AMD Ryzen (Summit Ridge) Benchmarks Thread (use new thread)

[.vodka]

He thought they were pitching Zen vs. 6700K.

Oh, that makes more sense.

Well, they did put a 4.5GHz 6700k vs Zen at one point.

 

[bjt2]

It seems 14lpp is not as Bad as expected

It seems 2. 8ghz is not base frequency and ipc is not between Sandy and Ivy...

The process is what it is. Zen has lower FO4 than INTEL's chip. This is the reason for higher clocks and lower power draw. I was saying this from weeks...

 

[krumme]

Lol it will be $1200 a pop, no wonder she's smiling.

She have been looking confident all of 2016. Like - tape out shows IPC is exceeded - confident.
Now she is more looking like jhh on stage. No wonder with a 3.4+ base to develop from. Its the smile of the hardest part beeing over.
Man what a mountain to climb with amd from where it was 3 years ago.

 

[Dresdenboy]

You still can, the market won't respond to this until tomorrow at the soonest. Then you can ride it up as the company continues to regain market share.

There is after hours and premarket trading (roughly matches AMD trading in Europe and stock futures). This year I often saw a lot happening during these phases, not only after ERs, but also after many other announcements scheduled at 4 p.m. ET or later.

But there are other funny effects, which look like the price is being driven down during the first hour, so that everyone could get into the boat. Good luck! But if this stock starts to run up tomorrow, it won't stop after a few cents anyway.

 

[witeken]

Nice, 95W TDP vs 140W TDP , both at 14nm and RYZEN is faster at least in those two scenarios.

TDP and process node are the two most meaningless numbers in all of the technology industry. So good job AMD... I guess.

 

[zinfamous]

TDP and process node are the two most meaningless numbers in all of the technology industry. So good job AMD... I guess.

well, if the rest of the presentation is true, they seem to be ahead in all of the other numbers in the industry so....well, huh.

 

[ElFenix]

3.4GHz base, impressive.

i've been figuring at 3.2 this should be fairly competitive for threaded workloads, 3.4 is some nice gravy. if it can ramp bits up into the 4+ range for lightly threaded workloads then this will be the biggest comeback since, well, trump beat clinton .

 

[sirmo]

There is after hours and premarket trading (roughly matches AMD trading in Europe and stock futures). This year I often saw a lot happening during these phases, not only after ERs, but also after many other announcements scheduled at 4 p.m. ET or later.

But there are other funny effects, which look like the price is being driven down during the first hour, so that everyone could get into the boat. Good luck! But if this stock starts to run up tomorrow, it won't stop after a few cents anyway.

You also have to account for the mountain of "analytical" traders who don't really know anything about tech or AMD, who just blindly "buy on the rumor, sell on the news". For anyone long in this stock this presentation looks really promising. If the Ryzen really has that big of an edge in perf/watt.. this is going to be great for AMD's bottom line.

 

[DidelisDiskas]

I'm cautiously excited about the potential performance, though not so much about the price premium that these cpu's might end up having seeing as intel left a lot of space there.

 

[witeken]

well, if the rest of the presentation is true, they seem to be ahead in all of the other numbers in the industry so....well, huh.

Suddenly people forgetting AMD's 40% IPC claim.

From the comparison's I have seen on this forum, 40% only lands you in Ivy Bridge territory.

 

[zinfamous]

i've been figuring at 3.2 this should be fairly competitive for threaded workloads, 3.4 is some nice gravy. if it can ramp bits up into the 4+ range for lightly threaded workloads then this will be the biggest comeback since, well, trump beat clinton .

this is like the band is on the field and spiking the ball into the trombone player's chest kind of a comeback.

 

[sirmo]

Suddenly people forgetting AMD's 40% IPC claim.

From the comparison's I have seen on this forum, 40% only lands you in Ivy Bridge territory.

In the presentation she said 40% was only the design goal, but that they have exceeded it.

 

[Doom2pro]

Suddenly people forgetting AMD's 40% IPC claim.

From the comparison's I have seen on this forum, 40% only lands you in Ivy Bridge territory.

Umm did you watch the event? Lisa clearly said they EXCEEDED 40%, straight from the horses mouth this time.

 

[bjt2]

Suddenly people forgetting AMD's 40% IPC claim.

From the comparison's I have seen on this forum, 40% only lands you in Ivy Bridge territory.

Lisa said they have exceeded the expectation...

 

[zinfamous]

Suddenly people forgetting AMD's 40% IPC claim.

From the comparison's I have seen on this forum, 40% only lands you in Ivy Bridge territory.

the presentation clearly showed greater than that, though. And they repeatedly claimed that they have beaten their 40% claims.

I'm not sure if you are responding to what is actually being presented, or just repeating things you assume are true?

Now...there is clearly more to be divulged here, but AMD has already made their data available and there are some pretty substantial, not-entirely-vague performance numbers to look at. So this strikes me as new territory compared to their previous previews.

 

[TheELF]

the presentation clearly showed greater than that, though. And they repeatedly claimed that they have beaten their 40% claims.

And brought it up to what, 100% increase?More?
Multithreaded tests without knowing how their SMT works are useless for figuring out the IPC.

it's right up there with the 4k gaming benchmarks where even the fx-9590 matches intel HEDT cpus.

 

[Abwx]

Hardware.fr article state that it s confirmed that there will be SKUs with base frequency higher than 3.4GHz, that mean at least 3.5 if not 3.6..

plus de 3.4 GHz en fréquence de base

More than 3.4GHz base frequency.

la seule information officielle est que le fer de lance aura une fréquence de base supérieure à 3.4 GHz.

The only official info is that the top SKU will have a base frequency higher than 3.4GHz.

http://www.hardware.fr/news/14880/ryzen-petite-partie-zen-devoilee.html

 

[witeken]

In the presentation she said 40% was only the design goal, but that they have exceeded it.

Umm did you watch the event? Lisa clearly said they EXCEEDED 40%, straight from the horses mouth this time.

Lisa said they have exceeded the expectation...

the presentation clearly showed greater than that, though. And they repeatedly claimed that they have beaten their 40% claims.

I'm not sure if you are responding to what is actually being presented, or just repeating things you assume are true?

Now...there is clearly more to be divulged here, but AMD has already made their data available and there are some pretty substantial, not-entirely-vague performance numbers to look at. So this strikes me as new territory compared to their previous previews.

No, peeps, I have not watched AMD's Zen presentation. I will just wait for the reviews to see where this ship strands, instead of being fed this marketing. I have watched the slides, though, and it wasn't there. One slide about the efficiency said something like "for entertainment purposes only".

Technically, 40.1% also exceeds 40%, mathematically speaking.

But thanks for your clarification.

 

[inf64]

It is evident that they have exceeded the target goal as Zen is performing at Broadwell-E level or better in two very different and both CPU heavy workloads which is outstanding. Even if we assume that Zen has 10% more efficient SMT engine (due to XYZ reasons) it would still imply that ST IPC is roughly in the same ballpark as Broadwell or Haswell. Which is in line with Lisa's comments on exceeding the IPC goals.

I'm impressed with the clocks they have achieved. I wrote previously that it is paramount for AMD to land in 3.3-3.7Ghz base-turbo range and this is exactly what they are saying they managed to do. I always assumed they will meet/exceed IPC goals since on paper Zen is 2x the XV core in almost every regard so it was basically a given. Adaptive Turbo and 95W TDP targets with those clocks they mention is simply outstanding piece of engineering work given how limited AMD is R&D wise (Vs intel).

 

[krumme]

Its not a K7 moment but its pretty darn close. It doesnt have the brute fpu force advantage k7 had over P5 - on the contrary one can say - but we get efficiency and especially perf/mm2 in spades.
This cpu will even save a few heads in mubadala they get value indirect via stock and especially because this looks like a savior for GF.

 

[zinfamous]

No, peeps, I have not watched AMD's Zen presentation. I will just wait for the reviews to see where this ship strands, instead of being fed this marketing. I have watched the slides, though, and it wasn't there. One slide about the efficiency said something like "for entertainment purposes only".

Technically, 40.1% also exceeds 40%, mathematically speaking.

But thanks for your clarification.

so, trolling basically.

k.

What part of "AMD has made their data immediately available" don't you understand? This wasn't a slide presentation, these were demos with detailed hardware specs and numbers. Not everything revealed obviously, but this was nothing like the marketing slides that you want to think they are.

You know, you could also read Ian's article on the front page about this, where this information is also detailed for you, if that is a trusted source more to your liking. This is based on the earlier demo and presentation given to reviewers ahead of this press event.

Again, I said this is new territory compared to what I and obviously YOU have come to expect from these events--especially with AMD--this is solid data. You do yourself a disservice by wearing your troll hat on your face with these content-less posts.

 

[.vodka]

No, peeps, I have not watched AMD's Zen presentation. I will just wait for the reviews to see where this ship strands, instead of being fed this marketing. I have watched the slides, though, and it wasn't there. One slide about the efficiency said something like "for entertainment purposes only".

Technically, 40.1% also exceeds 40%, mathematically speaking.

But thanks for your clarification.

Of course we all wait for actual benchmarks, but you should know it wasn't smoke and mirrors. Go watch the recorded stream.

They had live systems on a desk, not a powerpoint slide stating certain numbers, or a video of AOTS like in the Polaris reveal claiming stuff vs a 1080.

Different kind of show. Marketing, yeah, yet showing something tangible.



There's the Blender file ready for download... anyone can go try the livestream benchmark on an off the shelf stock, turbo enabled 6900k and see if they get the same results AMD got on stage.

 

[happy medium]

I think its great that Rizen performs well.

But I'm guessing they are gonna make money in the OEM bracket ?
I would think those individual customers that wanted 6 or 8 cores have had 6/8 cores for quite a while now. If performance is about equal between the 6/8 core Intel/AMD cpu's, what is the reason to switch to a Risen chip now?
I mean besides the holdout people in the " will only buy AMD" crowd?

I watched the video, I didnt see anything specific that stood out and said to me WOW that's new and much better than what I could have purchased for the past 2 years. Better IPC? no, higher clocks? no.

but overall , a good showing.

 

[F-Rex]

Its not a K7 moment but its pretty darn close. It doesnt have the brute fpu force advantage k7 had over P5 - on the contrary one can say - but we get efficiency and especially perf/mm2 in spades.
This cpu will even save a few heads in mubadala they get value indirect via stock and especially because this looks like a savior for GF.

K7 was not that better than p3. At least at first. It was its hability to climb the frequency scale which eventualy allowed k7 to succeed.

We still don't know turbob frequencies.

At least we know it's not3.2 GHz

 

[bjt2]

Hardware.fr article state that it s confirmed that there will be SKUs with base frequency higher than 3.4GHz, that mean at least 3.5 if not 3.6..



More than 3.4GHz base frequency.



The only official info is that the top SKU will have a base frequency higher than 3.4GHz.

http://www.hardware.fr/news/14880/ryzen-petite-partie-zen-devoilee.html

And now who is crazy?