AMD Ryzen 16 Core “Whitehaven” Enthusiast CPUs Leaked – 3.6GHz Clock Speed, Boatloads of Cache & Qua

In theory I think it will be possible for it to hit 3.8 but you better be on high end water cooling to sustain that overclock.
 
  • Like
Reactions: N4CR
like this
JustReason said:
So is there something wrong in the article? 3.6 instead of 3.1 isn't condemnable.
Click to expand...

Reporting the turbo clocks instead base clocks is their tactic to get their fanbase interested.

The socket is a derivation of SP4. The 10-core and 14-core configurations are suspicious because those would imply the use of dies with odd number of cores: 5, 7.

They reported 64MB of cache. After my complain (see #33) they have now corrected the table and report 32MB. They also add now the official TDP, which was missing before the update of the article. However, the 180W that they mention is the marketing rating, not the real TDP.
 
DooKey said:
If it can overclock up to 3.8 I'll buy one.
Click to expand...

I see no reason why it couldn't. The process node supports those clocks and this CPU is made of golden-class 1700 dies. However be prepared to use a >300W cooling system.
 
juanrga said:
Reporting the turbo clocks instead base clocks is their tactic to get their fanbase interested.

The socket is a derivation of SP4. The 10-core and 14-core configurations are suspicious because those would imply the use of dies with odd number of cores: 5, 7.

They reported 64MB of cache. After my complain (see #33) they have now corrected the table and report 32MB. They also add now the official TDP, which was missing before the update of the article. However, the 180W that they mention is the marketing rating, not the real TDP.
Click to expand...
Petty. So complaining about cache amounts makes you a genius? Well I beat you to it, 64mb didn't make any sense with the 8mb per CCX and how AMD reported these new CCXs scaled. As far as the 10c and 14c, it will be the same as the 4c and 6 core we have now, some bad cores removed. And lastly you gotta get off this TDP = Wattage crap. Never has never will. Just a smoke screen to hide behind when there is nothing left to complain about.
 
JustReason said:
And lastly you gotta get off this TDP = Wattage crap. Never has never will. Just a smoke screen to hide behind when there is nothing left to complain about.
Click to expand...

Bro, TDP does equal wattage, you know, unless you're AMD. Then it means something else entirely.

Which is why you have the poster above you saying that with a 180w 'AMD TDP', you'd better be prepared to cool 300w, which makes sense with AMD's interpretation.
 
IdiotInCharge said:
Bro, TDP does equal wattage, you know, unless you're AMD. Then it means something else entirely.

Which is why you have the poster above you saying that with a 180w 'AMD TDP', you'd better be prepared to cool 300w, which makes sense with AMD's interpretation.
Click to expand...

We already seen how much they lied with Ryzen, Polaris etc.
 
IdiotInCharge said:
Bro, TDP does equal wattage, you know, unless you're AMD. Then it means something else entirely.

Which is why you have the poster above you saying that with a 180w 'AMD TDP', you'd better be prepared to cool 300w, which makes sense with AMD's interpretation.
Click to expand...
And you both have no clue what it really means. TDP in AMD terms is the listing for the cooler necessary to dissipate heat for average consumer usage.
The thermal design power (TDP), sometimes called thermal design point, is the maximum amount of heat generated by the CPU that the cooling system in a computer is required to dissipate in typical operation. Rather than specifying CPU's real power dissipation, TDP serves as the nominal value for designing CPU cooling systems.
Click to expand...
You each keep showing max usage as if it is typical usage when it is far from it. My 8350 can pull up to 350W but typical is between 140W and 250W, surfing to gaming. Again this TDP crap spewing is to obfuscate the true fear of the Intel crowd that their years of bullying the AMD products and those interested in them has come to an end and in all likely hood the numbers of AMD users is actually going to grow and there is no real negative going with an AMD setup.
 
  • Like
Reactions: Zuul
like this
JustReason said:
And you both have no clue what it really means. TDP in AMD terms is the listing for the cooler necessary to dissipate heat for average consumer usage.

You each keep showing max usage as if it is typical usage when it is far from it. My 8350 can pull up to 350W but typical is between 140W and 250W, surfing to gaming. Again this TDP crap spewing is to obfuscate the true fear of the Intel crowd that their years of bullying the AMD products and those interested in them has come to an end and in all likely hood the numbers of AMD users is actually going to grow and there is no real negative going with an AMD setup.
Click to expand...

You just described AMDs ACP before they renamed it to TDP because they kept falling behind on the perf/watt curve.
 
IdiotInCharge said:
Cut the partisan bullshit.
Click to expand...
Little late for that now is it?
IdiotInCharge said:
Bro, TDP does equal wattage, you know, unless you're AMD. Then it means something else entirely.

Which is why you have the poster above you saying that with a 180w 'AMD TDP', you'd better be prepared to cool 300w, which makes sense with AMD's interpretation.
Click to expand...

You know that what you posted here and above shows that several of you really don't care for anything then derailing threads. Pick your facts pretend there wrong argue about nothing else as it would be the pinnacle of the topic and bury it with trivial mumblings.
I wish some of you could just keep it on the Intel side of things because that is all you want. Why bother coming in here in the first place but to spread such an amount of garbage.

And some of you complain that wccftech is bad ..
 
Actually, I came here with interest in a 3.6GHz 16-core CPU- only to find out that that's the 'burst' speed, because the thread author copied and pasted clickbait. 3.6GHz for that many cores would be exceptional, especially given how close that is to the speed of the top-end eight-core parts, and something that Intel hasn't yet shipped (a plus-core part running close to the speeds of their consumer parts).

If you post bullshit, expect it to be called, and if you defend bullshit, expect to be made light of. The OP couldn't even relay the rumor correctly.
 
Intel's comeuppance for all their years of sandbagging the HEDT products is coming and it comes right soon...:D
 
IdiotInCharge said:
Actually, I came here with interest in a 3.6GHz 16-core CPU- only to find out that that's the 'burst' speed, because the thread author copied and pasted clickbait. 3.6GHz for that many cores would be exceptional, especially given how close that is to the speed of the top-end eight-core parts, and something that Intel hasn't yet shipped (a plus-core part running close to the speeds of their consumer parts).
Click to expand...

You of all people should know if it does 3.6 boost, it can likely easily be OCd there.
Can't really think of anything that can't OC to its' 'boost' rating without much stress or enthusiast involvement (as long as the cooling solution is good).
 
N4CR said:
You of all people should know if it does 3.6 boost, it can likely easily be OCd there.
Can't really think of anything that can't OC to its' 'boost' rating without much stress or enthusiast involvement (as long as the cooling solution is good).
Click to expand...

After the struggles of the R7 and trying to hit >4.0GHz, I don't rate anything as 'very likely' when we're talking about doubling the package power draw on this architecture. Remember that 'burst' does not ever mean all cores in this case, and the possibility that average copies might not even be stable at 3.5GHz across all cores, if that, is certainly in the cards.

Or, AMD could have certainly improved their process by then, and maybe we'll see 4.5GHz on all cores, somehow?

But in light of the current struggles, I won't give a pass to an AMD zealot misstating rumors, or others defending such silliness with religious fervor. If we're going to discuss this stuff, let's at least be honest.
 
IdiotInCharge said:
After the struggles of the R7 and trying to hit >4.0GHz, I don't rate anything as 'very likely' when we're talking about doubling the package power draw on this architecture. Remember that 'burst' does not ever mean all cores in this case, and the possibility that average copies might not even be stable at 3.5GHz across all cores, if that, is certainly in the cards.

Or, AMD could have certainly improved their process by then, and maybe we'll see 4.5GHz on all cores, somehow?

But in light of the current struggles, I won't give a pass to an AMD zealot misstating rumors, or others defending such silliness with religious fervor. If we're going to discuss this stuff, let's at least be honest.
Click to expand...

Valid point. If I were to go out and buy a Ryzen today, I'd expect no less than 3.8 going balls out. I've had GPUs that wouldn't OC a few MHz, so OC disappointment is well under the bridge. People expecting 4.0 are buying FUD and BS as not many do without heaps of volts or a lucky chip. I don't get so stuck up on GHz numbers after P4 days either.
Same thing happened to Polaris. Sure next to a 1060 with stock volts it's not so flash (wide voltage bin), but undervolt them like for like and suddenly they are competitive on power fronts.... no one talks about that though.
Part of it is likely AMD, other part is people hyping shit up and the other manufacturing tolerances.

That said, 3.6GHz I think is well within possibilities with a solid cooler. Big air might but probably water.
8Rch6JF.png
 
It does seem that AMD's inability to enforce consistency in their process, or possibly the requirement to sell underperforming examples with the regular stock, has resulted in retail examples that have been set to run hot for no apparent reason. This does afford the mentioned opportunity to undervolt them afterward or possibly overclock at stock voltage more than would be expected for those willing to experiment, and in retrospect does paint AMD's engineers in a better light (and that follows the theory that it's been their management and marketing that has done more to limit their potential than their engineering on average over the years).
 
  • Like
Reactions: N4CR
like this
IdiotInCharge said:
...and in retrospect does paint AMD's engineers in a better light (and that follows the theory that it's been their management and marketing that has done more to limit their potential than their engineering on average over the years).
Click to expand...

That's been my thought for a long time. Their management appears to have improved somewhat of late, though. Their marketing department is still filled with morons, though.
 
IdiotInCharge said:
Actually, I came here with interest in a 3.6GHz 16-core CPU- only to find out that that's the 'burst' speed, because the thread author copied and pasted clickbait. 3.6GHz for that many cores would be exceptional, especially given how close that is to the speed of the top-end eight-core parts, and something that Intel hasn't yet shipped (a plus-core part running close to the speeds of their consumer parts).
Click to expand...

it's too early in the 14 nm process to expect much more than 4Ghz on the process. There will be process revisions, or mask revisions, but even Intel is up against 5Ghz. The process will benefit from better tooling, as in EUVL. Integrating that into process will produce smoother profiles and less current leak, the cause of heat. This is why monolithis dies are not efficient enough.When you pack complicated assemblies so closely together leakage and heat generation is exponential with clock speed. This is why multi chip packages are better. Better back end yield, and sacrificing little latency through interposers and package interconnects.
 
sorry but some of you arguing about the mention of 3.6Ghz in the title is asinine and petty.

How about speak of the articles contents.

Yes it runs at 3.1ghz with 3.6ghz turbo although not sure on how many cores and differing clock steppings ie 3.4ghz on 4 cores.

And as already mentioned the L3 cache is actually 32mb instead of 64mb as each CCX is 8mb, as stated by AMD and why the have so many core counts coming (scalable).

44 PCI-E lanes are rather nice and quad channel memory.
 
If it can sustain 3.6 GHz boost across all cores I'd say that's pretty damn good. My two 12c/24t workstations are running 2.9 GHz and 2.7 GHz (old/new respectively). With 4 extra cores and 8 threads that would be a serious upgrade for my workload.
 
IdiotInCharge said:
Actually, I came here with interest in a 3.6GHz 16-core CPU- only to find out that that's the 'burst' speed, because the thread author copied and pasted clickbait. 3.6GHz for that many cores would be exceptional, especially given how close that is to the speed of the top-end eight-core parts, and something that Intel hasn't yet shipped (a plus-core part running close to the speeds of their consumer parts).

If you post bullshit, expect it to be called, and if you defend bullshit, expect to be made light of. The OP couldn't even relay the rumor correctly.
Click to expand...

OMG yes i can see how you feel so outraged, having to suffer the pain of clicking on the link only to find that your assumptions were incorrect must be dealt with immediately. Who gives a fuck about such an insignificant detail. Besides, how quickly we forget the clock speeds and clock speed speculation of Ryzen engineering samples.

2D3101A8UGAF4_36/31_N 3600 3100 16 32 AMD Whitehaven
1D3101A8UGAF3_36/31_N 3600 3100 16 32 AMD Whitehaven
2D2701A9UC9F4_32/27_N 3200 2700 12 24 Alienware R3

This 16 core 32 threaded monster is exceptional at 3.1Ghz already. The best intel can manage is 3.3Ghz with 10 core so far according to the same leaks at similar TDPs.
 
WOW! http://wccftech.com/amd-ryzen-9-lineup-threadripper/

Looks like AMD is going to bring it down hard when it comes to 16 core monsters. Looks pretty legit. 3.5GHZ base for 16 core top of the line part! That is pretty damn impressive to say the least.

No wonder intel is in Hurry up mode. Bring on the competition.
 
NKD said:
WOW! http://wccftech.com/amd-ryzen-9-lineup-threadripper/

Looks like AMD is going to bring it down hard when it comes to 16 core monsters. Looks pretty legit. 3.5GHZ base for 16 core top of the line part! That is pretty damn impressive to say the least.

No wonder intel is in Hurry up mode. Bring on the competition.
Click to expand...

I can see it now; 3.5Ghz base, 3.6Ghz boost, lol. You must be dreaming if you believe a 16 core Ryzen chip will have a 3.5Ghz base with a 4.1Ghz boost when the 8 and 6 core chips can barely hit 4.0Ghz.

And what is up with with that Ryzen 9 name? AMD planning on copying Intel in naming?
 
Source being a random reddit with nothing but text. LOL.

Clickbaits win again.
 
ZeroBarrier said:
I can see it now; 3.5Ghz base, 3.6Ghz boost, lol. You must be dreaming if you believe a 16 core Ryzen chip will have a 3.5Ghz base with a 4.1Ghz boost when the 8 and 6 core chips can barely hit 4.0Ghz.

And what is up with with that Ryzen 9 name? AMD planning on copying Intel in naming?
Click to expand...

you must have read that wrong. 3.5 base and 3.9 boost for 16 core part.
 
ZeroBarrier said:
I can see it now; 3.5Ghz base, 3.6Ghz boost, lol. You must be dreaming if you believe a 16 core Ryzen chip will have a 3.5Ghz base with a 4.1Ghz boost when the 8 and 6 core chips can barely hit 4.0Ghz.

And what is up with with that Ryzen 9 name? AMD planning on copying Intel in naming?
Click to expand...

remember that boost clock isn't all the cores like Intel's so yeah a 4Ghz boost could happen but i'd say more likely to be around 3.8Ghz if they in fact go with a 3.5Ghz base clock but even that seems like a stretch, would probably have to be some insane cherry picked dies.

as far as the naming goes i'm not a huge fan of it because i always liked AMD's original naming schemes but honestly they don't work with modern processors. i think the way it is now does help so you know what you're comparing AMD processors to intel processors and vice versa.. so what ever simplifies that process for the consumer works for me. lets just say it works a hell of a lot better than when they tried to use the R9 R7 naming scheme with their graphic card lineup.
 
sirmonkey1985 said:
remember that boost clock isn't all the cores like Intel's so yeah a 4Ghz boost could happen but i'd say more likely to be around 3.8Ghz if they in fact go with a 3.5Ghz base clock but even that seems like a stretch, would probably have to be some insane cherry picked dies.

as far as the naming goes i'm not a huge fan of it because i always liked AMD's original naming schemes but honestly they don't work with modern processors. i think the way it is now does help so you know what you're comparing AMD processors to intel processors and vice versa.. so what ever simplifies that process for the consumer works for me. lets just say it works a hell of a lot better than when they tried to use the R9 R7 naming scheme with their graphic card lineup.
Click to expand...

Intel doesnt boost all cores either.
 
Shintai said:
Source being a random reddit with nothing but text. LOL.

Clickbaits win again.
Click to expand...

Until real announcements and reviews come, we're all just spitballing here, Shintai. I don't think anyone is saying this is gospel, just interesting stuff to bullsh*t about.
 
JustReason said:
Yes it runs at 3.1ghz with 3.6ghz turbo although not sure on how many cores and differing clock steppings ie 3.4ghz on 4 cores.
Click to expand...

3.6GHz is obviously the single-core turbo.

Peppercorn said:
OMG yes i can see how you feel so outraged, having to suffer the pain of clicking on the link only to find that your assumptions were incorrect must be dealt with immediately. Who gives a fuck about such an insignificant detail.
Click to expand...

The clickbait site that reports the single-core turbo instead the base clocks "gives a fuck". As does the people that got hungry when this fact was mentioned and corrected in this thread.

Peppercorn said:
This 16 core 32 threaded monster is exceptional at 3.1Ghz already. The best intel can manage is 3.3Ghz with 10 core so far according to the same leaks at similar TDPs.
Click to expand...

>200W is not similar to 160W n my book, and comparing base clocks, whereas ignoring all-core and single turbo cores can be misleading. E.g. Intel gets up to 4.5GHz whereas AMD only can hit 3.6GHz.
 
16 threads + another 16 threads + drone video post production equals OMG faster than an overclocked 8 core Zen 7.

I am going to move my 1700X to a mITX board once they actually hit the market. I will get a top of the line 16 core AMD, unless Intel has an equally priced chip at the time, and I will not have to wait nearly as long as I do for post products to be completed.

I am a little excited, reserved, but excited.
 
NKD said:
WOW! http://wccftech.com/amd-ryzen-9-lineup-threadripper/

Looks like AMD is going to bring it down hard when it comes to 16 core monsters. Looks pretty legit. 3.5GHZ base for 16 core top of the line part! That is pretty damn impressive to say the least.

No wonder intel is in Hurry up mode. Bring on the competition.
Click to expand...

It looks 'legit' AMD only can get 3.4--3.6GHz base on '95W' marketing envelope (real TDP 128W), but then joining two of those dies gives 3.5GHz on 155W, surely because the MCM package consumes negative amount of power. LOL

In real word a pair of 3.6GHz 8-core dies dissipate

128W + 128W = 256W

plus ~20W from packaging gives 276W. To fit into a 200W real envelope clocks of this 16-core CPU have to be reduced to

sqrt(200/276) * 3.6GHz = 3.065GHz

Is not terribly curious that the only known sample of Threadripper is a 3.1GHz 16-core with a real 200W TDP (marketing TDP of 180W)?

But hey let us trust anonymous forum leaks regurgitated by clickbait sites as WCCFTECH and fuel the hype train once again. All of us know how good it worked for RyZen: it has IPC 20% above Kabylake and can hit 5GHz on air as leaks promised...
 
Last edited:
Man some of the logic in this thread is totally without basis, basically pull it out your ass and then complain that others don't sniff it up.

The base or Zen module is 4 cores, the ability of those 4 cores by design or process will pretty much be the same and limited the same as in APU 4 core + APU, RyZen 8 core or RyZen 9 16 core. Power ratings will limit the clock speed rating in general with derivations on golden chips able to hit higher clock speeds at reduced voltages etc.

Meaning, hitting 4ghz should be similar to hitting 4ghz with a Ryzen 8 core except now you have more cores that will need to all hit that speed. As the process matures that will become more and more likely. So I fully expect some Ryzen 9's to hit 4ghz, albeit with the needed power and cooling but the cooling will have more surface area with 4 modules so will be relatively the same, the heat sink will just need to transfer that and dissipate it.

One thing many have not noted is that the scaling of RyZen is outstanding, adding more cores is rather linear in performance - I dare say better than Intel designs. So until we see both Intel Skylake with RyZen 9 performance much will be left up in the air.
 
At this point I just wish AMD would be honest about TDPs. With their consumer parts (or currently released ones), variations didn't matter too much because the delta was still well within the capacity of the average consumer cooling solution.

But we're talking >250W for 16 cores approaching 4.0GHz- that's decidedly beyond your average tower HSF or closed-loop cooler (120mm single/dual-fan variety).

Claiming that they'll hit 3.8GHz on 16 cores at 155W is just downright laughable.
 
IdiotInCharge said:
At this point I just wish AMD would be honest about TDPs. With their consumer parts (or currently released ones), variations didn't matter too much because the delta was still well within the capacity of the average consumer cooling solution.

But we're talking >250W for 16 cores approaching 4.0GHz- that's decidedly beyond your average tower HSF or closed-loop cooler (120mm single/dual-fan variety).

Claiming that they'll hit 3.8GHz on 16 cores at 155W is just downright laughable.
Click to expand...

Their process is very efficient to 3.5 to 3.6 after that you pay a higher and higher voltage penalty. What your reading is boost clocks and thinking its all cores and it wont be, likely 4 cores and down or even just two cores and xfr for single core operation. 3.5 and 3.6 are the highest all core clocks which seems right to me as for TDP Intel and AMD do it differently and you just have to live with that.
 
You must log in or register to reply here.
Back
Top