• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Question Speculation: RDNA3 + CDNA2 Architectures Thread

Page 159 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
TESKATLIPOKA said:
N33 has 32CUs or 16WGPs, the same as N23 and that one had 2048 stream processors. N33 will have the same number of them, but dual issue(dual pumped).
Performance will be higher than N23.

If It had only 16CU(8 WGP) adding up to 1024 dual pumped SP, then It wouldn't match N23, but It wouldn't need 200mm2 die either.
Click to expand...

-True, I think it was also initially rumored as a 4096 SP part that got scaled back to 2048 with dual issue, so I doubt we'll see any more surprises in SP counts.
 
GodisanAtheist said:
-True, I think it was also initially rumored as a 4096 SP part that got scaled back to 2048 with dual issue, so I doubt we'll see any more surprises in SP counts.
Click to expand...
Then why bring up useless 1024 SP hypothesis that never had any basis? lol
We've known for ages how many and there haven't been any changes to that, unlike N31 or N32.

Also I don't remember people calling AD102's 10754 being "scaled back" to 5376 after technical data was out. It's just that AMD chose to predominantly use the lower number. Which is why the lower number got traction in conversations. That's not "scaled" back, it's akin to semantics. Ampere and Lovelace is not Turing and RDNA3 is not RDNA2.

Once SMT became a mainstay of x86, we started counting in threads too, but we didn't stop counting cores.
Pardon the analogy lol
 
Kaluan said:
Then why bring up useless 1024 SP hypothesis that never had any basis? lol
We've known for ages how many and there haven't been any changes to that, unlike N31 or N32.

Also I don't remember people calling AD102's 10754 being "scaled back" to 5376 after technical data was out. It's just that AMD chose to predominantly use the lower number. Which is why the lower number got traction in conversations. That's not "scaled" back, it's akin to semantics. Ampere and Lovelace is not Turing and RDNA3 is not RDNA2.

Once SMT became a mainstay of x86, we started counting in threads too, but we didn't stop counting cores.
Pardon the analogy lol
Click to expand...
RDNA3 and Ada, Ampere are pretty similar in that they doubled FP32 units, while only half is capable of executing also INT32.
Yet, Ampere saw a lot bigger IPC(PPC) gain than RDNA3. Why? Is this truly only driver fault or Nvidia is using more HW.
 
Kaluan said:
Well, that was that was my whole point, can't compare game clocks to boost clocks (the former are more realistic in mobile setups anyway) or cut die (6600M) to full die (7600M/7700M?)
(again, under similar TGP config, less silicon needed to be powered up likely means higher clocks)
Didn't say your numbers are wrong, I said the reasoning is. Apples to oranges IMO.
That would also more easily explain the performance rumour/projection.
But we still don't know for sure if that leaker was talking about mobile for sure lol
Click to expand...
I compared boost full N23 vs boost full N33, because I thought he was talking about boost clock.

We still don't know If he was talking about boost or game clock, so you can't confidently say he was talking about game clock. With this in mind, you can't really say I was comparing apples to oranges.
We don't even know If It's 2700MHz, he said comparable to RX 6650XT AIB and I just guessed this frequency.

I will compare full N23 vs full N33:
Boost clock: 2416MHz vs 2700MHz (+12%)
Game clock: 2162MHz vs 2700MHz (+25%)
If I consider that he said desktop RX 6750XT level of performance on average, then this clock really should be game clock, It would be just a bit higher(+8%) than 2495MHz for RX 6750XT.
The rest has to come from the architecture ad that would be ~20%.
That's a lot.
 
Last edited:
TESKATLIPOKA said:
RDNA3 and Ada, Ampere are pretty similar in that they doubled FP32 units, while only half is capable of executing also INT32.
Yet, Ampere saw a lot bigger IPC(PPC) gain than RDNA3. Why? Is this truly only driver fault or Nvidia is using more HW.
Click to expand...

Strangely enough, from my findings, compute uplift and meta launch day gaming performance uplift (3Dcenter) shows almost identical gains for both Turing to Ampere and RDNA2 to RDNA3:

3090 to 2080Ti and 7900XTX to 6900XT:
~2,65x compute and ~47% raster 4K both, ~58% and ~61% 4K RT 3090 and 7900XTX.

Bare in mind both are competing with their previous gen w/ pristine drivers. Versus launch day 2080Ti and 6900XT performance would be different, but that's beyond the scope of my comment.

The main difference is RDNA3 arguably saw a bigger node jump from RDNA2 than than Ampere saw from Turing. And AMD clearly planned for bigger gains. Hopefully it's not all down to hardware mishaps and they can improve the situations by drivers.

TESKATLIPOKA said:
I compared boost full N23 vs boost full N33, because I thought he was talking about boost clock.

We still don't know If he was talking about boost or game clock, so you can't confidently say he was talking about game clock. With this in mind, you can't really say I was comparing apples to oranges.
We don't even know If It's 2700MHz, he said comparable to RX 6650XT AIB and I just guessed this frequency.

I will compare full N23 vs full N33:
Boost clock: 2416MHz vs 2700MHz (+12%)
Game clock: 2162MHz vs 2700MHz (+25%)
If I consider that he said desktop RX 6750XT level of performance on average, then this clock really should be game clock, It would be just a bit higher(+8%) than 2495MHz for RX 6750XT.
The rest has to come from the architecture ad that would be ~20%.
That's a lot.
Click to expand...
Yeah that's what it came down to for me, what clocks are we comparing.

I wonder if desktop N33 is the same silicon as these soon the be announced laptop chips or newer. And how much further could a 150-180W TGP SKU clock from the specs we'll see at CES.
 
Slightly off the current topic, but has anybody else been keeping track of stock for the 7900 XT and XTX?

The XT is coming back in stock pretty much every day. NowInStock shows new cards arriving all the time. However, the XTX is almost never coming in stock.

This suggest that the vast majority of cards being built are XTs. This is contrary to how most people though the 7900 XT only existed to upsell people to the XTX. But maybe the XT is priced so close because AMD knew that XTX cards were going to be manufactured in much lower quantities, so price the XT a bit higher to make up some of the difference?
 
Stuka87 said:
Slightly off the current topic, but has anybody else been keeping track of stock for the 7900 XT and XTX?

The XT is coming back in stock pretty much every day. NowInStock shows new cards arriving all the time. However, the XTX is almost never coming in stock.

This suggest that the vast majority of cards being built are XTs. This is contrary to how most people though the 7900 XT only existed to upsell people to the XTX. But maybe the XT is priced so close because AMD knew that XTX cards were going to be manufactured in much lower quantities, so price the XT a bit higher to make up some of the difference?
Click to expand...
Been on the hunt for a Sapphire Nitro XTX. Nowinstock hasn't been close to accurate for drops. There are more XT's available than XTX's but I know of no one that wants an XT. Ref XTXs and Merc310s pop in and out of stock all day, every day. I still have mercs sitting in my cart at bestbuy. Amazon has red devils today as well.

There's just not a lot of reason to buy an XT if you already have a 6800xt or better and if you're on something older there's no point in buying the XT over the XTX.
 
nurturedhate said:
Been on the hunt for a Sapphire Nitro XTX. Nowinstock hasn't been close to accurate for drops. There are more XT's available than XTX's but I know of no one that wants an XT. Ref XTXs and Merc310s pop in and out of stock all day, every day. I still have mercs sitting in my cart at bestbuy. Amazon has red devils today as well.

There's just not a lot of reason to buy an XT if you already have a 6800xt or better and if you're on something older there's no point in buying the XT over the XTX.
Click to expand...

My reasoning for edging towards the XT over the XTX is power consumption. A card with 3x8pin requires me to get a new PSU. Which makes the $100 prices increase of the XTX to actually be more like $200.

And I have been looking manually as well. The XTs do sell out, and then get new stock the next day. And while NiS has not been totally accurate for the XTX, manual searches show its also not getting replenished nearly as often.
 
Stuka87 said:
My reasoning for edging towards the XT over the XTX is power consumption. A card with 3x8pin requires me to get a new PSU. Which makes the $100 prices increase of the XTX to actually be more like $200.

And I have been looking manually as well. The XTs do sell out, and then get new stock the next day. And while NiS has not been totally accurate for the XTX, manual searches show its also not getting replenished nearly as often.
Click to expand...
Needing the Sapphire specifically due to case limitations, it's the shortest at 320mm. Going to need to replace my carbide 540 next build it seems.
 
nurturedhate said:
Needing the Sapphire specifically due to case limitations, it's the shortest at 320mm. Going to need to replace my carbide 540 next build it seems.
Click to expand...

I did not go with that case for a prior build specifically because of the maximum GPU length limits. Its a nice case otherwise. But really like my current O11.
 
Stuka87 said:
Slightly off the current topic, but has anybody else been keeping track of stock for the 7900 XT and XTX?

The XT is coming back in stock pretty much every day. NowInStock shows new cards arriving all the time. However, the XTX is almost never coming in stock.

This suggest that the vast majority of cards being built are XTs. This is contrary to how most people though the 7900 XT only existed to upsell people to the XTX. But maybe the XT is priced so close because AMD knew that XTX cards were going to be manufactured in much lower quantities, so price the XT a bit higher to make up some of the difference?
Click to expand...
Here in Finland there has been barely any XTX cards available (not many has been sold because the stock has been really low). Reference XT never sold out and people aren't buying.
 
nurturedhate said:
Been on the hunt for a Sapphire Nitro XTX. Nowinstock hasn't been close to accurate for drops. There are more XT's available than XTX's but I know of no one that wants an XT. Ref XTXs and Merc310s pop in and out of stock all day, every day. I still have mercs sitting in my cart at bestbuy. Amazon has red devils today as well.

There's just not a lot of reason to buy an XT if you already have a 6800xt or better and if you're on something older there's no point in buying the XT over the XTX.
Click to expand...
While availability for XTX is indeed a problem (my native country still has NO XTXs in stock since launch whatsoever), there's also the by-product of that, the pricing. For the most part XTs are hovering around MSRP, even saw 1 or 2 listings at under MSRP. While XTXs are constantly above MSRP, some even with almost pandemic-like inflated prices.

In relation to the 7900s, AMD needs to prioritize 2 things, drivers and stock outside the US (a AMD policy that should end anyway in a week after the stupid US tariffs kick in). With possibly discontinuing the MBA until the issues are fixed being a 3rd.
 
Ah AMD, how did they go from a very solid RDNA2 gen to a total dumpster fire of an RDNA3 generation...

I mean if they literally did nothing with the cooler... It would have been fine?
 
GodisanAtheist said:
Ah AMD, how did they go from a very solid RDNA2 gen to a total dumpster fire of an RDNA3 generation...

I mean if they literally did nothing with the cooler... It would have been fine?
Click to expand...
Now I know why I liked the RDNA2 cooler better than the new one.
It could actually cool properly. 😀
 
7900xt has been in stock constantly since launch on amd.com for Canada, while 7900xtx has been out of stock the entire time. This is from my own manual checking, so obviously there are gaps in my sampling, but I check about 5 times a week.
 
adamge said:
7900xt has been in stock constantly since launch on amd.com for Canada, while 7900xtx has been out of stock the entire time. This is from my own manual checking, so obviously there are gaps in my sampling, but I check about 5 times a week.
Click to expand...
10am est most weekdays are when they have xtx's for sale on amd.com
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top