Official AMD Ryzen Benchmarks, Reviews, Prices, and Discussion

[.vodka]

Well, i am looking at those Prime numbers and Physics results and keep thinking latency matters

Have you seen this?

Because Ryzen with this memory at hand

would paint a completely different picture on those two subtests vs the crappy DDR4 2133 C17 results going around.

This is a non issue anymore.


edit: at this point I'm repeating myself like a broken record, it seems hard to drive this point home.

 

[lobz]

Nah, it is new Phenom

Well, i am looking at those Prime numbers and Physics results and keep thinking latency matters

new phenom, yes, I see Juan's ghost haunts you and possessed your hands, in the last week you've bevome so cocky that it's starting to get awkward

 

[Teizo]

I imagine all the boards needed was a bios update to get the memory stuff worked out. The tech YouTuber Steven Walton of Hardware Unboxed said explicitly he was having trouble with the memory on his Crosshair board.

 

[Mockingbird]

Ryzen Wraith MAX preview(video):

It's a pretty beefy cooler with eatpipe construction fan.

This should ad up tu a benchmark as tu how a stock cooler should look like.

 

[inf64]

Corsair (on their YT account) posted a video of Ryzen running 4x 3Ghz DDR4 sticks.

 

[lobz]

Yes, but i meant that 2133C15 has LITERALLY lower latency than 2400C17. He meant that Ryzen posts a notably better latency result here. On which i can only note that it is done with that MSI X370 XPOWER or something board. Can't get much better than that.

I see... so now you say you assumed inf64 didn't know this. I rest my case then.

 

[lolfail9001]

Corsair (on their YT account) posted a video of Ryzen running 4x 3Ghz DDR4 sticks.

Hooray, that got fixed quickly.

new phenom, yes, I see Juan's ghost haunts you and possessed your hands, in the last week you've bevome so cocky that it's starting to get awkward

What arguments do you have against it being a new phenom. new Phenom II, if you will, even. Granted, this one is actually launched in the better time.

would paint a completely different picture on those two subtests vs the crappy DDR4 2133 C17 results going around.

Mobo replacement had as much impact on memory latency as replacing 2400c17 with 4266c12, consider that for a second.

I see... so now you say you assumed inf64 didn't know this. I rest my case then.

Did you?

 

[IEC]

Does anyone find it odd how there are continuously consistent goalposts in this anti-"anti-Ryzen narrative" narrative?

Cite same person saying that. You can start with me.

Wasn't talking about you, but if you really had insider knowledge you'd be breaking NDA, so I can safely discount your opinion.

Only two days until we find out whether the lower bound of expectations or the higher bound of expectations is accurate. My money (and preorder) is on it being somewhere in the middle

 

[Teizo]

It was always March 2, 28th Feb was some rumor conflating it with the other AMD event.

Or, due to the bios issues they had to push the embargo date back 2 days so board partners could get out bios updates to fix the memory problems so the press could do a proper review. AMD has allot at stake with this launch. They can't afford poor reviews. I don't think it was a rumor. Several sources stated the review embargo was Feb. 28th with March 2nd mass availability.

Oh well. Waited this long. What is two more days....

 

[IEC]

http://www.anandtech.com/show/11143...or-under-330-preorder-today-on-sale-march-2nd

We’ll tell you what our benchmarks say, with official retail processors. But you will have to wait until March 2nd. Sorry.

by Ian Cutress on February 22, 2017 9:00 AM EST

The date was March 2nd from the beginning.

 

[anhhq]

yeah 4x4GB 3000 MHz

Zen vs. 40 Years of CPU Development

https://www.pcper.com/reviews/Editorial/Zen-and-Art-CPU-Design

take some relax :

 

[OrangeKhrush]

some inside info, only one motherboard vender has not surpassed 3200Mhz RAM speeds, the Crosshair VI is guaranteed to hit 3800Mhz and with Ryzen it loves high frequency RAM, it is like nitrous to a top fueler

 

[Teizo]

I see. I guess there was a mix up then with AMD's event today. It was reported that the embargo was going to be lifted today coinciding with their event. Regardless, good thing it is March 2 then so board partners could get the bios updates out to fix the memory issues. That is all it is...Probably QVT stuff with some brands. It's not a big deal at all. Memory issues are common with early bios.

 

[lolfail9001]

with Ryzen it loves high frequency RAM

I just hope for this to be true just to destroy the speculations that this would not be the case. Damn this launch is fun.

 

[OrangeKhrush]

I just hope for this to be true just to destroy the speculations that this would not be the case. Damn this launch is fun.

The person spilling the beans initially said about 3ghz limits but corrected today saying that AMD very recently forced a firmware update that got the IMC support up to 3800mhz. He did say sadly now though the memory subsettings are gone and all you get is 4 subsets with the dividers to play around with. Basically AMD are pushing the engines up to 100% now.

 

[lolfail9001]

He did say sadly now though the memory subsettings are gone and all you get is 4 subsets with the dividers to play around with

!?!?!?!? Guess we'll see in 2 days.

 

[IEC]

3800MHz? I'll believe it when I see it.

I have a kit capable of 3600+ so I should be able to see it soon enough, if true...

 

[OrangeKhrush]

So I posted this, and I ran the math and the numbers checked out, the 1600 score is only at 3.7Ghz being the MAX T all core load for MT usage.

The reply

Score is correct and I can confirm this. 1730 ~ 1775 is right for 4GHz. Obviously don't know memory config etc as it is hard to see, but the numbers are exactly where they should be. Won't get more legitimate than this.

 

[lopri]

I actually would not mind using the new Wraith cooler for a 1700. Assuming that it performs semi-decently, it adds more value to the 1700's in comparison to the 1700X/1800X.

 

[OrangeKhrush]

3800MHz? I'll believe it when I see it.

I have a kit capable of 3600+ so I should be able to see it soon enough, if true...

Doubts is what has driven this fun ride, the best way into peoples bank accounts is to feed the doubt then #@$# their minds.

 

[unseenmorbidity]

I actually would not mind using the new Wraith cooler for a 1700. Assuming that it performs semi-decently, it adds more value to the1700's in comparison to the 1700X/1800X.

The 1700 comes with the wraith spire, which is maybe a $10-$15 cooler. Depending if you think leds are good.

 

[Dresdenboy]

The person spilling the beans initially said about 3ghz limits but corrected today saying that AMD very recently forced a firmware update that got the IMC support up to 3800mhz. He did say sadly now though the memory subsettings are gone and all you get is 4 subsets with the dividers to play around with. Basically AMD are pushing the engines up to 100% now.

Interesting tactic. Waiting for a new episode of "Intel engineers react to... an AMD firmware update"...

 

[PotatoWithEarsOnSide]

The MP Ratio on that CB bench was 10.85, so SMT is effectively giving an extra 35-36% worth of performance on a per core basis.
What would the MP Ratio be on Intel's top CPUs?

 

[OrangeKhrush]

Interesting tactic. Waiting for a new episode of "Intel engineers react to... an AMD firmware update"...

It is called Cannon Lake and Skylake X, Since AMD needs AM4 to drive it over the next 3 years, they need the tech up front. Now whether anyone has 3800Mhz is another question altogether.

 

[OrangeKhrush]

The MP Ratio on that CB bench was 10.85, so SMT is effectively giving an extra 35-36% worth of performance on a per core basis.
What would the MP Ratio be on Intel's top CPUs?

Which is even sweeter considering Juanrga claimed AMD had only SMT hiding IPC, but it is confirmed that AMD's IPC is far over what he claimed, like 12+ % which in some aspects could be 40% others 10% gains, either way he got it terribly wrong.