Question Speculation: RDNA2 + CDNA Architectures thread

Page 77 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
uzzi38

uzzi38

Platinum Member
Oct 16, 2019
2,650
6,084
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
 
  • Like
Reactions: MagnusTheBrewer, lightmanek, Tlh97 and 2 others
Sort by date Sort by votes
Glo.

Glo.

Diamond Member
Apr 25, 2015
5,723
4,594
136
eek2121 said:
So where is the 4th? I'm seeing nothing in driver source code and little mention of it anywhere. There may be a 4th eventually, but I suspect AMD is finally planning on retiring Polaris, so we'll see a backfill of even lower tier cards.

EDIT: Speaking of which, AMD just added (probably initial) Van Gogh support to Mesa 4 hours ago.
Click to expand...
Because it only taped out recently.

P.S. 20 secs. Twitter search:
https://twitter.com/x/status/1300865710621626373

Even GFX name.
 
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,027
136
Glo. said:
Because it only taped out recently.

P.S. 20 secs. Twitter search:
https://twitter.com/x/status/1300865710621626373

Even GFX name.
Click to expand...
Glo. said:
Yes.
Click to expand...

That's likely something a bit different then. Current next gen products are gfx1030. Very interesting that @_rogame thinks it's Navi 24. I wonder if it's a mobile variant?

EDIT: IIRC Mobile variants of the 5000 series were all gfx1010.

For the record:

Code: 
  case CHIP_NAVI10:
      return "gfx1010";
   case CHIP_NAVI12:
      return "gfx1011";
   case CHIP_NAVI14:
      return "gfx1012";
   case CHIP_SIENNA_CICHLID:
   case CHIP_NAVY_FLOUNDER:
   case CHIP_DIMGREY_CAVEFISH:
   case CHIP_VANGOGH:
      return "gfx1030";
 
Upvote 0 Downvote
Glo.

Glo.

Diamond Member
Apr 25, 2015
5,723
4,594
136
eek2121 said:
That's likely something a bit different then. Current next gen products are gfx1030. Very interesting that @_rogame thinks it's Navi 24. I wonder if it's a mobile variant?
Click to expand...
Right.

videocardz.com

AMD NAVI 23 graphics processor gets GFX1032 ID - VideoCardz.com

AMD Navi 23 is confirmed Decoding AMD graphics chips is not an easy task. The more products AMD releases, the more likely they are to overlap with different architectures. Not all products are always up to date, so despite being the latest they actually feature last-gen architecture (such as...
videocardz.com videocardz.com

Navi 23 is GFX1032.
N24 is GFX1034.

The number simply indicates its a different, discreet die, from the same GFX family of GPUs.
 
  • Like
Reactions: Tlh97
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,027
136
eek2121 said:
That's likely something a bit different then. Current next gen products are gfx1030. Very interesting that @_rogame thinks it's Navi 24. I wonder if it's a mobile variant?

EDIT: IIRC Mobile variants of the 5000 series were all gfx1010.

For the record:

Code: 
  case CHIP_NAVI10:
      return "gfx1010";
   case CHIP_NAVI12:
      return "gfx1011";
   case CHIP_NAVI14:
      return "gfx1012";
   case CHIP_SIENNA_CICHLID:
   case CHIP_NAVY_FLOUNDER:
   case CHIP_DIMGREY_CAVEFISH:
   case CHIP_VANGOGH:
      return "gfx1030";
Click to expand...

The 5500XT was gfx1012. I wonder if we could be looking at a 6300XT?
 
  • Like
Reactions: lightmanek
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,027
136
Glo. said:
Right.

videocardz.com

AMD NAVI 23 graphics processor gets GFX1032 ID - VideoCardz.com

AMD Navi 23 is confirmed Decoding AMD graphics chips is not an easy task. The more products AMD releases, the more likely they are to overlap with different architectures. Not all products are always up to date, so despite being the latest they actually feature last-gen architecture (such as...
videocardz.com videocardz.com

Navi 23 is GFX1032.
N24 is GFX1034.

The number simply indicates its a different, discreet die, from the same GFX family of GPUs.
Click to expand...

Then you would be implying that Navi 23 isn't Sienna Cichlid .

According to Mesa, all of the current Navi chips are gfx1030.
 
Upvote 0 Downvote
Glo.

Glo.

Diamond Member
Apr 25, 2015
5,723
4,594
136
eek2121 said:
Then you would be implying that Navi 23 isn't Sienna Cichlid .

According to Mesa, all of the current Navi chips are gfx1030.
Click to expand...
Yeah, they are 1030...

Just like Navi 14 was 1010? Or was it 1012?

N21 is 1030. N23 is 1032, N24 is 1034.

The number at the end simply indicates that it is completely different die from others.

Just like Navi 14 was different die from Navi 10. Why is it so hard to understand?
 
  • Love
  • Like
Reactions: Tlh97 and spursindonesia
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,027
136
Glo. said:
Yeah, they are 1030...

Just like Navi 14 was 1010? Or was it 1012?

N21 is 1030. N23 is 1032, N24 is 1034.

The number at the end simply indicates that it is completely different die from others.

Just like Navi 14 was different die from Navi 10. Why is it so hard to understand?
Click to expand...

I'm just saying if that's the case, AMD hasn't updated the source code. I've seen no evidence anywhere else of this either. I'm not saying what you say isn't true, I just want to know where the information came from.

Right now it says that all of the following GPUs are marked as gfx1030, and AMD JUST had a PR merged 4 hours ago where this was updated:

Sienna Cichlid
Navy Flounder
Dimgrey Cavefish
Van Gogh

EDIT: Oh you got it from VideoCardz. Navi 23 -> gfx1032 -> RX 6500 Series so Navi 24 -> gfx1034 -> RX 6400 or RX 6300 and gfx1033 is van gogh. Got it.
 
  • Like
Reactions: Tlh97 and spursindonesia
Upvote 0 Downvote
Glo.

Glo.

Diamond Member
Apr 25, 2015
5,723
4,594
136
eek2121 said:
I'm just saying if that's the case, AMD hasn't updated the source code. I've seen no evidence anywhere else of this either. I'm not saying what you say isn't true, I just want to know where the information came from.
Click to expand...
Which, SPECIFIC GPUs in AMD code use those GFXfamily IDs in AMD's source code?

I don't believe you understand what you are looking at, or what you imply. If Sienna Cichlid and Navy Flounder are marked as both using GFX1030 - they are based ON THE SAME DIE.

The same die.
 
Upvote 0 Downvote
A

andermans

Member
Sep 11, 2020
151
153
76
Them both being marked gfx1030 is very likely just simplicity. Note that this is just the "CPU name" for LLVM. If both GPUs have the exact same instruction set, why bother calling out all the revisions? Just do the simplest thing that works, which is reusing gfx1030. VanGogh is a clear example of this.
 
  • Like
Reactions: Tlh97 and spursindonesia
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,027
136
Glo. said:
Which, SPECIFIC GPUs in AMD code use those GFXfamily IDs in AMD's source code?

I don't believe you understand what you are looking at, or what you imply. If Sienna Cichlid and Navy Flounder are marked as both using GFX1030 - they are based ON THE SAME DIE.

The same die.
Click to expand...

According to VideoCardz:

  1. Navi 21: GFX1030 - RX 6900 XT Sienna Cichlid (I hate typing that)
  2. Navi 22: GFX1031 (Mac Only, RX 6600M)
  3. Navi 23: GFX 1032: RX 6500 series
  4. Van Gogh: GFX 1033 (Zen2, RDNA2)
The information VideoCardz has could be inaccurate, but if they are right, Navi24 -> GFX1034 -> RX 6400/RX 6300.

According to the drivers, right now everything is 1030.
 
Upvote 0 Downvote
Glo.

Glo.

Diamond Member
Apr 25, 2015
5,723
4,594
136
eek2121 said:
According to VideoCardz:

  1. Navi 21: GFX1030 - RX 6900 XT Sienna Cichlid (I hate typing that)
  2. Navi 22: GFX1031 (Mac Only, RX 6600M)
  3. Navi 23: GFX 1032: RX 6500 series
  4. Van Gogh: GFX 1033 (Zen2, RDNA2)
The information VideoCardz has could be inaccurate, but if they are right, Navi24 -> GFX1034 -> RX 6400/RX 6300.

According to the drivers, right now everything is 1030.
Click to expand...
Considering that they marked N22 as Mac only, I'd say pretty much is wrong.

As I have said.

80 CU die - X900 SKUs
60 CU die - X800 SKUs
40 CU die - X700/X600 SKUs
smaller than that - X500 SKUs.
 
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,027
136
andermans said:
Them both being marked gfx1030 is very likely just simplicity. Note that this is just the "CPU name" for LLVM. If both GPUs have the exact same instruction set, why bother calling out all the revisions? Just do the simplest thing that works, which is reusing gfx1030. VanGogh is a clear example of this.
Click to expand...

Well, they would identify the specific revisions for cases that involve chip specific changes. If the chip behaves exactly the same and needs no further optimization, sure, it is fine to identify them all as the same.

Also, an interesting comment in a PR submitted for review today:

/* It's wasteful to enable all CUs for PS if shader arrays have a different
* number of CUs. The reason is that the hardware sends the same number of PS
* waves to each shader array, so the slowest shader array limits the performance.
* Disable the extra CUs for PS in other shader arrays to save power and thus
* increase clocks for busy CUs. In the future, we might disable or enable this
* tweak only for certain apps.
*/

So it sounds like they are more aggressively power gating to allow for higher clocks. I believe I mentioned some time ago that we'd see this.
 
  • Like
Reactions: lightmanek
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,027
136
Glo. said:
Considering that they marked N22 as Mac only, I'd say pretty much is wrong.

As I have said.

80 CU die - X900 SKUs
60 CU die - X800 SKUs
40 CU die - X700/X600 SKUs
smaller than that - X500 SKUs.
Click to expand...

We'll see. You are basically saying that the 6700XT is the leaked card today and will feature no performance improvements at all. EDIT: Unless a 2.3-2.4 GHz clock is enough to overcome the limitations of a 192 bit bus size for memory. The RX 5600XT did happen to scale really well with it's bus, after all.
 
Upvote 0 Downvote
Glo.

Glo.

Diamond Member
Apr 25, 2015
5,723
4,594
136
eek2121 said:
We'll see. You are basically saying that the 6700XT is the leaked card today and will feature no performance improvements at all. EDIT: Unless a 2.3-2.4 GHz clock is enough to overcome the limitations of a 192 bit bus size for memory. The RX 5600XT did happen to scale really well with it's bus, after all.
Click to expand...
That 40 CU die will be formidable opponent for RTX 3070. VERY formidable ;).
 
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,027
136
Glo. said:
That 40 CU die will be formidable opponent for RTX 3070. VERY formidable ;).
Click to expand...

Not with a 192-bit bus it won't. At least definitely not at 4K. NVIDIA is using a smaller bus size, but they have GDDR6X to make up for that. By all reports AMD is still using GDDR6. By comparison, the 1080ti and 2080ti both used a 352-bit bus.

According to NVIDIA:

1080ti -> 11gbps
2080ti -> 14gbps
3070 -> 16 gbps

AMD? 8gbps?

EDIT The RX 5700 XT apparently had 14gbps memory, so the 40CU RX 6000 equivalent will have 10 or 12 gbps.
 
Upvote 0 Downvote
K

Konan

Senior member
Jul 28, 2017
360
291
106
Glo. said:
Navi 23 - RTX 2080 Super performance levels +10%, @150W TBP.
Navi 22 - 10-20% above RTX 2080 Ti performance levels @225 TBP.
Navi 21 - 40-50% above RTX 2080 Ti performance @ 275W TBP.

Those were performance targets for RDNA2, that I got hint few weeks ago.
Click to expand...

What is the CU count next to each of these?
 
Upvote 0 Downvote
Glo.

Glo.

Diamond Member
Apr 25, 2015
5,723
4,594
136
Glo. said:
Navi 23 - RTX 2080 Super performance levels +10%, @150W TBP.
Navi 22 - 10-20% above RTX 2080 Ti performance levels @225 TBP.
Navi 21 - 40-50% above RTX 2080 Ti performance @ 275W TBP.

Those were performance targets for RDNA2, that I got hint few weeks ago.
Click to expand...
Konan said:
What is the CU count next to each of these?
Click to expand...
N21 - 80 CU
N22 - 60 CU
N23 - 40 CU.

I should also specify that I was specifically told the CU count/performance targets, and GPU die names are my addition, based on what I knew at that time.

I can also add what I have been told just today.

Quote, and read it as I write: "RX 5500 XT replacement will be roughly 60% faster than that GPU".
 
  • Like
Reactions: lightmanek, Tlh97 and spursindonesia
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,027
136
Glo. said:
N21 - 80 CU
N22 - 60 CU
N23 - 40 CU.

I should also specify that I was specifically told the CU count/performance targets, and GPU die names are my addition, based on what I knew at that time.

I can also add what I have been told just today.

Quote, and read it as I write: "RX 5500 XT replacement will be roughly 60% faster than that GPU".
Click to expand...

Promise me this. If you are wrong, will you record yourself eating a hat for the forums?

EDIT: I will at least attempt to eat a MAGA hat if someone sends me one...or at least I'll do as Ian Cutress did with an Intel wafer did and bite one... ;) I hope he didn't catch something from that. :D

Shoot if I didn't have crazy nutso disabilities, I would actually eat one ;)
 
Last edited:
  • Haha
Reactions: lightmanek and Tlh97
Upvote -1 Downvote
K

Kepler_L2

Senior member
Sep 6, 2020
353
1,279
106
Wild speculation: AMD might have actually went out of their way to troll/mislead leakers.

I believe that Navi23 (Dimgrey Cavefish) is actually the 80 CU, 505mm², 384b model we were expecting.

Now there is one main problem with this theory and that is Sienna Cichlid (80 CU, HBM2?) which most people assumed was Navi 21.

However it's common for some products to show up in drivers and never be released. Alternatively, Sienna could be a spin off of Navi 23 for the professional market.

Evidence of this theory include the Navi 21 die circulating on AIBs
https://twitter.com/x/status/1303890583123968000
being much smaller than expected
https://twitter.com/x/status/1303986125225095169
Navi 21 being 256b and Navi 22 192b (https://twitter.com/_rogame/status/1306655000454725636) and finally the fact that Dimgrey hexcode is #696969 (https://www.colorhexa.com/696969) because obviously Navi 23 is the 6900XT! ;)
 
  • Wow
  • Like
Reactions: lightmanek and spursindonesia
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom