How AMD Radeon Sabotages Itself & Its Partners: Development Timelines & Failures

erek

[H]F Junkie
Joined
Dec 19, 2005
Messages
10,785
Still watching this myself, so no comment yet

"AMD's Radeon division keeps sabotaging itself and its partners, and to help illustrate how the VBIOS change does that, we've recapped the manufacturing process."

 
AMD are so shit at making a stable product.

Their CPU side is pretty rock solid and most of their GPUs are too. They blundered with the vBIOS update on this card and have some driver issues. The Navi lineup is otherwise fine.
 
He's reading the email in the video starting from the beginning, it's printed out in front of him.
 
AMD are so shit at making a stable product.

They are?

My Radeon 6990 was stable
My R9 290X's were stable, all 4 of them.
My 5700xt was stable
My 1700x was stable.
My 2600x was stable
My 1950x was stable
My 2950x was stable
My 3600 was stable
My 3900x was stable
My 3960x is stable


Again how is AMD unstable ? Or are you just spewing forth Youtuber propaganda like GN.
 
Last edited:
Not really sure what to think of this, since he labeled it as a bait and switch. That would be the case if you said it is one thing but then when you receive the product, it is less than what was listed, not more.
 
If you've been watching AMD from the beginning, this is something that they have a habit of doing; still, the 5600XT has objectively been one of their better products out of the gate, and the Zen 2 / Ryzen 3000 launch went pretty well overall.

So things are looking up?
 
Not really sure what to think of this, since he labeled it as a bait and switch. That would be the case if you said it is one thing but then when you receive the product, it is less than what was listed, not more.

Bait and and switch has never been solely defined by getting less. That said, I don't really know if it applies here. There is a decent performance margin between old and updated BIOS cards, but is it really enough to call it "bait and switch"? It might apply to AIBs though. They're the ones kind of getting hosed here by the panic update.

The script is someone else's whiny email.

"Wahhh! Mommy! Somone is saying something mean about my favorite company. Make them stop!"
 
They are?

My Radeon 6990 was stable
My R9 290X's were stable, all 4 of them.
My 5700xt was stable
My 1700x was stable.
My 2600x was stable
My 1950x was stable
My 2950x was stable
My 3600 was stable
My 3900x was stable
My 3960x is stable

That's a pretty short memory, despite the long list; further, all but the 3000-series had enough issues at launch to warrant delaying purchase well after release. The criticism may be a bit over the top, but it's not unfair.

And the HD6990... it might have been stable (if you weren't trying to play Skyrim!), but you couldn't pay me to game with one. I've already suffered enough from that generation of GPUs.
 
Or are you just spewing forth Youtuber propaganda like GN.

I'd recommend some restraint here. They're not perfect, but they do produce both their YouTube content and written editorial content, and they do perform some of the best testing currently available.

Personally I think that this situation is somewhat overblown, however, the frustration expressed both by the reviewer community and by AIBs is far from unwarranted given that AMD has changed the baseline at what is probably the worst time, and that baseline requires significant work on the part of reviewers, AIBs, and users, and further will likely both not apply to all parts already shipped as marked while also becoming a source of RMAs that AIBs are expected to cover.

On top of the damage such a release does to product differentiation in AIB lineups and the work they've put forward to get all of that straight.


As a consumer, the potential to get more performance for my money is nice, but in this case it's clearly not coming without a cost to others.
 
77 vulnerabilites patched just in November, hundreds before that. I think we are over 1000 now on Intel's amazing Skylake +++++++++++++++++++++++++++++++++ Silicon.

I'd rather the damn issues be at launch than find out 3 years later after your shit is stolen and your identity just bought a house in the Caribbean and your credit score is buried in the depths of hell its so low, and you have warrants for your arrest because your ID was used to commit felonies in other locations.

There's a difference between obscure vulnerabilities being exploited half a decade after an architecture is released, and obscure workarounds just to get it working in the first place.

You mentioned your personal experience with recent AMD CPU releases with respect to stability, which is already veering off topic, but I responded simply to provide some context. This post is out of context and off topic.
 
Bait and and switch has never been solely defined by getting less.

That is what it means. It means you're being baited with something desirable and then getting something less desirable.

It sounds like AMD is having driver issues right now with the newer cards but any complaints about the performance getting a last minute boost are stupid unless it's about having to redo benchmarks(which would be annoying).
 
You've wildly misread the situation.

I dislike Gamer's Nexus.

So you're breaking forum rules by attacking the source instead of discussing the topic at hand?

That is what it means. It means you're being baited with something desirable and then getting something less desirable.

It sounds like AMD is having driver issues right now with the newer cards but any complaints about the performance getting a last minute boost are stupid unless it's about having to redo benchmarks(which would be annoying).

Something to remember: The boost won't apply to every card on the market.
 
That is what it means. It means you're being baited with something desirable and then getting something less desirable.

It sounds like AMD is having driver issues right now with the newer cards but any complaints about the performance getting a last minute boost are stupid unless it's about having to redo benchmarks(which would be annoying).

You can't define someone as baiting and switching because something didn't live up to your own internal expectations.

If a certain car company says that their car makes 350 horsepower and you buy it and to you it doesn't feel like 350 horsepower, even if all the Dyno's show it is, did they bait and switch you?\

Same goes with the Impossible Whopper. So you hear it tastes just like beef, but you dont think so, are you going to sue Burger King because they baited and switched you?

Court will laugh at you and charge you the court fees for making such an ridiculous claim.

Gamers Nexus BETTER, and I mean with caps, better watch what they say. AMD is very rich, and very capable of dragging GN to lawsuit for libel, slander, and other violations of corporate law. GN will NOT win because AMD is loaded and the law is on their side. Im just saying. GN is walking a tight rope here.

Corporations may sue for defamation if they can show that the published material has caused them or is likely to cause them financial loss. If the required elements exist, a corporate plaintiff may recover presumed damages. This means that harm is presumed—no proof is required—in the matter, and a fact finder may assess an amount he or she deems is appropriate. Brown & William Tobacco Corp. v. Jacobson, 827 F.2d 1119, 1139 (7th Cir. 1987); Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749 (1985). - MincLaw.com
 
Last edited:
  • Like
Reactions: Mega6
like this
but any complaints about the performance getting a last minute boost are stupid unless it's about having to redo benchmarks(which would be annoying).

The problem is that the performance boost isn't certain. AIBs built to one spec, and AMD is raising the spec to include raising power draw and heat generation after the products have been produced, validated, and delivered at the original, lower spec.


Now, mitigating that fact is that most of the 5600 cards were built around the 5700XT, so many are likely overbuilt and perfectly capable of handling the increase, but it's still a poor move on AMDs part for everyone else involved, and that includes consumers that may have to deal with a GPU that becomes unstable after the update.
 
Bait-and-Switch

Bait-and-switch is a form of fraud used in retail sales but also employed in other contexts. First, customers are "baited" by merchants' advertising products or services at a low price, but when customers visit the store, they discover that the advertised goods are not available, or the customers are pressured by salespeople to consider similar, but higher-priced items ("switching").
 
Last edited:
Bait-and-Switch

Bait-and-switch is a form of fraud used in retail sales but also employed in other contexts. First, customers are "baited" by merchants' advertising products or services at a low price, but when customers visit the store, they discover that the advertised goods are not available, or the customers are pressured by salespeople to consider similar, but higher-priced items ("switching").

So what do we call it when customers are baited with a stable product, but after a launch-day update the product isn't?
 
Can somebody please make a condensed, tl;dw of this video?
 
Can somebody please make a condensed, tl;dw of this video?
  • AMD is providing for a BIOS update that will improve performance for most 5600 GPUs
  • In doing so, they basically screwed over everyone else, including customers that get a GPU that can't handle the update
(I didn't watch the video, but this issue isn't complicated)
 
You're telling me I can't call the source bad, for the reasons I said it was bad?

The Fucking Rules said:
Lately, it seems to have become fashionable to constantly disrupt threads from sources that some forum members here have issue with. This will end now.

If you disrupt discussions that you simply do not like because of the sources those are pulled from, you will be warned using the warning system here (as described in the HardForum Rules Thread). If you continue to present the behavior that has been warned against, your account will be banned for a limited time. Beyond that, if you continue with the behavior, you risk permanently losing your account.

Can somebody please make a condensed, tl;dw of this video?

Short version: The vBIOS went out during a holiday, making things rough for AIBs. Because parts are only validated for certain specs everything needs to be revaldiated and not everything is going to meet new validation. This could cause higher RMA rates for cards already in the channel and will mean there will be different SKUs using the old and new vBIOS.
 
There's a difference between obscure vulnerabilities being exploited half a decade after an architecture is released, and obscure workarounds just to get it working in the first place.

You mentioned your personal experience with recent AMD CPU releases with respect to stability, which is already veering off topic, but I responded simply to provide some context. This post is out of context and off topic.

There are no stability issues with AMD cpus. Lets just put that to rest lol. Even if one had their own opinion about boost thing that was going aroud. That still wasn't stability issue.
 
Short version: The vBIOS went out during a holiday, making things rough for AIBs. Because parts are only validated for certain specs everything needs to be revaldiated and not everything is going to meet new validation. This could cause higher RMA rates for cards already in the channel and will mean there will be different SKUs using the old and new vBIOS.

Is that fact or is that rambling guy's opinion?
 
You can't define someone as baiting and switching because something didn't live up to your own internal expectations.

If a certain car company says that their car makes 350 horsepower and you buy it and to you it doesn't feel like 350 horsepower, even if all the Dyno's show it is, did they bait and switch you?\

Same goes with the Impossible Whopper. So you hear it tastes just like beef, but you dont think so, are you going to sue Burger King because they baited and switched you?

Court will laugh at you and charge you the court fees for making such an ridiculous claim.

Gamers Nexus BETTER, and I mean with caps, better watch what they say. AMD is very rich, and very capable of dragging GN to lawsuit for libel, slander, and other violations of corporate law. GN will NOT win because AMD is loaded and the law is on their side. Im just saying. GN is walking a tight rope here.

Corporations may sue for defamation if they can show that the published material has caused them or is likely to cause them financial loss. If the required elements exist, a corporate plaintiff may recover presumed damages. This means that harm is presumed—no proof is required—in the matter, and a fact finder may assess an amount he or she deems is appropriate. Brown & William Tobacco Corp. v. Jacobson, 827 F.2d 1119, 1139 (7th Cir. 1987); Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749 (1985). - MincLaw.com

I was just saying that getting less is the definition of bait and switch, not that AMD actually was.

The problem is that the performance boost isn't certain. AIBs built to one spec, and AMD is raising the spec to include raising power draw and heat generation after the products have been produced, validated, and delivered at the original, lower spec.


Now, mitigating that fact is that most of the 5600 cards were built around the 5700XT, so many are likely overbuilt and perfectly capable of handling the increase, but it's still a poor move on AMDs part for everyone else involved, and that includes consumers that may have to deal with a GPU that becomes unstable after the update.

If it's causing stability issues then that's certainly a problem but it's a bad design decision not bait and switch.
 
There are no stability issues with AMD cpus. Lets just put that to rest lol.

Currently, I couldn't claim that there are -- but that's not to say that there haven't been or weren't. Also, no CPU exists without the platform under it, and well, AMD has a long history of falling short there.
 
Is that fact or is that rambling guy's opinion?

He's visited factories where these cards are produced. He has contacts with every major AIB on the market. EVERY card that goes out from AIBs has to be validated. MSI even said as such during one of their own streams. Just because you have a retarded bias against a source doesn't make them wrong.
 
If it's causing stability issues then that's certainly a problem but it's a bad design decision not bait and switch.

That's the thing: the issues that could arise are from an increase in spec for existing hardware. Even if nearly all of the hardware shipped can handle it, the question will still hang over the product.
 
He's visited factories where these cards are produced. He has contacts with every major AIB on the market. EVERY card that goes out from AIBs has to be validated. MSI even said as such during one of their own streams. Just because you have a retarded bias against a source doesn't make them wrong.

I'm asking a question, did he prove that as fact or did he just say he thinks that is what is going to happen? I never said I had a bias against him, though the dude does ramble and repeat his points 50+ times a video which is why I am asking for a condensed version so I don't have to sit through 30 minutes of him talking again. I care about the topic, I don't care to listen to this specific guy talk about said topic.
 
I'm asking a question, did he prove that as fact or did he just say he thinks that is what is going to happen? I never said I had a bias against him, though the dude does ramble and repeat his points 50+ times a video which is why I am asking for a condensed version so I don't have to sit through 30 minutes of him talking again. I care about the topic, I don't care to listen to this specific guy talk about said topic.

He has footage of MSI's validation labs and talked in detail about how it's done and even talked about why certain cards fail. What needs to be proved? That cards are only validated at the speeds they're advertised to ship with and never guaranteed to go faster? That's been the case for hardware since forever. That pushing the silicon beyond validation could result in increased RMAs? That's simple logic. Silicon can only be pushed so hard before it fails. MSI even made comments about how flashing is never a guaranteed process, which means they could be expecting some bricked cards from the vBIOS update process. Not really sure what you need specific proof for since every statement follows simple logical reasoning.
 
That's a pretty short memory, despite the long list; further, all but the 3000-series had enough issues at launch to warrant delaying purchase well after release. The criticism may be a bit over the top, but it's not unfair.

And the HD6990... it might have been stable (if you weren't trying to play Skyrim!), but you couldn't pay me to game with one. I've already suffered enough from that generation of GPUs.

I have never had any issues with Anything AMD since the days of the original 1ghz Athlon. Buldozer may not have been the benchmark winner... but it wasn't unstable or a shitty product either. I have had no real issues with any AMD or ATI video card ever. I did have a 5770 that died on my after 3 years or so that I fixed by baking it... in an oven. Lasted almost a year after that, then I baked it again and got a few more months out of it. lol

Too many people equate loosing a benchmark by 5-10% with being a crap product. AMD hasn't always been the top performer... but I can't remember when they where not the hands down best bang for the buck. I also can't say I remember any terribly unstable products either. Of course there have been issues here and there over the years with this game or that... but what company is immune to that. For every specific GPU you can point to and say it was a jet engine cause it ran at 100 degrees with any load on it... I we can find an example of the same from the green team. For every CPU someone can complain wasn't a true 8 core we can find a NV card that didn't have as much ram as it was advertised to have. For ever example of X or Y game didn't really run right between patch A and B.... no doubt an example of the other guys having the same type of issue on another game can be found. Hell how many times has Nvidia straight up been caught cheating in drivers over the last 10 years. AMD may have not been on the winning side of benchmarks for the majority of the last decade... but at least they never out right cheated users out of quality to pad their numbers. :)
 
They are?

My Radeon 6990 was stable
My R9 290X's were stable, all 4 of them.
My 5700xt was stable
My 1700x was stable.
My 2600x was stable
My 1950x was stable
My 2950x was stable
My 3600 was stable
My 3900x was stable
My 3960x is stable


Again how is AMD unstable ? Or are you just spewing forth Youtuber propaganda like GN.

I have 3 generations of AMD ATI, they were all unstable POS. Went to NVIDIA and never had an issue since.
 
I have 3 generations of AMD ATI, they were all unstable POS. Went to NVIDIA and never had an issue since.

I've had numerous AMD cards and never had any issues. On the other hand, I've had a few nvidia 8800gt/9800gt cards fail. Therefore, nvidia sux.
 
I have never had any issues with Anything AMD since the days of the original 1ghz Athlon. Buldozer may not have been the benchmark winner... but it wasn't unstable or a shitty product either. I have had no real issues with any AMD or ATI video card ever. I did have a 5770 that died on my after 3 years or so that I fixed by baking it... in an oven. Lasted almost a year after that, then I baked it again and got a few more months out of it. lol

It's not the CPUs, really -- but you can't have those without the motherboards, the chipsets on them, and the drivers they come with, and that has been a fight. With AMD, it's a fight pretty much every time they update the platform, with X570 recently being a notable and welcome exception.


Too many people equate loosing a benchmark by 5-10% with being a crap product.

I'll say that it matters much more in marketing and does have a real affect on both perceptions and on sales. I don't care unless I actually need that 5-10%, and that's pretty rare.

but I can't remember when they where not the hands down best bang for the buck.

That's a bit rose colored: AMD has shown time and again that if they feel they have the superior product they'll not hesitate to outprice their competition. That on average they're usually cheaper is an indication that on average, well, they're usually slower.

I also can't say I remember any terribly unstable products either.

For GPUs... honestly I've not dealt with as many since AMD bought ATI, but in single-card configurations I can't say that I've seen too many pervasive issues. More than Nvidia, sure, but that's to be expected considering the disparity of their operations.

Today I really have no preference in terms of drivers and software and how they contribute to stability. Nothing I run will care.

Hell how many times has Nvidia straight up been caught cheating in drivers over the last 10 years.

Eh, I'm less critical here; on the one hand, if you're not cheating, you're not trying, and on the other, developers do some stupid stuff, and hardware sales are heavily influenced by the results of benchmarks.

AMD may have not been on the winning side of benchmarks for the majority of the last decade... but at least they never out right cheated users out of quality to pad their numbers. :)

I'm sure there are examples out there, but given how poorly their drivers were for the majority of their releases since buying ATI, especially upon release, I'd say that they were too busy just trying to get stuff running to worry about optimizing around boneheaded development decisions.
 
Back
Top