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

[michty_me]

Is it not well known that and are well behind in the mobile CPU market anyway?

 

[humbug]

AMD is behind in the laptops chips with 15W-25W.

https://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-1065G7+@+1.30GHz&id=3466

https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+5+3500U&id=3421

The Ryzen 5 3500U is Last gen, (12nm Ryzen 2000)

Edit and mid range vs high range on the Intel side. that Intel chip is about 4X as expensive.

 

[jigger]

It may seem so but I am really not. You are free to try with the Ryzen 7 3700U. Its result is ~8000 Passmark points.
I don't have all the existing CPUs in my head - these are the first that I found.

Without security mitigation?. Apple are seeing 40% regression.

Basically, Intel are rubbish and AMD are, and I hesitate to use this word as it’s horribly overused, but... AMD are awesome.

 

[Orange Nexus]

Not sure. With or without the security mitigation?

In order for that i7 to be fully mitigated HT has to be disabled.

 

[Woodsta888]

It may seem so but I am really not. You are free to try with the Ryzen 7 3700U. Its result is ~8000 Passmark points.
I don't have all the existing CPUs in my head - these are the first that I found.

Enthusiasts forum...laptops, meh

 

[willhub]

Is it worth updating to PI 1.0.0.3 ABB?

 

[Lolik]

Even plebs on B350 got update Version 5216 2019/09/09 (showed up few min ago), fingers crossed

 

[beany_bot]

https://adoredtv.com/amd-pushes-out...ost-issues/amp/#click=https://t.co/Rh6ZJYqUeQ

 

[welshrat]

Is it worth updating to PI 1.0.0.3 ABB?

I would like to know this as well, given that we will have ABBA in around a month or so, is it worth upgrading from AB to ABB ? My system has been solid since AB so I guess I should stick ?

 

[LumpyVR]

Well this isnt ideal. Thought I'd get ready for my Ryzen 3700X upgrade I'm planning soon by getting my bios updated on my Gigabyte X370 Gaming 5. It was on bios F6 (yeah I know but it was working nicely) and went to F31 which is required before going to the latest (F42b).

Updated using Q-Flash and it appeared to work just fine, everything was at stock defaults and looked good. Before even attempting to re-apply my CPU overclock I tried putting my ram back on the standard XMP profile thats been working on the old bios since the beginning of time and imagine my surprise when it bluescreened windows during boot. And theres me thinking memory support had got better with newer bios versions not worse! Didnt have time to really play around with it, manually just entering 3200Mhz resulted in the same crash so I settled on 3000Mhz for now as that seems stable although not stressed it yet.

I'm guessing maybe entering the speed and primary timings manually is probably the first step to see if that works?

So annoying as I thought I'd been lucky with this ram on this board never having to do anything other than enable the XMP profile, I foolishly thought we were past these teething issues on X370!

 

[AStaley]

I would like to know this as well, given that we will have ABBA in around a month or so, is it worth upgrading from AB to ABB ? My system has been solid since AB so I guess I should stick ?

If it's not broke don't fix it. Unless there's a specific fix you need in the current firmware, leave well enough alone.

 

[willhub]

If it's not broke don't fix it. Unless there's a specific fix you need in the current firmware, leave well enough alone.

They're all broke, at least with 3700x, I can't hit 4.4Ghz but I want my CPU to work as intended, newer firmware seems to address this, ABB I guess isn't the one.

 

[TrixX]

They're all broke, at least with 3700x, I can't hit 4.4Ghz but I want my CPU to work as intended, newer firmware seems to address this, ABB I guess isn't the one.

Tell that to the 3700X I have running 4.5 and 4.4 GHz on the respective CCX's...

If you do manual overclocks with good cooling they run fine

 

[willhub]

Tell that to the 3700X I have running 4.5 and 4.4 GHz on the respective CCX's...

If you do manual overclocks with good cooling they run fine

Mine won't run at all near 4.4

 

[marms]

It may seem so but I am really not. You are free to try with the Ryzen 7 3700U. Its result is ~8000 Passmark points.
I don't have all the existing CPUs in my head - these are the first that I found.

The intel i7 1065g7 costs $426.00, about the same price as a laptop with an r5 3500u processor, AMD haven't brought out any mobile processors with the zen 2 cores yet because the desktop range are so good and such great value that they're using up all the 7nm chiplets they can produce.

 

[AStaley]

They're all broke, at least with 3700x, I can't hit 4.4Ghz but I want my CPU to work as intended, newer firmware seems to address this, ABB I guess isn't the one.

Your'll be wanting the BIOS with AGESA ABBA which is supposed to be the one to help with boosts clocks, AMD have said it's with motherboards manufacturers now and they expect BIOS's based off it to be available within the next 3 weeks depending on the manufacturers testing processes. Personally I'd still hold off installing it for a week and keep an eye on the forums for your board to see how it fairs.

 

[DragonQ]

AMD is behind in the laptops chips with 15W-25W.

https://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-1065G7+@+1.30GHz&id=3466

https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+5+3500U&id=3421

Mobile chips are a different beast to desktop because Intel is a step ahead there (10nm Ice Lake on mobile vs 14nm++++ Comet Lake on Desktop), whereas AMD is a step behind (12nm Zen+ on mobile vs 7nm Zen 2 on desktop).

Also performance numbers are really muddy these days because Intel is going for stupidly low base clocks with high boost clocks. You are going to see massively different performance for a 10 second benchmark versus a 10 minute one, for example. There's a reason the base clock is so low.

 

[Lolik]

Seems to be working ok (small hick up memory profile was showing corrupted on usb stick , luckily enough still had a picture with Ryzen mem calc settings for 3444 fast ) big difference noticed so far : temps, never went above 66 C with all testings (noctua D15s ), but still early days; playing with pbo, xfr atm.

Update: PBO(on the first page) works ok , increased multicore boosting from 4 to 4.15-4.2 ( not 100% sure, but behavior of cores looks different, instead of hitting 4.2+ at the beginning of R20 mp and going down as temps get higher (as before), now it does more core freq switching and it is not a downtrend all the time, some of them go back to 4.2 again , points wise it increased from 4700 stock to 4900 with pbo. Temps went up to 75C . Max volts seen so far 1.44 .
Update2: Pbo in overclocking settings ( when leaving rest auto) does ... all (nothing, dunno maybe cause you can't change PPT,TDC,EDC there) anyways haven't noticed any changes in that 1.
Update3: none of the other changes on advanced overclocking page in bios (xfr, nbio etc) made any difference (or my testing is too short and light ) , didn't try changing ram timings there.

 

[Silent_Scone]

Crosshair updates can be downloaded here.

https://www.dropbox.com/s/rjge7gan7h8pb2z/1001.rar?dl=0&file_subpath=/1001

 

[4K8KW10]

Mobile chips are a different beast to desktop because Intel is a step ahead there (10nm Ice Lake on mobile vs 14nm++++ Comet Lake on Desktop), whereas AMD is a step behind (12nm Zen+ on mobile vs 7nm Zen 2 on desktop).

Also performance numbers are really muddy these days because Intel is going for stupidly low base clocks with high boost clocks. You are going to see massively different performance for a 10 second benchmark versus a 10 minute one, for example. There's a reason the base clock is so low.

I think it just shows that AMD's performance per watt graph peaks at 65W and is worse in the beginning of the curve.
Ryzen 5 2500U base clock (frequency at idle) is also 1.47GHz, despite that officially it's claimed as 2.0GHz.