Poll: AMD Keeps Screwing Up (HUB video) - do they?

[humbug]

My thoughts

- the 7900xt(x) launch disaster with the vapour chamber and cards catching fire, more notworthy this given the comments amd amd made about the 4090 fire hazard.
I don't remember, but sorry if it was the case, but I thought the vapour chamber issue just caused throttling. I don't think they caught fire did they? And it only impacted the reference model. Whereas the 4090 and cable issue were melting across the board. I think we can probably say the 4090 issue in comparison, from a safety pov, is/was worse.

Its never caught fire, obviously that's a ridiculous thing to say and he knows it, this is him looking for reactions...

 

[arc@css]

I suppose I would say more "yes" than "no".

The price of the 7900XT was way too high to begin with, as was/is the 7600 range and the 7700XT.
All considered, I don't think the cost of the 7800XT or 7900GRE is too bad. I'd still like to see them priced a good bit lower, though!

 

[jokerguv]

Their continual dip in market share says otherwise

Makes me laugh when people suggest it is not AMD's problem/fault.

Radeon segment yes. Their GPU segment overall is doing more than fine actually.

But then again thats not because marketing didn't get those launches right. There's a whole host of other factors involved there including a ruthless and a very good competitor, and them really having to do with what they have. Though they seem to have acknowledged this and are working towards improving the software/feature side of things. Better late than never I guess.

Still I want the Radeons to stay in the race despite them looking like a lost cause at this point because the idea of having only Nvidia as the option for gaming dGPUs is not very enticing.

 

[sparkymark75]

- lack of RT focus with rdna 2, could be excused at launch but as shown, the cards are aging pretty poorly now with all the RT games coming out and even with the writing on the wall, they didn't properly address this with rdna 3 so now we have rdna 3 matching 4 year nvidia gpus in RT games
To me, that is probably the only downside at the moment - if RT is important to you. However, I just ran the Black Myth Wukong benchmark, which mrk informed me is using RT (Lumen), and my RDNA 2 card still managed respectable FPS at 1440p.

That's the point, if RT is important to you. The vast majority of games will work via rasterisation, the same can't be said for RT. nVidia released it when they didn't have the hardware to deliver it so used software trickery to make it playable and then their marketing convinced everyone they needed it

 

[TNA]

Radeon segment yes. Their GPU segment overall is doing more than fine actually.

But then again thats not because marketing didn't get those launches right. There's a whole host of other factors involved there including a ruthless and a very good competitor, and them really having to do with what they have. Though they seem to have acknowledged this and are working towards improving the software/feature side of things. Better late than never I guess.

Still I want the Radeons to stay in the race despite them looking like a lost cause at this point because the idea of having only Nvidia as the option for gaming dGPUs is not very enticing.

That's what this thread is about. Why would it be about anything else?

There is a huge list of fails by them if you scroll up. They need to sort all that out and price their cards properly going on. None of this we are premium stuff and just giving us 50-100 quid off the looney priced nvidia stuff.

 

[TNA]

That's the point, if RT is important to you. The vast majority of games will work via rasterisation, the same can't be said for RT. nVidia released it when they didn't have the hardware to deliver it so used software trickery to make it playable and then their marketing convinced everyone they needed it

Marketing never convinced anyone, raw results did. Well maybe it convinced Grim who loved dlss 1.0

 

[humbug]

Mt problem with all of this is its based on the assumption that given how many segments AMD operate in retail GPU's are at the top of that priority, it isn't, i'll tell you where i think it is.

Its about....

#######
#######
#######
#######
#######
#######
#######
#######
#######
Here
#######

 

[Murphy]

Radeon segment yes. Their GPU segment overall is doing more than fine actually.

Did you mean to say CPU segment because their GPU segment pretty much consists entirely of Radeon.

If you did then yes historically their CPU segment has been doing fine but they did drop the ball a bit with the Zen 5 launch, i hope they don't continue down that road but i suspect they will seeing the sorry state their main competitor is in ATM.

 

[humbug]

Mt problem with all of this is its based on the assumption that given how many segments AMD operate in retail GPU's are at the top of that priority, it isn't, i'll tell you where i think it is.

Its about....

#######
#######
#######
#######
#######
#######
#######
#######
#######
Here
#######

I'll add to that ^^^^ do you know why that is? Because we don't care, we only care if AMD's actions result in cheaper Nvidia GPU's and AMD know it.

 

[jokerguv]

I'm going to go out on a limb and say you've probably not watched the video and/or picked up on the general opinion of AMD marketing practices. Because we're not talking about one's interpretation/understanding of how those products should be marketed, we're talking about AMD's interpretation/understanding of how those products should be marketed.

If your marketing sets unreasonable exceptions for an upcoming product people will be disappointed if/when the product doesn't meet the expectations your marketing department set.

Maybe its a personal thing but considering the adjectives used within the first minute i.e. "disappointing, downright embarrassing, misleading potential customers, blunders, poor launches, underwhelming" and the overall tone thereafter, to me seem very emotional in a video basically about 1st party marketing % percentage differences and pricing opinions.

I'd expect these terminologies to be used where somebody has blatantly lied on record.

Anyways I'm actually gonna go out on a limb and say if HUB really believed AMD has misled customers they should be taking them to court. Payday probably gonna be higher than the youtube revenue from this video. Short of this honestly this video would be considered emotional ranting from a cool mind/fresh set of eyes.

You said it yourself, consumers make buying decisions based on the results. They make decisions based on the price of the product vs the performance. If your marketing gives consumers high expectations and your product doesn't meet those expectations you'll have issues because people are expecting to pay £900 for £900 worth of performance.

That's bourn out of the facts, it's not "emotional". AMD themselves end up dropping prices across a wide range of products at times only 3-4 months after the product has launched.

You missed the bit about independent third parties' results. I don't think consumers in this space just look at the product page and buy stuff based on that.

If somebody does that then that's a fair point the product page should be reflective of the actual performance, which I'm certain it mostly is with all the small-print and asterisk references nobody reads .

 

[Nexus18]

My thoughts

- the 7900xt(x) launch disaster with the vapour chamber and cards catching fire, more notworthy this given the comments amd amd made about the 4090 fire hazard.
I don't remember, but sorry if it was the case, but I thought the vapour chamber issue just caused throttling. I don't think they caught fire did they? And it only impacted the reference model. Whereas the 4090 and cable issue were melting across the board. I think we can probably say the 4090 issue in comparison, from a safety pov, is/was worse.

- each new fsr version being hailed as the dlss killer only to take 1 step forward then 2 steps back and intel being able to get their version on par with dlss in comparison despite being new to the dGPU market
It's taken a while yes, but FSR 3.1 is probably where AMD want to be (finally). And lets not forget that this works on all makes of cards, so it also benefits Nvidia users with older cards. Personally I'd be saying thanks AMD!

- anti lag getting people banned, major oversight by amd this
Tricky one. But can we fully lay this on AMD? I bet AMD and likewise Nvidia can't test every single driver change with every game and involve every dev team. I personally think some blame lies with VAC system. I mean, did Nvidia not know about Windows 11 22H2 and do any internal testing in advance before release? https://www.tomshardware.com/news/windows_11_22H2_nvidia_GPU_woes - I'd think this was a bigger oversight, impacting more people and more games than just CS2.

- pricing their gpus at launch in same bracket as nvidia counterparts despite not offering the same complete overall package as their competition
Nvidia stuff is overpriced, and usually Nvidia launch lesser cards first before the "proper" cards come out later - so no way AMD should be trying to emulate any of this. However, I don't know if general costs have gone up so hence price rises. Could all be fixed for all we know lol.

- the claims of how efficient rdna 3 was going to be and better than ada yet come launch, power guzzling like mad and issues with idle power consumption with dual monitors, amd quickly removed some of their slides from the website as well to hide these false claims (also, this wasn't pr marketing but comments direct from Lisa too)
Again I think the dual monitor issue was quite specific to high resolutions and high refresh rates. I'm not 100%, but, I think it is now under control (or at least improved). RDNA 3 performance per watt vs previous AMD cards (not Nvidia here) is actually an improvement. Toms Hardware did a comparison, "...the RX 7700 XT ultimately ends up delivering about a 20% improvement in performance per watt over the 6700 XT, and a 30% increase compared to the RX 6750 XT." Neat!

- knee jerk reaction to DLSS frame gen and making it out like it was in the works for months/years only to release in 2 awful titles and with a lot of missing features
Not sure what you mean on this one. lol.

- lack of RT focus with rdna 2, could be excused at launch but as shown, the cards are aging pretty poorly now with all the RT games coming out and even with the writing on the wall, they didn't properly address this with rdna 3 so now we have rdna 3 matching 4 year nvidia gpus in RT games
To me, that is probably the only downside at the moment - if RT is important to you. However, I just ran the Black Myth Wukong benchmark, which mrk informed me is using RT (Lumen), and my RDNA 2 card still managed respectable FPS at 1440p.

There were reports from people that their gpu popped because of the lack of water to cool i.e. vapour chamber issues. I can't recall if it impacted all but I believe it was only the 7900xtx affected, amd said it was a small batch but still the point stands, marketing/making fun of the competition only to face a similar problem yourself. Wasn't the 4090 issues in the end pointed down to user error (wasn't it gamer nexus who came to this conclusion, the issue was the connector wasn't providing a good enough feedback to show the cable had been fully inserted) and mostly affecting AIB models along with people using custom cables? I think if using the supplied power cable with the FE, it wasn't an issue? AMD handled it well though tbf to them.

FSR 3.1 as shown by DF etc. still is not anywhere near DLSS and barely improved over FSR 2.x versions (upscaling side of things). Frame generation is much better and very good now although question of consistentancy is still up in the air. As shown by the poll created here a while back, most people don't use FSR includng amds own customers so whilst it is is good and probably will eventually get there, it's not a great outcome still if most of their own user base aren't using it.

Given AMD always like to say about what they do is to work best for consumers but also their game dev partners, you would imagine they would be working closely to test such things like this. Yes new features like this should be tested by anyone who is going to be involved, the injection method of AFMF is done by amd, not by game devs though so no one to blame here but amd (unless they did pass it on to the game devs to test and they also missed it in their testing.....) hence why they pulled it and have now relaunched it. Microsoft releasing an update which harms nvidia (or any of the dgpu brands for that matter) is not on nvidia, of course, microsoft should be working alongside with dgpus vendors if they are going to be making changes which will impact gpu drivers. Personally can't say I had any issues with that update and my 3080 though and this is more a driver issue i.e. can't apply to all, same way AMD ran great for me throughout the decade but not so for others and vice versa.

Prices are 100% fixed and no doubt Jenson and Lisa are both behind it together.

It is still a far cry from what Lisa/AMd were saying before launch though.... But power efficiency no longer matters since RDNA 3 came out

This YT covers it well:

It then came out in 2 of the worst games to date, forgotten and immortals and was completely useless, didn't work with VRR, frame pacing issues, you had to be hitting your screen refresh rate for it to be smooth and using vsync so then getting insane latency among other issues i.e. kneejerk reaction to dlss 3

That's the point, if RT is important to you. The vast majority of games will work via rasterisation, the same can't be said for RT. nVidia released it when they didn't have the hardware to deliver it so used software trickery to make it playable and then their marketing convinced everyone they needed it

Software trickery that basically every tech press confirm to be of value and deem to be worthy of having.... Again, why are amd investing millions into their software department?

 

[Rogueywon]

That's the point, if RT is important to you. The vast majority of games will work via rasterisation, the same can't be said for RT. nVidia released it when they didn't have the hardware to deliver it so used software trickery to make it playable and then their marketing convinced everyone they needed it

You're right that Nvidia released the tech before it was really in a position to benefit users, but I disagree on the rest.

The 2000-series was not a great demonstration of either RT or DLSS (and I say that as somebody who owned a 2080Ti). RT implementations were pretty minimal, the performance cost was huge and using DLSS1 to claw back framerate came with a huge hit to image quality. What it did do was get developers excited that the tech was out there and get them to start making games that supported those new features, but from the customer's perspective, you were best off skipping the generation. Then the 3000-series arrived, DLSS got significantly better and (on the 3090 and 3080 at least) RT could actually be worth using and, for the first time, there were games that looked much better with it (Cyberpunk probably the first). That trend continued and accelerated for the 4000-series. It doesn't matter whether DLSS and frame generation are "software trickery"... the fact is that in many situations, they work.

For my money, the biggest harm to the reputation of RT came from Microsoft and Sony's decisions to make RT a box-tick feature for the Playstation 5 and Xbox Series X. Those consoles are about as good at it as the 2000-series was, which is to say, not very good at all. I think a lot of people outside of the PC enthusiast sector have become quite jaded about it as a result.

 

[Journey]

"Better launches with more accurate marketing" is said in the first few sentences of the video, why make a topic on a video then try and deviate away from the core point, the marketing.

There is no such thing as accurate marketing, it is still first party data and cannot be trusted. I mean if you want to start a thread without the pointless video linked and pointless poll, and just say "Do AMD fail to launch products in ideal conditions more often than not in the recent past?" Answer = Yes.

So rename to the thread, delete the poll. or stick to topic, MARKETING!

 

[K1LaW8tt5]

Maybe with the 8000 series we will see Taylor Swift, Danny devito, Ryan Reynolds and Rachael Gunn presenting the new architecture.

That should make the GPUs worth at least another £100

 

[Nexus18]

You're right that Nvidia released the tech before it was really in a position to benefit users, but I disagree on the rest.

The 2000-series was not a great demonstration of either RT or DLSS (and I say that as somebody who owned a 2080Ti). RT implementations were pretty minimal, the performance cost was huge and using DLSS1 to claw back framerate came with a huge hit to image quality. What it did do was get developers excited that the tech was out there and get them to start making games that supported those new features, but from the customer's perspective, you were best off skipping the generation. Then the 3000-series arrived, DLSS got significantly better and (on the 3090 and 3080 at least) RT could actually be worth using and, for the first time, there were games that looked much better with it (Cyberpunk probably the first). That trend continued and accelerated for the 4000-series. It doesn't matter whether DLSS and frame generation are "software trickery"... the fact is that in many situations, they work.

For my money, the biggest harm to the reputation of RT came from Microsoft and Sony's decisions to make RT a box-tick feature for the Playstation 5 and Xbox Series X. Those consoles are about as good at it as the 2000-series was, which is to say, not very good at all. I think a lot of people outside of the PC enthusiast sector have become quite jaded about it as a result.

Agree except for this part:

For my money, the biggest harm to the reputation of RT came from Microsoft and Sony's decisions to make RT a box-tick feature for the Playstation 5 and Xbox Series X. Those consoles are about as good at it as the 2000-series was, which is to say, not very good at all. I think a lot of people outside of the PC enthusiast sector have become quite jaded about it as a result.

We can blame devs for not doing a better job of optimising and still supporting raster (although as shown, that is quickly changing now, much sooner than I predicted tbh), metro ee, avatar, spiderman 2 are all examples of RT only which run incredibly well on consoles.

"Better launches with more accurate marketing" is said in the first few sentences of the video, why make a topic on a video then try and deviate away from the core point, the marketing.

There is no such thing as accurate marketing, it is still first party data and cannot be trusted. I mean if you want to start a thread without the pointless video linked and pointless poll, and just say "Do AMD fail to launch products in ideal conditions more often than not in the recent past?" Answer = Yes.

So rename to the thread, delete the poll. or stick to topic, MARKETING!

Don't like, you know where the door is. For such a pointless thread, you seem to enjoy posting in it and discussing the points raised....

Marketing and execution i.e. releasing in a good state go hand in hand.

 

[JediFragger]

One thing is for sure, the marketing ***holes who had a hand in creating the RDNA3 and Zen5 slides need firing or moving to a lower position!!! Awful awful awful, both of them.

 

[Murphy]

Maybe its a personal thing but considering the adjectives used within the first minute i.e. "disappointing, downright embarrassing, misleading potential customers, blunders, poor launches, underwhelming" and the overall tone thereafter, to me seem very emotional in a video basically about 1st party marketing % percentage differences and pricing opinions.

I'd expect these terminologies to be used where somebody has blatantly lied on record.

Anyways I'm actually gonna go out on a limb and say if HUB really believed AMD has misled customers they should be taking them to court. Payday probably gonna be higher than the youtube revenue from this video. Short of this honestly this video would be considered emotional ranting from a cool mind/fresh set of eyes.

You think a Youtube review channel with 3 members of staff is going to take on a multinational company that's worth in excess of $200 billion? And for what, they wouldn't get a penny because like you say they potential mislead customers not HUB.

I suspect it is a personal thing for you, most others look at it rather more objectively.

You missed the bit about independent third parties' results. I don't think consumers in this space just look at the product page and buy stuff based on that.

If somebody does that then that's a fair point the product page should be reflective of the actual performance, which I'm certain it mostly is with all the small-print and asterisk references nobody reads .

No i did not, what do you think "based on the results" and "the performance" means? You seem to be intentionally missing the salient point of what you replied to and the HUB video, the point that the performance AMD claim before launch do not typically match what the actual performance is, and the price on launch does not match the actually value of the product.

If you're saying the product page should be reflective of the actual performance then clearly you don't understand what you're talking about, if you did you would've said the product page does not reflect the actual performance, if you did you'd know that even when third-parties match the system specs in the small print exactly they often find that AMD's performance numbers are out by sometimes multiple factors.

E.g Before the 7900XTX launch AMD were saying it matched the performance of a 4090 and only cost $1k, when third-parties got their hands on it they discovered it was closer to the 4080 and after 6 months the price had fallen by $150-200.

 

[Gerard]

E.g Before the 7900XTX launch AMD were saying it matched the performance of a 4090 and only cost $1k, when third-parties got their hands on it they discovered it was closer to the 4080 and after 6 months the price had fallen by $150-200.

Where did they say this, as that comes off as one of those random rumours that became *insert company name* said this or that after a few months. I remember an AMD rep saying they could have built a card to compete with 4090 on raster but power requirements would be too high, the quote in the article was via google translate as it was a foreign article in the first place.

Technically, it is possible to develop a GPU with specs that compete with theirs (NVIDIA). However, the GPU developed in this way was introduced to the market as a "graphics card with a TDP (thermal design power) of 600 W and a reference price of $1,600 (about ¥219,000)" and was accepted by general PC gaming fans. After thinking about it, we chose not to adopt such a strategy.

AMD intentionally held back from developing a 600 W, US$1,600 RDNA 3 equivalent for the RTX 4090, feels mainstream gamers can instead spend savings on other PC parts

AMD could have made an RTX 4090 competitor but chose not to in favor of a trade-off between price and competitive performance. AMD's rationale is that a 600 W card with a US$1,600 price tag is not a mainstream option for PC gamers and that the company always strived to price its flagship GPUs in...

www.notebookcheck.net

Some guy modded one and water-cooled it and performance jumped up quite a bit.

AMD's RX 7900 XTX Matches an RTX 4090, While Using 700W of Power

It takes 700W for AMD to defeat Nvidia's flagship

www.tomshardware.com

 

[G J]

Their continual dip in market share says otherwise

Makes me laugh when people suggest it is not AMD's problem/fault.

Years people where screaming that we need competition and if people did not buy AMD they would go under and we would be paying £££ for cards, market share tanks even more and now it's AMD's fine bro you dont know what your talking about.

As things stand in the PC DIY space, a company releases a product, all the information about how it performs is already available via numerous independent third parties and the consumers make buying decisions based on the results of those third parties and not how the company marketed it. Or at least that's how I suppose people usually buy their stuff.

What numerous independent third parties are these as anyone who get's samples, early access, has to sign NDA's, does sponsorship, etc are not independent.

 

[G J]

Maybe with the 8000 series we will see Taylor Swift, Danny devito, Ryan Reynolds and Rachael Gunn presenting the new architecture.

That should make the GPUs worth at least another £100

I'd be down for special edition's as a 8xxx/5xxx Tay Tay sounds better than a boring Ti/XT/XTX/Super