• 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 ***

Soldato
Joined
19 Apr 2012
Posts
5,184
I tend to just opt for the tried and tested route that most folk have gone for really.
I did notice the R5 3600 in the spec list and would probably be looking at similar. I haven't really thought much about budget really. It's been 8 years since I upgraded so will have to probably spend a decent wedge.
 
Associate
Joined
5 Mar 2009
Posts
1,138
Location
Essex
^^ similar boat here, the 3600 looks like it will be the bang / buck champ. My boy is just at an age where hes getting into PC gaming and my old gaming PC is starting to show its age.
 
Joined
2 Jan 2019
Posts
617
I do recall seeing somewhere an enhanced view of the traces on that die shot above. IIRC, the enhanced image seemed to suggest a crossbar toplogy with what appears to be a thick horizontal interconnect separating 4 cores on either side, and a thick vertical interconnect leading to the other chiplet. Since we're expecting the L3 to be on the IO die, the complexity of cross-CCX communication shouldn't impact latency in unpredictable ways; all L3 checks will be going to the IO irrespective of which core's L1 and L2 it was previously held in.
The question appears to be whether going to the IO for L3 adds greater latency overall than the increased size L1 and L2 are reducing the need to even go to L3.
 
Associate
Joined
14 Dec 2016
Posts
958
I'm hoping we get some better motherboards this time, x370 etc were ok, not great with a couple of exceptions, my CH6 is decent mind.

x470 were a bit better but not hugely it seems..

This time round I'm hoping there's some really great B550 boards, I think this time I'm going to go for the B series and put the saved cash to other areas of the PC
 
Soldato
Joined
24 Feb 2003
Posts
4,198
Location
Stourport-On-Severn
I'm hoping we get some better motherboards this time, x370 etc were ok, not great with a couple of exceptions, my CH6 is decent mind.

x470 were a bit better but not hugely it seems..

This time round I'm hoping there's some really great B550 boards, I think this time I'm going to go for the B series and put the saved cash to other areas of the PC

It's never worth "cheaping out" on a mobo if you want the best from a cpu. The CH6 was and still is a top class AM4 mobo, as a quick look at the Cinebench R20 thread will prove. As it stands, none of us knows for certain if Zen2 actually will need a 570x mobo to perform at it's best. I for one will keep my CH6 until it's proved beyond doubt that a 570x will actually perform better than a CH6.
 
Soldato
Joined
22 Mar 2014
Posts
3,956
It's never worth "cheaping out" on a mobo if you want the best from a cpu. The CH6 was and still is a top class AM4 mobo, as a quick look at the Cinebench R20 thread will prove. As it stands, none of us knows for certain if Zen2 actually will need a 570x mobo to perform at it's best. I for one will keep my CH6 until it's proved beyond doubt that a 570x will actually perform better than a CH6.
My motherboard only cost me £80 and has been holding a 3570k and then this 3770k @4.5ghz for many years, I agree it's best not to go too cheap but the CH6 is overkill for most.
 
Soldato
Joined
30 Jan 2009
Posts
17,185
Location
Aquilonem Londinensi
I do recall seeing somewhere an enhanced view of the traces on that die shot above. IIRC, the enhanced image seemed to suggest a crossbar toplogy with what appears to be a thick horizontal interconnect separating 4 cores on either side, and a thick vertical interconnect leading to the other chiplet. Since we're expecting the L3 to be on the IO die, the complexity of cross-CCX communication shouldn't impact latency in unpredictable ways; all L3 checks will be going to the IO irrespective of which core's L1 and L2 it was previously held in.
The question appears to be whether going to the IO for L3 adds greater latency overall than the increased size L1 and L2 are reducing the need to even go to L3.

They've used slightly troubling terms like doubling throughout in certain tasks (paraphrasing).

General compute at home is not massively parallel operations that fit within L2, for example.

I'm cautiously optimistic but the cynic (realist?) in me thinks Zen2 will be a mixed bag again
 
Permabanned
Joined
2 Sep 2017
Posts
10,490
They've used slightly troubling terms like doubling throughout in certain tasks (paraphrasing).

General compute at home is not massively parallel operations that fit within L2, for example.

I'm cautiously optimistic but the cynic (realist?) in me thinks Zen2 will be a mixed bag again

It will be. First attempt with a brand new design and execution, must be not polished.
 
Back
Top Bottom