Discussion RDNA4 + CDNA3 Architectures Thread

Page 273 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

DisEnchantment

Golden Member
Mar 3, 2017
1,754
6,631
136





With the GFX940 patches in full swing since first week of March, it is looking like MI300 is not far in the distant future!
Usually AMD takes around 3Qs to get the support in LLVM and amdgpu. Lately, since RDNA2 the window they push to add support for new devices is much reduced to prevent leaks.
But looking at the flurry of code in LLVM, it is a lot of commits. Maybe because US Govt is starting to prepare the SW environment for El Capitan (Maybe to avoid slow bring up situation like Frontier for example)

See here for the GFX940 specific commits
Or Phoronix

There is a lot more if you know whom to follow in LLVM review chains (before getting merged to github), but I am not going to link AMD employees.

I am starting to think MI300 will launch around the same time like Hopper probably only a couple of months later!
Although I believe Hopper had problems not having a host CPU capable of doing PCIe 5 in the very near future therefore it might have gotten pushed back a bit until SPR and Genoa arrives later in 2022.
If PVC slips again I believe MI300 could launch before it

This is nuts, MI100/200/300 cadence is impressive.



Previous thread on CDNA2 and RDNA3 here

 
Last edited:

maddie

Diamond Member
Jul 18, 2010
5,029
5,306
136
If AMD is so confused on pricing, after what I will assume was a well considered original price reflecting their place in users mindshare, cost to produce and acceptable margin considering the mentioned realities, then they are tacticians playing at strategy. They will lose it all for some extra upfront $. Fools.

Typical of most modern Western businesses.
 

Meteor Late

Senior member
Dec 15, 2023
266
291
96
I was comparing the perf/dollar of the 5070 vs the 4070 and 4070 super, not vs AMD. You said that AMD expected the 5070 to be more than $600. The 5070 at $650 would be worse than it's predecessor. I don't see how AMD could have been expecting that the 5070 to be worse than the 4070 super in perf/dollar.

Because AMD probably expected the 5070 to have much higher performance than what Blackwell seems to be.
 

GTracing

Senior member
Aug 6, 2021
275
645
106
4080 was worse than 3080 in perf/$
That's fair. Though I'd argue that the 4080 isn't in the same price tier as the 3080. They have a $500 price difference, not just $50. And high end cards are generally worse performance per dollar.

Nevertheless I agree that prices don't always go down.
 

adroc_thurston

Diamond Member
Jul 2, 2023
4,714
6,501
96

Josh128

Senior member
Oct 14, 2022
612
1,001
106
This doesnt hold and is contradictory in all directions, that s the other way around
as the 9070XT was supposed to be a 7800XT successor, so with something like 20% better perf at best.

Since that doesnt seem to be the case you can conclude that it s not AMD who was caught off guard, you can be sure that the 5070 was priced in this perspective, that is to face a 500$ card performing 20% better than the previous 500$ one.

Edit : The only pressure they could have is a price hike pressure if ever mlid leak has some substance.
It makes perfect sense and is not contradictory in any direction. How could Huang be caught off guard?! He went first, lol. Huang was not caught off guard whatsoever, he just chose not to milk the SKUs that barely see any hardware gains in favor of pricing the $5090 at $2K.

I never said anything about a $649 5070 at all, some other poster here claimed I said that, but I did not. I said AMD likely planned to price the 9070 non-XT at $599, which would have looked very good against a $599 5070 as it beats it quite handily in most perf. Likewise, a the 9070XT at $799 would have still looked fantastic against a 5070Ti at $799 (last gens price) or $849 (possible Huang milk).

If the above is not the case, why else would AMD cancel its 9070 announcement at CES last minute? It makes very little sense that they cancelled the whole event so they could jack the prices up, much more sense that they panicked when they saw the gen on gen drop in prices from Huang and pulled their announcements last minute.
 
Jul 27, 2020
22,298
15,554
146
I don't get it. The only way this works is if AMD holds an event the day after 5070 non-Ti/Ti reviews go online and tells everyone, hey surprise! The moment this event ends, retailers are gonna start taking orders and shipping the cards ASAP! And they decide the price point once they have done a comparative price/perf analysis of NV cards with their own. I just don't have faith in them that they can do a great job of making sure everything goes as planned. Unless, they know for sure that NV cards will go OOS in minutes and it may take weeks to replenish them which would give them a window for gamers who will figure, hey, 9070 XT is available now and it mostly does what is needed so why wait until who knows when the NV cards will be back.
 

Josh128

Senior member
Oct 14, 2022
612
1,001
106
That doesn't make sense to me. If Nvidia raised the price on the 5070 to $650 then it would have worse performance per dollar than the 4070 super.
  • The one benchmark Nvidia released shows a 21% lead for the 5070 over the 4070 in RE4 with RT on at 1440. (This is probably a best case scenario).
  • TPU's 4070 super review says that the 4070 is 13% slower than the 4070 super at 1440.
  • $650 is 8.3% more than $600.
Multiplying those together, a $650 RTX 5070 would have 3% worse performance per dollar.
I never said anything about pricing the 5070 at $650. If the price remained unchanged, AMD could have matched it and had quite better raster perf and still looked good. AMD would have absolutely LOVED to see 5070 @ $649, and even that would not have been an unreasonable expectation as it has marginally better raster and better AI BS.

Since when has Huang cared about losing the perf/$ by a few %?? Apparently not until this gen unless I missed something.
 
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/    |