• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD Zen 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***

CuriousTomCat said:
Some of the issue with games is optimisation. The original Ryzen only came out two years ago this February. Considering most triple A games take 2 or 3 years to make, every new game released today was started prior to Ryzen's debut so any changes for AMD had to be added in as an afterthought.
Click to expand...

It's not only a case of games optimisation, it very heavily revolves around Windows Schedular as well. Until Microsoft actually give AMD a level playing field on that count, whatever little games devs do will mean next to nothing.
 
Potatowithearsontheside said:
The 3300G will be Zen+ and not Zen2. AMD already announced that on the eve of CES.
Click to expand...

No, that's the laptop parts with H and U suffixes. Nothing with a G suffix has been announced.

Truth be told I don't know if 3000 series APUs will be based on Zen+ or Zen 2. The package design does allow itself for a graphics chiplet to be paired with a CPU chiplet, which is what AdoredTV was theorising about with Navi anticipated to be a small die. But look at the announcement date: Q3 2019. That's when Navi is anticipated to show up. And if the 3000 APUs use Navi graphics, not Vega graphics, it strikes me as a little odd that they'd choose to use Zen+ for the CPU side.

But then the laptop parts are using Zen+ with Vega - purely to get the ball rolling?
 
CuriousTomCat said:
But people will be saying how rubish AMD are, even if they are only 1% slower. I'm forever defending AMD in the comments section of YouTube videos where there's a comparision between the 9900K and 2700X etc. People are obsessed with small differences in FPS. This is about the IPC crown and it's so important for the image of AMD. Naive Intel fanboys still have a common misconception that AMD make subpar products for people who cant afford anything decent and that needs to change.
Click to expand...

+++++absoutely. If ever there was a time to shove it up the fanboys.................now is it.
 
Legion said:
"AMD stated that, at this time, there will be no version of the current Matisse chiplet layout where one of those chiplets will be graphics."
Click to expand...

That is as insightful as "7nm is coming to gamers"; it could mean anything :p

So if we take it as face value and we're not getting the Matisse layout for APUs, then the 3000 series APUs will be a monolithic die. But I wonder if it'll be Zen+ with Vega graphics or Zen 2 with Navi graphics? Navi is a 7nm design, I have no idea if it could be produced on 12nm.

I supposed July-September timeframe gives AMD time to combine Zen 2 cores, I/O and Navi graphics into a 7nm monolith, but realistically it's just going to be Zen+ and Vega on 12nm like the laptop parts, isn't it.
 
LePhuronn said:
That is as insightful as "7nm is coming to gamers"; it could mean anything :p

So if we take it as face value and we're not getting the Matisse layout for APUs, then the 3000 series APUs will be a monolithic die. But I wonder if it'll be Zen+ with Vega graphics or Zen 2 with Navi graphics? Navi is a 7nm design, I have no idea if it could be produced on 12nm.

I supposed July-September timeframe gives AMD time to combine Zen 2 cores, I/O and Navi graphics into a 7nm monolith, but realistically it's just going to be Zen+ and Vega on 12nm like the laptop parts, isn't it.
Click to expand...
Actually it was Lisa Su who confirmed Matisse won't have an APU design based off the 3xxx series. For an APU they will be using a different layout which they would designate a different CPU entirely. So semantics aside the Matisse layout is not compatible with a GPU instead of a CPU module.
 
TrixX said:
Actually it was Lisa Su who confirmed Matisse won't have an APU design based off the 3xxx series. For an APU they will be using a different layout which they would designate a different CPU entirely. So semantics aside the Matisse layout is not compatible with a GPU instead of a CPU module.
Click to expand...
We already know the APUs won't be Matisse, they will be Renoir. We don't know yet what the package will be like.
 
NinjaCool said:
Possible 12-Core 24-Thread Third-Gen Ryzen ES CPU spotted at UserBenchmark!
https://twitter.com/TUM_APISAK/status/1088264161191022593?s=09
https://www.tomshardware.co.uk/amd-matisse-third-gen-ryzen-benchmark,news-59832.html
Click to expand...
Interesting. Using 2666 MHz DDR4, and with a 3.7 GHz maximum boost as suggested by the codename, it gets a single core score of 116 in this benchmark. Linearly scaled up to 5.0 GHz, that would be around 157. Realistically it'll be a bit under this, but then the RAM is slow so that balances out somewhat.

For reference, an i9-9900K, which boosts to 5.0 GHz on a single core, seems to get around 155 in this test. It's also worth noting that the disparity between the integer and floating point scores is rather large compared to Intel's chips (with the integer score being weaker on the AMD part).

For another comparison, the R7 2700X boosting to 4.2-4.3 GHz gets around 126, which'd be about 148 at 5.0 GHz. Ultimately this data point suggests the IPC improvement is not as strong in this kind of workload as multithreaded Cinebench.
 
Last edited:
DragonQ said:
Interesting. Using 2666 MHz DDR4, and with a 3.7 GHz maximum boost as suggested by the codename, it gets a single core score of 116 in this benchmark. Linearly scaled up to 5.0 GHz, that would be around 157. Realistically it'll be a bit under this, but then the RAM is slow so that balances out somewhat.

For reference, an i9-9900K, which boosts to 5.0 GHz on a single core, seems to get around 155 in this test. It's also worth noting that the disparity between the integer and floating point scores is rather large compared to Intel's chips (with the integer score being weaker on the AMD part).

For another comparison, the R7 2700X boosting to 4.2-4.3 GHz gets around 126, which'd be about 148 at 5.0 GHz. Ultimately this data point suggests the IPC improvement is not as strong in this kind of workload as multithreaded Cinebench.
Click to expand...

Its hard to read anything from this stuff, when it is out you'll know, my gen 1 scores around 126 @ 4.2, we'll only know the score once in some properly optimized systems. ​
 
TrixX said:
Actually it was Lisa Su who confirmed Matisse won't have an APU design based off the 3xxx series. For an APU they will be using a different layout which they would designate a different CPU entirely. So semantics aside the Matisse layout is not compatible with a GPU instead of a CPU module.
Click to expand...

I doubt Lisa confirmed anything, a CEO does not confirm or deny to leave the window open in case situations change and opportunities arise, e.g. "7nm is coming to gamers" but was never confirmed she explicitly and directly meant Navi, and thus an opportunity for Radeon 7 presented itself.

*** Let's not get personal thank you **
 
Last edited by a moderator:
In a way, as excruciating as the wait is for zen2 to be released, i am glad that they are taking their time and properly testing everything to make sure its the best it can be. its all well and good throwing in cores and clock speeds, but if the optimisation is poor it'll be poor overall.

Judging by how the test sample worked at ces, my hopes remain high, and after reading the article that was posted a few posts above, anyone who thinks zen2 will be **** (basically just the intel fanboys) are grossly mistaken
 
It's funny to me that Intel have had 2 years of Ryzen, plus the hype for ~6 months before that to try and counter it and still nothing is being shown other than more cores more cores more cores.
 
You must log in or register to reply here.
Back
Top Bottom