Ryzen 3000 listed online early on russian site.

[Wavetrex]

https://videocardz.com/newz/amd-ryzen-9-3800x-matisse-listed-with-16-cores-and-125w-tdp

Holy mother of 16 cores on AM4 !!! @ 3.9->4.7 Ghz

Seems that Jim might be (almost) correct with his leak. SMOKING !!!

 

[Vayra86]

4.7 Turbo

is it real?

 

[Wavetrex]

Threadripper 2950X on 12nm can reach 4.4 Ghz by default... I don't see why 7nm can't do 4.7 or higher.

I personally don't doubt the frequency that could be reached with Zen2, but the core counts... well, still unsure, seems TOO GOOD to be true.
But then, the market has stagnated for too long, it's really the time to see something like this !

 

[IceScreamer]

Oh boy, now to see what trickles down the line to the more budget offerings.

 

[Vayra86]

Threadripper 2950X on 12nm can reach 4.4 Ghz by default... I don't see why 7nm can't do 4.7 or higher.

I personally don't doubt the frequency that could be reached with Zen2, but the core counts... well, still unsure, seems TOO GOOD to be true.
But then, the market has stagnated for too long, it's really the time to see something like this !

Agreed, its just that this is the magical edge Ryzen needed for gaming. The missing 10%. Was waiting for this since it launched.

 

[Hellfire]

As always, taking it with a pinch of salt, but this looks VERY promising. Especially since I want a 3850X

 

[silentbogo]

If that's true, then leaks about 8 cores per CCX were also accurate. We'll probably see some bad-ass 220-250W Threadreaper chips w/ 32C/64T!

 

[Vya Domus]

4.7 Turbo

For a single core turbo it's pretty underwhelming actually , people managed 4.5 on 2700Xs.

 

[Wavetrex]

If that's true, then leaks about 8 cores per CCX were also accurate.

That's not leaks, that's actual real chips which AMD demoed in the Rome chip.
64 cores, 8 chiplets x 8 cores.

The question is... will they use the same for Ryzen 3000, in combination with an I/O hub ? Or a more integrated design in a single larger die.
My gut tells me it will be a small I/O hub with 2 chiplets, that makes the most sense as they can fab all the 8-core chiplets with the same process and then just package differently for various markets.

 

[IceScreamer]

Apparently there are even more listings:
R3 3300: http://www.e-katalog.ru/AMD-3300.htm
R3 3300X: http://www.e-katalog.ru/AMD-3300X.htm
R5 3600: http://www.e-katalog.ru/AMD-3600.htm
R5 3600X: http://www.e-katalog.ru/AMD-3600X.htm
R7 3700: http://www.e-katalog.ru/AMD-3700.htm
R7 3700X: http://www.e-katalog.ru/AMD-3700X.htm
R9 3800X: http://www.e-katalog.ru/AMD-3800X.htm

 

[adulaamin]

For a single core turbo it's pretty underwhelming actually , people managed 4.5 on 2700Xs.

All core turbo would be ...

 

[phill]

I will leave my amazement until I see this in reviews... Then it'll be a case of upgrade!!

 

[silentbogo]

Wow. 6-core Ryzen 3!
I think I'll delay my next upgrade for a few months...

 

[Wavetrex]

To be honest even if top chip only does 4.5 and has 12 cores only, I'll still upgrade.

Using a 6-core Broadwell-E 6800K which costed plenty of money about 2.5 years ago, and can only do 4.2 Ghz safely (voltage wall at 4.3, crap chip).
12c @ 4.5Ghz would be an amazing upgrade for me, without the price premium of Threadrippers. (Intel is not even in the discussion anymore)

 

[xkm1948]

Hmm these looks to be 100% the same as AdoredTV’s speculations. Coincidence?

 

[Vayra86]

For a single core turbo it's pretty underwhelming actually , people managed 4.5 on 2700Xs.

This is stock, out of the box. Let's pull out the 8700K

Given the fact Zen was always more conservative with clocking than Core... this is quite a leap.

 

[ShurikN]

For a single core turbo it's pretty underwhelming actually , people managed 4.5 on 2700Xs.

A handful of people. 99.9% couldn't get more than 4.3

 

[Deleted member 158293]

Might be time to repurpose my "old" 2700x as a freenas VM server and get a 16 core 3800x onboard!

Since AdoredTV seems to be right on, maybe I could stretch it to AMD's anniversary edition.

 

[cdawall]

Still waiting to see real world for the TR4 lineup..

 

[Vya Domus]

A handful of people. 99.9% couldn't get more than 4.3

Sure but the point is that for a 7nm chip 4.7 is not a lot. And GloFo's 14nm wasn't that great.

 

[cdawall]

Sure but the point is that for a 7nm chip 4.7 is not a lot.

The process matters a lot more than die size. People on this page forget that basically every single time something is posted it seems.

 

[R0H1T]

For a single core turbo it's pretty underwhelming actually , people managed 4.5 on 2700Xs.

Final clocks can be adjusted upwards, or downwards, even till the last moment, before the chip's launched.

 

[Vya Domus]

Final clocks can be adjusted upwards, or downwards, even till the last moment, before the chip's launched.

Not really, it takes a lot of time to properly bin these chips. AMD made a step forward in making this step easier by going for this chiplet architecture but it's still not something they can change on whim.

 

[phanbuey]

16 cores is amazing... but i think that 12 core 5ghz turbo is the best one of the bunch.

At some point i have to ask... do i really need moar cores?

 

[R0H1T]

Not really, it takes a lot of time to properly bin these chips. AMD made a step forward in making this step easier by going for this chiplet architecture but it's still not something they can change on whim.

Is this the best bin they have? I'm fairly certain they'll be reserved for EPYC & TR, so if TR3 isn't gonna be launched anytime soon they'll probably have enough incentive to clock them higher & sell the best bins. There's also the thing about early chips, on a new node, clocking lower ~ so it's possible that 4.7 isn't the real limit for single core (turbo) on Zen2 just yet.