AMD Ryzen 5 3000 Series Lineup Detailed

[NdMk2o1o]

I said the specs are there, and "supposed" doesn't change that.

Well yes, it makes a big difference as to a statement being fact or not, which was my point that wasn't aimed at you in the first place anyway though you felt the need to make it yours

 

[cucker tarlson]

They are matching Intel 5.0 GHz with 4.4GHz.

Pics or didn't happen.

 

[londiste]

Pics or didn't happen.

Cinebench R20:

Single-thread boost clocks:
9700K 4.9 GHz
3700X 4.4 GHz +1%
9900K 5.0 GHz +2%
3800X 4.5 GHz +3%

 

[Darmok N Jalad]

Pics or didn't happen.

Edit: Ninja’d by londsite.

 

[SL2]

The 9900K runs at 4,7 GHz turbo all core.

Well yes, it makes a big difference as to a statement being fact or not, which was my point that wasn't aimed at you in the first place anyway though you felt the need to make it yours

You replied to Shatun_Bear who in turn replied to TheLostSwede, so you're just as guilty for making it yours as I am.

 

[stimpy88]

Interesting to see all the bile and spite thrown at AdoredTV (pretty much the only tech journalist left that's not living cosily in the pocket of a tech company) here. It's very interesting to note that he has ALWAYS said to take the 5GHz info with a bucket of salt, and doubted it highly, but that does not stop the Intel police from losing their minds over what AMD has just achieved. Infact I would say it's even better that a lower clocked AMD part can beat a higher clocked Intel part, adds a little more venom to the bite, does it not? lol

I'm currently running an i7 4770k, after being a loyal Intel customer for more than 12 years. Not any longer, especially as I've lost more than 20% (nearly a whole CPU core) of the IPC my CPU can give, due to Intel's security leaks. Intel offers nothing that is competitive for my wallet.

I, for one, are looking forward to jumping ship.

 

[cucker tarlson]

Edit: Ninja’d by londsite.

Ah that's what you meant,I get it now.I thought it can match it core for core across the board,gaming included.That's brute forcing higher cinebench scores with more cores.
What is 3800x core config? 8c or 10c?

 

[londiste]

The 9900K runs at 4,7 GHz turbo all core.

Depends on testing conditions. It should definitely not run Cinebench R20 (with AVX2) at this high frequency.

Ah that's what you meant,I get it now.I thought it can match it core for core across the board,gaming included.That's brute forcing higher cinebench scores with more cores.
What is 3800x core config? 8c or 10c?

8c/16t.
The bunch of graphs on the left are single-core results (supposedly with maxed boost clocks).

 

[SL2]

Depends on testing conditions. It should definitely not run Cinebench R20 (with AVX2) at this high frequency.

You're right. I was trying to point out that the 9900K couldn't run at 5 GHz all core, but like you say, it's probably lower than that.

 

[cucker tarlson]

Depends on testing conditions. It should definitely not run Cinebench R20 (with AVX2) at this high frequency.
8c/16t.
The bunch of graphs on the left are single-core results (supposedly with maxed boost clocks).

Then it's really good.
Although I just checked 2700x vs 9900k and tpu shows 9900k only 5 percent faster,that'd make 3800x 8 percent faster than 2700x

 

[Zyll Goliat]

Do you guys think that AMD potentially could pull out from the hat also a 10 core Ryzen 3000 cpu?

 

[stimpy88]

Do you guys think that AMD potentially could pull out from the hat also a 10 core Ryzen 3000 cpu?

Struggling to see the point... But I would think it's technically possible...

 

[R0H1T]

No, 10/20/30 cores aren't happening with the CCX based designs.

 

[HD64G]

Do you guys think that AMD potentially could pull out from the hat also a 10 core Ryzen 3000 cpu?

Ofc not since they will use 6 or 8 cores per chiplet as already shown by their line up.

 

[Metroid]

...with 15% IPC lift and at 65W. They are matching Intel 5.0 GHz with 4.4GHz. And the 3900X boosts to 4.6GHz at 105W.

Also, EPYC is likely eating all the best chiplets right now anyway. Better margins there, I’m sure.

We still need to see reviews that confirm that, one thing to consider, you said that amd is matching intel 5ghz with 4.4ghz, that is 600mhz difference, that is 15% and yet amd released a chart stating 1 to 4% faster single thread performance. If all you said turns out to be true in july 7 then we can see a 16% better single thread performance if we compare 3700x x 9700k at this moment.

 

[TheLostSwede]

Do you guys think that AMD potentially could pull out from the hat also a 10 core Ryzen 3000 cpu?

Could yes, will, maybe not. A 16 core part will be available at some point, as that's been seeded to the board makers.

 

[B-Real]

So like me and a few other sensible posters had been saying, only to be derided, the AdoredTV hype train chart from December that everyone was whipped up into a frenzy about was total BS.

The sanity check should have been applied when this 'leaker' (did Adored make it up himself?) revealed that the 12-core had a 4.2Ghz BASE clock and the 16-core had a 4.3Ghz base clock. It turns out the purported (fanfiction) base clock of the 12-core was only 400Mhz lower than it's real BOOST clock!

- His SKUs were all wrong, core counts weren't moved down a stack
- APU info was totally wrong, prices were off massively (there is no miracle $99 6-core, there is one that is just...double that price)
- Clocks are not just slightly lower but 400Mhz off in base and boost in some cases...
- And to round it all off, his info on Navi from the same hype December video was completely wrong.

Ban Adored as a credible news source from now on please TPU, or at least, please do not make news stories on his info. Thanks



This is a monster. Totally wipes the floor with the only Intel equivalent.

If you wait a performance based on a leaker based on rumours, it's your problem.

3700X with 9700K and 3800X with 9900K gaming performance for less power consumption and $100 cheaper, that is very very nice.

 

[TheLostSwede]

And here are some 16 core details... Yes, more rumours if you'd like...

AMD Ryzen 9 16-core overclocked processor scores 4346 in Cinebench - VideoCardz.com

It seems that 16-core Ryzen 9 was already demoed at Computex. AMD Ryzen 9 16-cores: 3920X, 3950X? Since the 12-core Ryzen 9 3900X borrowed the naming from Ryzen Threadripper, we are simply not sure how the next SKU would be called. The Ryzen 9 with 16-cores would utilize both dies, as opposed to...

videocardz.com

 

[illli]

lol someone is salty by adoredtv. Calling it a 'news source' is laughable. It was rumor and speculation, it was said numerous times take it with a grain of salt. Anyone triggerd and bitter and calling 'hype train' and calling for a ban (lol) is maybe a little too emotionally invested.

 

[SL2]

And here are some 16 core details... Yes, more rumours if you'd like...

I fail to see the reasoning behind the model numbers. 3700X and 3800X are both 8C, 3900X is 12C, which doesn't leave much room for the 16C.
The two 8C is farther apart than the 12C and the 16C, in numbers. It should have been 3700X & 3750X, 3800X, and 3900X.

Yeah, I don't really care that much, the performance is more important, but still, it's kind of weird.

 

[R0H1T]

We still need to see reviews that confirm that, one thing to consider, you said that amd is matching intel 5ghz with 4.4ghz, that is 600mhz difference, that is 15% and yet amd released a chart stating 1 to 4% faster single thread performance. If all you said turns out to be true in july 7 then we can see a 16% better single thread performance if we compare 3700x x 9700k at this moment.

15% isn't just for single threaded workloads, also IPC isn't a fixed number ~ it changes with application, processors (same uarch different cache size like 8700k/9900k) & finally HT, in some edge cases memory speeds as well.

 

[Metroid]

15% isn't just for single threaded workloads, also IPC isn't a fixed number ~ it changes with application, processors (same uarch different cache size like 8700k/9900k) & finally HT, in some edge cases memory speeds as well.

Just to be clear, they said ipc 15% on tests they conducted x zen+. As I said before we will have to wait for reviews to have a broad understanding about it. Single thread performance is what I most care about, if I wanted more cores aka multi-thread i would go for server cpus.

 

[R0H1T]

Just to be clear, they said ipc 15% on tests they conducted.

Yes & most of them would be MT runs as long as the application supports it, you don't purposefully limit your application to a single core.

Single thread performance is what I most care about, if I wanted more cores i would go for server cpus.

ST or single core? There's a slight difference, I also believe AMD have taken the lead in ST tasks.

 

[Metroid]

Yes & most of them would be MT runs as long as the application supports it, you don't purposefully limit your application to a single core.
ST or single core? There's a slight difference, I also believe AMD have taken the lead in ST tasks.

There are many old apps and games that are limited to a single core, their exe was never updated to support multicores, most because the engine as a whole has a single thread limitation. So for those apps of games single thread performance is what is important, for example, simulators, I play simcity 4 a lot and that is limited to a single thread performance, billions of cores/threads wont help that.

Also just to clear things up, single core might have one or 2 threads due to hyperthreading or smt, application like the example i gave above, simcity 4 will take just one thread to work with so, hyperthreading wont help it and in many cases will make it worse cause the thread performance has resources shared between the threads, reason why i applaud 9700k, I have no idea how is done on amd, with popularity growing on amd, I hope if they dont have a way yet, a function to disable smt altogether, i really think I should have the choice to choose if I want smt or not.

 

[Shatun_Bear]

Mate you're embarrassing yourself, for a start tpu have never quoted adoredtv as fact or news so stop throwing around libellous accusations stating that they have, for another thing you don't know what amd are going to release on 7th July along with the rest of us so stop attacking other members who quite frankly seem to be in the know a lot more than you are, though you can't seem to see that. Put the phone down, turn off the pc and go for a walk outside and stop throwing rants at people on the Internet

Don't embarrass yourself, they have used him as a source, just look at the reply to your post.

Interesting to see all the bile and spite thrown at AdoredTV (pretty much the only tech journalist left that's not living cosily in the pocket of a tech company) here. It's very interesting to note that he has ALWAYS said to take the 5GHz info with a bucket of salt, and doubted it highly, but that does not stop the Intel police from losing their minds over what AMD has just achieved. Infact I would say it's even better that a lower clocked AMD part can beat a higher clocked Intel part, adds a little more venom to the bite, does it not? lol

No, NO HE DIDN'T! I can't believe he has so many defenders. He was literally hinting that we should be so excited about Ryzen 3000 because of the 5Ghz+ clockspeeds, he believed it very much, so you're spouting nonsense to defend him.