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

[LePhuronn]

Hasn't AMD already promised "outstanding gaming performance"? So anything less and people are obviously gonna be upset, though saying something like that can mean equal the competition or above, it's not so clear yet.

Such is corporate linguistics.

I like how Lisa Su carefully words things to drop hints and never makes absolutes: "7nm is coming to gamers", "you'd expect us to fill space that" (CES Ryzen demo). But "outstanding gaming performance" came from Papermaster and I think such an absolute is dangerous territory right now.

We'll see.

 

[Double07]

My CH6 dosn't and never has needed any CPU LLC. That's for the 1700 i had and my current 2700x. Like yours, my VRM's have always been cool. The hottest i have seen them was around 54c last summer when the ambient temp was in the mid 20c.

yeah its not like they (and others) have ever promised that before and fallen flat on their faces. Hopefully not this time but time will tell.

 

[Potatowithearsontheside]

I think the danger for AMD is that the 6c variant ends up being the biggest seller again, as that likely means lower ASPs. Admittedly, they may be expecting a very significant increase in sales volume that lower ASPs don't impact their profitability in any meaningful way.
It'll be interesting either way, since this round almost certainly goes to AMD.

 

[4K8KW10]

I think the danger for AMD is that the 6c variant ends up being the biggest seller again

Every product lineup sells the lowest-end members in the largest quantities. I don't see where the news is.

 

[TrixX]

Every product lineup sells the lowest-end members in the largest quantities. I don't see where the news is.

Not always. As per the Subaru WRX the performance part is sold more than the stock part, looking at the 9900K sales vs the rest of the product stack it's the high value part outperforming the mid-range to low end as the Ryzen parts have dominated that sector.

 

[4K8KW10]

Not always. As per the Subaru WRX the performance part is sold more than the stock part, looking at the 9900K sales vs the rest of the product stack it's the high value part outperforming the mid-range to low end as the Ryzen parts have dominated that sector.

I guess those figures may come only from rich markets, global sales would show otherwise.

 

[SiDeards73]

I have a hunch that the 8c 16t chips may end up the fastest, if they are the single CCX ones, I think the 14nm IO chip will add latency to AMs design and any of the singular CCX chips may end up the fastest for gaming, if this happens I also expect this chip to replace the 1600/2600x chips as the best mid range priced chip.

And let's be honest 8c/16t is still going to under utilised in games until Zen3.

My money's on this chip being the star in the portfolio.

Normally I'd buy whatever the top end chip is, but like I say I have a sneaky suspicion this time round it's going to be this chip

 

[Potatowithearsontheside]

Every product lineup sells the lowest-end members in the largest quantities. I don't see where the news is.

The 2600 certainly wasn't the lowest-end Zen+ CPU, yet it was consistently the best seller.
This time around though, 6c may end up being the low-end, which likely brings lower pricing than the current 6c mid-tier 2600.
As I said, it may be that the increase in volume ultimately overrides any pricing concerns within AMD.

 

[Journey]

Normally I'd buy whatever the top end chip is, but like I say I have a sneaky suspicion this time round it's going to be this chip

Yes, top end isn't always the best product especially if you re willing to run it out of spec. A bit like my old Celeron 300A clocked at 450MHz, which due to having a 128KB on chip L2 cache was faster in a lot of things than the Pentium II 450MHz with 512KB of half speed L2 cache that was not on-die. You could spend less than 1/4 of the cost and get the same or better performance, it was ace.

 

[kitfit1]

Yes, top end isn't always the best product especially if you re willing to run it out of spec. A bit like my old Celeron 300A clocked at 450MHz, which due to having a 128KB on chip L2 cache was faster in a lot of things than the Pentium II 450MHz with 512KB of half speed L2 cache that was not on-die. You could spend less than 1/4 of the cost and get the same or better performance, it was ace.

Of course it was...................................then. It's different now though. At that time neither AMD or Intel had any idea that we clocked and even if they thought about it, didn't care. Both of them are fully aware of it now though, which is why Intel in particular have been milking the market for years for high clocked parts.

 

[Journey]

Of course it was...................................then. It's different now though. At that time neither AMD or Intel had any idea that we clocked and even if they thought about it, didn't care. Both of them are fully aware of it now though, which is why Intel in particular have been milking the market for years for high clocked parts.

No disagreement there, it is very different now. However, I think things like CCX latency with the I/O die and such could play a big factor in what is best, in what situation, especially if you can clock them the same. Some might happily spend £650 on a hypothetical 16c/32t CPU at 4.8GHz, and it could end up being slow than the hypothetical 8c/16t CPU at 4.6GHz in certain circumstances, and then if you level the clock speed you are ahead. Interesting times ahead, for a change.

 

[CuriousTomCat]

Of course it was...................................then. It's different now though. At that time neither AMD or Intel had any idea that we clocked and even if they thought about it, didn't care. Both of them are fully aware of it now though, which is why Intel in particular have been milking the market for years for high clocked parts.

Well Intel must have known about overclocking because the 300A had a locked multipler to stop people from changing it. AMD chips were fully unlocked of course. You're right though, Intel didn't milk the market like they do now.

 

[CuriousTomCat]

No disagreement there, it is very different now. However, I think things like CCX latency with the I/O die and such could play a big factor in what is best, in what situation, especially if you can clock them the same. Some might happily spend £650 on a hypothetical 16c/32t CPU at 4.8GHz, and it could end up being slow than the hypothetical 8c/16t CPU at 4.6GHz in certain circumstances, and then if you level the clock speed you are ahead. Interesting times ahead, for a change.

Unless they save the fully working 8c dies for the higher models and use 2 partially working 8c dies to make an 8c CPU. Although the engineering sample at CES only had one die, it may not reflect the final version.

 

[EsaT]

Whilst I agree with the aesthetics over functionality is becoming worrisome, and frankly annoying. VRMs are not anything i've had a problem with Ryzen builds. Though I have always chosen X370 and X470 boards. The CH6 VRMs are insanely good.

Why not just use more describing term assthetics?
Because marketroids are selling fancy looking excrement for those thinking with their arse.

And X470 TUF has same garbage design VRM as other Asus boards below Prime X470 Pro.
VRM cooling is less sabotaged, but that design is still going to run really hot if powering fully loaded 2700X.
That VRM is basically what should be found from low end motherboards, not mid range.

https://www.hardwareluxx.de/community/f12/pga-am4-mainboard-vrm-liste-1155146.html

 

[kitfit1]

Well Intel must have known about overclocking because the 300A had a locked multipler to stop people from changing it. AMD chips were fully unlocked of course. You're right though, Intel didn't milk the market like they do now.

A locked Multi then still didn't stop you from clocking it. Even then..........FSB was your friend . As long as your mobo could do it.

To be honest, of course Intel knew about clocking......................they just chose to ignore it at the time.

 

[BongoHunter]

Of course they knew about it - it's a company with a lot of hardware nuts in it (or at least used to be)

Dual slot 1 / socket 370 celerons was one of the best bang for buck platforms going 20ish years ago - that was another hardware 'hack' that Intel did not officially support

 

[Scougar]

Something has to hit 5GHz in some fashion, otherwise the ill-informed and misguided will automatically cry at it being another failure.

Lol. I figure they have to get to at least 4.5ghz. crying.... Doubtful.

 

[CuriousTomCat]

Anything below 5GHz will lead to a lot of people in the 9900k thread laughing at us saying "I told you so". The fact we'll beat them on performance wont matter because they're blinded by clockspeed.

 

[LePhuronn]

Lol. I figure they have to get to at least 4.5ghz. crying.... Doubtful.

Intel shills will laugh at AMD still not hitting 5GHz (despite AMD being the first to do it), the clockspeed-obsessed will see sub 5GHz as a failure, and those saying it's not an issue will be lambasted as AMD nuthuggers.

There will be lots of petulant crying.

 

[easyrider]

Anything below 5GHz will lead to a lot of people in the 9900k thread laughing at us saying "I told you so". The fact we'll beat them on performance wont matter because they're blinded by clockspeed.

This is such a divisive post...I mean what is your agenda? Do you really think that people with 9900k at 5ghz be bothered that a so called 3600x running at say 4.5ghz is beating it in cinebench?

Are you ten? I mean your post doesn't seem to come from any place of maturity at all...

And FYI from your short time on these forums...There are people in that 9900k thread with years of experience in overclocking and getting on with posting real world data not posting pointless posts like yours. Many of them also post in the AMD overclocking threads....

So wind it in....

Intel shills will laugh at AMD still not hitting 5GHz (despite AMD being the first to do it), the clockspeed-obsessed will see sub 5GHz as a failure, and those saying it's not an issue will be lambasted as AMD nuthuggers.

There will be lots of petulant crying.

I doubt it...Its clear from my years on this forum that the AMD sycophants have the issues

You haven't had anything worthy to talk about as nothing from AMD since Opterton 170 was worth bothering with...

Its like the AMD crowd are always looking for conflict all the time and its its boring as ****