Baseless Next Generation Rumors with no Technical Merits [post E3 2019, pre GDC 2020] [XBSX, PS5]

[chris1515]

This is not inside a console too powerful but very interesting probably Navi 21 and help understand how Microsoft reach 12 Tflops RDA. I post it in Navi speculation thread but it is interesting about RDNA 2 potentital at least for Xbox Series X

https://www.reddit.com/r/Amd/comments/emhb8f/about_the_amd_gpu_beating_2080ti_by_17_leak/fdpdvta

Very interesting the benchmark Navi 21?

One of the dev of openVR benchmark said it can't be the laptop with an AMD integrated GPU and a TRX 2060

• OpenVR Benchmark is a GPU Benchmark, so being CPU-limited is basically impossible in there. The 4800H would definitely be fast enough to get a good result out of a GPU that is 30% faster than a 2080 Ti.
• I can confirm the AMD engineering sample result is in the leaderboard
• OpenVR Benchmark is still quite new, and every software can have bugs. But there's also almost 2000 leaderboard entries already, and if there would be something majorly buggy, that would be known (and fixed) by now. I do not think any bug is involved here.
• It's definitely not the result of a RTX 2060, the performance is way too good. The fact that the user who submitted the result is using an engineering sample CPU also makes it very unlikely that this is "fabricated by someone in their free time", because that would involve a lot of time, and someone who has an engineering sample CPU is likely using it at work, for testing purposes.

Other interesting tidbit

It hasn't really been mentioned here, but that best performing 2080 Ti in the leaderboard is already a really good performing 2080 Ti. As usual, there's many GPUs that perform somewhere in the middle, with few being faster and few being slower. The majority of 2080 Ti actually perform more around ~80 fps in this benchmark, so the unnamed GPU we talk about here is around 29% faster than a regular 2080 Ti, while being 17% faster than the highest overclocked 2080 Ti.

Very interesting

 

[eloic]

Gadgets like these?

WTF is that?

 

[fehu]

Nami from One Piece has a GPU now. Nice!

Yes, but it runs pretty hot

 

[Proelite]

Yes, but it runs pretty hot

She gains one teraflop every season.

 

[chris1515]

I think at 99.99999% Xbox Series X will be more powerful or nearly as powerful than Ps5. Ps5 will just have a faster SSD nothing more. Microsoft push at max what is possible in a console. Like Tom Warren said Sony need a miracle and walk on water to be better.

We begin to have misterxmedia conspiracy theory level on the side of Sony fan, dual GPU shit. Don't know if it will be 9 against 12 or 10 against 12 but I think Sony will be behind.

The Ps5 will probably be cheaper at 399 dollars but less powerful like the Ps4 Pro against the Xbox One X.

Just hope Sony will not forget to provide enough memory bandwidth this time not like the PS4 Pro more like the PS4.

 

[Proelite]

I think at 99.99999% Xbox Series X will be more powerful than Ps5. Ps5 will just have a faster SSD nothing more. Microsoft push at max what is possible in a console. Like Tom Warren said Sony need a miracle and walk on water to be better.

We begin to have misterxmedia conspiracy theory level on the side of Sony fan, dual GPU shit. Don't know if it will be 9 against 12 or 10 against 12 but I think Sony will be behind.

The Ps5 will probably be cheaper at 399 dollars but less powerful like the Ps4 Pro against the Xbox One X.

Just hope Sony will not forget to provide enough memory bandwidth this time not like the PS4 Pro more like the PS4.

This is treason against Klee.

 

[PSman1700]

Klee is banned i heard.

 

[Proelite]

Klee is banned i heard.

Can't get banned by mods if you ban yourself.

 

[chris1515]

This is treason against Klee.

Seeing the die size of the Xbox Series X at 380 to 400 mm2, the only possibility to do better is to use another process node. And this is impossible or Sony need to launch the PS5 in 2021 in 7nm+.

The best Sony can do is been the same power range I would say at best they can be on the same power range. After we don't know what better means it can be 0.1 Tflops. But Sony launch one console if they have the choice they will launch at 399 dollars.

For example we have much more condename for PS than Xbox Series X maybe Sony study multiple solution and will decide to go against the 499 dollars solution.

 

[Xbat]

Hey, stories always have three acts! I don't know how it ends!

In fact lube would be a luxury

 

[PSman1700]

Can't get banned by mods if you ban yourself.

He felt the heat coming, had his time atleast. That ppl fell for his lies is beyond me.

 

[chris1515]

Imo Jim Ryan told to Investor they want to have a faster transition from PS4 to PS5 than any other before and this is impossible at 499 dollars.

 

[Proelite]

Seeing the die size of the Xbox Series X at 380 to 400 mm2, the only possibility to do better is to use another process node. And this is impossible or Sony need to launch the PS5 in 2021 in 7nm+.

The best Sony can do is been the same power range I would say at best they can be on the same power range. After we don't know what better means it can be 0.1 Tflops. But Sony launch one console if they have the choice they will launch at 399 dollars.

For example we have much more condename for PS than Xbox Series X maybe Sony study multiple solution and will decide to go against the 499 dollars solution.

Multiple solutions of the same APU possibly if you go by steppings. Designing an APU is hundreds of millions of dollars. If Sony had multiple APUs we would be hearing about multiple PS5 skus.

My guess:
Same # CUs at higher clock.
Fully activated CUs at same clock.
Fully activated CUs at higher clock.

This is why in my wager thread, I had 40CUs at up to 2.2ghz to give my self some wiggle room. 11.3TF and 12TF are essentially the same.

There is a 30% difference between 2080 and 2080TI paper, but the difference in games is 15-20%. Tha's imperceptible at 4k.

No matter where Sony comes in at, they're fine perf wise, as long as they have comparable RT and VRS.

Xbone had 40% less compute, 50% less ROPS, and 50% less shader engines, 80% less ACE, and a huge bandwidth deficit. PS5 won't be in that boat even if it's 8.5 vs 12.

 

[chris1515]

https://www.anandtech.com/show/1532...-apus-7nm-8core-on-both-15w-and-45w-coming-q1

Very interesting the 4000 mobile series looking CPU we will have in consoles only have PCIE3 support.

It should be noted that for both U-series and H-series, the chips only support PCIe 3.0 - I suspect that PCIe 4.0 might be too power hungry for these form factors, and for modern graphics PCIe 3.0 is sufficient.

 

[AbsoluteBeginner]

Ok, I think I understand entire Shakespearean thingy about upcoming PS5, so let me recap my thoughts.

PS5 seems to be using Shakespeare themed codenames. That is, all codenames associated with PS5 are characters from two Shakespeare's plays, Tempest and Midsummer's Night Dream.

Ariel/Gonzalo/Prospero - characters from Shakespear's Tempest (early hardware, dev kit)

• Ariel - early chip
• Gonzalo - platform that got benchmarked (containing Ariel)
• Prosper - development kit

Oberon/Flute - characters from Shakespear's Midsummer's Night Dream (retail hardware)

• Oberon - retail chip
• Flute - platform that got benchmarked (containing Oberon)

Now that we got that out of the way, lets start with basics...

JANUARY 2019 - Ariel and Gonzalo 1.0

Ariel chip was "unearthed" back in late January 2019 by users APISAK and KOMACHI, who found codename of Gonzalo which contained following id :

2G16002CE8JA2_32/10/10_13E9

This chip was ES (engineering sample) back in January and its code indicated console semi custom chip consisting of Zen2 and 13E9 iGPU running at 1.0GHz

https://pci-ids.ucw.cz/read/PC/1022

From PCI ID list we can see 13E9 chip is range for iGPU (13FF max range).


APRIL 2019 - Ariel and Gonzalo 2.0

Gonzalo has surfaced again in April, now containing following ID :

ZG16702AE8JB2_32/10/18_13F8

We can see this chip was now QS (qualification sample) and iGPU is same one as Ariel (13FF range), but now running at 1.8GHz

JUNE 2019 - Flute benchmark

So we come to summer of 2019 and interesting thing pops up on UserBenchmark. Flute benchmark.

This one contained in following ID : AMD Eng Sample: 100-000000004-15_32/12/18_13F9

8 core Zen2 chip with 1/4th of L3 cache (likely to bring down APU die size, as L3 in Zen2 is big and in PC is used to hide latency. In SOC, it might not be of utmost importance)
iGPU - 13F9
Socket - BL5 (indicating semi custom)
16GB of RAM on 256 bit bus with ~530GB/s of BW.

Well, well...13F9. Same chip as Ariel, but clearly different revision. What could it be? Lets se...

August 2019 - Oberon

Comes August 2019 and Komachi leaks certain SOC codenamed OBERON, which contains PS4 and PS4 Pro BC modes. Certainly looks like PS5 chip...but I thought its Ariel that is codename of PS5 chip?

December 2019 - AMD Github Leak (Ariel > Oberon)

In December shit hit the fan when user _Rogame leaked Oberon A0/B0 specifications. Internet being internet found out the source, and what was found was incredible amount of data from internal AMD testing of upcoming chips (among them Ariel, Oberon, Arden, Sparkman and Renoir).

From the leak we found out that Ariel was early codename for PS5 chip, that later (sometimes during May 2019) was renamed - Oberon. Both chips sport same amount of CUs, and look like seemingly same chips, except for the fact that Oberon has higher memory BW on same bus, indicating faster memory chips being used.

To tie it all up, what was Flute benchmark iGPU range? 13F9

Heh...Looks familiar. 13F9 : OBR : A0 stepping. So Flute used 13F9 iGPU which is now identified as Oberon.

The conclusion therefore points to Ariel being early codename of a chip. This chip was benchmarked in Gonzalo leak (20K+ in firestrike).

Gonzalo was benchmarked in April and from Wired, we know Mark Cerny said dev kits will start coming out shortly after (indicating May/June). Since we know first APU based dev kits (V style) went out in June, these most likely sported Gonzalo/Ariel.

Ariel (spirit character) is also a servant to Prospero (magician in tempest, dev kit name), while Oberon is king of fairies in Midsummer's Night Dream and is found in later benchmark - Flute. ie - chip made for mass production.

As we know that Ariel had A0 and B0 stepping, and Oberon had A0 and B0 stepping, Komachi's "fifth stepping" would actually mean Oberon being in C0 revision, and not E0, because Ariel and Oberon share same iGPU ID (therefore : Ariel A0/B0 > Oberon A0/B0/C0 = 5)

Thats the hypothesis at least

 

[ultragpu]

He felt the heat coming, had his time atleast. That ppl fell for his lies is beyond me.

That is absolutely not the reason why he banned himself, don't spread lies. He's just one of those who has info but gets overly abused by some rude Era peeps and doesn't matter whether your info is right or wrong, you just can't deal with it at some point. He's just being polite by saying his time is over there, I don't get why you're dead set on he's having an agenda. We'll see tho, if he's a fraud then he's a fraud.

 

[AbsoluteBeginner]

Multiple solutions of the same APU possibly if you go by steppings. Designing an APU is hundreds of millions of dollars. If Sony had multiple APUs we would be hearing about multiple PS5 skus.

My guess:
Same # CUs at higher clock.
Fully activated CUs at same clock.
Fully activated CUs at higher clock.

This is why in my wager thread, I had 40CUs at up to 2.2ghz to give my self some wiggle room. 11.3TF and 12TF are essentially the same.

There is a 30% difference between 2080 and 2080TI paper, but the difference in games is 15-20%. Tha's imperceptible at 4k.

No matter where Sony comes in at, they're fine perf wise, as long as they have comparable RT and VRS.

Xbone had 40% less compute, 50% less ROPS, and 50% less shader engines, 80% less ACE, and a huge bandwidth deficit. PS5 won't be in that boat even if it's 8.5 vs 12.

They will be lucky to hit 2.0GHz with 36CUs. I have no doubt they will be deactivating 4CUs in retail machine. As for hitting 2.2GHz, well, we have people doubting 2.0GHz even on improved node, let alone 2.2GHz.

 

[Tkumpathenurple]

WTF is that?

A Friday night at @London-boy's house.

 

[Globalisateur]

Seeing the die size of the Xbox Series X at 380 to 400 mm2, the only possibility to do better is to use another process node. And this is impossible or Sony need to launch the PS5 in 2021 in 7nm+.

The best Sony can do is been the same power range I would say at best they can be on the same power range. After we don't know what better means it can be 0.1 Tflops. But Sony launch one console if they have the choice they will launch at 399 dollars.

For example we have much more condename for PS than Xbox Series X maybe Sony study multiple solution and will decide to go against the 499 dollars solution.

I disagree. This is not impossible in 2020 If they had planned (gambled even) from the start on 7nm EUV since the first roadmaps from AMD / TSMC.

Like the way they planned early on, some would even say gambled, on GDDR5 in the previous gen.

But yes at this point they are either releasing a 8tf (not 9tf) machine at 7nm or a >10tf machine at 7nm+. They won't overclock a 7nm DUV APU at 2ghz. That's not happening.

 

[Deleted member 13524]

Imo Jim Ryan told to Investor they want to have a faster transition from PS4 to PS5 than any other before and this is impossible at 499 dollars.

People are expecting to pay the exact same amount for a new console between 2013 and 2020?
That would put the PS5 as much cheaper than the PS4, if you adjust for inflation. It wouldn't make much sense, since the PS4 was already considered pretty cheap when it came out.

Very interesting the 4000 mobile series looking CPU we will have in consoles only have PCIE3 support.

The 4000 mobile series are TDP limited because they're mobile. The consoles aren't mobile.
Regardless, PCIe 4.0 will be used if it makes sense. It could range from PCIe 3.0 to a custom IF link that is 2 times faster than PCIe 4.0. No one knows.

That is absolutely not the reason why he banned himself, don't spread lies. He's just one of those who has info but gets overly abused by some rude Era peeps and doesn't matter whether your info is right or wrong, you just can't deal with it at some point. He's just being polite by saying his time is over there, I don't get why you're dead set on he's having an agenda. We'll see tho, if he's a fraud then he's a fraud.

Because if he's right then our local B3D Messiah @Proelite (who AFAIK has zero track record of industry connections or older predictions, though be free to correct me) has to be wrong, and we B3D'ers need to stick together and demonize all "insiders" to promote our own.