• Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
Status
Not open for further replies.

Albert Penello

Verified
Nov 2, 2017
320
Redmond, WA
Well, Albert Penello seems to think that they could possibly use GCN adjusted TF:



100% no. I never said that, nor do I have any reason to believe that the reported numbers are adjusted on either side.

I was wondering IF they SHOULD adjust the numbers at launch given how casual folks will compare them to the Gen8 numbers without understanding the generational improvements.

I've seen this posted a few times, so please let's clear this up.
 

Dokkaebi G0SU

Member
Nov 2, 2017
5,922
yep. MS didnt bother giving away the clock speeds of the cpu, the tflops count of the gpu and the type of ram at the xone reveal.

whoever, be it sony or ms, loses the tflops war, or cpu clock war or bandwidth war, will simply not mention them.

xbox-720-hardware-ports.jpg


Sony just released this in their pdf.

ps4-specs.jpg



i get banned every few weeks. mostly for making stupid jokes about console wars.

i am pretty sure im a ban away from a perma because of previous infractions.

remember me.
tenor.gif
 

Expy

Member
Oct 26, 2017
9,861
We're going to find out the details in the coming months, and then we'll find out nothing has changed.
 

SeanMN

Member
Oct 28, 2017
2,185
Well, Albert Penello seems to think that they could possibly use GCN adjusted TF:

Albert's comment, to me, seems to be more in regard to Lockhart, where it could be perceived as having lower capability than the 1X.

I think MS will handle this in a similar fashion as they do today. They tout the 6TF capability of the 1X, for the 1S they say "the best value in games and entertainment." This is what I expect them to do next gen.

Example:
"Meet the Project Scarlett family of consoles featuring 4 times the CPU power, and SSDs that are 40x faster. Choose best in class performance and capability with the 12 TF Anaconda, or the best value in next gen gaming with Lockhart"
They won't hide Lockhart's specs, but where it would make it look weak, they definitely won't advertise it.
 

BradGrenz

Banned
Oct 27, 2017
1,507
Do you have any implementation examples of NVMe using 1 or 2 lanes? I tried finding some but only some esoteric stuff in older documentation for server gear (2 lanes... cant find anything on one)

2 lane nVME drives are part of the M.2 standard. Not that it matters since you can do whatever configuration you want in a console.

The current HDD inside the Xbox is 100MB/s or higher but let us stick with 100 for easy math if you take the 40 times better performence as "faster" than you are at 4GB/s already.

We don't actually know what they were measuring with that 40x quote. What if it was latency, not bandwidth?

Just to say something I think makes sense. Most developers have reportedly not even had Scarlett dev kits, while some have gotten PS5 kits.
It would make sense then if any information regarding their relative power in regards to one another, should be looked at with a questioning eye.

Logically, we might conclude that the developers confident in sharing their impressions on the performance differences were the ones with access to both and ability to make such a judgement.

Could the 12TF figure be when compared to X1X in GCN rather than RDNA flops? with the RDNA figure possibly being closer to 10TF?

Converting between GCN flops versus RDNA flops is not a real thing anyone in the industry does. It's a behavior entirely invented in these threads.
 

Deleted member 1589

User requested account closure
Banned
Oct 25, 2017
8,576
He was talking in terms on how to do public relations. The first sentence was already about how the difficulty of explaining to the general public. Then he gave 3 examples on how the PR can be written.

It wasn't about 'is it GCN or RDNA" but rather "how do we explain to people who have no clue even though the teraflop number doesn't explain the difference"

So you go to Brad Sams article on Windows Central for example, on how he did it.

- give out the teraflops number, and compare it to a product of this gen

- explain in an another paragraph about how this doesn't really show how much better it is compared to this gen.
 

anexanhume

Member
Oct 25, 2017
12,913
Maryland
100% no. I never said that, nor do I have any reason to believe that the reported numbers are adjusted on either side.

I was wondering IF they SHOULD adjust the numbers at launch given how casual folks will compare them to the Gen8 numbers without understanding the generational improvements.

I've seen this posted a few times, so please let's clear this up.
They've never adjusted them before. I'm sure the comparison between the two matters more for what will control the debates. If one manufacturer chooses a different metric, it could get interesting.
 

foxbeldin

Member
Oct 27, 2017
747
100% no. I never said that, nor do I have any reason to believe that the reported numbers are adjusted on either side.

I was wondering IF they SHOULD adjust the numbers at launch given how casual folks will compare them to the Gen8 numbers without understanding the generational improvements.

I've seen this posted a few times, so please let's clear this up.

Thanks for putting an end to this madness.
 

Gemüsepizza

Member
Oct 26, 2017
2,541
100% no. I never said that, nor do I have any reason to believe that the reported numbers are adjusted on either side.

I was wondering IF they SHOULD adjust the numbers at launch given how casual folks will compare them to the Gen8 numbers without understanding the generational improvements.

I've seen this posted a few times, so please let's clear this up.

Thanks for the clarification! And sorry, it wasn't my intention to misrepresent your post, probably read a bit too fast + not a native speaker.

I doubt Era servers would survive that ;)

I don't think they will survive either way tbh, lol.
 
Last edited:
Nov 11, 2017
2,744
100% no. I never said that, nor do I have any reason to believe that the reported numbers are adjusted on either side.

I was wondering IF they SHOULD adjust the numbers at launch given how casual folks will compare them to the Gen8 numbers without understanding the generational improvements.

I've seen this posted a few times, so please let's clear this up.
.
 

Expy

Member
Oct 26, 2017
9,861
Perhaps MS surprised Sony who thought they would have a larger advantage back in 2018, but now are closer. Now both targeting ~12TF.
 

Chamon

Member
Feb 26, 2019
1,221
The problem with the FLOPGATE theory is the fact that Lockhart is being reported as a 4tf console in the same story. Why would Anaconda be old flops and Lockhart new flops. Do people believe that Lockhart is 4 old GCN troops? I suppose it's possible but doesn't see likely. Seems like some folks are reaching here.
This makes sense.

Edit: but if it is in fact 4 GCN Tflops for Lockhart, that would match with those that were saying that it is similar to a PS4 Pro...
 
Last edited:

VX1

Member
Oct 28, 2017
7,000
Europe
One of the guys (Karmaprof) over at B3D made a pretty good point regarding unified memory architecture cause now it seems both PS5 and Anaconda will have 256-bit 16GB memory bus, i didn't think about this much before:

"If this the real Endproduct /Retail then good Night. It is the same Bottleneckarchitecture like on the Ps4. What People forget with an UMA Architecture is , you cannot have the full 256 Bit Bandwidth for the GPU because you must share it with the CPU , and a Zen2 CPU at 3 or 3,5 Ghz , 16 x Multithreading , need more Bandwidth then the oldJaguar CPU at 1,6 Ghz. It is not only about Gbytes/s . And Gddr6 is a high Latency Memory not really good for CPU Operations. So in real you have only 192 Bit for the GPU and the Rest goes to the CPU , not enough for native 4K gaming and 60 Fps , and not enough for the Zen2 CPU to utilize her full strength. In such Bottleneckarchitecture a Zen2 CPU is far more limited as in a PC Architecture/Motherboard.

So yes 16 Gbyte Ram and a 256 Bit Bus is good for the Buisness , but bad for the Games, bad for the Progress, and bad for my Eyes. So what have they done in the 7 Years at the R&D Headquarter?"
 
Feb 10, 2018
17,534
can we realistically deduce how much 8, 10 and 12tf rdna gpus with hardware raytracing are going to cost at the time of next gen consoles launch?
 

Tappin Brews

#TeamThierry
Member
Oct 25, 2017
14,869
Teraflops were always a horrible power metric. They're used for PR purposes because they're the biggest number available since they're just theoretical maximums. That doesn't tell you anything about what actual utilization looks like, and just as importantly, it doesn't tell you anything about the efficiency of the computation being done. Doing a bunch of math and later discarding the results shouldn't be cause for envy by anyone, but it all folds in under that nice, simple, quotable top line number. Sure, they can give you some rough idea of how two chips with an otherwise identical architecture might perform. Otherwise, they're marketing nonsense designed to grab headlines rather than inform.

Trying to convert TF numbers between architectures, even between iterations within an architectural family, is utter folly. Any engineer who got information from a console vendor quoting theoretical maximums after multiplied by some arbitrary conversion factor to make them look better would just shake her head in wonder at the stupidity of it all and look for something more meaningful. The theoretical maximum is still the theoretical maximum. It doesn't change. 12TF GCN is actually equivalent to 12TF RDNA for the purpose it was intended to measure in the first place.

What changes is how efficiently a given piece of code is going to run, and there's no fixed multiplier available that will hold true for all code. Each stage in each rendering pipeline will need to be re-evaluated to see whether decisions made because they were the fastest approach still hold true, and once you've started changing on GCN hardware compared to RDNA hardware, it becomes even less clear how to describe the relative difference in results. Yet that's all that really matters: results. What you can get on the screen in the 16 or 33ms you have to render a frame. Nobody can boil that down to "2x the power!" and have that statement hold water across a wide range of experiences.

... and yet that's what the market wants to hear. So you get cherry-picked numbers that show the best case, or the maximum theoretical, or some other nonsense. Don't worry about it. Most of us aren't going to have the time and background required to become experts and come to meaningful conclusions from marketing specs alone. Focus on the gaming experiences they enable, because then you can judge for yourself what's worth your hard earned dollars.

TLDR


how many TF are we talking here?
 

sncvsrtoip

Banned
Apr 18, 2019
2,773
One of the guys (Karmaprof) over at B3D made a pretty good point regarding unified memory architecture cause now it seems both PS5 and Anaconda will have 256-bit 16GB memory bus, i didn't think about this much before:

"If this the real Endproduct /Retail then good Night. It is the same Bottleneckarchitecture like on the Ps4. What People forget with an UMA Architecture is , you cannot have the full 256 Bit Bandwidth for the GPU because you must share it with the CPU , and a Zen2 CPU at 3 or 3,5 Ghz , 16 x Multithreading , need more Bandwidth then the oldJaguar CPU at 1,6 Ghz. It is not only about Gbytes/s . And Gddr6 is a high Latency Memory not really good for CPU Operations. So in real you have only 192 Bit for the GPU and the Rest goes to the CPU , not enough for native 4K gaming and 60 Fps , and not enough for the Zen2 CPU to utilize her full strength. In such Bottleneckarchitecture a Zen2 CPU is far more limited as in a PC Architecture/Motherboard.

So yes 16 Gbyte Ram and a 256 Bit Bus is good for the Buisness , but bad for the Games, bad for the Progress, and bad for my Eyes. So what have they done in the 7 Years at the R&D Headquarter?:
yep, 256bit but is small for 12tf and zen2 >3ghz, maybe they will use 18gbps as in some old rumour
 

Welfare

Prophet of Truth - You’re my Numberwall
Member
Oct 26, 2017
5,912
Holy shit, no one is going to convert RDNA and GCN flops in these comparisons.

Anaconda and Lockhart are being reported as 12 and 4. That's the final on paper number. They aren't secretly 10 and 3 or some shit.
 
Status
Not open for further replies.