Question Speculation: RDNA2 + CDNA Architectures thread

[uzzi38]

All die sizes are within 5mm^2. The poster here has been right on some things in the past afaik, and to his credit was the first to saying 505mm^2 for Navi21, which other people have backed up. Even still though, take the following with a pich of salt.

Navi21 - 505mm^2

Navi22 - 340mm^2

Navi23 - 240mm^2

Source is the following post: https://www.ptt.cc/bbs/PC_Shopping/M.1588075782.A.C1E.html

 

[eek2121]

Because audiences are monetized. I am not walking into that bleak topic, but if you want ad revenue you need views. To get views you need followers and you see where I am going here. It’s FUD culture and the noise drowns out the legitimate concerns too. Let’s just say that the current state of social media/media online does not impress me one bit.

Edit:
Re-read my post and I walked into it. Sigh.

YouTube is a cesspool of people competing for views and pennies. People who religiously follow folks like JayZ2cents should reexamine life.

 

[eek2121]

We have...
gc_num_se * gc_num_sh_per_se * gc_num_cu_per_sh
4 * 10 * 2 => 80 CUs for Sienna Cichlid (Navi21=GFX1030)
2 * 14 * 2 => 56 CUs for Arden (Arden=Anaconda=GFX1020)
2 * 10 * 2 => 40 CUs for Navy Flounder (Navi22=GFX1031)
1 * 12 * 2 => 24 CUs for Dimgrey Cavefish (Navi23=VN=MR=LH=GFX1032=GFX1033=GFX1040)
1 * 5 * 2 => 10 CUs for Navi24 (GFX1034)

Are you disputing AMD on the matter? Because AMD says you are wrong. Nostra, I love you, but this is why we ignore you.

The firmware was leaked. Oh sure, it could change, but the firmware is less than 2 weeks old and it was leaked less than a week ago.

 

[NostaSeronx]

Are you disputing AMD on the matter?

Um, those are what AMD is saying on this matter.

 

[Konan]

More than one person has corroborated the die size. AquariusZi is not the only person who says ~500mm^2 for Navi21. If he was, we'd wouldn't be so sure of it being this size.

Who is the 2nd person?

 

[A///]

Well, if AMD has 20% market share AND the limitless, insatiable, nay infinite demand for the latest nvidia products is what is causing their supply woes - unless AMD has 20% of infinity ready I think they are screwed. /s

I never go on Twitter, but when I do it’s because I follow links here. I did read his latest comments about how there were listening to feedback, etc. Realistically, what could the they really do? Either the containers are essentially already at customs/domestic ports now allowing for delivery in the next four weeks or they are allocating more resources to air freight because they think there is blood in the water. Flipping 2080ti type consumers for at least a year with Big Navi would be quite a coup and worth a ton of advertising bucks in terms of influencers on forums like these. IDK what to read into that.

I would think that if they thought they would be more likely to be competitive in the lower price tiers there isn’t as much margin to go crazy...

Here’s to hoping they have cards worthy of the hype, and more than a handful of them to sell!

The way I see it is AMD had some idea months or a long time ago of NVidia's progress and what was happening. Even if AMD had a 3090 killer they seem to take a relaxed approach to all of this. The lack of info id deafening now that we're roughly one month away. Perhaps they had foresight into the competition and knew they should order more.

I will be pleasantly surprised if AMD manage to launch a competitive card for slightly less that matches or beats NVidia at their own game sans fancy software. Drivers with minimal bugs, too. Card that runs cooler and with less watts. I'm reaching deeper than the stars here. The coming months should speak for themselves as NVidia supplies normalize over an extended period of time and we can better evaluate, as a community, whether NVidia delivered a solid product with no compromise or sold us a promise and left the next morning, taking out wallet with them.

Or AMD's Radeon group puts their head in a guillotine again.

 

[A///]

Lisa's gonna have his jewels if he can't back up what he tweets.

The comment about scrapping the 16k gaming was amusing.

She's already scalped his hair... Bad joke but yeah if AMD can't deliver enough product that will be bad. It really depends on how good RDNA2 is, of course. Sadly, that 16K gaming was take quite literally by a few people online. Wish people used common sense. I'd be more impressed if either company can provide a GPU in the future that does native 8K flawlessly at 80 FPS, rarely dipping below that. Now if I could afford an 8K television or monitor.

It was considered that AMD has no contention this time around by who, exactly?

S...heads of course, and by that I mean tech vloggers who reach real deep into themselves for their "sources."

 

[GodisanAtheist]

The way I see it is AMD had some idea months or a long time ago of NVidia's progress and what was happening. Even if AMD had a 3090 killer they seem to take a relaxed approach to all of this. The lack of info id deafening now that we're roughly one month away. Perhaps they had foresight into the competition and knew they should order more.

I will be pleasantly surprised if AMD manage to launch a competitive card for slightly less that matches or beats NVidia at their own game sans fancy software. Drivers with minimal bugs, too. Card that runs cooler and with less watts. I'm reaching deeper than the stars here. The coming months should speak for themselves as NVidia supplies normalize over an extended period of time and we can better evaluate, as a community, whether NVidia delivered a solid product with no compromise or sold us a promise and left the next morning, taking out wallet with them.

Or AMD's Radeon group puts their head in a guillotine again.

-Not sure I'd call AMD's approach relaxed at all. The level of guardedness they're approaching this launch with would make intelligence agencies jealous. Little scraps coming out of linux kernals here or there, strategically leaked to pique interest but not give too much away. Its really something else, reminiscent of the opsec involved in the Evergreen/HD5xxx series launch.

Honestly NV's launch this time around, despite or because of all the carefully staged damage control leaks, was uncharacteristically sloppy. They clearly were not happy with the product they were putting out to market, they knew it, and it showed.

Guess the real question is can AMD's best meet Nvidia's worst?

 

[A///]

-Not sure I'd call AMD's approach relaxed at all. The level of guardedness they're approaching this launch with would make intelligence agencies jealous. Little scraps coming out of linux kernals here or there, strategically leaked to pique interest but not give too much away. Its really something else, reminiscent of the opsec involved in the Evergreen/HD5xxx series launch.

I think I used the wrong word. I meant relaxed as in they're not collectively wetting the bed and spilling out like they did historically. I am not sure if this is a company wide directive issued by Dr. Su herself or a new AMD as a whole as they got rid of a lot of their old marketing people who put the influence one at their competitors who snatched them up, mainly NVidia and Intel. Prior to the Red Hat Linux kernel leak from this last summer, the only time I'd heard anything about Epyc was much, much earlier in the year before everything truly kicked off. Leaks on Ryzen have been scarce and they're so questionable I'm not willing to place much faith in them being true, even by people with a decent track record such as Wallossek.

Honestly NV's launch this time around, despite or because of all the carefully staged damage control leaks, was uncharacteristically sloppy. They clearly were not happy with the product they were putting out to market, they knew it, and it showed.

Oh good. Someone else who didn't think those leaks we had for months were actual leaks. It always felt like NVidia were controlling the narrative to get people really hyped up for their product.

Guess the real question is can AMD's best meet Nvidia's worst?

Indeed. There's been so much back and forth one what the cards are capable of doing, people convinced it's junk and DOA, or it'll be amazing. Honestly, while the rumors are very fluid the reality of it all isn't. We know nothing. It could also all be misdirection by AMD as some have said these last few months. If I really wanted to keep people on their toes and I was a CEO of a company like AMD, I'd definitely do that. Someone also stated that AMD is throwing out these canary traps to catch leakers in their company. Microsoft still does something like this or did numerous years ago under Ballmer.

Zen was leaky. I don't remember Zen+ being leaky. Zen 2 was fairly leaky and we knew a decent amount long before it was announced. RDNA was leaky. Zen 3 and RDNA2 is as silent as an extinct bird.


OT: Do you still use your Q9550?

 

[beginner99]

So instead of using high density libraries to increase density by ~25-50% to Renoir levels which can be used for more CUs, ROPs... they use high performance libraries to increase the frequency by ~20%? That's not a very good tradeoff in my opinion.

You are forgetting heat density. On renoir not really a problem with 8CU Vega. But Radeon VII and N10 already had some heat density issues. So going dense doesn't help if you then can't cool it. HP libraries will help with clocks and heat density.

And then what was speculated about before. The bigger die size could also be due to big caches which themselves might actually also help with heat density.

 

[GodisanAtheist]

OT: Do you still use your Q9550?

- Can't DM you for some reason so:

Nope, retired it from my wife's build last year with a Ryzen 2200G so she can have a "1 fan pc" that's quiet but still play some light sim games and power through office work.

I keep it and every other processor I've run either in a case or on my desk as a little paperweight. Those old Athlon 64's can hold down a ream of paper in a hurricane I tell you what.

 

[Kenmitch]

Honestly NV's launch this time around, despite or because of all the carefully staged damage control leaks, was uncharacteristically sloppy. They clearly were not happy with the product they were putting out to market, they knew it, and it showed.

Blasphemy! /s ?....unsure.

Guess the real question is can AMD's best meet Nvidia's worst?

Nvidia's worst? Isn't it the best they could do in reality? I guess Samsung takes the rap for any shortcomings?

Wait and see. It would be strange, but welcomed if AMD managed to pull off a perfect launch this go around.

 

[Timorous]

-Not sure I'd call AMD's approach relaxed at all. The level of guardedness they're approaching this launch with would make intelligence agencies jealous. Little scraps coming out of linux kernals here or there, strategically leaked to pique interest but not give too much away. Its really something else, reminiscent of the opsec involved in the Evergreen/HD5xxx series launch.

Honestly NV's launch this time around, despite or because of all the carefully staged damage control leaks, was uncharacteristically sloppy. They clearly were not happy with the product they were putting out to market, they knew it, and it showed.

Guess the real question is can AMD's best meet Nvidia's worst?

Ampere is no FX. Ampere is more like Fermi but not late.

Logically it would be sensible for AMD to go as dense as possible because that means fewer wafers needed to supply RDNA2 parts and more for Zen3 parts or APUs.

I cannot reconcile 500mm being needed for 80CUs and a large cache + narrow memory bus being faster than a wider bus + more CUs.

I think AMD are leaking nonsense.

 

[A///]

- Can't DM you for some reason so:

Nope, retired it from my wife's build last year with a Ryzen 2200G so she can have a "1 fan pc" that's quiet but still play some light sim games and power through office work.

I keep it and every other processor I've run either in a case or on my desk as a little paperweight. Those old Athlon 64's can hold down a ream of paper in a hurricane I tell you what.

I need to look into that. It's probably a setting.

Oh nice. I don't remember much about those quad cores and knew one person who had it. I knew a fair amount of people who'd bought the Q6600. I was in awe around that time. I got a late quad core, and then more quads. Went to hex and then dropped back to 4 because multicore wasn't really worth it around that time. I wish I could predict the future in terms of core use so I can make a sound purchase decision...

Blasphemy! /s ?....unsure.

There was talk of just how bad things were at Samsung prior to last summer (2019). Few believed in the goss but as time wore on it kept on being updated. I was hoping the goss was overblown and it seemed as much prior to the in depth testing on the 3080. The 3090 takes the cake when it comes to wasted energy though.

 

[Timorous]

You are forgetting heat density. On renoir not really a problem with 8CU Vega. But Radeon VII and N10 already had some heat density issues. So going dense doesn't help if you then can't cool it. HP libraries will help with clocks and heat density.

And then what was speculated about before. The bigger die size could also be due to big caches which themselves might actually also help with heat density.

When overclocked Renoir has higher power/mm² than any other current AMD product so I think not.

 

[TESKATLIPOKA]

You are forgetting heat density. On renoir not really a problem with 8CU Vega. But Radeon VII and N10 already had some heat density issues. So going dense doesn't help if you then can't cool it. HP libraries will help with clocks and heat density.

And then what was speculated about before. The bigger die size could also be due to big caches which themselves might actually also help with heat density.

I though about the heat density, but I am not sure about It. Is heat dissipation worse for my 120CU chip than the 80CU chip when it has comparable die size and most likely lower power consumption?

HP libraries will increase clocks but with It also power consumption.

I can understand that big caches can help with heat concentration, but what about performance? Can 128MB L2 cache compensate for a missing 128-256bit bus? You can't really use It as a framebuffer, It's not nearly big enough.
3840x2160 * (64bit(HDR color dept)+32bit(z-buffer)) * 3(Tripple buffering) /8 = ~300MB (If you want MSAA then you multiply It by 2x, 4x, 8x)

 

[TESKATLIPOKA]

I don’t care about die sizes. Right now we have actual proof of specs from AMD regarding 3 of the 4 chips.

EDIT: AMD leaked the firmware of the actual cards via ROCm and a Linux PR and then later tried to retract (via git, rofl). Thus happened just a couple days ago.

But I do care about both die size and CU count.
And how do we know If what was written in firmware is actually correct and It wasn't a deliberate move from AMD?

 

[beginner99]

And how do we know If what was written in firmware is actually correct and It wasn't a deliberate move from AMD?

That would be next-level mind-games by AMD. Tell the engineers to intentionally put wrong stuff into the commits to confuse the competition and leakers. I like it actually. Would be funny if true.

 

[coercitiv]

And how do we know If what was written in firmware is actually correct and It wasn't a deliberate move from AMD?

You know because they do this with a purpose, and that purpose is linked to engineering needs, not marketing. AMD determined silence is their desired strategy this time around, this excludes purposely disseminating false information.

No poor Volta, no overclocking dream, no borderline fake efficiency demo, no 150W TDP promise, just awkward silence and emoticons.

 

[DisEnchantment]

That would be next-level mind-games by AMD. Tell the engineers to intentionally put wrong stuff into the commits to confuse the competition and leakers. I like it actually. Would be funny if true.

Nobody will allow to upstream it. And it also means you miss the kernel merge windows
Linus Torvalds will not spend his weekend to merge some fake commits from any Vendor.
The ramifications for this would be that AMD will never be allowed to upstream and they can only work on their stuff downstream. Like NVIDIA.

For the end user it means, for any distro they use, it won't work out of box.
They would need to install some debian/rpm packages from somewhere to make things work.
And this defeats the whole purpose of Open Source. Might as well go NVidia way and just deliver rpms/debs/binaries.

It also means other non AMD contributors to AMD source code cannot work.
Valve, Intel, Google, Microsoft, Oracle, Redhat, etc contribute to AMD code, imagine if you push fake things.

 

[coercitiv]

Occam's razor ruining all the fun...

 

[TESKATLIPOKA]

Ok, then 80CU is max.

 

[A///]

Even then the open source world monitoring people do doesn't paint much of a picture. At least you can buy a foot warmer this Christmas if you fancy NVidia more.

You know because they do this with a purpose, and that purpose is linked to engineering needs, not marketing. AMD determined silence is their desired strategy this time around, this excludes purposely disseminating false information.

No poor Volta, no overclocking dream, no borderline fake efficiency demo, no 150W TDP promise, just awkward silence and emoticons.

I sometimes feel this was heavily Raja influenced. A few weeks ago he posted a photo of him holding a bottle of pure capsicum gel. You know. Spicy. The moment certain people left AMD's marketing, the BS died down. Intel and NVidia hired those people.

 

[maddie]

You are forgetting heat density. On renoir not really a problem with 8CU Vega. But Radeon VII and N10 already had some heat density issues. So going dense doesn't help if you then can't cool it. HP libraries will help with clocks and heat density.

And then what was speculated about before. The bigger die size could also be due to big caches which themselves might actually also help with heat density.

I think the heat density issue is a CU level problem to solve, not a total die area problem. Heat density doesn't increase with more CU as the circuit density is the same. More heat over a proportionally greater area is not a cooling problem. The arrangement of high load/power circuits close together is what causes hot spots. This is independent of CU numbers.

 

[Hans Gruber]

Look at the power usage chart. Notice the 5700 and 5700xt compared to the 3080. If AMD comes in anywhere close to those plus 50w. That will be a win against the 3080 on power consumption alone. Nvidia set the bar very low for excessive power usage on the 3080.

NVIDIA GeForce RTX 3080 Founders Edition Review - Must-Have for 4K Gamers

NVIDIA's new GeForce RTX 3080 "Ampere" Founders Edition is a truly impressive graphics card. It not only looks fantastic, performance is also better than even the RTX 2080 Ti. In our RTX 3080 Founders Edition review, we're also taking a close look at the new cooler, which runs quietly without...

www.techpowerup.com

 

[kurosaki]

Valve, Intel, Google, Microsoft, Oracle, Redhat, etc contribute to AMD code, imagine if you push fake things.

So You're saying that AMD has managed to conspire together with Valve, Intel, Google, Microsoft, Oracle, Redhat AND Linus Torvalds etc to contribute with phony AMD code? This might be huge! 240 CU, 150TF, 6900XT confirmed!

#The_Linux_Conspiracy_2020