Judge Allows Lawsuit Over AMD's FX Processors to Continue

[GoldenLag]

2 minutes ago, 2Buck said:

Yes, they share resources, but there are, in fact, 8 physical cores.

well. there are 8 exectution units. and those are typically called cores. 

 

for where i draw the line personally it is a 8 execution unit CPU and not 8 classical cores. though that is just me diffirentiating them. 

 

there are 8 "cores" doing stuff on it. and calling it a 4 core would be less correct imo

[mxk]

They list very carefully that my CPU has "6 processing cores" but don't mention the fact that 2 of them are for the APU. Technically, they're right, but it's misleading to people who don't do their research (a lot of people, sadly).

[dalekphalm]

3 minutes ago, mxk. said:

They list very carefully that my CPU has "6 processing cores" but don't mention the fact that 2 of them are for the APU. Technically, they're right, but it's misleading to people who don't do their research (a lot of people, sadly).

See now that would be a somewhat valid thing to sue AMD over.

 

That's not what they're suing about.

[GoldenLag]

2 minutes ago, mxk. said:

They list very carefully that my CPU has "6 processing cores" but don't mention the fact that 2 of them are for the APU. Technically, they're right, but it's misleading to people who don't do their research (a lot of people, sadly).

AMD themselves?

 

i mean if they define the CUs as cores. which they can, bu ti just noticed that they write it:

Quote

Graphics Core Count

256

so.........

 

were do we draw the line?

 

do we draw the line?

[mxk]

1 minute ago, GoldenLag said:

AMD themselves?

 

i mean if they define the CUs as cores. which they can, bu ti just noticed that they write it:

so.........

 

were do we draw the line?

 

do we draw the line?

too late, I guess.

[Taf the Ghost]

19 minutes ago, GoldenLag said:

while bulldozer was hot. AMD didnt exactly do anything wrong in calling them cores. Because they sort of are, but sort of arent. 

If they'd had reversed the CMT approach, Bulldozer would have also aged better. 

 

Heck, if they'd launched it in 2009, it'd have been brilliant. 

[Curufinwe_wins]

For historicity sake, the point of the lawsuit is that AMD went from an architecture to another one where the performance of a single 'core' (defined logically or physically it doesn't matter actually) dramatically decreased in many if not most consumer workloads (even without the literally halved FPU performance). That combined with falsely, in years of advertising, equating their 'cores' with that of both their own previous cores and their competitors misled millions of consumers about the levels of performance they were getting for the price. They were exceedingly negligent in informing consumers about the weaknesses of their new architecture, and deliberately tried to obfuscate the dramatic real-world performance losses these chips resulted in.

 

The lawsuit has all required merit and should be awarded properly.

 

 

For example, I had friends as recently as 2016 who still believed (directly as a result of AMD's marketing materials and claims) that the 8350 was a far better 8-core cpu than the 5820k 6-core.

 

If Intel had been falsely marketing their products like this for the 6 years AMD did, everyone would be absolutely losing their minds. But since AMD did it, nah man it's all good. Bulldozer's marketing was/is one of the largest scams in the history of computing products. And only general consumers paid the price, because everyone else knew better (or found out quickly).

 

Intel BTW had to settle a similar an extremely class action lawsuit over Pentium 4's deceptive marketing (directly surrounding GHz), and how they made the original very slow Pentium 4s appear better in marketing materials than both Athlon and Pentium 3.

[RejZoR]

AMD's 8 cores were actual real 8 cores. There was no hyperthreading involved. It's just that they had separate INT units, but shared FP unit. And that's seen by some as cheating...

[Subway]

FX 8350 is still pretty good in 2019 consider it is 8 years old processor! 

[GoldenLag]

4 minutes ago, Taf the Ghost said:

If they'd had reversed the CMT approach, Bulldozer would have also aged better. 

 

Heck, if they'd launched it in 2009, it'd have been brilliant. 

could have, should have, didnt. 

 

4 minutes ago, Curufinwe_wins said:

They were exceedingly negligent in informing consumers about the weaknesses of their new architecture, and deliberately tried to obfuscate the dramatic real-world performance losses these chips resulted in.

are Companies supposed to show the limitations to their customers and avdertise it?

 

it would be like intel informing their CPUs are a little hot at times and are a little bandwhidt starved at AVX workloads etc and that ringbus has limits to how well it can be scaled.

 

i know what you mean, but still, marketing is marketing. Now that i think about it, if im really lucky i might be able to pull of a lawsuit using my own country's laws. though i would need to scheme through everything AMD has said. and the lawsuit would have to take place here. 

8 minutes ago, Curufinwe_wins said:

The lawsuit has all required merit and should be awarded properly.

isnt the lawsuit over if they are actually cores. marketing is marketing. 

[GoldenLag]

2 minutes ago, Subway said:

FX 8350 is still pretty good in 2019 consider it is 8 years old processor! 

its good as long as you arent doing anything singlethreaded. 

 

in more multithreaded loads like modern games it can hold its own quite well equal to around a 2600k. (according to hardwareunboxed i believe)

[Subway]

Just now, GoldenLag said:

its good as long as you arent doing anything singlethreaded. 

 

 in more multithreaded loads like modern games it can hold its own quite well equal to around a 2600k. (according to hardwareunboxed i believe)

It is good in 1080P gaming only. 

[GoldenLag]

2 minutes ago, Subway said:

It is good in 1080P gaming only. 

no, it can do 4k just fine. 

 

just remmebr in any game to turn off graphic settings that involve the CPU (remember to do this at any resolution. )

[Taf the Ghost]

2 minutes ago, Subway said:

It is good in 1080P gaming only. 

It's high-refresh on anything single-thread demanding it can't do.

[Sakkura]

4 minutes ago, Subway said:

It is good in 1080P gaming only. 

That's a silly thing to say. Resolution barely affects CPU workload at all; if anything, higher res makes a fast CPU less necessary as you'll be more GPU-bottlenecked anyway.

[Subway]

4 minutes ago, GoldenLag said:

no, it can do 4k just fine. 

  

just remmebr in any game to turn off graphic settings that involve the CPU (remember to do this at any resolution. )

 

3 minutes ago, Taf the Ghost said:

It's high-refresh on anything single-thread demanding it can't do.

 

Oh shit, so it can do 4k fine as long CPU is not a demand?! I thought it can do 1080P only because there are a lot of videos on Youtube where testers test 1080P with GTX 1070 and it seems like FX 8350 will bottleneck it without OC, and even OC will reduce bottleneck to the point it is not very noticeable. GTA 5 for example will hit FX 8350 a lot. 

[Taf the Ghost]

Just now, Sakkura said:

That's a silly thing to say. Resolution barely affects CPU workload at all; if anything, higher res makes a fast CPU less necessary as you'll be more GPU-bottlenecked anyway.

Games aren't CPU demanding like Rendering, which is something that gets lost on most of the tech sphere. What they leverage is extremely fast, small calculations, especially in DX11. What screws up any of the CMT parts isn't really the CMT, but the interior parts of the pipeline. That's the bit that gets lost about what happened with Piledriver.

[Taf the Ghost]

1 minute ago, Subway said:

 

 

Oh shit, so it can do 4k fine as long CPU is not a demand?! I thought it can do 1080P only because there are a lot of videos on Youtube where testers test 1080P with GTX 1070 and it seems like FX 8350 will bottleneck it without OC, and even OC will reduce bottleneck to the point it is not very noticeable. GTA 5 for example will hit FX 8350 a lot. 

The x300 parts will bottleneck between 80 to 100 FPS in most game engines. The 5 Ghz parts go a little higher, but not that much. Unless you put a 2080 Ti in one, you're not getting that at 4K. That's why it doesn't bottleneck there. At least, not as much.

[ThePD]

21 minutes ago, Curufinwe_wins said:

For example, I had friends as recently as 2016 who still believed (directly as a result of AMD's marketing materials and claims) that the 8350 was a far better 8-core cpu than the 5820k 6-core.

 

If Intel had been falsely marketing their products like this for the 6 years AMD did, everyone would be absolutely losing their minds. But since AMD did it, nah man it's all good. Bulldozer's marketing was/is one of the largest scams in the history of computing products. And only general consumers paid the price, because everyone else knew better (or found out quickly).

It is our duty as consumers to call out companies for false marketing. People are not glossing over this because it is AMD. Why would people do that for a product line that was considered some of the worst to come from AMD? Intel paid for false third party benchmarks to be revealed ahead of the actual hardware reviewers... But it was 'Nah man its all good' because Intel was 'better at gaming'.

Hold AMD accountable for things they do wrong, as well as Intel/Nvidia. AMD put enough execution units on the cpus to match the core count. It was the internal design that was flawed so the performance never scaled accordingly.

[Curufinwe_wins]

18 minutes ago, GoldenLag said:

could have, should have, didnt. 

 

are Companies supposed to show the limitations to their customers and avdertise it?

 

it would be like intel informing their CPUs are a little hot at times and are a little bandwhidt starved at AVX workloads etc and that ringbus has limits to how well it can be scaled.

 

i know what you mean, but still, marketing is marketing. Now that i think about it, if im really lucky i might be able to pull of a lawsuit using my own country's laws. though i would need to scheme through everything AMD has said. and the lawsuit would have to take place here. 

isnt the lawsuit over if they are actually cores. marketing is marketing. 

'Failure to disclose' information that would have dramatically impacted the decision to buy a product is legally misrepresentation and fraud thus subject to FTC Code 52 and 54, even in the US (with some of the weakest consumer protection enforcement in the developed world).

 

Lawsuits from 'failure to disclose', however, are harder to win, as demonstrative proof of ill-consequence or intent to deceive needs to be provided. Also, many times courts simply refuse to hear them.

 

Again. Intel lost (well agreed to settle once it was evident they were going to lose) a class action suit over similar marketing with regards to the Pentium 4 lineup amidst the 'Gigahertz War'.

[Subway]

10 minutes ago, Taf the Ghost said:

The x300 parts will bottleneck between 80 to 100 FPS in most game engines. The 5 Ghz parts go a little higher, but not that much. Unless you put a 2080 Ti in one, you're not getting that at 4K. That's why it doesn't bottleneck there. At least, not as much.

What is x300? 

[Curufinwe_wins]

6 minutes ago, ThePD said:

It is our duty as consumers to call out companies for false marketing. People are not glossing over this because it is AMD. Why would people do that for a product line that was considered some of the worst to come from AMD? Intel paid for false third party benchmarks to be revealed ahead of the actual hardware reviewers... But it was 'Nah man its all good' because Intel was 'better at gaming'.

Hold AMD accountable for things they do wrong, as well as Intel/Nvidia. AMD put enough execution units on the cpus to match the core count. It was the internal design that was flawed so the performance never scaled accordingly.

And Intel saying their P4 is 2 GHz isn't directly wrong either. But the misuse of that material fact to mislead consumers into forming a false impression of the relative performance between P3 and P4 and Athlon still yielded a successful class action lawsuit.

[GoldenLag]

4 minutes ago, Curufinwe_wins said:

'Failure to disclose' information that would have dramatically impacted the decision to buy a product is legally misrepresentation and fraud thus subject to FTC Code 52 and 54, even in the US (with some of the weakest consumer protection enforcement in the developed world).

 

Lawsuits from 'failure to disclose', however, are harder to win, as demonstrative proof of ill-consequence or intent to deceive needs to be provided. Also, many times courts simply refuse to hear them.

 

Again. Intel lost (well agreed to settle once it was evident they were going to lose) a class action suit over similar marketing with regards to the Pentium 4 lineup amidst the 'Gigahertz War'.

which is why you might be able to win the lawsuit against AMD by attacking from that angle. 

 

though the "core" lawsuit seems a bit silly to me. 

[Taf the Ghost]

1 minute ago, Subway said:

What is x300? 

FX 6300, 6350, 8300, 8320, 8350 and I think there's an 8370 in there.

[Curufinwe_wins]

3 minutes ago, GoldenLag said:

which is why you might be able to win the lawsuit against AMD by attacking from that angle. 

 

though the "core" lawsuit seems a bit silly to me. 

The lawsuit argues that as part of their brief. This source article is literally picking up just tiny bits of it instead, because it is easier to frame a narrative when you cherry pick.