Intel Fires Shots At AMD For False Marketing Of Boost Clocks

Grim5 · 30 Aug 2019 at 20:35

Bluntwrapped said:
On the subject of false marketing, has anyone suggested Intel review their interpretation of TDP?

It’s not a lie - Intels tdp is their base clock tdp. Auto boost is self overclocking and raises power consumption

Changing a single bios setting will make your 9900k stick to its 95w tdp. Nothing you can do in your bios will make your 3900x hold 4.6ghz

Bluntwrapped · 2 Sep 2019 at 13:57

Grim5 said:
It’s not a lie - Intels tdp is their base clock tdp. Auto boost is self overclocking and raises power consumption

Changing a single bios setting will make your 9900k stick to its 95w tdp. Nothing you can do in your bios will make your 3900x hold 4.6ghz

Who buys an unlocked CPU to run it at the base clock though?

Nghtmare · 2 Sep 2019 at 13:59

Bluntwrapped said:
Who buys an unlocked CPU to run it at the base clock though?

I have a friend that bought a 7700K and said he doesn't want to overclock it because its good enough.....

I don't agree with that but there seems to be people that indeed buy K processors and not overclock them.

humbug · 2 Sep 2019 at 15:37

Wrong thread

Nghtmare · 2 Sep 2019 at 15:41

That is a pretty big difference.
14% on lows and 10% on average? Correct me if I messed up the math lol

hyperseven · 2 Sep 2019 at 17:03

Only 5.6% of 3900X reach advertised boost speeds :eek:

.

Maybe grounds for another class action for AMD incoming for telling porkies, especially after their recent one resulted in a huge pay out?

https://amp.tomshardware.com/news/a...ing-advertised-boost-speeds-survey,40291.html

TrM · 2 Sep 2019 at 17:16

hyperseven said:
Only 5.6% of 3900X reach advertised boost speeds .

Maybe grounds for another class action for AMD incoming for telling porkies, especially after their recent one resulted in a huge pay out?

https://amp.tomshardware.com/news/a...ing-advertised-boost-speeds-survey,40291.html

sorry but i must have mised something hehe what huge payout for amd telling lies? if your talking baout the 12+m payout for bulldozer lets be honest here amd or any othegr company will not learn from that 12m payout is nothing and if the story is to belived its going to = $30 dollors a person for so much effort and time and how many years have they waited?

secondly i know this is a multinational forum userbase but its tied a lot to the uk and since we can only just bring classaction lawsuites vers companies i belive it was introduced in 2015 but none of real account has gone though yet uk people have missed out on so many in the tech world in the uk usa thats a differnt story

but do u think 12m will teach a lesson to a company the size of amd or just be anotther expense? vs sales on fx processors i bet its a drop in the ocean for them

jigger · 2 Sep 2019 at 17:20

TrM said:
sorry but i must have mised something hehe what huge payout for amd telling lies? if your talking baout the 12+m payout for bulldozer lets be honest here amd or any othegr company will not learn from that 12m payout is nothing and if the story is to belived its going to = $30 dollors a person for so much effort and time and how many years have they waited?

secondly i know this is a multinational forum userbase but its tied a lot to the uk and since we can only just bring classaction lawsuites vers companies i belive it was introduced in 2015 but none of real account has gone though yet uk people have missed out on so many in the tech world in the uk usa thats a differnt story

but do u think 12m will teach a lesson to a company the size of amd or just be anotther expense? vs sales on fx processors i bet its a drop in the ocean for them

I don’t think AMD have a lesson to learn from the FX payout. The chip had 8 cores.

TrM · 2 Sep 2019 at 17:26

jigger said:
I don’t think AMD have a lesson to learn from the FX payout. The chip had 8 cores.

i only used that alwsuitre as it was the amd payout all over the news.

but saying that i dont think the fx was a truew 8 core cpu though in terms definded by intel and amd all the way back to pentuim d and athon duel core though but tahts how i see it and i dont expect every one to agree

jigger · 2 Sep 2019 at 17:34

TrM said:
i only used that alwsuitre as it was the amd payout all over the news.

but saying that i dont think the fx was a truew 8 core cpu though in terms definded by intel and amd all the way back to pentuim d and athon duel core though but tahts how i see it and i dont expect every one to agree

Would it have been fair if AMD seen the cases through to end and sued for 12 million.

In this case I wouldn’t be bothered about 50mhz of performance on a single core from 12. If people are then they should return the chips for a full refund not £20 in 5 years time.

hyperseven · 2 Sep 2019 at 18:02

jigger said:
Would it have been fair if AMD seen the cases through to end and sued for 12 million.

In this case I wouldn’t be bothered about 50mhz of performance on a single core from 12. If people are then they should return the chips for a full refund not £20 in 5 years time.

'Fair' doesnt come into it they are a business and answer to the board/shareholders :confused:

Costs were mounting up so they conceded and paid out.

jigger · 2 Sep 2019 at 19:39

hyperseven said:
'Fair' doesnt come into it they are a business and answer to the board/shareholders Costs were mounting up so they conceded and paid out.

Of course fair has to be a consideration when coming to an agreement.

Jacky60 · 3 Sep 2019 at 11:53

https://www.guru3d.com/news-story/d...en-3000-chips-does-not-reach-boost-clock.html
Seems pretty widespread!

humbug · 3 Sep 2019 at 12:33

Jacky60 said:
https://www.guru3d.com/news-story/d...en-3000-chips-does-not-reach-boost-clock.html
Seems pretty widespread!

Again, while i agree AMD's marketing is somewhat misleading Der8auer's piece is an exercise in misinformation.

The fact is Ryzen 3000 does hit the advertised boost clocks, just not when and where one might reasonably expect that, and this is the trap Der8auer himself has fallen into, him and many of us who bought one were expecting them to boost to the advertised boost speeds in single threaded workloads, no matter what that application is, this is simply not how they work and AMD never said this is how they work, mine never hits 4.2Ghz in Cenebench, or any game, it does boost to 4.2Ghz pretty much all the time while apparently idle, of course a CPU is never actually idle, windows is constantly making requests of the CPU, so it is running at its advertised boost speeds for Windows background services.

Having said that i have run AIDA64 quite a lot recently as i'm still messing about with tuning my RAM and i see it run at 4.2Ghz in that too.

I think a good tech journalist would do a better job informing the public studying what is actually going on and explaining it, this instead of falling into the same trap normies do.

LePhuronn · 3 Sep 2019 at 12:44

Grim5 said:
It’s not a lie - Intels tdp is their base clock tdp. Auto boost is self overclocking and raises power consumption

No, but Intel say 5GHz boost and 95W TDP. Both are accurate independently of each other, but not together. Intel also do not explicitly say that 95W TDP is for base clocks and it'll go higher when the CPU boosts itself, therefore rarely staying at its advertised TDP. Now yes, you can BIOS lock it down, but the Venn diagram of users who have the knowledge to enter the BIOS yet do not understand Intel's TDP rating is a very, very small crossover.

So Intel are lying just as much as AMD here, i.e. both fully or both not at all.

humbug · 3 Sep 2019 at 12:57

LePhuronn said:
No, but Intel say 5GHz boost and 95W TDP. Both are accurate independently of each other, but not together. Intel also do not explicitly say that 95W TDP is for base clocks and it'll go higher when the CPU boosts itself, therefore rarely staying at its advertised TDP. Now yes, you can BIOS lock it down, but the Venn diagram of users who have the knowledge to enter the BIOS yet do not understand Intel's TDP rating is a very, very small crossover.

So Intel are lying just as much as AMD here, i.e. both fully or both not at all.

Its worse than that, some reviewers, PCPer for example, on reviewing the 9900K did it with MCE on (so 5Ghz all core) when performance benchmarking, and then switched to a motherboard that had a 95 Watt lock on it for power consumption tests concluding it a 95 Watt CPU.

Who knows if they did this off their own volition or by request of Intel, all i know is Ryan shrout now works for Intel and is one of those people advocating "real world applications in performance benchmarks"

LePhuronn · 3 Sep 2019 at 12:59

Now that's just outright deceit.

Burn the shills, humbug, burn the shills. I wonder if Ryan is actually paid by Intel for his role? Strikes me he'd be happy to do it for free given Senpai finally noticed him...

humbug · 3 Sep 2019 at 13:06

I never did like Ryan Shrout, can you tell? when comparing G-Sync to Free-Sync he deliberately used a Screen known to be broken on the Free-Sync side and of course didn't have anything to say about it that was positive, when called out on his choice of dodgy screen he put a vague disclaimer in the comment section of the Video. :rolleyes:

There's more Ryan Shrout shenanigans i could tell you about....

4K8KW10 · 3 Sep 2019 at 13:22

jigger said:
I don’t think AMD have a lesson to learn from the FX payout. The chip had 8 cores.

If it indeed had 8 cores, why didn't AMD prove it and dismiss the class-action lawsuit claim by the customers?

humbug · 3 Sep 2019 at 13:24

Last one i promise...

Remember AMD's Mantle API? Its a direct to metal API that works far more efficiently than DX, basically it can greatly increase CPU performance where the CPU is the bottleneck.

So Ryan Shrout reviewed this API my using a refrence R9 290 none X and an FX 8350 vs a very overclocked top on the line Intel CPU, can't remember which one... he did all this in Single Player campaign mode BF4,where the CPU is pretty much irrelevant, especially with a throttling reference 290.

And concluded, words pretty much to the effect "Mantle is crap, it doesn't do anything"

I made this video to prove otherwise.