• 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 ***

lobz said:
You're wrong.

https://optocrypto.com/amd-zen-3-milan-processors-will-not-use-ddr5/amp/

it's just one link but you'll find the same information dozens of places on the interwebz
Click to expand...

Huh?

Nothing about AM4 vs AM5 in that link.

Secondly talks about Milan anyway.

Need to provide better evidence mate.

AMD have fulfilled their promise on AM4. Nothing obliging them to use AM4 next. Giving people such certain comments when you have no evidence will mean they make bad purchasing decisions off your falsely placed confidence.


24:30 minutes in the pertinent conversation starts. No commitment left after Zen 2.
 
muon said:
Huh?

Nothing about AM4 vs AM5 in that link.

Secondly talks about Milan anyway.

Need to provide better evidence mate.

AMD have fulfilled their promise on AM4. Nothing obliging them to use AM4 next. Giving people such certain comments when you have no evidence will mean they make bad purchasing decisions off your falsely placed confidence.


24:30 minutes in the pertinent conversation starts. No commitment left after Zen 2.
Click to expand...
They did say that there were no immediate plans to change socket. They also said that DDR5 would necessitate a socket change.
Unless you're suggesting they'll have a new socket next year, and then another the year after, then AM4 for 4th gen Ryzen seems most logical assumption.
 
Potatowithearsontheside said:
They did say that there were no immediate plans to change socket. They also said that DDR5 would necessitate a socket change.
Unless you're suggesting they'll have a new socket next year, and then another the year after, then AM4 for 4th gen Ryzen seems most logical assumption.
Click to expand...

I'm saying you can't promise anyone that they won't have a new socket next year.

Like he said, it would be something that would change the pin count of the CPU that triggers it. He said DDR5 "could" be the cause of the that only because he was forcibly asked that. Who knows, DDR5 might work with AM4.

lobz said:
we

don't expect it, we know it, because it's 100% confirmed by AMD that Zen 3 is DDR4 + PCIe4 and AM4
Click to expand...

This guy just pulled something out of his arse, in answering a prospective buyer's question.
 
Martini1991 said:
If 4XXX is the same as 2XXX in terms of improvement, it's basically pointless for people looking at upgrading unless you're increasing your core count.

I bought a 2700 from a 1700.
Wasn't really worth it lol.
I did gain 200MHZ though. Woo.
Click to expand...
There won't be a Zen 2+, but Zen 3 is due next year. I would not expect IPC improvements akin to Zen 2, maybe more like 5% max with some tweaks. However, the process should be more mature so clocks might rise slightly a la Zen -> Zen+.

Renoir (8c/16t APU) is probably more interesting than Zen 3.
 
DragonQ said:
There won't be a Zen 2+, but Zen 3 is due next year. I would not expect IPC improvements akin to Zen 2, maybe more like 5% max with some tweaks. However, the process should be more mature so clocks might rise slightly a la Zen -> Zen+.

Renoir (8c/16t APU) is probably more interesting than Zen 3.
Click to expand...

8C APU is definitely more interesting.
Wish AMD had 6C APU's now.

We've just bought a bunch of 2400G PC's, but could quite have easily gone i5's. I pushed for the Vega GPU on the APU for the workloads.
 
DDR5 is supposed to be ready for next year. I can see AMD developing both memory controller versions just in case it needs to revert back to DDR4/AM4 if DDR5 runs into trouble. Otherwise being an early adopter of DDR5 will help with performance gains over Intel (assuming it meets spec from day one).
 
DragonQ said:
There won't be a Zen 2+, but Zen 3 is due next year. I would not expect IPC improvements akin to Zen 2, maybe more like 5% max with some tweaks. However, the process should be more mature so clocks might rise slightly a la Zen -> Zen+.
Click to expand...
I think it has been announced that they are using 7nm+ for Zen 3 so it's a new node albeit not that much denser.

"Initially, semiconductors produced with TSMC's 7nm+ will contain only a few layers of masks using EUV, masks being the design templates used in semiconductor manufacturing.
Despite the limited use, the report claims the transition to a 7nm+ EUV lithography will increase transistor density in Zen 3 by 20 percent, while reducing power consumption by 10 percent. That is an impressive leap. If I had to guess, Zen 3 probably will not provide the same jump in performance over Zen 2 that Zen 2 will have over Zen+, as both are still on a 7nm process node. However, it sets the groundwork for future iterations."
 
That link last page with 3600 at 4.2 ac and 1.4v at 70c, I wonder what cooler as it seems very unlikely a 12c part would sustain anything like 4.5 unless on water. That's unless he's using low end air.
 
muon said:
AMD have fulfilled their promise on AM4. Nothing obliging them to use AM4 next.
Click to expand...
They made am4 work with completely revolutionary change to chiplets and even with original chipsets.

It did sound that they had to do clever tricks in packaging and routing to adapt chiplets to am4
But now thats solved.

There is no reason do do anything as different until ddr5 and maybe pcie5
 
muon said:
Huh?

Nothing about AM4 vs AM5 in that link.

Secondly talks about Milan anyway.

Need to provide better evidence mate.

AMD have fulfilled their promise on AM4. Nothing obliging them to use AM4 next. Giving people such certain comments when you have no evidence will mean they make bad purchasing decisions off your falsely placed confidence.


24:30 minutes in the pertinent conversation starts. No commitment left after Zen 2.
Click to expand...
What? So you're saying now, that AMD of all companies will keep the same socket for Zen 3 server CPUs but create a new socket on the desktop for the same DDR4 and PCIe 4?
I don't think I have the mental strength for arguing over this :D
 
LePhuronn said:
Don't forget about the SMT upgrade to 3 or 4 threads
Click to expand...
For Zen 3? Anything more than mere speculation about that? I wonder if they would disable that in desktop parts if it makes no sense outside of servers!
I've only seen SMT beyond 2T per core in server parts before I think!
7nm+ should give them the die space for free in one sense but there's the validation and the extra cost of EUV lithography.
 
lobz said:
Despite you being so nice, I'll remain polite and break it to you, that you basically just said, Forest Norrod pulled this information out of his arse.
Click to expand...

https://www.anandtech.com/show/14568/an-interview-with-amds-forrest-norrod-naples-rome-milan-genoa

Show me where Forrest Narrod did that. Doesn't even look like his job has anything to do with consumer ryzen products.

You are influencing someone who is spending hundreds of pounds based on something you've made up.

Also on you're other post. You have any evidence to suggest that a server socket change is in sync with a desktop socket change and vice versa? Or is that more guesswork presented as a 100% fact. To me it just sounds like a desperate attempt to defend your supposed 100% confidence statement.
 
Last edited:
Drollic said:
That link last page with 3600 at 4.2 ac and 1.4v at 70c, I wonder what cooler as it seems very unlikely a 12c part would sustain anything like 4.5 unless on water. That's unless he's using low end air.
Click to expand...

I don’t know if I reading this right, but the 3600 isn’t a 12 core. I’m pretty sure it’s just a 6.
 
muon said:
https://www.anandtech.com/show/14568/an-interview-with-amds-forrest-norrod-naples-rome-milan-genoa

Show me where Forrest Narrod did that. Doesn't even look like his job has anything to do with consumer ryzen products.

You are influencing someone who is spending hundreds of pounds based on something you've made up.

Also on you're other post. You have any evidence to suggest that a server socket change is in sync with a desktop socket change and vice versa? Or is that more guesswork presented as a 100% fact. To me it just sounds like a desperate attempt to defend your supposed 100% confidence statement.
Click to expand...
I see you're not kidding. That is sad. I wish that guy you're influencing now the best of luck.
 
lobz said:
I see you're not kidding. That is sad. I wish that guy you're influencing now the best of luck.
Click to expand...

Well I certainly ain't the one promising him future compatibility with Zen 3 with 100% certainty when really all I have are some inferences from some comments on server products.

lobz said:
we

don't expect it, we know it, because it's 100% confirmed by AMD that Zen 3 is DDR4 + PCIe4 and AM4
Click to expand...
 
You must log in or register to reply here.
Back
Top Bottom