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

[pandem0nium]

Raid using two slots from the chipset using PCIe 4.0 nvme ssds will probably be pointless.
There'll be the same problem Intel currently has with raid 0 in that the performance will be limited by the 4 lanes the cpu uses to communicate with the chipset.
Cpu lanes m.2 paired with one m.2 from the chipset should be fine though.

 

[pcfarrar]

Why?

I'm upgrading to the 3700x purely for the better efficiency vs my 2700x. Less heat so I can lower fan speeds etc.

 

[gavinh87]

I'm upgrading to the 3700x purely for the better efficiency vs my 2700x. Less heat so I can lower fan speeds etc.

Gotta save that £5 per year

 

[d_brennen]

Gotta save that £5 per year

That's being too generous Idle power won't be far different. It's like people who buy a £25k car because it has slightly better MPG than their perfectly fine car.

Looks like there'll be a decent performance uplift though

 

[Thekwango]

This is the problem with amd and its fans. Products get hyped to the moon and when they dont meet expectations people **** them off.

imo, the real problem is the bs'ers posting their 'leaks' and people believing them - whether that's because they wanted to believe them or because they didn't know any different. you take all the utter ******* ***** that was posted as 'leaks' and bin it out of the equation and what amd have released is a pretty decent line up that appears to be pretty close or better, potentially, than intels equivalent offerings but at a lower price point - that would normally be greeted with applause. but then you add in the insignificant ******* with their youtube channel and 'ohhh super brilliant exclusive ryzen leak' headlines and suddenly the less informed are getting all frothy round the gusset at the prospect of a 6.5ghz, 32core cpu for £2.99. then the reality hits and those same folk who fell for the high production value 'leaks' are distraught and start crying foul at amd when they really should be turning their ire towards these idiot tech guru/knowitalls who know **** all the rest of us with half a brain and a calculator couldn't have worked out for ourselves.
Ryzen 3xxx appears to be a solid product at a decent price point - that is what people should be taking from this. not falling into the trap of listening to nobodies who make a living out of peddling their own brand of bs.

of course amd could have nipped a lot of these leaks in the bud at computex but they didn't - they seem happy to let the nonsense guessing and prevarication continue.

 

[humbug]

Plus 14% puts 3800x just above the 8700K @103.39

GT-V is not the absolute worst but also not a great example for 2700X vs 9900K, but its a fair'ish overall representation and i'll take it.

9900K: 106
2700X: 90

Its a difference of 18%

We can also take @Orange Nexus PUBG

9900K: 171
2700X: 163

A difference of 5%.

You know, its swings and roundabouts, 5% in one 18% in another, what i said was an average of 15%, its really not a lot for Ryzen 3000 to make up, Zen+ is already a very good way to a 9900K.

 

[Drollic]

I don't mind paying a little extra for a motherboard to be an early adopter of new tech. The next round of Intel boards are hardly going to be cheap.

 

[KentMan]

Will pee myself if AMD has sandbagged OC performance on these, notice nothing has been said officially about PBO and XFR...

If at stock they are matching the 9900k etc and have room to OC there will be plenty of Intel defence league on these forums eating some large portions of humble pie

Judging by the amount of motherboard models, something tells me the manufacturer's know something that the armchair specialists on these forums do not...

Will be hilariously interesting if AMD take the performance crown to watch the drivel from the usual suspects here trying to downplay and naysay as usual

 

[N19h7m4r3]

9900K: 171
2700X: 163

A difference of 5%.

You know, its swings and roundabouts, 5% in one 18% in another, what i said was an average of 15%, its really not a lot for Ryzen 3000 to make up, Zen+ is already a very good way to a 9900K.

Personally I'd only be looking at the minimum percentile frames for gaming performance.

Going by that the 2700X is already giving a comparable gaming experience ( In PUBG ).

It'll be interesting to see how Ryzen 3000 does there for me in games compared to the rest.

 

[humbug]

Personally I'd only be looking at the minimum percentile frames for gaming performance.

Going by that the 2700X is already giving a comparable gaming experience ( In PUBG ).

It'll be interesting to see how Ryzen 3000 does there for me in games compared to the rest.

A couple % better than the 8700K even.

I think it all maters.

 

[Zeed]

Why?

So he can never EVER EVERRR make money back from swap

 

[DragonQ]

Will pee myself if AMD has sandbagged OC performance on these, notice nothing has been said officially about PBO and XFR...

Doubt it'll be mind-blowing but we'll see.

If at stock they are matching the 9900k etc and have room to OC there will be plenty of Intel defence league on these forums eating some large portions of humble pie

Well regardless of anything else, I keep coming back to that CES result. That has to have been the 65W R7 3700X and it matched the i9-9900K in Cinebench. Not gaming, but gaming performance will not be that far away. The R7 3800X is faster still, although probably only a tiny bit.

 

[N19h7m4r3]

A couple % better than the 8700K even.

I think it all maters.

I'd like to see some IPC tests where they run some games at exact same clock rates, along with other benchmarks.
AMD haven't been too inaccurate with their Ryzen benchmark presentations since their original launch, so even with some salt their Computex showing was darn fine.

Even if they're 5% behind in gaming compared to Coffeelake, it doesn't matter. As a 3900X overall would be a significant upgrade over my 5820K that can't exceed 4.0Ghz, or best even the 9900K for project rendering with those threads alone.

I'd sure be even happier if it does best the 9900K in gaming though

 

[Stephanie Peterson]

I still think xfr will add on at least a hundred mhz, prob a bit more on one or two cores.

 

[Zeed]

I'm upgrading to the 3700x purely for the better efficiency vs my 2700x. Less heat so I can lower fan speeds etc.

EEEEMMMM You are better of investing in 240mm AIO if You do it for Quiets mate... Calculate it using money saved on electricity ITS NEVER GONNA PAY ITSELF BACK

 

[Zeed]

I'd like to see some IPC tests where they run some games at exact same clock rates, along with other benchmarks.
AMD haven't been too inaccurate with their Ryzen benchmark presentations since their original launch, so even with some salt their Computex showing was darn fine.

Even if they're 5% behind in gaming compared to Coffeelake, it doesn't matter. As a 3900X overall would be a significant upgrade over my 5820K that can't exceed 4.0Ghz, or best even the 9900K for project rendering with those threads alone.

I'd sure be even happier if it does best the 9900K in gaming though

It's pointless. Only real test will be 9900k@5ghz vs 3800x/3900x at max overclock whatever that will be. I cant think of single pwerson buying 9900k and not overclocking it...

 

[beany_bot]

They wont have "hidden" or "sandbagged" overclocking performance. That's ridiculous. Why hide and loose performance they could ship with? You realise a very VERY small proportion of people actually overclock. and youtube benchmarks are always done at stock. They will ship with as much clockspeed as guarantees stability. I think Jim from Adored is really clutching at straws now.

 

[beany_bot]

I cant think of single pwerson buying 9900k and not overclocking it...

The you are out of touch with reality. 90%+ of people buying the 9900k will not be overclocking it.
Obviously people in this forum / community will. But we are a drop in the ocean.

Having said that I want to see overclocked 9900k vs overclocked 3x00X too because I plan to overclock.

 

[humbug]

Doubt it'll be mind-blowing but we'll see.


Well regardless of anything else, I keep coming back to that CES result. That has to have been the 65W R7 3700X and it matched the i9-9900K in Cinebench. Not gaming, but gaming performance will not be that far away. The R7 3800X is faster still, although probably only a tiny bit.

I watched Hardware Unboxed last night with Tim defending his calling out the TDP listing on AdoredTV slide, which turned out to be right. He was explaining that it didn't look right because 65 Watt vs 105 Watt with only a 100Mhz difference wouldn't make any sense.

He didn't realize, and still doesn't realize that difference in TDP is nothing to do with the clocks that are written on the slide, the difference is in XFR boosts, with a higher TDP the 3800X is setup for a higher all core boost, the difference in performance, out of the box at least is going to be more than "tiny"

 

[Kashinoda]

They wont have "hidden" or "sandbagged" overclocking performance. That's ridiculous. Why hide and loose performance they could ship with? You realise a very VERY small proportion of people actually overclock. and youtube benchmarks are always done at stock. They will ship with as much clockspeed as guarantees stability. I think Jim from Adored is really clutching at straws now.

I mean he is but, XFR is actually pretty good at getting close to maximum OC out of the box?