps5.
i m confident if ps5 was going with 9.2 TF plan they would be screaming every where “most powerful console ever”
most glorious thing would be dealer and colteast walk back the importance of power narrative that they have been going hard at for the past few days about series Xlevel of mental gymnastics will be a sight to behold
Pretty great analysis on AMD chip manufacturing as well as some info on the Ps5 SOC. Can't wait till we all get final specs Console part at 18:15
The dual GPU code in Linux OS is most likely for PC development of applications -- NOT games. The film companies want dual GPUs because the 12G/16G of RAM isn't enough for the assets. I very seriously doubt you will see dual GPUs on a console.
The dual GPU code in Linux OS is most likely for PC development of applications -- NOT games. The film companies want dual GPUs because the 12G/16G of RAM isn't enough for the assets. I very seriously doubt you will see dual GPUs on a console.
Long live KenMakes sense with the shape of the dev kit. One Oberon on the left and one on the right. Cell 2 in the middle.
This dual GPU delusion needs to stop
Except for the insiders saying ps5 being the same or slightly more powerful than, xsx and pachters $1000 ps5 quote; everything lines up for 9.2 and 12 tf. But as Adored says, the ps5 chip looks to be “a broken design”.
it does make sense that Sony would go for the smaller chip size (316-340) vs MS (enormous 420-450) because MS has 4tf Lockhart in the pipeline and Sony is cheap AF. This would mean the insiders are all wrong and Sony has designed a broken chip that is 30% weaker, yet hotter and more power hungry. Possible? Sure. But i would prefere a theory that did not contradict either github or our insiders and make Sony engineers look like moronic stumblebums.
adored mentions chiplets being possible, and if they are feasible, 2X36 would not be crazy expensive, ps5 chip design would not be broken, insiders would be right, and github would not be contradicted.
__________
otoh, the above rests on xsx being 12tf. If it was 9tf, and was upclocked recently to get a 1tf bump to 10; then we dont need dual gpus or chiplets to explain anything. Xsx was less powerful last summer, now it is more powerful, and the power delta is less than 10%. Pachters $1000 ps5 comments can refer to the need for massive cooling + power and maybe a giant ssd. Everything works in this scenario.
What about the studios that already started making games for PS5 and Xbox? Aren’t they supposed to know by now?
Don’t bring sense into this thread please.The dual GPU code in Linux OS is most likely for PC development of applications -- NOT games. The film companies want dual GPUs because the 12G/16G of RAM isn't enough for the assets. I very seriously doubt you will see dual GPUs on a console.
This dual GPU delusion needs to stop![]()
Except for the insiders saying ps5 being the same or slightly more powerful than, xsx and pachters $1000 ps5 quote; everything lines up for 9.2 and 12 tf. But as Adored says, the ps5 chip looks to be “a broken design”.
it does make sense that Sony would go for the smaller chip size (316-340) vs MS (enormous 420-450) because MS has 4tf Lockhart in the pipeline and Sony is cheap AF. This would mean the insiders are all wrong and Sony has designed a broken chip that is 30% weaker, yet hotter and more power hungry. Possible? Sure. But i would prefere a theory that did not contradict either github or our insiders and make Sony engineers look like moronic stumblebums.
adored mentions chiplets being possible, and if they are feasible, 2X36 would not be crazy expensive, ps5 chip design would not be broken, insiders would be right, and github would not be contradicted.
__________
otoh, the above rests on xsx being 12tf. If it was 9tf, and was upclocked recently to get a 1tf bump to 10; then we dont need dual gpus or chiplets to explain anything. Xsx was less powerful last summer, now it is more powerful, and the power delta is less than 10%. Pachters $1000 ps5 comments can refer to the need for massive cooling + power and maybe a giant ssd. Everything works in this scenario.
This dual GPU delusion needs to stop![]()
AMDs press conference is earlier than sonys. Could be that they figured that in that order
Sonys PS5 talk later is put into context regarding their GPU ...
.
give me a scenario that makes sense of insiders, git hub, and sony engineers not being idiots.![]()
It's either 9.2 tflops as per github or the consoles are pretty close as per insiders since the github leak has some issues with it. You definitely should go with the second option so you can have a piece of mind.
give me a scenario that makes sense of insiders, git hub, and sony engineers not being idiots.![]()
Double gpu is never happening man. Why double the cost when you can add one more stack of 20 cu to the same apu for much much cheaper ..
give me a scenario that makes sense of insiders, git hub, and sony engineers not being idiots.![]()
If a 9 Tflop PS5 would hurt the baseline, wouldn't a Lockhart console at 4 Tflops be even worse if these numbers are to be true4k isnt the only problem. sony devs will be forced to use 4kcb and thats ok. the problem is all the extra stuff that the cpu will allow. destruction, changes in geometry, cg quality visual effects, ray tracing, an unprecedented levels of interactivity will all require the gpu to render everything. yes, 9 tflops is good for running current gen games, but next gen games will be far more complex in the graphical department and a 9 tflops baseline will hurt both the ps5 and the xbox series x.
games on the xbox still look phenomenal. if you believe rdr2 is the best looking game this gen, it runs at identical settings as the ps4 on the x1 albeit at 900p. games still look fantastic on the x1, just a tad blurry. so yeah, its not the end of the world, but then again the 1.3 tflops x1 isnt the end of the world either. its a just a massive disappointment that held back an entire gen.
Lockhart would be targeting 1080p while PS5 4K, big difference.If a 9 Tflop PS5 would hurt the baseline, wouldn't a Lockhart console at 4 Tflops be even worse if these numbers are to be true
If a 9 Tflop PS5 would hurt the baseline, wouldn't a Lockhart console at 4 Tflops be even worse if these numbers are to be true
I see. It'll be interesting but I'm going to go out on a limb and say that both consoles will achieve similar results only through different waysLockhart would be targeting 1080p while PS5 4K, big difference.
The main concern i have with the 2GHZ is heat. I simply see how that would fit in a consumer device. I fully agree with 44 to 52 CU near the 1500gz theory as that would be easier to achieve. We will soon find outIf Sony were targeting 9.2ish TFlops they would be far better off with something like 2x Navi 14. 44CUs at 1.7Ghz will be far more efficient than 36CUs at 2Ghz.
I just don't believe in 2GHz unless RDNA2 has some changes to help it clock higher and if that is the case everything changes anyway.
It's either 9.2 tflops as per github or the consoles are pretty close as per insiders since the github leak has some issues with it. You definitely should go with the second option so you can have a piece of mind.
If a 9 Tflop PS5 would hurt the baseline, wouldn't a Lockhart console at 4 Tflops be even worse if these numbers are to be true
Not only that . Yield for 2ghz gpu has to be insanely high which makes it impractical. I m sure final ps5 is wider and clocked lower than 2 ghz.The main concern i have with the 2GHZ is heat. I simply see how that would fit in a consumer device. I fully agree with 44 to 52 CU near the 1500gz theory as that would be easier to achieve. We will soon find out
Soon? Sony are unlikely to reveal the specs till at least E3 and maybe even not then. You will just have to rely on various leaks and theories until then.The main concern i have with the 2GHZ is heat. I simply see how that would fit in a consumer device. I fully agree with 44 to 52 CU near the 1500gz theory as that would be easier to achieve. We will soon find out
Soon? Sony are unlikely to reveal the specs till at least E3 and maybe even not then. You will just have to rely on various leaks and theories until then.