Official AMD Ryzen Benchmarks, Reviews, Prices, and Discussion

[Dresdenboy]

To the moon you say?

... and beyond. Nah, for now I expect about 2-5% points higher SMT yield on top of the 1T IPC gain.

 

[psolord]

The OC tool is called Ryzen Master.

I laughed my ass off!

Jesus Christ, that's brilliant!

 

[SPBHM]

probably the best they looked since 2006 before Core 2 was launched, the CPUs announced today are still on the very expensive side, but the 1700 non X certainly looks very appealing, the fact that AMD can ask $499 again for a CPU without looking stupid (thinking of you FX 9590 and friends) just confirms this is their best since when K8 was the best thing around, now the real damage will be the i5 priced 6 core I guess, a shame it's not being released with the 8 core ones...
newegg lists an AM4 board as low as $79, you also don't want 4 sticks.. so you can buy 8 core Ryzen 1700 + b350 MB for so much less than an x99 alternative, Broadwell-e (for desktop users) is looking really dead now (they have that 10 core part, but the price is so outrageous that it's the same as not existing), huge price drops incoming from Intel, or are they going to pretend nothing happened?

this is really well beyond any expectations I had for AMD since 2011.

 

[Ryzenenthusiast]

Something I've been thinking about and haven't seen anyone address really.

Recently some material has been leaked from the AMD's ISSCC presentation: https://videocardz.com/65774/more-details-about-amd-zen-cpu-core-revealed-at-isscc

The curious tidbit is this:

The material claims that AMD has a local LDO voltage regulator. LDOs are typically linear but in other article I've seen a mention of the term "digital LDO", it's kind of a weird name but anyways I think this is a switching regulator.. as linear regulators would make no sense (they waste too much power).

Why am I bringing this up? Well anyone who's worked with electronics can tell you that voltage regulators/LDOs or whatnot have a dropout voltage. I think one of the reasons we've been seeing higher than usual voltages in various screenshots (1.87v in that 5.2Ghz LN2 OC and 1.35v in that Ryzen Master screenshot for 3900Mhz) is because of the dropout voltage needed to compensate for this LDO.

The voltage being reported is the voltage at the Vin or at the VRM.. but the cores need a certain margin to compensate for the dropout in each LDO regulator.

So this is something to keep in mind when thinking of voltages.

I do not have an information about how much volts getting lost on VRM. Can you do a rough estimation about what 1,87v and 1,35v means how many Volts in Intel or old FX CPU's?

 

[lopri]

So the "full stack" at launch was just BS and this is mostly a paper launch? with 2/3rds of the line not being released??

This is disappointing i was looking froward to thr 6 and 4 core OC results from the initial batches.

I am disappointed about that as well because I really do not need performance at the moment but I sure do want to play with it. A 1100 or a 1300X would have been nice.

 

[sirmo]

@Ryzenenthusiast "I do not have an information about how much volts getting lost on VRM. Can you do a rough estimation about what 1,87v and 1,35v means how many Volts in Intel or old FX CPU's?"

Hard to tell.. But 0.3V - 0.6V are possible figures.. I think 0.3V would make sense in this case.

So like 1.35V may mean your core is really operating at 1.05V as an example.

We would need confirmation from AMD.. I do hope one of the reviewers find out.

 

[Sanguineus]

Was that 4.66 8C16T LNC or...?

Hmm, I assume this "16C/16T" for RyZen 7 is just a wrong reading or product description? Otherwise it would be strange...

 

[Phynaz]

Has anybody noticed that AMD is using Cinebench for their benchmark? I think that's weird, I mean I've been told on this forum so many, many times that you can't use Cinebench to compare between AMD and Intel because it's biased towards Intel.

So what's up with that? Do you think AMD designed Ryzen to do good in Cinebench?

 

[lopri]

Considering that Zen's integer unit is much stronger than the floating-point unit, I'd say AMD is underplaying its hands still.

 

[Maverick177]

Has anybody noticed that AMD is using Cinebench for their benchmark? I think that's weird, I mean I've been told on this forum so many, many times that you can't use Cinebench to compare between AMD and Intel because it's biased towards Intel.

So what's up with that? Do you think AMD designed Ryzen to do good in Cinebench?

Probably they want to show Zen is superior even in Shilltel biased benchmark.

 

[IEC]

Has anybody noticed that AMD is using Cinebench for their benchmark? I think that's weird, I mean I've been told on this forum so many, many times that you can't use Cinebench to compare between AMD and Intel because it's biased towards Intel.

So what's up with that? Do you think AMD designed Ryzen to do good in Cinebench?

AMD was known for underperforming in an area. What better way to announce a return to competitiveness than by demonstrating they now trade blows with the best (Intel) while offering a much better value proposition?

 

[Thunder 57]

Jesus Christ, that's brilliant!

I feel like I'm missing something funny about "Ryzen Master"?

 

[Crono]

I feel like I'm missing something funny about "Ryzen Master"?

Zen Master

 

[Conroe]

Ricin or raisin? I like raisins. Ricin not so much. I might have to get me some of these tasty raisins.

 

[Crono]

Ricin or raisin? I like raisins. Ricin not so much.

I substist on ricin beans.

 

[Allio]

AMD was known for underperforming in an area. What better way to announce a return to competitiveness than by demonstrating they now trade blows with the best (Intel) while offering a much better value proposition?

Reminds me a bit of the original Conroe demonstration. Best way to silence the critics is to flatter your opposition rather than minimising it. The fact that Intel overclocked the FX-60 they were benchmarking against, and still whooped it, has stuck with me even 10 years later.

 

[Doom2pro]

I would have preferred Zen Maester

 

[Crono]

I would have preferred Zen Maester

I would have preferred it if the 1800X came with Leighton Meester, but that's neither here nor there...

 

[Charlie22911]

Review kits are in the wild!
https://videocardz.com/66321/amd-ryzen-review-kit

Hopefully this hasn't been posted yet...

 

[Rifter]

I am disappointed about that as well because I really do not need performance at the moment but I sure do want to play with it. A 1100 or a 1300X would have been nice.

It usually wouldnt bother me as its become the norm to paper launch nowadays with stock being very hard to come by in the first weeks.

What irks me about it as they made the effort to come out and say they are launching with a full stack of SKU's on launch day and then after doing that still paper launch it anyways.

 

[vissarix]

Isnt strange that the i7 6900k used in the comparison with Ryzen by Amd have the lowest score ever recorded on any review?

is about 100 points lower, not a big deal but still thats a cheap move from Amd...

 

[thepaleobiker]

So here's the really important question that seems to have been missed:

Is it pronounced 'risen' or 'rye-zen'? I defaulted to the former but Lisa Su seems to disagree and I don't want to be making a fool out of myself in public (in the imaginary world where I have friends that actually care about the cpu market anyway).

Its the latter. Rye-Zen. A Rye that's become totally Zen. I agree, even the "Risen" pronunciation would have deep philosophical meaning in the context of AMD's rise back to relevance/dominance.

Regards,
Vish

 

[Sven_eng]

Isnt strange that the i7 6900k used in the comparison with Ryzen by Amd have the lowest score ever recorded on any review?

is about 100 points lower, not a big deal but still thats a cheap move from Amd...

What about the ST score, how does that measure up vs other reviews? Highest in any review?

 

[frozentundra123456]

It was created as an amd bashing thread by a troll that devotes countless hours of his life to hate amd.

I bet if I go look at that old thread I will see that Fritz benchmark that was proven fake almost immediately after they were posted.

The perpetrators mostly went to troll another forum, and it is a far better environment because of it. I imagine the shame will keep them away for awhile.

EDIT: Ohh look there it is...

PS: idk why you want to stir that pot though. Just let it go.

You are the one that needs to "let it go". Your post is the one filled with venom.

 

[MajinCry]

AMD was known for underperforming in an area. What better way to announce a return to competitiveness than by demonstrating they now trade blows with the best (Intel) while offering a much better value proposition?

Nope, the reason why is that Cinebench used the Intel Compiler, back when it would detect the AuthenticAMD and VIA VIA VIA cpuid string, which in turn made the program use crippled SSE2 code.

https://www.ftc.gov/sites/default/files/documents/cases/091216intelcmpt.pdf

Page 11, #64