Video Card Rumours

[Thinder]

6 hours ago, EightyDuce said:

I'm willing to spend upto $1000 and get the best card for that budget. I'm no where near willing to pay $1600 (if you're lucky to get msrp, 2000 in his area by his own admission) for a 4090. I'm not looking for the best, I'm looking for the best I can get for my $1000 budget; so are others.

Agreed. I am running an EVGA GEFORCE GTX 1080Ti that I purchased second hand (two years warranty) at CEX (Britain) and I'm very happy with it, I need to upgrade it to play 4K VR, so I'm shopping around with a clear idea of the amount of money I'll splash into the card that will allow me to do just that, not "THE BEST", just a competitive one within my budget.

[VR Flight Guy in PJ Pants]

Again, I would wait for hard numbers, nonetheless, on paper, 7900XTX is very attractive: lower price, smaller form factor, less power hungry, and possibly, less heat waste, with a performance expected to surpass 4080...

Baseline, if the XTX runs faster than 3090TI, which sells for £1149.00 right now,I can wait and get the XTX instead.

Edited November 6, 2022 by VR Flight Guy in PJ Pants

[Pilotasso]

I  love this guy

But bear in mind, Im am not biased on my GPU choice, if these cards have issues like stuttering or driver issues, I'll got 4090.

[Pilotasso]

18 hours ago, SkateZilla said:

I foresee a 7900XTX 2X (dual Navi31s on infinity fabric) 

I was thinking in lines of 7950X with vertically stacked cache, but if both happen that would be wild.

[VR Flight Guy in PJ Pants]

Box UK (on my another thread) has some price points for RTX 4080. Kinda not good looking unless supported by benchmarks. If the R7900XTX beats the performance of  4080 at £1200 print point, I am getting one.

Edited November 11, 2022 by VR Flight Guy in PJ Pants

[SkateZilla]

RTX4080s are already inflated to 1200USD, and up to 1600 USD

Edited November 11, 2022 by SkateZilla

[VR Flight Guy in PJ Pants]

It is around £1,400.00 here, according the Box UK, it may be even marked higher by other vendors, like OCUK

Price aside, if you are to trust a YT channel with out 12K subscribers, here is some benchmarks agains some games for your discretions:

 

Edited November 11, 2022 by VR Flight Guy in PJ Pants

[VR Flight Guy in PJ Pants]

4080 reviews are flooding and they are not pretty...

[LucShep]

16 minutes ago, VR Flight Guy in PJ Pants said:

4080 reviews are flooding and they are not pretty...

And they shouldn't be pretty...
In Europe the RTX 4080 16GB is going to sell between 1600 and 2200 €uros (serves also to force sales of RTX3000 series old stock)... 
And same will happen to the AMD RX7000 series (likewise, also to force sales of RX6000 series old stock), you'll see.

We thought it was the mining craze and the lockdown creating the price gouging on GPUs.... BS!  This is now beyond stupid, sorry.
Nvidia and AMD can go fly a kite with these new GPUs (as good as they may be). 

Used Market, for higher end Nvidia RTX3000s and AMD RX6000s, is where it's at (while you can). Be patient and look for those from trusted sellers and at the right price. 

Edited November 15, 2022 by LucShep

[SkateZilla]

23 hours ago, LucShep said:

And they shouldn't be pretty...
In Europe the RTX 4080 16GB is going to sell between 1600 and 2200 €uros (serves also to force sales of RTX3000 series old stock)... 
And same will happen to the AMD RX7000 series (likewise, also to force sales of RX6000 series old stock), you'll see.

We thought it was the mining craze and the lockdown creating the price gouging on GPUs.... BS!  This is now beyond stupid, sorry.
Nvidia and AMD can go fly a kite with these new GPUs (as good as they may be). 

Used Market, for higher end Nvidia RTX3000s and AMD RX6000s, is where it's at (while you can). Be patient and look for those from trusted sellers and at the right price. 

 

As a 25yr system builder, I dont see any issue with AMD's Pricing Model, as long as it doesnt get super inflated.

Add in nVidia's competitor for that 7900XT/X is technically underwhelming, it only makes AMD look better, being better and cheaper. and using less power.

Edited November 16, 2022 by SkateZilla

[VR Flight Guy in PJ Pants]

5 minutes ago, SkateZilla said:

Add in nVidia's competitor for that 7900XT/X is technically underwhelming, it only makes AMD look better, being better and cheaper. and using less power.

I am not very sure about the "being better" part, I am waiting for the hard numbers...

[SkateZilla]

1 minute ago, VR Flight Guy in PJ Pants said:

I am not very sure about the "being better" part, I am waiting for the hard numbers...

Me too, NDA ended on reviewers, so now it's just a matter of days.

[LucShep]

21 hours ago, SkateZilla said:

As a 25yr system builder, I dont see any issue with AMD's Pricing Model, as long as it just get super inflated.

Add in nVidia's competitor for that 7900XT/X is technically underwhelming, it only makes AMD look better, being better and cheaper. and using less power.

Come on now.... "being better" ?   

If you're a 25yr system builder (like I am as well) then you're aware of these three realities:

• AMD is running on 10 years of overhype and disapointment (if these fail, it'll become 12+ years then). Passing a blank cheque to NVIDIA regarding top performance and drivers quality (basically saying "we can't compete there"), allowing NVIDIA to make their own products more expensive.
• People want AMD GPUs to compete, sure - so that NVIDIA is forced to lower prices on their own GPUs.
• NVIDIA drivers have been, and probably will keep on being, MUCH better. That's a huge part of a GPU ownership experience (FWIW, just the NVIDIA drivers team is bigger than the whole AMD corporation!)

I agree that we need to see the reviews first, and then maybe we can say if "it's better", or not.
Better yet, let's see the real prices on the market (not "MSRP" BS), and concrete feedback on user experience with these new GPUs after a couple of months. Then we can see if they're really better. 
 

Edited November 16, 2022 by LucShep

[SkateZilla]

1 hour ago, LucShep said:

Come on now.... "being better" ?   

If you're a 25yr system builder (like I am as well) then you're aware of these three realities:

• AMD is running on 10 years of overhype and disapointment (if these fail, it'll become 12+ years then). Passing a blank cheque to NVIDIA regarding top performance and drivers quality (basically saying "we can't compete there"), allowing NVIDIA to make their own products more expensive.
• People want AMD GPUs to compete, sure - so that NVIDIA is forced to lower prices on their own GPUs.
• NVIDIA drivers have been, and probably will keep on being, MUCH better. That's a huge part of a GPU ownership experience (FWIW, just the NVIDIA drivers team is bigger than the whole AMD corporation!)

I agree that we need to see the reviews first, and then maybe we can say if "it's better", or not.
Better yet, let's see the real prices on the market (not "MSRP" BS), and concrete feedback on user experience with these new GPUs after a couple of months. Then we can see if they're really better. 
 

 

AMD's FX/Polaris Era is over.

[AngleOff66]

GN may have found the answer with the cable.

 

[EightyDuce]

13 hours ago, LucShep said:

Come on now.... "being better" ?   

If you're a 25yr system builder (like I am as well) then you're aware of these three realities:

• AMD is running on 10 years of overhype and disapointment (if these fail, it'll become 12+ years then). Passing a blank cheque to NVIDIA regarding top performance and drivers quality (basically saying "we can't compete there"), allowing NVIDIA to make their own products more expensive.
• People want AMD GPUs to compete, sure - so that NVIDIA is forced to lower prices on their own GPUs.
• NVIDIA drivers have been, and probably will keep on being, MUCH better. That's a huge part of a GPU ownership experience (FWIW, just the NVIDIA drivers team is bigger than the whole AMD corporation!)

I agree that we need to see the reviews first, and then maybe we can say if "it's better", or not.
Better yet, let's see the real prices on the market (not "MSRP" BS), and concrete feedback on user experience with these new GPUs after a couple of months. Then we can see if they're really better. 
 

 

10 years of overhyped disappointment?That's a bit dramatic.

Zen has hardly been a disappointment. RDNA has hardly been a disappointment. Hawaii (290/290X), hardly a disappointment.

 

People want competition so that they don't have to sell their kidney for a trash product. Competition breeds innovation and combats price gouging by a monopoly. Saying people want competition from AMD only so that they can buy an Nvidia card at a lower price is a bit disingenuous. People want competition so that they can get the best value for their hard-earned money. I'd wager that if Burger King launched a graphics card that beat or matched AMD/Nvidia offerings at a cheaper price, people would line up to have it their way....

 

As far as drivers, I do agree Nvidia has had overall good drivers. As you mentioned, their driver division is huge. However, AMD drivers have hardly been a dumpster fire they are being perpetuated to be, at least in recent years. It appears that they can't shake the stigma of years passed that gets parroted as still being a thing. Again I can only speak from personal experience for the Hawaii and RDNA2 generations. 

 

Out of curiosity, what is your metric for measuring success or failure of AMD? 

[LucShep]

5 hours ago, EightyDuce said:

10 years of overhyped disappointment?That's a bit dramatic.

Zen has hardly been a disappointment. RDNA has hardly been a disappointment. Hawaii (290/290X), hardly a disappointment.

Obviously not refering to processors (Zen is a success, and a much welcome one by AMD). I'm refering only to AMD graphics cards, which indeed have been disapointing, even if they have insisted and improvements have been done. It just hasn't been enough to bring them up to parity, hardware or software wise. 

The Hawaii 290/290X and Tahiti 280/280X (GCN1.0) were less of a disapointment than the two following generations, but they too were trumped by Nvidia (GTX780, GTX780Ti, GTX Titan).
RDNA 1 (and I had a RX5700XT) was a better return to form after VEGA and RDNA4.0 before but, gosh, it was a god awful experience for quite a while (over eight months of the worst graphics drivers and workarounds I've ever had the displeasure of handling).
Everybody I've successfully convinced to invest in an AMD GPU even before GCN4.0 (RX470, RX580) to RDNA2 (RX6800, RX6900XT) have regreted it and already gone back to Nvidia, and all gladly so, unfortunately.

Look at these forums and you too may sense there's a good ammount of people using them, who would have prefered Nvidia after the experience (even if better these days). 
 

5 hours ago, EightyDuce said:

People want competition so that they don't have to sell their kidney for a trash product. Competition breeds innovation and combats price gouging by a monopoly. Saying people want competition from AMD only so that they can buy an Nvidia card at a lower price is a bit disingenuous. People want competition so that they can get the best value for their hard-earned money. I'd wager that if Burger King launched a graphics card that beat or matched AMD/Nvidia offerings at a cheaper price, people would line up to have it their way....

Oh, absolutely.
I agree with you, yet the huge lines of fools paying thousands for single units of GPUs that actually sell out, and helping what used to be considered a problem become the norm, seem to disagree with us......
 

5 hours ago, EightyDuce said:

As far as drivers, I do agree Nvidia has had overall good drivers. As you mentioned, their driver division is huge. However, AMD drivers have hardly been a dumpster fire they are being perpetuated to be, at least in recent years. It appears that they can't shake the stigma of years passed that gets parroted as still being a thing. Again I can only speak from personal experience for the Hawaii and RDNA2 generations. 

I'm glad you didn't went through the VEGA generation and the early 8 months of RDNA 1.
Regret and "wasted time that I'll never get back" wouldn't even start to describe the feeling...
 

5 hours ago, EightyDuce said:

Out of curiosity, what is your metric for measuring success or failure of AMD? 

The same metric of any other competing brand for whatever type of product - to make a product that can compete at the same level or better.
If the drivers keep being "OK, but not as good", and if they're unable to create a direct competitor to the rival's flagship (as has been for the last 10 years), it just means what it is - defeat. They're a second tier.

Which means the competitor (Nvidia) won't (re)adjust pricing on their semi-perceived "superior" products (because there's no need to), and have now adopted strategies of making hugely cut down versions of the flagship at utter hilarious prices (hello RTX4080 16GB... ) unlike previous generations, while laughing all the way to the bank.

Edited November 17, 2022 by LucShep
spelling(?)

[SkateZilla]

14 hours ago, EightyDuce said:

10 years of overhyped disappointment?That's a bit dramatic.

Zen has hardly been a disappointment. RDNA has hardly been a disappointment. Hawaii (290/290X), hardly a disappointment.

 

People want competition so that they don't have to sell their kidney for a trash product. Competition breeds innovation and combats price gouging by a monopoly. Saying people want competition from AMD only so that they can buy an Nvidia card at a lower price is a bit disingenuous. People want competition so that they can get the best value for their hard-earned money. I'd wager that if Burger King launched a graphics card that beat or matched AMD/Nvidia offerings at a cheaper price, people would line up to have it their way....

 

As far as drivers, I do agree Nvidia has had overall good drivers. As you mentioned, their driver division is huge. However, AMD drivers have hardly been a dumpster fire they are being perpetuated to be, at least in recent years. It appears that they can't shake the stigma of years passed that gets parroted as still being a thing. Again I can only speak from personal experience for the Hawaii and RDNA2 generations. 

 

Out of curiosity, what is your metric for measuring success or failure of AMD? 

He's Talking about FX Series Processors and the GCN GPUs, It wasn't 10 Years, as FX launched in 2011, and Retired in 2017/2018.

Between FX and GCN GPU's, AMD just couldn't compete, and priced everything stupid low.

Jim Keller Designed both Zen's initial Design and the First Revision. (he's also listed as a consultant/contributor on the revisions there after), as well as offering consults for rDNA.

For the Record, Jim Keller ahs never designed a Processor that Flopped.

Also GCN when launched was not a bad design, the problem is, when the market and demands evolved they refused to move from GCN, and it was kept alive way too long.


Lisa Su took over AMD and the first thing she did was bring in Jim Keller, and course correct the roadmap.

AMD was Content with riding the GCN and FX Wave, and just undercutting market price by 50%-70%, but both architectures were on life support for years, and it was obvious they needed to be let go.

had Lisa Su not taken over AMD, they'd be gone today, in 2015 they were hemorrhaging in every way a business can, 
Lisa came in, stopped the bleeding and treated the problem.

Jus to note, in 2012, my AMD Shares were less than $.75 (Cents), now that same share is worth $75~ today (and spiked to $~150 ish during the late 2021.

 

 

Also 280s/290s were GCN 1.1

7970/7950/7800 etc were GCN1.0

And I'm still using my R7970 Lightning 12 years later, so that's an argument for how good the GCN1.0/1.1 GPUs are,
GCN 2.0 onward is the question mark.

Edited November 17, 2022 by SkateZilla

[LucShep]

3 hours ago, SkateZilla said:

He's Talking about FX Series Processors and the GCN GPUs, It wasn't 10 Years, as FX launched in 2011, and Retired in 2017/2018.

HEH??   Did you even bother to simply read the title of the thread, and whats on topic all this time?  (SPOILER ALERT ---> Video Card Rumours)
Who the frak is talking about CPUs?  (not me, that's for sure...!)

...is it just me sensing some AMD fanboyism there?   hmmm

3 hours ago, SkateZilla said:

 

Also 280s/290s were GCN 1.1

7970/7950/7800 etc were GCN1.0

I didn't even remember that very important detail, thanks for refreshing my memory! The decade old disapointment comment is about GCN1.1 being the start of the disapointment era, exactly because it was merely a soft update (rebranding, really) on what was there prior to it, which was utterly pathetic. The stupid high temps and wattage on the R9 7970s was/is legendary, and it was kept onto the following (rebranded) models .... and only to disapoint again with the R9 Fury, and again in summer 2016 (after huge hype going for months) with the top GCN 4.0 model (RX480) that, in fact, could only compete with GTX1060 6GB mid range model of Nvidia. ...and then over a year later the forever hyped and delayed "Polaris", the Vega 56, Vega 64 and (...omg...) the Radeon VII...!  *SHIVERS*  Please, let's stop now.

Edited November 17, 2022 by LucShep

[SkateZilla]

36 minutes ago, LucShep said:

HEH??   Did you even bother to simply read the title of the thread, and whats on topic all this time?  (SPOILER ALERT ---> Video Card Rumours)
Who the frak is talking about CPUs?  (not me, that's for sure...!)

...is it just me sensing some AMD fanboyism there?   hmmm

I didn't even remember that very important detail, thanks for refreshing my memory! The decade old disapointment comment is about GCN1.1 being the start of the disapointment era, exactly because it was merely a soft update (rebranding, really) on what was there prior to it, which was utterly pathetic. The stupid high temps and wattage on the R9 7970s was/is legendary, and it was kept onto the following (rebranded) models .... and only to disapoint again with the R9 Fury, and again in summer 2016 (after huge hype going for months) with the top GCN 4.0 model (RX480) that, in fact, could only compete with GTX1060s 6GB mid range model of Nvidia. ...and then over a year later the forever hyped and delayed "Polaris", the Vega 56, Vega 64 and (...omg...) the Radeon VII...!  *SHIVERS*  Please, let's stop now.

 

I Put FX and GCN in the same category, so CPU's were included by victim of association.

As for the GCN Details:
250,260(X),270(X),280(X) were rebrands of HD7000 and GCN1.0 still
290(X) was GCN1.1

GCN1.0 was great, but the rebrand and GCN1.1 cards were a mistake, and that mistake continued with GCN1.1a / 295X2, 390(X), And All APUS Containing GCN1.1a, then GCN1.2 / M280(X) etc, and GCN 1.3 RX400 Series and APUs, and GCN 1.4 RX500 Series and APUs, GCN 1.5 VEGA Series

Granted GCN1.4 / RX500 Series weren't AS BAD performance numbers wise, but they were behind the competition, and used more power.

Edited November 17, 2022 by SkateZilla

[LucShep]

Bitwit has the best review of the RTX4080 so far:
 

 

Edited November 22, 2022 by LucShep

[LucShep]

Yes, rumours for yet to be announced future GPUs have started already...

NVIDIA RTX 50 Series ‘Blackwell’ 
Source: https://wccftech.com/rumor-nvidia-rtx-50-series-blackwell-gpus-will-bring-biggest-performance-leap-in-nvidias-history/

"Previous reports have indicated that Blackwell GPUs will be fabricated on TSMC's 3nm process.

The rumor comes from RedGamingTech and they have received some new information regarding the architecture. Firstly, Blackwell will feature an entirely new SM structure. Considering the underlying micro architecture is shifting to an MCM design with Blackwell - this is not surprising. Also, Blackwell will leverage a hyperspeed bus that will interlink the various SM and chiplets. A denoising accelerator will also be part of the ray tracing pipeline (modern path tracing setups don't actually trace the full sequence, they do it partially and a denoiser handles the rest) which should result in significantly improved RT performance.

There still seems to be no word on specifications although the source notes that there are various Blackwell GPUs being considered and a lot of the binning will depend on how AMD's current RDNA3 and future RDNA4 offerings perform. Finally, RGT leaves us with the following teaser: "biggest perf leap in NVIDIA's history".

NVIDIA Hopper was the world's fastest 4nm GPU at launch and the world's first with HBM3 memory. It featured higher specifications than even the NVIDIA RTX 4090 (which contains 16,384 CUDA cores) at a net total of 18,432 CUDA cores. Blackwell will provide significant generational improvement over Hopper (as has always been the case).
Four NVIDIA Blackwell GPUs have already been confirmed in a prior leak."
 

 

AMD RDNA4 
Source: https://wccftech.com/amd-rdna4-rumored-specifications-monstrous-129-fp32-tflops-gpu-gddr7-3-5-ghz-clock-rates-and-more/

"One of the things that was lacking with AMD’s RDNA3 architecture was the ray tracing and AI/ML performance and AMD is looking to fix that with RDNA4. Caches will get big upgrades and 3rd Generation Infinity Cache is going to be rolled out along with a new data prefetch system. WMMA (Wave Matrix Multiply-Accumulate) V2 will also enable 2x the performance per CU for (presumably) matrix instructions allowing full SIMD lane usage. Even though the total increase in CUs is around 50%, the performance increase generation over generation is 100% - which is superb to hear. Given below are the separate dies:

AMD Navi 41 – 144 CUs

Here is where things get even crazier, Navi 41 is touted to have up to a massive 32 GB of vRAM – probably of the GDDR7 variety. A second configuration also exists with 24 or 48 GB of GDDR7 and 6 MCDs (the 32 GB variant has 4 MCDs). The Navi 41 will have 144 CUs, which clocked at 3.5 GHz and at a calculation of 128 ALU per CU should yield a whooping 129 TFLOPs of FP32 performance. Needless to say that is a Godzilla level of performance right there.

AMD Navi 42 – 96 CUs

Navi 42 will be slightly muted and have a net total of 12,288 CUs which clocked at 3.5 GHz should yield around 86 TFLOPs of FP32 performance. 96 CU is also exactly the count of the RDNA3 flagship, the Radeon 7900 XTX, so this SKU will represent the apples to apples generation over generation comparison between RDNA3 and RDNA4 (or more accurately, as apples to apples its ever going to get). As RGT notes in their slides, specifications like these *can* and do change before the final revision.

AMD Navi 43 – 48 CUs

Finally, we have Navi 43, which is going to have 48 CUs. Clocked at around 3.5 GHz, this will yield around 43 TFLOPs of FP32 performance. This will probably be the lower end of AMD’s RDNA4 lineup and just goes on to show how much performance is being packaged in this upcoming RDNA4 architecture where the “lower end” easily exceeds 40 TFLOPs. The memory standard is still going to be GDDR7 and it is going to come with 2x MCDs. This is also the only SKU where RGT notes that the GPU might not be of an MCM design – although they are being told this is currently the case."

 

Edited December 21, 2022 by LucShep

[Pilotasso]

Speculation is fun but don't hold your breath waiting for it. In my own experience I will never again adopt the 1st gen of a new tech, been there with powerVR and the original geforce. Both overpriced and could only use them on tech or gaming demos. Never did significant actual gaming on them that couldn't be done with a normal 3DFX. Eventually got one that I milked for years. When 2000 series came along I stayed put. Who played ray tracing games on them?

[Dangerzone]

2 hours ago, LucShep said:

Finally, RGT leaves us with the following teaser: "biggest perf leap in NVIDIA's history".

Eeeek... Given the last 2 releases, I can't help but read that 'leap' as in pricing!

[Lange_666]

4090 Titan:

https://www.guru3d.com/news-story/photo-of-alleged-rtx-4000-titan-prototype-shows-thick-heatsink-and-two-12vhpwr-connectors.html