AMD advertisement claims Ryzen Pro capable of 5GHZ. You must be doing it wrong.

[pgdeaner]

Y'all must be doin it wrong.

 

[Zerileous]

Really it implies more than 5GHz. This is getting ridiculous. It's like the Bulldozer settlement only encouraged them.

 

[Bustos]

I couldn’t hear it and there weren’t any caption options (deaf). Based on what I saw...technically, it doesn’t state it can reach 5.0GHz. 5.0GHz is the max of that meter.

Lmk if it actually stated it can reach 5.0GHz.

 

[ShrimpBrime]

Now hold on a second.

The guy in the video on said blazing fast speeds.... nothing about 5ghz.\

The illustration showed 5ghz. It really doesn't mean it will do that......
So the car you drive with the 4 banger (VW jetta 2.0) has a speedo that reads up to 160. I can tell you from experience even down a hill with the wind at it's back.... ain't doing no 160.... SUING VW NOW!!

HOWEVER!!!!
The 3950X has a max boost of 4.7ghz (probably single core boost).
Could be possible to reach 5ghz under certain circumstances?? (?LN2?) =/-
edit: Maybe SMT off and knocked down to 8 cores ambient 5ghz????

 

[Zerileous]

OK good point, but it's still a bit deceptive. I guess we're used to reading the needle, and the needle is under the 5GHz mark. However we're also used to reading what's moving, and when I first saw it, I thought the needle was representing 5GHz and the tape was representing speed exceeding 5GHz. Before you say "that's ridiculous" that is actually common in some vehicles, older Chevy pickups used tape dials instead of needles.

Maybe everyone does this stuff in their ads anyway

 

[bignazpwns]

Now hold on a second.

The guy in the video on said blazing fast speeds.... nothing about 5ghz.\

The illustration showed 5ghz. It really doesn't mean it will do that......
So the car you drive with the 4 banger (VW jetta 2.0) has a speedo that reads up to 160. I can tell you from experience even down a hill with the wind at it's back.... ain't doing no 160.... SUING VW NOW!!

HOWEVER!!!!
The 3950X has a max boost of 4.7ghz (probably single core boost).
Could be possible to reach 5ghz under certain circumstances?? (?LN2?) =/-
edit: Maybe SMT off and knocked down to 8 cores ambient 5ghz????

The Speedo on my Camaro says 200 and at the Texas mile it did 218.4mph

 

[Woomack]

I guess that most users would be more than happy with 4.5GHz while barely anyone will see fully stable all cores at more than 4.3GHz. Can only check companies that sell pre-tested Ryzens. Looking that what they sell I feel like my 3700X is special and it can only make a bit above 4.3GHz

 

[ShrimpBrime]

We should be happy with Ryzen the way it is.
At this small 7nm, we could have been stuck in the 3ghz range.
An old timer once told me core count will go up and speed will go down while IPC will be gained either way. This man was around before X86 was invented and has recently passed away.

What Im trying to say is the chips are fast even if lacking that wonderous 5ghz.

I for one am interested in effeciency. Low power consumption and lots of IPC. I think AMD has made a fine processor and is now very competitive. Wouldnt have it any other way.

 

[EarthDog]

Really it implies more than 5GHz. This is getting ridiculous. It's like the Bulldozer settlement only encouraged them.

I really feel the same... hilarious they even have 5ghz in an image, nonetheless an arrow inferring it can be reached. Good times.

This is like a speedometer, except it shows the needle at 200 mph (5ghz).

 

[Woomack]

We should be happy with Ryzen the way it is.
At this small 7nm, we could have been stuck in the 3ghz range.
An old timer once told me core count will go up and speed will go down while IPC will be gained either way. This man was around before X86 was invented and has recently passed away.

What Im trying to say is the chips are fast even if lacking that wonderous 5ghz.

I for one am interested in effeciency. Low power consumption and lots of IPC. I think AMD has made a fine processor and is now very competitive. Wouldnt have it any other way.

I remember a bit different statement from some engineers maybe 15 years ago. It was more like max clock for mass sales won't be higher than 5GHz because it's hard to stabilize it regardless of used process. In fact there was only one Xeon at 5.2GHz or something and 2 cores which was in limited quantities and for special order. It was the only Intel CPU available at 5GHz+ till 9900K was released that I remember. Even with 9900K, CPUs clock is not 5GHz but CPU's mode let to run single threads at 5GHz, depends on many factors. It's clear that Intel has also a big problem with stabilizing 5GHz.
Because of problems with higher frequency, all went with more cores and improving cache or IPC.

I wasn't checking it directly but I feel like Ryzen 3000 has worse efficiency than 1000/2000. It uses more power and heats up more. I wouldn't call even Ryzen 3600 low power CPU when it's passing 200W under full load and barely any, less expensive cooler can keep it under 95°C. AMD couldn't even deliver stock coolers that would keep their CPUs under the throttling point. On paper it's 65W TDP for 3600/3700x, in real it's at least twice as much. Even cooler manufacturers started to change their compatibility and TDP lists as it was far from reality.

I only wonder about that Ryzen Pro and 5GHz as Pro is for business computers and workstations and supposed to support ECC too. Would be nice to see better binning but I guess that architecture simply won't let it.

For sure AMD marketing is strong right now. They sell a product which isn't perfect but no one is talking about its disadvantages and no one really cares.

 

[ShrimpBrime]

I believe the misconception is real.
Leaving Intel out of this discussion though.

So the misconception is that the Ryzen chips are rated at say 85w for w/e example.
I dont not believe this includes the boost frequency and voltages.
So my 2700x is well over that tdp rating of 105w when PBO is enabled. However, turning it off which is not default to the motherboard... Its actually well within that TDP spec at 3.7ghz 1.212v.
But the boards post stock with boosting enabled and thats what people want cause thats what is advertised. And so then they are running that 1.4v and wonder where the heat comes from.

But who am I to say the Boost Range should be within TDP spec. It may exceed that TDP spec and thats ok. But then its all misleading.

I wont have a 3000 series chip for a while. I just dont need it. Ive found all the ways to make my 2700x nice and effecient. It really doesnt need 1.4v to run 4ghz either like the boards seem to push them to.
But in experience, I think the 3000 series could be more effecient if the bios on boards where better rounded off with Boosting in Tiers and less temperature dependant.

The white papers for the first two generations actually recommend a temp of 70c and lower. It is a high temp warning limit in fact, I imagine the 3000 series chips are the same way.

Earthdog did mention and I fully agree that the size and voltage plays a big role in the heating issues, but to further add to this I am in belief the IHS plates are too small or too thin. With a little more mass to them, overall temps could be lower.
I only say this from experience while Ive delidded mine. I have a variety of cold plates including one made of silver. The plate makes a difference believe it or not.

 

[Zerileous]

We should be happy with Ryzen the way it is.

I agree with this completely. I just don't think AMD sees it / markets it that way. They keep doing things that imply unreasonable clocks, and the way they set up their boost is suspect too. While enthusiasts wanted a CPU to boost to the moon and not be held back by imposed clock speed limits (like Zen+). Every CPU should hit the clocks advertised on the box if cooling is good enough. Extending / removing that range should be an optional enthusiast / OC option. They in fact already have this with Auto OC (+200MHz max boost). They could have set up their CPUs to have a max boost 100MHz lower out of the box, and made AutoOC a +400MHz max boost and made everyone happy.

Regardless of what clocks you reach etc, they are matching / tying Intel at this point, so why advertise clocks they won't hit? Well the mass consumer won't be looking at benchmarks and reviews, of course. To them, the only way to communicate that "we're as good as Intel," is to show the same clock speeds as Intel. Sadly. I want AMD to succeed, badly. I know you don't want to bring up Intel, but honestly it is annoying that they have locked processors, all kinds of stuff to create artificially bad CPUs to sell at budget / moderate price points and charge insane amounts of money for the same old stuff without those limitations. They need good competition to keep them innovating, and that is what AMD is providing. However, none of this justifies dishonest marketing.

 

[ShrimpBrime]

Its only dishonest because of a gif on a slide?
Dont recall ever seeing 5ghz actually advertised to be the norm.
The OP like many nit picked a singular video.
Is that trying to start some flame war???

 

[Bustos]

I had a similar thought when I first saw the OP and the video. ‘Is this an attempt to troll?’

 

[Kenrou]

What am i missing ?

 

[EarthDog]

Nothing?

It wasnt private before... suddenly it is...


...maybe they are correcting the 5 GHz thing?

 

[ShrimpBrime]

Interesting. Bait troll thread gets called out and video disappears.

Now to figure out which one of you made it.

 

[EarthDog]

Interesting. Bait troll thread gets called out and video disappears.

Now to figure out which one of you made it.

 

[ShrimpBrime]

 

[EarthDog]

Here is the new video they put up...

... and a screenshot of the same 5 GHz portion...