CPU 'Marketing'

[Hans de Vries]

A questionable "news-source" talks of an AMD scandal and claims it might result in a law suit. The author claims that AMD misleads customers into thinking that Ryzen PRO is capable of 5GHz. It is however the author himself (or his sources) who are highly misleading.

It is all about an AMD marketing Video published just yesterday. Note WCCFtech's (manufactured) image below has a speedometer pointing all the way up to 5 GHz and the author falsely claims: "AMD's Ryzen Pro promo showcases clock speeds of up to 5GHz "

This is in no way what AMD's real video shows. In the real video the speed fluctuates but never goes up to 5 GHz and varies around the frequencies as we know them.




This following video below seems to be some related discrediting marketing material: It is a professional Ray (path) traced video that accuses AMD of 5 GHz lies. The production of this video within such a short time raises the question if there is something bigger behind this discrediting article as just the author?





Below is an actual photo of AMD's marketing video: The speedometer fluctuates but stays well below 5 GHz. Nowhere in the video it goes up to 5 GHz and nowhere in the video is claimed that speeds of 5GHz would be possible.

If i try to post the image below on WCCFtech it won't get published but is marked as spam instead.... The same author was recently responsible for the claim that Lisa Su was leaving AMD claiming that his sources had whispered this to him (which?) and refused to retract the claim in any way. I'm sure this article to discredit AMD will stay there as well. The video is here.

It will be interesting to see if there will be more news-sources who will publish this in the same way and then refuse to take it down if debunked.

 

[Kenmitch]

Did you watch and listen to the video the image was pulled from? I don't even remember any clock speeds mentioned throughout the video at all.

If I was on a jury and watched the video I wouldn't be able to come to the conclusion that it says we can hit 5 GHz. I view it as a speed chart and you've got the needle bouncing around, the upper max speed achieved illustration and the 5 GHz target. I view it as 5 GHz on the chart is the completion of the full 1/2 circle of the progress bar. If AMD was out to mislead then they'd most likely of had the needle slapping just above the 5 in 5 GHz.

 

[Hans de Vries]

If I was on a jury and watched the video I wouldn't be able to come to the conclusion that it says we can hit 5 GHz.

Yes, you get the video if you click on the image. I'll add an extra link and some more comments.

Unfortunately the (intentionally) highly misleading image which they have on their front page will mislead most of the tens of thousands of their readers. Note that the meter in their misleading image points right to the end of the 5 GHz scale. This is not what AMD's video shows.

 

[Kenmitch]

You view it as the end of the scale, while I see it as the max clock achieved. Who's right?

 

[Markfw]

You view it as the end of the scale, while I see it as the max clock achieved. Who's right?

I did not see it hit the end, just close. Any they never said it, and there is no delineations on the scale. Yes, its marketing, but if it not in print or verbal then it makes users ASSUME something. In Intels advertisement they said it hit 5 ghz (on that 28 core water chilled demo), they just neglected to say its overclocked, and what it took to do it. Did Intel get sued ? No. Will AMD in this case ? I doubt it.

That is marketing,.

 

[Thunder 57]

What if my car isn't capable of hitting the 150 on the speedometer, is that false advertising?

 

[joesiv]

You view it as the end of the scale, while I see it as the max clock achieved. Who's right?

I'm torn, if the speed in the text changed (went up as the bar was filled), I would say they were saying it would hit 5ghz. But it's strange the bar moves, but the needle is already bouncing around, at the same spot, at the same time as the 5Ghz appears on the screen. Also the 5Ghz is so bold, it's what you're eyes are drawn to, and without any markings on the bar which fills...

It's just a confusing graphic TBH. Marketing I guess...

As for WCCF, yeah that seems like a pretty obvious clickbait, likely intentionally misleading article.

To be honest, the Call of Duty Ray Traced article also seems very clickbaity, and more of a paid advert. I think the new engine looks great, but it has more to do with the way they're rendering and generating the scenes than raytracing. They've finally got a next gen engine!

 

[Vattila]

AMD has edited out the "5 GHz" in the video. It now says only "GHz".

 

[DrMrLordX]

That speedometer is just stupid marketing. But yeah the needle doesn't actually go there.

With the edit it just looks equally stupid.

 

[Hans de Vries]

I just knew the dirty smear campaign was coming: This is from John Bonini, Intel's Vice President, General Manager Desktop, Workstations.

https://twitter.com/x/status/1167493914502127616

https://fudzilla.com/news/pc-hardware/49309-intel-gm-of-desktop-slams-ryzen-5ghz-claim

This is how it usually works:

- First he prepared:

1) The faked image suggesting that AMD's video claims 5 GHz for Ryzen pro.
2) A special 5 GHz i9-9900k in "regular" nitrogen slide for the campaign.
3) Very likely the profesional "AMD 5 GHz lies" video as well. (see the opening post)

- Next he used some of the click-bait tech-press to spread out the lies. (Now you know one of these WCCFtech / Fudzilla's / TweakTown "sources" )

- The final step is then to "refer" to these "press-articles" in all kinds of marketing material.

In this way their lawyers assure that they can't be hold liable for spreading lies, even if they invented the lies in the first place. It has always been like this.

 

[ehume]

Why is this such a dirty smear? I had my 8700k OC'd all six cores and twelve threads at 5GHz with no decrement for AVX, cooled on air; and I don't think I was alone. Thus I find claims that the 9900k can be all-core OC'd to 5GHz on air to be credible. Besides, from my recollection, AMD CPU's have always run at a lower clockspeed than Intel's. Lower clockspeed would not keep an AMD CPU from matching an Intel CPU in productivity.

It's AMD's fault for trying to make a statement about clockspeed. Although you can look to clockspeed to compare your relative OC to another's OC on a comparable CPU (AMD vs AMD, Intel vs Intel), trying to compare clockspeeds between AMD and Intel is not legit. Their architectures differ, after all.

 

[Hans de Vries]

Why is this such a dirty smear? ........
It's AMD's fault for trying to make a statement about clockspeed. Although you can look to clockspeed to compare your relative OC to another's OC on a comparable CPU (AMD vs AMD, Intel vs Intel), trying to compare clockspeeds between AMD and Intel is not legit. Their architectures differ, after all.

That's how this kind of marketing works.

Intel's only selling point left is the 5 GHz. So what they do is fabricate a fake story in which AMD seems to "desperately" lie to their business customers that Ryzen Pro can do 5GHz. Next they spread this story via via in a sleazy way over the entire internet shaming AMD for their so called lies. Then, finally, after shaming AMD they triumphantly come up with the "best of the best", their new 5GHz 9900KS.

Of course the 9900KS is just a marginal little step forward and will not help them to compete in productivity applications with the 3900X, let alone the 3950X.

The dirty smear here is is that they claim that AMD is lying to its business customers and the sleazy way this fabricated fake story is spread over the internet.

 

[Thunder 57]

...Besides, from my recollection, AMD CPU's have always run at a lower clockspeed than Intel's..

Useless trivia: K7 ran faster than the PIII. Pretty sure Intel caught up with Coppormine, then lost it again when Thunderbird came out. Also I'm pretty sure that the A64 X2 was faster (clockspeed) than Core 2 Duo. Of course that was irrelevant since it was outperformed.

EDIT

Actually, I'm not sure if the P-III was ever faster than the Athlon. The Athlon was the first to 1GHz, and that was before Thunderbird.

 

[Markfw]

Useless trivia: K7 ran faster than the PIII. Pretty sure Intel caught up with Coppormine, then lost it again when Thunderbird came out. Also I'm pretty sure that the A64 X2 was faster (clockspeed) than Core 2 Duo. Of course that was irrelevant since it was outperformed.

EDIT

Actually, I'm not sure if the P-III was ever faster than the Athlon. The Athlon was the first to 1GHz, and that was before Thunderbird.

While I currently favor AMD as the one to buy, the core2duo was faster than any Athlon, and overclocked like no tomorrow. I remember a friend and I both had 1.86 ghz stock C2D's overclocked to 3.5 ghz. Even at 1.86 they beat the X2's. I actually have one, but its not turned on anymore.(the X2's). But it was Intel, then AMD, then Intel, then AMD, then Intel.... Until 2017 and the Ryzen started to be competitive, and now, the 3000 series is a just way better value, and performance king, except the edge cases that the 9900k wins.

 

[Thunder 57]

While I currently favor AMD as the one to buy, the core2duo was faster than any Athlon, and overclocked like no tomorrow. I remember a friend and I both had 1.86 ghz stock C2D's overclocked to 3.5 ghz. Even at 1.86 they beat the X2's. I actually have one, but its not turned on anymore.(the X2's). But it was Intel, then AMD, then Intel, then AMD, then Intel.... Until 2017 and the Ryzen started to be competitive, and now, the 3000 series is a just way better value, and performance king, except the edge cases that the 9900k wins.

The question was strictly regarding clock speed. My very next sentence was "Of course that was irrelevant since it was outperformed.", meaning C2D outperformed A64 despite having a lower clock speed. I never had a C2D as I was too invested into AM2. In the end I remember my Phenom II at 3.0GHz was similar to my brothers C2Q at 2.6GHz. We both got 3570k's after those which he's still rocking while I'm back to AMD on a 2600X.

 

[zinfamous]

What if my car isn't capable of hitting the 150 on the speedometer, is that false advertising?

Mine goes up to 185 but is limited to 155.

Time to sue VW! er...I mean, how has no one else thought of doing that???? $$$$$

 

[Thunder 57]

Mine goes up to 185 but is limited to 155.

Time to sue VW! er...I mean, how has no one else thought of doing that???? $$$$$

I think mine actually goes to 160, pretty sure 80 is at the top. Just sort of picked a number ,

Because we are used to not maxing out the speedo, kind of like not redlining all the time? Or maybe because everyone would switch to ugly, horrible digital speedometers.

 

[Markfw]

I think mine actually goes to 160, pretty sure 80 is at the top. Just sort of picked a number ,

Because we are used to not maxing out the speedo, kind of like not redlining all the time? Or maybe because everyone would switch to ugly, horrible digital speedometers.

The RX-8 got it right. The main instrument, and analog is the RPM. Redline@8500, but without the limiter, will do 14,000. I have only had mine up to 130 mph, but based on RPM and only being in 5th gear out of 6, I would say my top end is most likely 160-170, but since the speedo is digital, who knows.

Anyway, I don't know why everybody is making such a big deal of the MHZ of the CPU, its the performance that matters. @stock all 3 of mine run between 4000 and 4100 all cores@100%load. 24/7 and they take less juice than my 2500 mhz Xeons. Time for the kill-a-watts. I will post back after I have actual numbers.

Edit: I know the 3900x has more IPC than the E5-2683v3 Xeons, so I calculated "work done" by multiplying threads*mhz.. 97.7 for the 3900x at 227 watts. 70 for the Xeons@185 watts. The 3900x puts out 50% more work for the same wattage.

Just so you know the parameters. Both were running win 10 build 1903. Load was all threads, 100% CPU usage of WCG. Since the Ryzen 3900x has better IPC than the Xeon, the numbers should sway even more, not sure how to test better, suggestions welcome.

 

[DrMrLordX]

Why is this such a dirty smear?

It looks like he edited the AMD "5 GHz" spedometer graphic to make the needle go all the way to the end. If you saw the original video, you'll see that AMD never let the needle go there. The clear implication was that it goes close to 5 GHz but doesn't quite reach it.

 

[Hitman928]

It looks like he edited the AMD "5 GHz" spedometer graphic to make the needle go all the way to the end. If you saw the original video, you'll see that AMD never let the needle go there. The clear implication was that it goes close to 5 GHz but doesn't quite reach it.

It's not an edit but a dishonest screen grab. In the video the needle is there as the scale to 5GHz grows. So at one point the scale crosses the needle before growing past it to show the needle doesn't fully reach 5GHz. If you screenshot the video at that right point you could dishonestly say AMD is claiming 5 GHz speeds. If you screenshot just before that moment, the needle would be past 5 GHz, if you screenshot just after that moment, the needle would be just short of 5 GHz as it was intended to be.

 

[DrMrLordX]

It's not an edit but a dishonest screen grab. In the video the needle is there as the scale to 5GHz grows. So at one point the scale crosses the needle before growing past it to show the needle doesn't fully reach 5GHz. If you screenshot the video at that right point you could dishonestly say AMD is claiming 5 GHz speeds. If you screenshot just before that moment, the needle would be past 5 GHz, if you screenshot just after that moment, the needle would be just short of 5 GHz as it was intended to be.

Ah okay. So more of a selective screengrab than an edit.

 

[TheELF]

In the video the needle is there as the scale to 5GHz grows. So at one point the scale crosses the needle before growing past it to show the needle doesn't fully reach 5GHz.

All this happened in less than a second in the AMD video...not enough time for a brain to process all of this.
All you would see is the big fat 5Ghz while the jittery needle on the gauge would just be background noise for your eyes.
While the screenshot does capture the worst possible moment the video itself wasn't all that non-miss-leading either.

 

[DrMrLordX]

While the screenshot does capture the worst possible moment the video itself wasn't all that non-miss-leading either.

Welcome to marketing. Enjoy your stay!

 

[Hitman928]

All this happened in less than a second in the AMD video...not enough time for a brain to process all of this.
All you would see is the big fat 5Ghz while the jittery needle on the gauge would just be background noise for your eyes.
While the screenshot does capture the worst possible moment the video itself wasn't all that non-miss-leading either.

I agree it was done in a way that many people could see it as advertising 5 GHz. It still doesn't change the fact though that,

1) A careful inspection showed that it didn't advertise 5 GHz speeds.
2) The tweet and following articles in question are using a dishonest screenshot to try to say that the advertisement was claiming 5 GHz speeds.

 

[TheELF]

to try to say that the advertisement was claiming 5 GHz speeds.

The video WAS claiming 5Ghz,it wasn't claiming that the pro would reach that but it was claiming 5Ghz that's the whole problem,that's why AMD edited out the 5 part and left only Ghz standing in the video.
Edited official video: