Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)

[DisEnchantment]

Speculate at will

 

[moinmoin]

He's not a source of anything.

Not even of entertainment?

 

[RnR_au]

AMD hype trains are the best!

 

[itsmydamnation]

Danilo ? Lol.
He's not a source of anything. Useless

we need to get Francois P point of view on Zen5

 

[Markfw]

AMD hype trains are the best!

What I find interesting is that there are NO negative views or leaks that I have seen on Zen 5. Unlike Intels latests that are marginally decent, but not even the best !

 

[adroc_thurston]

Zen 5 hype train is starting to leave the station, for better or worse. Lol

That's no clue city.

 

[poke01]

What I find interesting is that there are NO negative views or leaks that I have seen on Zen 5. Unlike Intels latests that are marginally decent, but not even the best !

Arrow Lake is meh. Lunar is great but that is low power not desktop. AMD will own desktop for a few years more imo.

 

[Saylick]

What I find interesting is that there are NO negative views or leaks that I have seen on Zen 5. Unlike Intels latests that are marginally decent, but not even the best !

Well, it’s got the same IOD as Zen 4, which is arguably a bad/sad thing. Probably means Zen 5 can come faster to market but could have had some improvements to the memory controller if they wanted to invest resources.

 

[StefanR5R]

Regarding Strix Halo (intended for laptop), rumors say it'll be 16C Zen5 + 40 RDNA 3.5-spec compute units. So essentially close to DT variants of Zen5 in CPU perf (but likely with somewhat slower clocks), and extremely much higher iGPU perf than DT. This is assuming Zen5 on DT will only be max 16C and have same IOD (and iGPU) as Zen4.

So this means for Zen5, DT will not have much higher MT CPU perf than laptop [...]?

In workloads which are actually able to utilize all thirty-two hardware threads, the Zen 5 based successor to 7950X will certainly make better use of the (default) socket power budget compared to 7950X. In other words, the zone of diminishing returns from an increased power limit should sit somewhere higher on Zen 5 than on Zen 4. Consequently, there should be a computing throughput gap between the 7950X successor and Strix Halo in such well parallelized workloads, if Strix Halo is only going to be available in form factors which are more constrained than desktop PCs WRT power supply and cooling performance.

 

[SiliconFly]

...Arrow Lake is meh...

But still better than the stupid Zen 5.

 

[TESKATLIPOKA]

But still better than the stupid Zen 5.

Says who? You? 😀

 

[coercitiv]

the stupid Zen 5.

stupid yes, but let's see it run

 

[Timorous]

But still better than the stupid Zen 5.

Arrowlake will be better at just one metric.... watts / frame.

 

[branch_suggestion]

But still better than the stupid Zen 5.

Conroe was pretty stupid as well, the memory controller was still on the chipset, very primitive compared to the sophisticated uncore of K8.
Still beat the everloving snot out of it, which is what matters.
It will take more than magical process leaps to make LNC competitive, the actual core design needs a massive overhaul vs their current cores.
IDC might be the stupid ones after all.

 

[Schmide]

Danilo ? Lol.
He's not a source of anything. Useless

It's almost like they're punk'ing reviewers. Sillycon Gang even got Steve.

lets go back a bit and laough at the bias

https://twitter.com/x/status/1574452493756420096

 

[Jan Olšan]

Well, that's a guy that IIRC had no problem calling Zen 4 housefire when its competitor is PL1=PL2* era Intel.

*unless it's 300W

 

[Panino Manino]

Zen 5 hype train is starting to leave the station, for better or worse. Lol

https://twitter.com/x/status/1749473720182419515

https://twitter.com/x/status/1749455220894036097

Aaaaah... it begins, the cycle of over expectations and forced disappointment when AMD "betrays the promises".

 

[Glo.]

Aaaaah... it begins, the cycle of over expectations and forced disappointment when AMD "betrays the promises".

When has AMD on CPU side "betrayed the promises" in recent times?

Its the opposite, actually.

 

[Elfear]

When has AMD on CPU side "betrayed the promises" in recent times?

Its the opposite, actually.

I think that's why @Panino Manino put that in quotes. There are some people who listen to the hype and then are disappointed in AMD because reality isn't as awesome as the rumors (rather than being disappointed in the rumormongers themselves).

 

[adroc_thurston]

I think that's why @Panino Manino put that in quotes. There are some people who listen to the hype and then are disappointed in AMD because reality isn't as awesome as the rumors (rather than being disappointed in the rumormongers themselves).

Plebs have not even the slightest idea on how fast the thing really is so...

 

[moinmoin]

Plebs have not even the slightest idea on how fast the thing really is so...

androc_thurston on introducing the meta hype train.

 

[adroc_thurston]

I think that's why @Panino Manino put that in quotes. There are some people who listen to the hype and then are disappointed in AMD because reality isn't as awesome as the rumors (rather than being disappointed in the rumormongers themselves).

There's a general expectation mismatch between what people expect (another Zen3 at maximum) and what it actually is.

 

[coercitiv]

androc_thurston on introducing the meta hype train.

Hype train? Where we're going, we don't need hype trains.

 

[gdansk]

Hold me to it: I'll buy an X3D part if there's even a 15% 1T improvement. But I expect 25% 1T improvement like Zens of yore.

 

[Saylick]

There's a general expectation mismatch between what people expect (another Zen3 at maximum) and what it actually is.

Idk, seems to me that peeps on Twitter know about 30%.

 

[adroc_thurston]

Idk, seems to me that peeps on Twitter know about 30%.

30% what