Lets gooo!

According to techpowerups GPU database Navi 32 goes up to 64 compute units (up from Navi 22 40 compute unit) though the 6800 XT was a 72 compute unit GPU so I would guess that 64 compute units just won't cut it for a 7800 XT. Though the 7800 non XT could be based on Navi 32 with an upgrade to 64 compute units from 60. If there was a big enough performance jump from RDNA 2 to RDNA 3 was in the 50-70% range then I could see them doing a 64 compute unit 7800 XT and getting away with it though things didn't work out that way.

7800XT will most possibly be a further cut-down N31. The successor of 6800 non-XT. And will be a bit faster than 6900XT. Price will make or brake it.

I'm thinking these are Navi31, heavily chopped down.

Navi32 wasn't even supposed to launch yet, and there's strong evidence it has been delayed from the original schedule of ~September so we might not even see it until Christmas.

Everything that everyone says is purely speculation. Some people have stated that there would be no 7800XT when the 7600XT was the only card AMD was launched. We don't know the specs so we might as well wait for official information from AMD.

Am I the only one that thinks that the Igor's Lab simulation, while fascinating for how it was contrived, is kind of misleading for several reasons?

I haven't looked into if this is still the case, but pro parts used to run at a fixed clock. On top of that, they're likely using the parts with lowest leakage (the opposite of what they need for a gaming part).

I understand it's a rough approx., and value that he does stuff like this, but still.

IMO, and I could be wrong, Navi32 will be clocked pretty high (relatively speaking). I think ideally they would like/would have liked 2900mhz on 5nm. Who knows what it is ACTUALLY using. Let me explain.

5nm has shown to yield somewhere >2500-<2850mhz @ ~.88-.99v for a gpu reliably before power consumption goes bananas.

I think AMD was planning for 1.1v(~+) scaling (like Apples's 3.24ghz) on 5nm. So was nVIDIA, but it's more-so important for AMD.

People chastise AMD for the '3ghz+' design, but I truly think this philosophy included 5nm scaling with voltage/power consumption to 1.1v with a planned shrink to N4P/X (1.2v). Look at die size for a clue: If N31/N32 utilized these processes they would be the smallest 256-bit/384-bit designs possible (192mm for N32, just like RV670). The clock speed/pipeline is already in place, no need for extra decap; just a shrink to capitalize on better power characteristics and voltage scaling (22% better power scaling according to TSMC, which probably equates to 11% higher on the curve). I think AMD planned for 1.1v and then refreshing to 1.2v. In reality they got 1v for decent power scaling, N4P may ACTUALLY be 1.1v (or so), and N4X 1.2v for GPU/HPC designs.

While N4X may or may not come too late, N4P has been in production for quite some time. Apple achieved what would equate to 2930mhz @ 1v (efficient) and 3460mhz @ 1.2v (max).

If we are to believe what TSMC is saying, N4X would smooth the line to be something like 3000mhz @ 1v, 3300mhz @ 1.1v, 3600mhz @ 1.2v.

It would make sense if 5nm doesn't yield 2900+ reliably with decent power scaling/yield to put N32 on N4P, as it would then not only be more power efficient/yield better but actually, you know, be possible.

This way, they could use the same chip for both ~2900/20000 and/or as high as they can clock it reliably paired with 24000 GDDR6 (probably the intention for refresh on both parts).

For reference this is my personal belief on optimal 'pairing'. Numbers are approximate, but pretty close wrt bandwidth limitations:

12288sp/96MB L3/384-bit/20000: 2720mhz. Theoretically a very achievable clock/yield on 5nm. It is, in-fact, the Ada stock clock. In reality (MBA) 7900XTX was 2631 avg according to Wiz' review.
12288sp/192MB L3/384-bit/20000: As high as 5nm was ever going to clock. Cancelled because it didn't clock/yield (scale with decent power consumption) well-enough to need it.
12288sp/96MB L3/384-bit/24000: 3264mhz. This would work if 5nm scaled well-enough OR for a N4P/N4X refresh, depending on time table of manufacturing. Tons of options to get there. The 'ideal'.
10240sp/80MB L3/320-bit/20000: The same ratio as full N31 because 5/6 design. Very similar to a full AD103 but using larger bus/cache instead of faster ram. We never got this part.
10752sp/80MB L3/320-bit/20000: 2591mhz. This is close to what we got with 7900 XT. Probably set up this way because clock yields and/or power consumption/clock scaling was bad.
7680sp/64MB/256-bit/20000: 2900mhz. This works theoretically both as best-case (realistically) on 5nm or worse-case on N4P/X. Since it is a smaller chip, makes sense to put it high on the curve or test with.
7680sp/64MB/256-bit/24000: 3480mhz. A clock in this vicinity would have always required N4P/N4X. It would be very similar to a 4070ti/6950XT.

I think when you look at it like that, it all starts to make sense...especially if you figure that designs would've been plotted before they knew what 5nm would actually achieve.

I'm not trying to start rumors, get anyone's hopes up with copium/hopium, nor be a conspiracy theorist. Only trying to put things in perspective. This is the only way it makes sense.

It's possible N4P too doesn't perform to expectations for the high-end; it's possible N4X came too late compared to N3E (and what's possible there with early yields). It's possible N32 ended up a successful or unsuccessful pipe-cleaner for N31. It's possible it's not. It's possible they could get N5 to the ideal 20000 N32 level with what they believe is acceptable power/yield. It's possible the part will be clocked low; in reality while ~4070ti/6950xt is an ideal for the design, they really only NEED an ~6800xt/4070 (a PC equivalent to the PS5pro) at an equal/lesser and sustainable price, which they can't do with N21. nVIDIA made this clear with how they positioned the 4070 in performance. AMD may make this clear with price (in which a lower avg perf = lower cost to make). It's sad, but I think everyone knows this chip has to compete in price with 4060 ti in price even though it will perform much better.

Who knows what will happen, but I hope AMD is able to capitalize on the promise of the design before it becomes irrelevant, because it *could* be quite clever if it ever works out, process/time permitting, and prove their design methodology. Small chips, high clocks, but not at the expense of ridiculous power consumption. Split processes, same performance tier, and cheap. Sadly, allowing the lower price ratio people will pay for AMD GPUs with similar performance to one from nvidia (essentially they have to compete one tier down).

That timetable too is arguable, especially since they just released a 6nm GPU and are releasing a 7nm CPU which do (7600 being similar to a PS5 gpu) and likely will (allow budget stock AM4 users to use a high-end GPU) fulfill their purpose. Perhaps those refreshes/performance levels will still happen with these designs, simply later. One must admit, it would be quite welcome to see at AMD's price ratios, whenever it could occur.

Thank you for coming to my Ted Talk.

HD64G7800XT will most possibly be a further cut-down N31. The successor of 6800 non-XT. And will be a bit faster than 6900XT. Price will make or brake it.

Real price is key, the whole RTX4000 lineup was ruined due to pricing.

Xex360Real price is key, the whole RTX4000 lineup was ruined due to pricing.

yeah but sadly AMD is just joining in really rather then offering true competition.
So these cards...probably disappointing.

and stupid Intel is also not making moves which is a really bad sign, the rx7600 and upcoming rtx4060 should have caused a bit of a panic over at Intel...but no......they basically dont seem to care.

ZoneDymoyeah but sadly AMD is just joining in really rather then offering true competition.
So these cards...probably disappointing.

and stupid Intel is also not making moves which is a really bad sign, the rx7600 and upcoming rtx4060 should have caused a bit of a panic over at Intel...but no......they basically dont seem to care.

I think Intel just gave up, the 770 is as big a 3070 made on a better and more expensive node yet it cannot beat it, they are just cutting their losses.

AMD has an opportunity to gain lots of market share, but at the same time given nVidia's hold, and nVidia fanboyism among reviewers such as Digital Foundry it's difficult for them to compete.

I strongly suspect we're dealing with a cutdown Navi31, especially since the W7800 is just that, and the reason we've been waiting a while since RDNA3 launch is so AMD can harvest enough dies that didn't make the grade as a 7900/7900XT. Perhaps they originally intended for Navi32 to handle this task, but it wasn't going to be competitive enough.

kapone32Everything that everyone says is purely speculation. Some people have stated that there would be no 7800XT when the 7600XT was the only card AMD was launched. We don't know the specs so we might as well wait for official information from AMD.

The 7600XT card hasn't been launched.

JAB CreationsThe 7600XT card hasn't been launched.

That 's right it's only the 7600.

kapone32That 's right it's only the 7600.

We need the replacement for the 6700/XT to drive the price of the 6700XT down. I saw 6600s for as low as $240 Canadian.

Honestly by the time this card comes out I'll have spent the cash waiting for the Bethesda to arrive.

The question is how much faster than a 6900/50xt will it be, or not? Then there is a 6800xt with its ridiculous price to consider as well. Buying the latest gen gpus isn't the smartest choice nowadays, and it becomes a norm.

Chrispy_I'm thinking these are Navi31, heavily chopped down.

Navi32 wasn't even supposed to launch yet, and there's strong evidence it has been delayed from the original schedule of ~September so we might not even see it until Christmas.

I agree with the Nav31 being chopped down. I also think that these chips are ones that did not pass inspection and defective, the ones that are partially made on the edges of the silicon wafers. If my memory serves me correctly this is a common practice of getting better yields from those wafers.

This is why they went with the chiplet/infinity fabric. Better yields. Maximum Profit.

However IMHO it is because of the Infinity Fabric there are known latency issues because of the concept. IMHO this is why the chiplet/core count has not changed. IMHO AMD is at their limits in their use with the Infinity Fabric.

Infinity Fabric concept was great when it was first conceived in AMD R&D department. But this technology is over 10 years old from its concept. And because of that and placing profitability overall performance, they gave Intel a chance to stay in the race.

I've owned every Ryzen Generation CPU except the current 7000 series. Because overall the cost is just not there when it comes down to wattage/performance.

IMHO their best generations was 1000 series and 5000 series. I've sold my 2000 series and 3000 series with the exception of the Ryzen 3600.

The same goes with the 5700 series of video card. Good card that is relevant today. Now that the 6000 series of video card has come down in price on where it should have been in the first place I just might buy one.

But right now I am not buying their 7000 video card series with the same reason why I am not buying a Ngreedia video card.

You are paying too much money for what you are getting.

VeseleilThe question is how much faster than a 6900/50xt will it be, or not? Then there is a 6800xt with its ridiculous price to consider as well. Buying the latest gen gpus isn't the smartest choice nowadays, and it becomes a norm.

I think 7800XT will match 6950XT in performance.

HD64GI think 7800XT will match 6950XT in performance.

It better be, as it'll probably be priced above $700 mark.

VeseleilIt better be, as it'll probably be priced above $700 mark.

Close to $600 more probably.

HD64GClose to $600 more probably.

Only if it's a cut-down 7900XT based on Navi31 with 4/5ths the performance of an $800 7900XT
If it's a super-clocked Navi33 with 16GB then it's going to be hard to charge more than $400.

JAB CreationsThe 7600XT card hasn't been launched.

The "7600" already got the full Navi 33 with all 2048 shaders. I don't see a hypothetical "XT" be anything else than a rebrand with slightly higher clocks.
There is no logic to use disabled Navi 32 because it is an MCM with much more complex design and execution - it won't fit.

The best would have been if AMD was brave and aggressive enough to call it 7500 and charge as little as 150 bucks for it, not more.

Chrispy_Only if it's a cut-down 7900XT based on Navi31 with 4/5ths the performance of an $800 7900XT
If it's a super-clocked Navi33 with 16GB then it's going to be hard to charge more than $400.

Navi 32, Navi 33 is a different chip, much smaller and less capable.

Chrispy_Only if it's a cut-down 7900XT
If it's a super-clocked Navi33 with 16GB then it's going to be hard to charge more than $400.

This isn't Navi32. At least it's incredibly unlikely, see my earlier posts - Navi32 is delayed, watch the Navi31 launch interviews explaining the issues with stability and artefacts at the originally-intended clocks. AMD themselves said that Navi 32 has been revised/respun in silicon because they couldn't fix it in software.

Even before the delay, Navi32 wasn't scheduled until September and the delay will likely add 3-6 months.

Chrispy_This isn't Navi32. At least it's incredibly unlikely, see my earlier posts - Navi32 is delayed, watch the Navi31 launch interviews explaining the issues with stability and artefacts at the originally-intended clocks. AMD themselves said that Navi 32 has been revised/respun in silicon because they couldn't fix it in software.

Even before the delay, Navi32 wasn't scheduled until September and the delay will likely add 3-6 months.

Why do you quote yourself?

First, information says that Navi 31 was delayed by a whole year. Which means its original plan was December 2021.
It got launched 1 full year later but still on the first silicon - revision A0, as the information goes on...

So, if you state that Navi 32 is further delayed, why would that be? The Navi 32 project lagged significantly behind Navi 31 and/or there are some schedule misarrangements at TSMC?
I don't get it.

Because they had 1 full year between December 2021 and December 2022, we are now in late June 2023, and nothing is fixed, and there are no positive news that anything will be fixed.
No timeline, nothing.

We should see RX 7950 (fixed Navi 31, newer revisions), and RX 7800 (Navi 32).

BTW, this is Navi 32 and it looks like ready for launch but artifically blocked by AMD.

I wonder, how much of an upgrade this card might be versus my 6900xt. Something tells me, not much of an upgrade.

ARFThe "7600" already got the full Navi 33 with all 2048 shaders. I don't see a hypothetical "XT" be anything else than a rebrand with slightly higher clocks.

Well hot-damn, AMD f#$%ed that up even more than I previously thought. :wtf: