This type of behavior is always for the same audience.You realize those people do this shit for the clicks and only a very small handful do it here just to irritate people right?
This type of behavior is always for the same audience.You realize those people do this shit for the clicks and only a very small handful do it here just to irritate people right?
I like that you see everything as console war nonsense.
So, he started accusing others of being who he is.
Phil Spencer's same victimhood
Before the PS5 and SX came out, everything about Xbox was about power... "Only we have Full RDNA2 and sustainable performance."
As soon as the consoles came out and the SX lost in EVERY comparison, Phil Spencer took to the internet to victimize himself, saying he was "against console wars."
It's amazing how you and Phil Spencer behave in the same way.
Now that you haven't been able to fool anyone with your lie, you've started accusing others of starting console wars.
![]()
Still repeating the script that MS, with dog whistles and manipulation, trained you?
More importantly, who gives a shit what it's called when we have all the information we need about them? If it were about upcoming devices, knowing if they would be RDNA5 or RDNA6 would be relevant. This whole discussion about RDNA1 vs RDNA2 is just pointless.Have you guys ever look at it from this perspective?
Well Mark Cerny did state that it takes roughly four years to develop a new console.I would assume PS does it close to the same but I know Xbox would sit down even before a current system launches and start getting ideas of approx what specs they could get for a set price for next gen which includes ideas from AMD
So loose specs even 5 years out would be quite common, this stuff doesn't happen overnight
Not saying these specs are correct because frankly its over my head in what most of this stuff means, just saying I do believe this presentation took place in 2023 for PS6 hardware launching 5 years later
I just think we're fighting the boredom of not having more leaks.More importantly, who gives a shit what it's called when we have all the information we need about them? If it were about upcoming devices, knowing if they would be RDNA5 or RDNA6 would be relevant. This whole discussion about RDNA1 vs RDNA2 is just pointless.
![]()
Still repeating the script that MS, with dog whistles and manipulation, trained you?
So, he started accusing others of being who he is.
Phil Spencer's same victimhood
Before the PS5 and SX came out, everything about Xbox was about power... "Only we have Full RDNA2 and sustainable performance."
As soon as the consoles came out and the SX lost in EVERY comparison, Phil Spencer took to the internet to victimize himself, saying he was "against console wars."
It's amazing how you and Phil Spencer behave in the same way.
Now that you haven't been able to fool anyone with your lie, you've started accusing others of starting console wars.
I would love it even if it was a little bit more. But other then that got to have a good gpu and cpu. Even if the rumored xbox to have 11 core cpu and sony stuck with 8 surely they would include some kind of 3D cache to its cpu for the lack of cores in the cpu.Well Mark Cerny did state that it takes roughly four years to develop a new console.
I can see them spending a few months discussing what they would want in the PS6 after the presentation.
MLiD saying that the CUs can increase is the good example of the presentation being just a proposal.
Magnus having 68CU and the PS5 Pro having 60CUs, makes it a possibility that the PS6 may have 60 as well.
We are just doing a bunch of speculation to fight the boredom. But like you said, the gates are open for more leaks.
I bet you they won't but it's an extremely safe bet to make considering the last 11 years of AMD. I hope I'm wrong or Intel surprises us with celestial.My point is the amount of progress AMD made with RDNA4 and FSR4, even with CDNA4. AMD has a very good chance of matching Nvidia next gen.
I would love it even if it was a little bit more. But other then that got to have a good gpu and cpu. Even if the rumored xbox to have 11 core cpu and sony stuck with 8 surely they would include some kind of 3D cache to its cpu for the lack of cores in the cpu.
Btw what is your power estimation of leaked ps6 specs vs ps5pro and next xbox ?RDNA5 is a pretty forward looking uarch. They have every feature Blackwell has and then some.
I would change the zen6c cores and have all 8 zen6 full fat cores probably have an advantage over rumored magnus as it has zen and zen c cores and would be good for cpu related games..My updated PS6 specs post MLiD recent leaks
2-4 Zen6 big cores, 4-8 Zen6c cores. Chiplet.
8GB DDR5 OS and SSD cache on 128bit bus, 70GB+/s.
44/48 CUs at 3.5ghz, 40 teraflops fp32. AT3 chiplet.
24GB of 32gpbs GDDR7 on 192 bit bus. 768GB/s.
2TB SSD
$599
Dense Geometry Format (basically HW level Nanite): WO2025085121 DENSE GEOMETRY FORMAT WO2025085120 INTERSECTION TESTING ON DENSE GEOMETRY DATA USING TRIANGLE PREFILTERINGWhile I fully expect it to match Blackwell do you have any idea what "then some" could be? RT wise I know that Cerny patent exists and DX12 is getting native support for OMM and SER with DXR 1.2 (which I assume RDNA5 will support, and hence Sony will have their own API equivalent), but I'm not sure what else is on the horizon? Cooperative vectors? But I believe Blackwell supports that as well.
PS6 looks around ~9070 XT and Xbox around ~5080Btw what is your power estimation of leaked ps6 specs vs ps5pro and next xbox ?
I was going to comment on the pattern of PS4 and PS5 leaks...
But better not to comment, I say that Phill and his fanboys do not learn from past mistakes
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
This is going to be fun to watch. Xbox fans will be beating their chests about power, once again, while their system is probably going to be $300 more expensive than the PS6. No difference in power is going to make up for that kind of difference for people who mainly game on consoles. Not to mention everyone's libraries are in the PS ecosystem, now. Can't wait. Got the popcorn ready.
Sounds like ps6 would be the new 9070XTX possibly due to more ram and the use of gddr7 and not gddr6 and xbox the new version of amd 9080XTDense Geometry Format (basically HW level Nanite): WO2025085121 DENSE GEOMETRY FORMAT WO2025085120 INTERSECTION TESTING ON DENSE GEOMETRY DATA USING TRIANGLE PREFILTERING
Streaming Wave Coalescer (out-of-order execution): US20250068429 STREAMING WAVE COALESCER CIRCUIT US20250130811 Spill-After Programming Model for the Streaming Wave Coalescer
Workgroup self-launch (reduced CPU and GPU-frontend bottlenecks): WO2025144455 LOCAL LAUNCH IN WORKGROUP PROCESSORS
Many improvements to RT cores (beyond current uarchs): US20250182377 CONFIGURABLE RAY/EDGE TESTING FOR CONVEX POLYGON GROUPS US20250200890 PRISM VOLUMES FOR DISPLACED SUBDIVIDED TRIANGLES US20250200865 RAY TRACING OF DISPLACED MICRO MESHES USING A BOUNDING PRISM HIERARCHY WO2025144454 SYSTEMS AND METHODS FOR DETECTING RAY INTERSECTIONS WITH DISPLACED MICRO-MESHES
PS6 looks around ~9070 XT and Xbox around ~5080
I'll keep this comment in mind when the actual specs are confirmed.You two jackasses are the only ones doing it in the thread. Knock it off smh.
People have already been laughing at PS for years porting to PC, MLB on Xbox and now Helldivers 2 on Xbox.This is going to be fun to watch. Xbox fans will be beating their chests about power, once again, while their system is probably going to be $300 more expensive than the PS6. No difference in power is going to make up for that kind of difference for people who mainly game on consoles. Not to mention everyone's libraries are in the PS ecosystem, now. Can't wait. Got the popcorn ready.
Damn. That's a motherload of info.Dense Geometry Format (basically HW level Nanite): WO2025085121 DENSE GEOMETRY FORMAT WO2025085120 INTERSECTION TESTING ON DENSE GEOMETRY DATA USING TRIANGLE PREFILTERING
Streaming Wave Coalescer (out-of-order execution): US20250068429 STREAMING WAVE COALESCER CIRCUIT US20250130811 Spill-After Programming Model for the Streaming Wave Coalescer
Workgroup self-launch (reduced CPU and GPU-frontend bottlenecks): WO2025144455 LOCAL LAUNCH IN WORKGROUP PROCESSORS
Many improvements to RT cores (beyond current uarchs): US20250182377 CONFIGURABLE RAY/EDGE TESTING FOR CONVEX POLYGON GROUPS US20250200890 PRISM VOLUMES FOR DISPLACED SUBDIVIDED TRIANGLES US20250200865 RAY TRACING OF DISPLACED MICRO MESHES USING A BOUNDING PRISM HIERARCHY WO2025144454 SYSTEMS AND METHODS FOR DETECTING RAY INTERSECTIONS WITH DISPLACED MICRO-MESHES
PS6 looks around ~9070 XT and Xbox around ~5080
Not as hard as laughing at your posts in the power king thread.People have already been laughing at PS for years porting to PC, MLB on Xbox and now Helldivers 2 on Xbox.
The best version of Sony first party games is to actually get them on a good PC and amp up the settings how you like, which blows away any PS or Xbox version of a game.
And at the time everyone was talking specs in 2020 is the same year Sony started porting to PC. A platform that has a million configs can play games better than on home turf PS. And now Sony's biggest selling game in years Helldivers 2 is coming to Xbox. You can have old games like Gears 1 coming the same day.Not as hard as laughing at your posts in the power king thread.
thought it was a slam dunk for Microsoft.
Yet still the ps5 is gonna sell 100 million+, whilst Xbox consoles head to similar relevancy of what I pooped out this morning.And at the time everyone was talking specs in 2020 is the same year Sony started porting to PC. A platform that has a million configs can play games better than on home turf PS. And now Sony's biggest selling game in years Helldivers 2 is coming to Xbox. You can have old games like Gears 1 coming the same day.
And a Honda Civic sells more than a BMW model. Helldivers 2 will be sold on Xbox, just like MLB games the past 4-5 years since they ported over are as good as the PS version.Yet still the ps5 is gonna sell 100 million+, whilst Xbox consoles head to similar relevancy of what I pooped out this morning.
Using the info from MLiD.PS6 looks around ~9070 XT and Xbox around ~5080
I would change the zen6c cores and have all 8 zen6 full fat cores probably have an advantage over rumored magnus as it has zen and zen c cores and would be good for cpu related games..
44/48 i would change it to the same as ps5 pro 60CUs @ 3ghz
30GB of 35gpbs GDDR7 on a 256 bit bus @ 850GB/s or lower
Also some kind of v cache between cpu n gpu
I wonder if Sony surprises us with the clocks again. It's looking like RDNA5/UDNA + N3P is shaping up to do well on clocks already.
I recall everyone being shocked by 2233MHz on the PS5, most were expecting 1500-1800, perhaps with a slightly bigger die. Some time back I was expecting 2.8-3.0GHz with 72CU for around 27TF Single Issue FP32 (+architectural gains), but maybe we see something well into the 3GHz range. If the clocks are good, the thermal solution will still likely be cheaper than a big die.
The Switch 2 market is hard to beat.
It's one device that quickly and hassle free, switches between being a handheld and being a console at a single price point.
Sony having two separate devices isn't going to effect the Switch 2 market.
If Canis some how turns out to be the PS6 controller, where it's in every box. Than Sony has a chance to gain market share there.
It opens new and unique ways to play and doesn't have to be the most powerful console either.
I can see marketing as:
- Second screen (chat, hud, inventory, etc.) for a clean TV screen experience.
- Quickly swap to stream mode (PS5/PS6) mode to play anywhere around the house with direct WiFi or WiFi Hotspots away from home.
- Native mode for all PS4 games and select PS5/PS6 games.
This is how I see them tapping into the Switch 2 market. Many people aways forget the Switch is a hybrid console.
Not a handheld or Ninty gamer, but that's one thing about Switch which is awesome. You can flip flop docking and everything runs seamlessly.The Switch 2 market is hard to beat.
It's one device that quickly and hassle free, switches between being a handheld and being a console at a single price point.
Had to ask ChatGPT to summarize all this.Dense Geometry Format (basically HW level Nanite): WO2025085121 DENSE GEOMETRY FORMAT WO2025085120 INTERSECTION TESTING ON DENSE GEOMETRY DATA USING TRIANGLE PREFILTERING
Streaming Wave Coalescer (out-of-order execution): US20250068429 STREAMING WAVE COALESCER CIRCUIT US20250130811 Spill-After Programming Model for the Streaming Wave Coalescer
Workgroup self-launch (reduced CPU and GPU-frontend bottlenecks): WO2025144455 LOCAL LAUNCH IN WORKGROUP PROCESSORS
Many improvements to RT cores (beyond current uarchs): US20250182377 CONFIGURABLE RAY/EDGE TESTING FOR CONVEX POLYGON GROUPS US20250200890 PRISM VOLUMES FOR DISPLACED SUBDIVIDED TRIANGLES US20250200865 RAY TRACING OF DISPLACED MICRO MESHES USING A BOUNDING PRISM HIERARCHY WO2025144454 SYSTEMS AND METHODS FOR DETECTING RAY INTERSECTIONS WITH DISPLACED MICRO-MESHES
Nice post, you'd think Microsoft would want more than a 20% advantage? That's similar to series x and ps5Using the info from MLiD.
PS6: (48CUs × 64 × 2 × 3GHz) ÷ 1000 = 18.43 TFLOPs
The 5080 is ~20% faster than the 9070XT.
Magnus TFLOPs:
Assuming TFLOPs are linear with performance.
20% more TFLOPs than 48 CU
= 1.2 × 18.43 = 22.12 TFLOPs
Magnus Clock Speed:
22.12 = (68CUs × 64 × 2 × f) ÷ 1000
f = (22.12 × 1000) ÷ (68 × 64 × 2)
f ≈ 2.54 GHz
Assuming CUs are increased to PS5 Pro CUs:
20% slower than Magnus:
22.12 × 0.8 = 17.7 TFLOPs
Clock Speed:
17.7 = (60 × 64 × 2 × f) ÷ 1000
f = (17.7 × 1000) ÷ (60 × 64 × 2)
f ≈ 2.31 GHz
To summarize:
PS6: 48 CU @ 3 GHz = 18.43 TFLOPs
PS6: 60 CU @ 2.31 GHz = 17.7 TFLOPs
Magnus: 68 CU @ ~2.54 GHz = 22.12 TFLOPs
(20% faster)
Bonus:
PS6: 60 CU @ 3 GHz = 23.04 TFLOPs
(Mark Cerny being a mad scientist)
I don't think we will see anything clocking below 3GHz, also you should use 2x FP32 rate for the calculations.Using the info from MLiD.
PS6: (48CUs × 64 × 2 × 3GHz) ÷ 1000 = 18.43 TFLOPs
The 5080 is ~20% faster than the 9070XT.
Magnus TFLOPs:
Assuming TFLOPs are linear with performance.
20% more TFLOPs than 48 CU
= 1.2 × 18.43 = 22.12 TFLOPs
Magnus Clock Speed:
22.12 = (68CUs × 64 × 2 × f) ÷ 1000
f = (22.12 × 1000) ÷ (68 × 64 × 2)
f ≈ 2.54 GHz
Assuming CUs are increased to PS5 Pro CUs:
20% slower than Magnus:
22.12 × 0.8 = 17.7 TFLOPs
Clock Speed:
17.7 = (60 × 64 × 2 × f) ÷ 1000
f = (17.7 × 1000) ÷ (60 × 64 × 2)
f ≈ 2.31 GHz
To summarize:
PS6: 48 CU @ 3 GHz = 18.43 TFLOPs
PS6: 60 CU @ 2.31 GHz = 17.7 TFLOPs
Magnus: 68 CU @ ~2.54 GHz = 22.12 TFLOPs
(20% faster)
Bonus:
PS6: 60 CU @ 3 GHz = 23.04 TFLOPs
(Mark Cerny being a mad scientist)
Most of the AI improvements already exist on MI350/MI400, RDNA5 will just adopt a more area-efficient version of them since there's no need for *that* much AI performance.These are some crazy good improvements.
You know of any patents on AI/ML?
I don't think we will see anything clocking below 3GHz, also you should use 2x FP32 rate for the calculations.
The info MLiD have, has it being a chiplet design looking like RDNA3. So it can still be monolithic in a sense.My guess is that Sony will not go with chiplets - that's a Microsoft style move so they get the economies of scale from using identical chiplets that are used in discrete GPUs. 99% sure Sony will go with a traditional APU design. I'm also 100% sure they will have a refined version of the variable clock speeds from this gen as it seems to have worked a treat, there have been no murmurs of software dev difficulties etc which was the "concern" before the ps5 launched.
It does for RDNA5I think using dual issue would confuse everyone.
it's not like it has any tangible performance impact on any game
I never thought we would see 68UCs clocked @3+ GHz from Microsoft.I don't think we will see anything clocking below 3GHz, also you should use 2x FP32 rate for the calculations.
N3P is a good node, around ~20% higher fmax than N4P at same power.I never thought we would see 68UCs clocked @3+ GHz from Microsoft.
Is RDNA5 really that energy efficient?
It does for RDNA5
It's amazing how you start making accusations when you have nothing else to say.![]()
You think that everyone not praising enough or (god forbid) talking negatively about Sony = Xbox fanboy?
There is a good chance that next Xbox will be more powerful than PS6. Will this help them with sales? I doubt it.
I'm trying to help you, Xbox fans, not make the same mistake again.You two jackasses are the only ones doing it in the thread. Knock it off smh.
Did Xbox first party studios use those features? Vrs yes but don't recall any of them using sfs even mesh shaders
We could do that or I could ask you if you aren't tired yet? Tired of pushing forward the same trolling arguments from years and years ago and trying to rile people up about PS5 Pro which is also low level bait from years ago.
I'm analyzing the behavior of a group led by Phil Spencer.are you actually this stupid or what's happening here?
edit:
ok, fuck me, there's actually something seriously wrong with you... like damn.
the following is actually not a joke:
you need help... speak to a psychiatrist man
N3P is a good node, around ~20% higher fmax than N4P at same power.
I'd argue that the info MLiD has confirms chiplets for Magnus, but he has nothing for Canis/Orion to suggest that. The AMD presentation to Sony was to try and get them on board with using chiplets (obviously AMD want this as it allows them to take advantage of re-using stuff they already have in flight) - my guess is that Sony will go with a single chip. MS are 99% going with the chiplets as they will not put in enough orders to get economies of scale by themselves - but if they take an off-the-shelf chiplet that is already going into a GPU...The info MLiD have, has it being a chiplet design looking like RDNA3. So it can still be monolithic in a sense.
CPU+GPU APU with MCDs
![]()