Question Speculation: RDNA3 + CDNA2 Architectures Thread

Page 178 - 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
2,705
6,427
146

leoneazzurro

Golden Member
Jul 26, 2016
1,051
1,711
136
It’s just that if it doesn’t offer a significant perf/price jump then the reason to upgrade is simply not there, even if there is new encoding features.

So I think AMD will make the switch to N32 at some point, they’re simply not in a hurry because it is going to be roughly the same performance as the N21. Or they might as the speculation goes try to do respin of RDNA3. Which I’m less incli to believe.

You are completely right if you put yourself in the perspective of the user who needs to upgrade, but N32 has probably reasons to exist from the perspective of AMD/OEMs because it could have lower costs/more availability than N21 dies/cards. Also, rumors say that we should see a mobile version of N32, which was not possible with N21.
 
Reactions: Tlh97 and biostud

Kaluan

Senior member
Jan 4, 2022
504
1,074
106
That's... disappointing. I hope it at least has better perf/W then.
At a glance, RT capabilities seem to see the biggest gains. In the mildly ray-traced/hybrid Port Royal test, the 28 RA 7600S (clock limited?) @ ~100W TDP (in excess of it's official max TDP) is ~16% faster than than ~100W TDP 32 RA 6800S.

Not exactly overwhelmingly positive, but the somewhat positive synthetic results point towards it underperforming due to drivers in the actual games.

Hopefully we'll see more mobile N33 SKUs tested soon, preferably alongside Phoenix too.

Sad that the TUF A16 is the only AMD Advantage this time around. Maybe H2 2023 will see a newer crop of these types of designs.

Right now the GPU and laptop markets aren't exactly cheering me up. At least on desktop I finally settled on my core build.
7800X3D, EK-AIO 360 D-RGB, ASUS B650E-E Strix and 2R DDR-6000. Planning on keeping my Corsair Obsidian big tower and TX750M for now.
By the time I get them and put them together, RX 7800XT may be around the corner.
I have zero faith nVidia will ever offer any value for my hard earned money.
 
Reactions: Tlh97 and scineram

jpiniero

Lifer
Oct 1, 2010
15,176
5,717
136
You are completely right if you put yourself in the perspective of the user who needs to upgrade, but N32 has probably reasons to exist from the perspective of AMD/OEMs because it could have lower costs/more availability than N21 dies/cards. Also, rumors say that we should see a mobile version of N32, which was not possible with N21.

The costs of N32 should be lower than N21... but AMD might still not be thrilled about what the price would have to be.
 
Reactions: Tlh97 and Joe NYC

Hans Gruber

Platinum Member
Dec 23, 2006
2,303
1,216
136
Price probably comes out to about the same. If the rumors about a major hardware bug and delayed N32 were correct then expect June/July launch.

No sense in rushing it out before it's ready.
I think AMD answered the known hardware bug. They use firmware not software to disable certain features on cards that they intend to enable or use on future graphics cards. They may never enable some of those features on current models of the 7 series cards. I think they said they were intended for future RDNA 3 cards.

The sleuths found it and logic and reason escaped them. So they jumped to conclusions and said the hardware was flawed or broken all because certain hardware features were disabled by AMD.
 

Mopetar

Diamond Member
Jan 31, 2011
8,110
6,754
136
It doesn't make a lot of sense to disable hardware on your flagship silicon that will be enabled on the mid-range parts and conceivably increase the overall performance. That's idiotic.

Don't expect AMD to come out and directly say that they screwed up and that you're buying some defective silicon that needs to have circuitry disabled so it functions properly. Who is going to buy a product if they were told that?

We'll certainly know for sure whether those rumors have any merit by the middle of the year. If we see N32 parts that clock quite a bit higher or seem to punch above their weight-class relative to N31 then it should be pretty obvious that there was a major hardware bug and that AMD decided to scrap the original N32 parts to do another spin.
 

Leeea

Diamond Member
Apr 3, 2020
3,795
5,549
136
Don't expect AMD to come out and directly say that they screwed up and that you're buying some defective silicon that needs to have circuitry disabled so it functions properly. Who is going to buy a product if they were told that?
As long as it functions properly and is priced for its performance level it is not going to matter that much.

But yea, I suspect every piece of silicon for the last 10 years has some of that going on.
 

Mopetar

Diamond Member
Jan 31, 2011
8,110
6,754
136
As long as it functions properly and is priced for its performance level it is not going to matter that much.

It still does matter because the implication that eventually the bug will be fixed will get some people to put off a purchase. Better not to even acknowledge it publicly until you can actually fix it lest you create another Vega magic driver situation.

AMD gains nothing by talking about it unless and until it is fixed.
 

jpiniero

Lifer
Oct 1, 2010
15,176
5,717
136

MyDrivers claims no additional desktop cards until June at the earliest.
 
Jul 27, 2020
19,876
13,621
146
MyDrivers claims no additional desktop cards until June at the earliest.
They badly want the RX 6000 inventory to dry up so they can launch the new cards at the absurd price points they prefer. I'm guessing this is what they want:

7800 XT $800
7800 $700
7700 XT $600
7600 XT $500
7600 $400
7500 XT $300
7400 $200
 

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
29,576
24,453
146
Waiting until summer for any new launches is savvy. They already said demand is expected to be low this Q and next. If you are going to get an uptick in consumer purchasing here in the U.S., a big market, it will be when the kids get out for the summer and have more time to game. Parting them from their summer job money is a good play too.
 

Kepler_L2

Senior member
Sep 6, 2020
528
2,151
136
They badly want the RX 6000 inventory to dry up so they can launch the new cards at the absurd price points they prefer. I'm guessing this is what they want:

7800 XT $800
7800 $700
7700 XT $600
7600 XT $500
7600 $400
7500 XT $300
7400 $200
There's zero chance the 7600 XT is more than $400, and even that is pushing it. It will probably slot in at current N23 prices. Also 7500/7400 won't be a thing this gen.
 
Reactions: Tlh97 and scineram

Aapje

Golden Member
Mar 21, 2022
1,506
2,060
106
They badly want the RX 6000 inventory to dry up so they can launch the new cards at the absurd price points they prefer.

It's actually way more important for previous gen inventory to dry up if they don't ask absurd prices for the new, because then the old generation is no longer attractive at prices they want to ask.

For the 6900 XT this is less of an issue since the margins are so good that they can offer deep discounts. And of course, because the 7900 is not that good.
 

Mopetar

Diamond Member
Jan 31, 2011
8,110
6,754
136
I think it's less of a matter of AMD being strategic to clear out old inventory or that they want to catch a rising market wave, but just that they had an issue with N32 and decided to do a respin which has added several months to the release schedule to fix the issue, change the masks, get new silicon back to validate it, and all of the other steps that add up to a summer release.

I'm sure they would have much preferred that everything went to plan and that N31 was performing better and that N32 was already on shelves.
 

Aapje

Golden Member
Mar 21, 2022
1,506
2,060
106
@Mopetar

It seems way more likely that N32 is delayed due to a N31 respin than that (only) N32 needs and gets a respin.

@jpiniero

I think that you presume wrong. The cost is probably around $30 (CPU) + $80 (VRAM) + $20 (mobo) + $20 (cooler) + construction ($20). So that's $170, but then you also have shipping and other costs, plus the profit for the shop. So the minimum price is probably a bit above $200.
 

Mopetar

Diamond Member
Jan 31, 2011
8,110
6,754
136
N31 might also be getting a respin, but AMD couldn't have nothing at all for the end of 2022. I would imagine that the AIB partners might be slightly miffed about N32 being delayed if they'd done a considerable amount of work for those cards or had ordered parts on the assumption of them being launched by now and having to sit on that inventory.

The performance issues sound substantial enough to warrant a respin for N31 and AMD can always just use it as an excuse to refresh the existing cards. 7900 XT and XTX just replaced with 7950 models and AMD just lets their partners sell custom cards well above MSRP.
 
Reactions: Kaluan

HurleyBird

Platinum Member
Apr 22, 2003
2,759
1,455
136
It doesn't make a lot of sense to disable hardware on your flagship silicon that will be enabled on the mid-range parts and conceivably increase the overall performance. That's idiotic.

If something is bugged on the flagship but not on the lower end parts, then it makes perfect sense to have it enabled on the later. Unless you intentionally want those parts to perform worse or it's not worth putting in the resources into driver development that won't benefit the flagship.

There is precedent for lower end parts in the same series having more advanced features by virtue of being designed later. The most extreme example that comes to mind is HD 7790, a GCN 2 part (Bonaire) existing alongside an otherwise GCN 1 lineup, until Hawaii came out and it was repurposed into the R7 260/X.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |