Question Speculation: RDNA2 + CDNA Architectures thread

Page 52 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

uzzi38

Platinum Member
Oct 16, 2019
2,627
5,929
146
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
 

wanderica

Senior member
Oct 2, 2005
224
52
101
And they are going to do it again. AMD is going to squeeze every last dollar worth of perf out of their silicon. If you think they are going to give you a huge die that runs at low voltage to save power in a desktop setting, you are dreaming. Opportunity cost is too high when they can use those 7nm wafers for CPUs with much smaller dies and no board, RAM, or cooler to pay for.

I"m not so sure. While it's true AMD has done this often in the past (Hawaii - 14nm Vega), I don't get that feeling this time. I'll give you, that's a lot of history to erase, but so was Bulldozer (by the way, such a shame that those ended up so bad. What an awesome codename for a chip!) to Ryzen. I hold no illusions that AMD is going to give us amazing overclockers like Maxwell, but at the same time, if the new consoles are any indication, then this is a seriously efficient architecture. It may be pushed to its efficiency limit, but it won't be pushed past its limit like AMD has done in the past. Besides, however AMD decides to do it this round, it will look just fine compared to Nvidia's efficiency offerings. And that hasn't happened in a very long time.
It may be a bit of an unpopular opinion given AMD's history, but I take their silence as confidence. The last time AMD was silent about a launch, we got Ryzen. Hopefully, history repeats itself here, and we (finally!) get some competition at the high end. I'm due for an upgrade. My 1080Ti is getting pretty long in the tooth. I'd love to go back to AMD this time, but I'll follow the best deal, and Nvidia might be hard to beat this round with the 3080.
 

Saylick

Diamond Member
Sep 10, 2012
3,141
6,357
136
Someone mentioned here before that AMD's old Radeon marketing team, the one that brought us "Fury is an overclocker's dream" and "Poor Volta", is not the same marketing team today, which jives with Lisa's management style of "speak softly and carry a big stick", i.e. if you got the big guns, then one does not need to boast because the product will do the talking. We don't see AMD teasing Zen 3 even though Intel is still the clear market leader, do we? Same goes for Radeon. In the last few years, AMD has been nothing but a tight ship and they've been hush-hush more so than ever on their future products. I'd argue that being as secretive as possible is more important when you're on the cusp of launching a competitive product, and even more so if you are the market leader. You can boast all you want if you're an underdog and it won't be taken seriously, but it's when you can be an immediate threat or don't want to lose an advantage that you want to underplay your hand.
 

senseamp

Lifer
Feb 5, 2006
35,787
6,195
126
Like Mike Tyson said, everyone has a plan until they get punched in the face. Once you are in the fight and the main lever you got is voltage, your power is going to be what it needs to be for your competitiveness.
 
  • Love
Reactions: lightmanek

Kenmitch

Diamond Member
Oct 10, 1999
8,505
2,249
136


I hope this is something and not nothing...


**** Please let it be the big navi reveal...

What are the odds of it partially coming true? That would make me almost a fortune teller.

I'll guess annonced 9/9 with 9 total skus split between cpu/gpu. Available 10/10 with a 10th hail mary jaw dropper sku....lol

And no I won't be holding my breath!
 

kurosaki

Senior member
Feb 7, 2019
258
250
86
But Frank Azor is also an acronym for: -Redacted- Razor = -Redacted- RYZEN MEGA BEAST!
There you have it! This and HL:3 will be revealed today. Bundled with RDNA2 big Navi and Valve index 2, retina 10k resolution wireless VR-headset. The bundle will cost 299 usd.

Source

Profanity is not allowed in the tech forums.

Daveybrat
AT Moderator
 
Last edited by a moderator:

blckgrffn

Diamond Member
May 1, 2003
9,123
3,064
136
www.teamjuchems.com
Wow, solid twitter drop. Easy graphic, no BS. Pre order link even.

Now to decide if I want to line up via pre order or not... I wish there was even a iota of info on whether the One series TV tuners were supported. Makes all the difference to me, sadly.

Come on AMD, let's see it!
 

Kenmitch

Diamond Member
Oct 10, 1999
8,505
2,249
136
woow $499 for the whole console... so i can expect AMD prices to drop acorderly and people stop repeating "7nm is expensive BS" when?

I thought the console APU's are low margin, very little profit, in it for the R&D dollars, offerings.....At least that's the way it went on the forum when Nvidia was arsed out of it. Ignoring their week SOC part of it of course.

We'll have to wait and see what AMD drops and then decide. I'm pretty sure you can't predict the price point of their dGPU's based on the cost of a console.

No idea what a 7nm wafer costs in the end. Most end users don't factor it in when making a buying decision.
 

Martimus

Diamond Member
Apr 24, 2007
4,488
152
106
Microsoft caved in. Guess it's Sony's turn now. How this effects AMD's announcements is unknown. I guess maybe a time frame?


That's awesome! I thought I would have to drop $600 on it, so I'm happy with the reduced price. I'll probably never use it, but I need to replace the XBOX One S for my daughter after my ex took it when she moved out.
 

Konan

Senior member
Jul 28, 2017
360
291
106

leoneazzurro

Senior member
Jul 26, 2016
927
1,452
136
Yeah.

Navi 23 will be just behind RTX 3070, and Navi 22 will be just behind RTX 3080. That is correct.

Guys, use your logic. Look at RX 5700 XT and its performance per ALU/CU compared to Turing, add redesigned caches, which increases internal bandwidth massively, add 25% core clocks(2.3 GHz) and 10% IPC.

Then use those equations for 50% bigger GPU, with 60 CUs, and then for 2x bigger GPU in the mix.

Im sometimes baffled that even if people have those things so plain, in their faces, they fail to see them.

And Im not going to post where does the info about bus width, and memory sizes comes from.

If 40 CU RDNA1 GPU was able to trade blows with 40 SM Turing GPU, why would 40 CU GPU(RDNA2) with increased IPC lose to a 44 SM GPU that LOST IPC(Ampere)? Especially with masively increased clock speeds, on RDNA2 GPU?

P.S. I think we should stop calling ALUs in GPUs cores, and call SM's, CU's, EU's cores. That way will be factually correct, and easier for laymans to understand the performance of GPUs, instead of marketing gibberish that all of companies are trying to sell us.

According to a scaling model of the RDNA architecture I made based on techpowerup reviews numbers , a theoretical 2,2 GHz 80CU Navi21 should be around 60% faster than 2080Ti, a 2,2 GHz Navi 22 20% faster than a 2080 Ti, the Navi 23 at same clockspeed should be in 2080 territory.