AMD Ryzen 9 3950X Overclocked To 5 GHz Across All 16 Cores On LN2

Status
Not open for further replies.
The Notebookcheck article claims that it is an ES, but both the hwbot submission and the Geekbench database prove it isn't an ES. :rolleyes:



I don't known anyone expecting 6GHz 10nm Intel parts, but I know people believed AdoredTV 'leak' about a 5.1GHz 16C Zen2 $499 chip.



You are quoting that I wrote in on other threads about other chips. I have explained in this thread that this chip is neither an ES nor an QS. It is a retail chip with the Ordering Part Number "100-000000033-01".



You are making up stuff again. I explained in #169 why this "AMD 100-000000033-01" chip isn't an engineering sample. And I added in #192 that the codenames 100-000000... are the Ordering Part Numbers for the retails chips.



I posted a leak with the full Rome lineup that will be announced soon. It isn't in any article. The table proves that 100-0000... codenames are the OPN of the actual chips.

Still rekt
 
The Notebookcheck article claims that it is an ES, but both the hwbot submission and the Geekbench database prove it isn't an ES. :rolleyes:



I don't known anyone expecting 6GHz 10nm Intel parts, but I know people believed AdoredTV 'leak' about a 5.1GHz 16C Zen2 $499 chip.



You are quoting that I wrote in on other threads about other chips. I have explained in this thread that this chip is neither an ES nor an QS. It is a retail chip with the Ordering Part Number "100-000000033-01".



You are making up stuff again. I explained in #169 why this "AMD 100-000000033-01" chip isn't an engineering sample. And I added in #192 that the codenames 100-000000... are the Ordering Part Numbers for the retails chips.



I posted a leak with the full Rome lineup that will be announced soon. It isn't in any article. The table proves that 100-0000... codenames are the OPN of the actual chips.

The leaked table doesn't mean shit and proves nothing! First, because it is a leak that hasn't been proven accurate. Two, because it is about Rome, and we are talking about Ryzen 2, and third, because OPN or sku's have nothing to do with revisions of piece of hardware/product, be it cpu's, motherboards, or the drugs you seem to be taking that seem to have disconnected you from reality. :D :D
 
Last edited:
The Notebookcheck article claims that it is an ES, but both the hwbot submission and the Geekbench database prove it isn't an ES. :rolleyes:



I don't known anyone expecting 6GHz 10nm Intel parts, but I know people believed AdoredTV 'leak' about a 5.1GHz 16C Zen2 $499 chip.



You are quoting that I wrote in on other threads about other chips. I have explained in this thread that this chip is neither an ES nor an QS. It is a retail chip with the Ordering Part Number "100-000000033-01".



You are making up stuff again. I explained in #169 why this "AMD 100-000000033-01" chip isn't an engineering sample. And I added in #192 that the codenames 100-000000... are the Ordering Part Numbers for the retails chips.



I posted a leak with the full Rome lineup that will be announced soon. It isn't in any article. The table proves that 100-0000... codenames are the OPN of the actual chips.

"proves"
 
The Notebookcheck article claims that it is an ES, but both the hwbot submission and the Geekbench database prove it isn't an ES. :rolleyes:

I mean, the person with the actual CPU in hand says it's an ES and I trust them a hell of a lot more than I do you on that topic. (*HINT* You actually need to scroll a bit on the HWBOT submission page to see that part.)
 
Juanrga, give this up. Notebookcheck says its an ES. That should be enough. Furthermore, AMD engineering samples do not show up in CPU-Z and other applications as "ES" models consistently, the way Intel CPU's have for more than a decade. I've had plenty of them and NONE of them were easily differentiated from retail samples.
 
Last edited:
Why so desperate to prove that it ISN'T an ES model anyway? Using speculative leaks to argue against other leaks.... Who cares? Either way the chip is showing a super-massive gain relative to Ryzen 2xxx. He claims Cinebench is AMD-centric and Geekbench is a toy bench, but the numbers between a ryzen 2xxx and ryzen 3xxx alone on the same 'toys' should paint the picture just fine. I bet a 3800x @ 4.7ghz is going to put out nearly the same if not slightly better performance than a 9990k@5ghz in almost everything while costing much less money. What a tremendous evil AMD is unleashing upon the earth come July.

There's no way to turn this into any FUD or Intel shilling.

Lastly, a consumer, mainstream chip is whatever they say it is as long as it sells to mainstream consumers (as opposed to professional consumers). We've had $1,000 extreme edition chips on mainstream boards from Intel going back ages. I paid $750~ for my Coppermine Pentium III 850mhz almost 20 years ago now, ignoring inflation... On my mainstream Slot 1 motherboard...

[feeding the troll.gif]
 
Last edited:
You are quoting that I wrote in on other threads about other chips. I have explained in this thread that this chip is neither an ES nor an QS. It is a retail chip with the Ordering Part Number "100-000000033-01".

So when it's a slower clocked version of the retail chip, with features possibly disabled it's an ES.

When it's a full clocked chip, with all features, but before release, it's a QS

When it's a down clocked chip, with all features, but before release, it's now a retail chip?
 
You would rather save face then stop these two guys from systematically controlling the narrative of every AMD thread. Every single thread gets derailed. It’s a detriment to the forum.

I’m sure I’ll be banned soon for speaking the truth. Soviet tactics up in here.
 
You would rather save face then stop these two guys from systematically controlling the narrative of every AMD thread. Every single thread gets derailed. It’s a detriment to the forum.

I’m sure I’ll be banned soon for speaking the truth. Soviet tactics up in here.
Bye Natasha.
 
The irony is it isn’t saving face. It’s jusf following rules. XD

On the topic. Is it really even that relevant if it’s ES QS or Retail? All 3 are frankly pretty damn close to each other. Sure “blah blah ES is lower clocked than retail. End of the day retail is all that matters to end users (i.e. retail). And overclocking a chip is overclocking a chip. ES or Retail nothing is created equal. One can easily OC better than the other. Chance has it all.
 
The best way to win this false inequality is to use intel's own definition.

https://www.intel.com/content/www/us/en/support/articles/000005719/processors.html

Engineering sample (ES) processors are also known as qualification sample processors. They are pre-production processors Intel loans to original equipment manufacturers (OEMs), original device manufacturers (ODMs), and independent software vendors (ISVs) to be used in the product design cycle before product launch.

These processors often include more features than production processors for customer pre-production evaluation and test purposes. ES processors:

  • Produced by Intel are the sole property of Intel.
  • Produced by Intel are Intel Confidential.
  • Are provided by Intel under nondisclosure and/or special loan agreement terms with restrictions on the recipient's handling and use.
  • Are not for sale or resale.
  • May not have passed commercial regulatory requirements.
  • Are not covered under Intel warranty and are generally not supported by Intel.
 
Last edited:
The leaked table doesn't mean shit and proves nothing! First, because it is a leak that hasn't been proven accurate. Two, because it is about Rome, and we are talking about Ryzen 2, and third, because OPN or sku's have nothing to do with revisions of piece of hardware/product, be it cpu's, motherboards, or the drugs you seem to be taking that seem to have disconnected you from reality. :D :D

The table reveals the codenames are OPNs. AMD uses the same codenames for server, and desktop (I have a decoder for decoding server/desktop/mobile). We have now a table that explains what is the codename and we have decoders for the codenames for engineering samples. Your 'theory' is that this chip is an ES, but AMD changed the codename for... unknown reasons.

I mean, the person with the actual CPU in hand says it's an ES and I trust them a hell of a lot more than I do you on that topic. (*HINT* You actually need to scroll a bit on the HWBOT submission page to see that part.)

Thanks. I have scrolled downs and read the comment. He writes

on X470 internal reference board. Board has no bus clock tuning capability. Ryzen 9 3950X engineering sample used.

So he didn't use a full board, as I guessed in a former post: "This R9 chip could be downlocked for lots of reasons, including cooling or motherboard stability issues."

He claims he used an engineering sample, but he didn't prove it and the codenames in the screenshots he submitted show his chip is not an ES. I already mentioned in a older post in this thread how the codename would look if this was a ES.

Juanrga, give this up. Notebookcheck says its an ES. That should be enough. Furthermore, AMD engineering samples do not show up in CPU-Z and other applications as "ES" models consistently, the way Intel CPU's have for more than a decade. I've had plenty of them and NONE of them were easily differentiated from retail samples.

Notebook simply parrots what says an anonymous user named Blueleader in a comment. I am analyzing the screenshots. Geekbench has a consistent records of showing ES, QS, PC, and other kind of chips. I can give you Geekbench entries for true engineering samples of Zen2 chips.

So when it's a slower clocked version of the retail chip, with features possibly disabled it's an ES.

When it's a full clocked chip, with all features, but before release, it's a QS

When it's a down clocked chip, with all features, but before release, it's now a retail chip?

No. It is an ES when the codename says it is an ES, it is an QS, when the codename says it is an QS, and it is a retail chip, when de codename shows it. I have given examples of codenames for Zen and Zen2 chips.

The irony is it isn’t saving face. It’s jusf following rules. XD

On the topic. Is it really even that relevant if it’s ES QS or Retail?

None to me. What matters is the performance the chip demonstrated for the quoted frequencies.

The best way to win this false inequality is to use intel's own definition.

https://www.intel.com/content/www/us/en/support/articles/000005719/processors.html

That is wrong. Engineering samples and qualification samples aren't the same. In this post I gave an explanation of the differences between both and I gave codenames for Zen. For instance 1D2801A2M88E4_32/28_N was a (first batch engineering sample of the 1800X, whereas ZD3601BAM88F4_40/36_Y was the qualification sample of the 1800X.
 
Dude with the chip says its an ES.
So somehow you know better than a person with access to AMD ES chips and AMD's own motherboards.
I seriously don't see how anyone could buy your spin.
Also, you decide that intel.com is wrong about their own definition of what they define as ES and QS?

Still rekt.
 
It reaches comedy levels at how much mental gymnastics is needed to make no real point.

You can be a brandboy and that is fine but when you see nothing wrong then you are reaching obsessive fanboy levels
 
I think for a desktop gaming rig, 6C is the sweet spot right now for strictly a budget gaming box in terms of money spent on a CPU. But with that said, you probably do not want to be "multitasking" very much while gaming. Of course memory footprint is probably going to come more into play there than CPU while actually gaming. It is really hard to argue not spending the money to put at least 8C/16T CPU in your box if you are building/buying new today. When I put this 6C/6T in this box, I really wanted to see just how it would work out for daily usage, as my previous employer had specifically asked me some questions about that. No better way to figure it out for yourself.
With both Sony and Microsoft putting 8 core 8 thread gen 2 Ryzen processors in their next consoles you can bet the next round of game engines (developed for consoles) will be more happily fed by 16 thread CPUs. The ball will move forward and quad and hex core CPUs won’t be up for the task. This is a year or two out though I’d assume.
 
With both Sony and Microsoft putting 8 core 8 thread gen 2 Ryzen processors in their next consoles you can bet the next round of game engines (developed for consoles) will be more happily fed by 16 thread CPUs. The ball will move forward and quad and hex core CPUs won’t be up for the task. This is a year or two out though I’d assume.

I would expect it's further out than that. Did the previous generation consoles do this yet? I think one reason this is not a big problem is console processors are usually much slower per core than high end desktop processor so even if the game engines are tuned to 8C or greater the same engine would likely not need 8C to run on a modern PC.
 
I would expect it's further out than that. Did the previous generation consoles do this yet? I think one reason this is not a big problem is console processors are usually much slower per core than high end desktop processor so even if the game engines are tuned to 8C or greater the same engine would likely not need 8C to run on a modern PC.

That was certainly the case with the PS4 and XBOne and their Jaguar-derived cores, but both next-gen consoles are slated to use Zen 2. That should represent a massive leap in CPU resources and we can only hope that will be reflected on the PC side.
 
MS and Sony will not allocate 8 cores to the games. They will reserve at least two cores and a chunk of memory for OS, recording and streaming + whatever new features they might dream up.

Edit: Just guessing. I am not a self proclaimed expert.
 
Last edited:
I expect the IPC to be much better now with Zen2 based consoles. The frequency probably will still be reduced. We shall see..
 
The table reveals the codenames are OPNs. AMD uses the same codenames for server, and desktop (I have a decoder for decoding server/desktop/mobile). We have now a table that explains what is the codename and we have decoders for the codenames for engineering samples. Your 'theory' is that this chip is an ES, but AMD changed the codename for... unknown reasons.

No, my comment that started this discussion was we don't know if this chip was an ES chip or not. It has already been proven that it is (yet you are still in denial) but that is besides the point. You are arguing about the number, saying that the number AMD 100-0000000033-1. Shows it is not an ES. Which you have already, thru your own statements proven yourself wrong. My point, not a theory, is the AMD 100-000000033-1 doesn't mean jack whether it is a ES chip or not, and IF your leak is correct verifies that because OPN / SKU is not any indication of what sample or revision a cpu is. Specially since the specification name in CPU-Z (CPUID) doesn't show revision number.

Even if AMD 100-000000033-1 is the real OPN of the RYZEN 3950x, it defines nothing more than using the label "Ryzen 3950x". In other words the name "Ryzen 3950x" and "AMD 100-000000033-1" are interchangable and mean the same thing and are identical in what information they give on whether a chip is ES or not.. The information either of those names, OPN, SKU, codenames, or what ever YOU want to call it gives us about being ES is absolutely NOTHING as they have nothing to do with revision number.

Even with a retail CPU, CPUID does not show revision number in it's specificarion window, not for my Intel Xeon, my Ryzen 2700x, or my son's i7 2600. To get the revision number, I have to look elsewhere. The same thing hold true with this cpu.

So, my "theory" is and was trying to demonstrate to you, is your argument, which is all based off that specification number means nothing, no matter what it says because AMD can call it what ever it wants, as it has nothing to do with the revision of a chip, be it ES or not. And was simply stating at the beginning that we didn't know if this was a ES chip or not. We now know, or at least have information that support that it is a ES, which that information out weighs everything you have tried to argue about the ID number of AMD 100-000000033-1, specially since you are now saying it is the OPN number, which is different than a revision ID number. Proving yourself wrong.
 
Last edited:
Dude with the chip says its an ES.
So somehow you know better than a person with access to AMD ES chips and AMD's own motherboards.
I seriously don't see how anyone could buy your spin.
Also, you decide that intel.com is wrong about their own definition of what they define as ES and QS?

Still rekt.

Dude can claim anything. He can claim he got 9Ghz or a score of a million of points. What matters is the verification image he submitted. And the verification image he submitted proves he didn't use an ES, neither a QS, but a retail chip.

Technical terminology isn't open to what the marketing department of a company claims in a marketing webpage. Moreover the Intel website means nothing to the codenames that AMD uses to identify ES, QS, PC, and retails chips. I have given examples of codenames for ES, QS, and final chips for Zen and Zen2, but you guy continue ignoring the evidence and pretending that ES and QS are the same for AMD. LOL.

Good lord.
By any measure, its early silicon.

It is final silicon, because he is using a retail chip. Moreover, the CPU-Z screenshot reveals B-grade silicon; this is final silicon. Engineering samples use A-grade silicon.

No, my comment that started this discussion was we don't know if this chip was an ES chip or not. It has already been proven that it is (yet you are still in denial) but that is besides the point. You are arguing about the number, saying that the number AMD 100-0000000033-1. Shows it is not an ES. Which you have already, thru your own statements proven yourself wrong. My point, not a theory, is the AMD 100-000000033-1 doesn't mean jack whether it is a ES chip or not, and IF your leak is correct verifies that because OPN / SKU is not any indication of what sample or revision a cpu is. Specially since the specification name in CPU-Z (CPUID) doesn't show revision number.

I already explained my point, gave data, and answered your criticism. No need to do it all again.
 
Last edited:
So Intel never binned a processor for better clocks over a similar part?

8086k...<cough, cough> 8700k

I posted a link to R9 "special binning" (AMD's own words), and you interpret that I am negating that the 8086K was a higher binned 8700K

megafacepalm.jpg
 
Last edited:
Or we could interpret his post as claiming that binning is common in the chip industry, so your news is no news.

EDIT: rekt
 
Or we could interpret his post as claiming that binning is an common in the chip industry, so your news is no news.

EDIT: rekt

It wasn't a news about binning, but about AMD using "special binning" for the R9-3950X.
 
Keep digging. Love how people think they know so much about an unreleased product on a new process that may or may not follow existing naming conventions.
Could be the case that it's using standard bin retail sillicon hence the reporting, but still technically an ES because it's not final bin of sillicon for that SKU.
But no, AMD sucks, everyone is wrong and Intel is faultless, every time.
 
Riveting discussion, guys.

Back on page 1 before this turned into a dumpster fire there was concern about the OC headroom since it takes LN2 to get from 4.7 to 5.25, but has anyone addressed that this might be due to TDP? The article seems pretty light on actual useful information, and I recall the [H] review of Precision Boost indicating that TDP is something to consider since it will limit how far PB will take the chip. As such, I'm looking at either the 3600X or 3800X simply due to high TDP with fewer cores.

Obviously PB isn't in play at 5.25 under LN2, but my questions would be:

(1) Is TDP a limiting factor or am I a moron? (I'm not seeing TDP being mentioned)

(2) How can we overcome any TDP limitations present on Zen/2 when overclocking manually?

(3) Which X570/X470 boards have/will have the features required for (2)?

(4) Is it possible to use PB to overclock higher by reducing the impact of TDP via (2) on boards in (3)?
 
Last edited:
Riveting discussion, guys.

Back on page 1 before this turned into a dumpster fire there was concern about the OC headroom since it takes LN2 to get from 4.7 to 5.25, but has anyone addressed that this might be due to TDP? The article seems pretty light on actual useful information, and I recall the [H] review of Precision Boost indicating that TDP is something to consider since it will limit how far PB will take the chip. As such, I'm looking at either the 3600X or 3800X simply due to high TDP with fewer cores.

Obviously PB isn't in play at 5.25 under LN2, but my questions would be:

(1) Is TDP a limiting factor or am I a moron?

(2) How can we overcome any TDP limitations present on Zen/2 when overclocking manually?

(3) Which X570/X470 boards have/will have the features required for (2)?

(4) Is it possible to use PB to overclock higher by reducing the impact of TDP via (2) on boards in (3)?

My opinion is that none of us bickering have sufficient information, knowledge or understanding to answer your on-topic and valid questions.

Edit: And anyone who could have answered your questions have probably been so turned off from this thread that you'll never get an answer unless you make a post in the AMD forum. But that one might get derailed aswell.
 
Riveting discussion, guys.

Back on page 1 before this turned into a dumpster fire there was concern about the OC headroom since it takes LN2 to get from 4.7 to 5.25, but has anyone addressed that this might be due to TDP? The article seems pretty light on actual useful information, and I recall the [H] review of Precision Boost indicating that TDP is something to consider since it will limit how far PB will take the chip. As such, I'm looking at either the 3600X or 3800X simply due to high TDP with fewer cores.

It doesn't have anything to do with TDPs. Once you overclock you are running outside of spec and the stock TDP rating is irrelevant. LN2 was needed because the voltage had to be pushed above 1.6V to achieve 5GHz clocks on that process node. Those high voltages produce excessive heat (Power ~ V³) and one has to use extreme cooling to get the system under control and running.

High TDP with fewer cores doesn't imply better OC headroom. It can mean that those SKUs are using worse silicon and need higher TDP to achieve similar clocks than SKUs on better silicon.
 
Last edited:
It doesn't have anything to do with TDPs. Once you overclock you are running outside of spec and the stock TDP rating is irrelevant. LN2 was needed because the voltage had to be pushed above 1.6V to achieve those clocks on that process node. Those higher voltages produce excessive heat and one has to use extreme cooling to get the system under control.
So as soon as I select "OC manually" in literally any X470/X570 BIOS, it's like TDP doesn't exist? Editing to make the next question more specific. If the goal is highest manual overclock, do you take the 3700X (4.4 boost, 65W TDP) because the lower TDP means the cores are better binned for efficiency, or do you take the 3800X (4.5 boost, 105W TDP) because it's guaranteed to boost higher? Editing due to your edit - sounds like you'd take the 3700X.
 
Last edited:
Status
Not open for further replies.
Back
Top