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

[itsmydamnation]

That's pretty disappointing, if true. We had been told before that it was the lower-end chips shipping in March, not the entire set. I guess if they have SMT bugs they haven't been able to sort out, they need more delays . . .



I sincerely wish they would keep to their earlier timeline of selling their high-end chip in January! How are they going to have a Ryzen giveaway on Jan 18th when the chips don't even ship until March? Do you get a voucher for a future chip or what?

SMT and Uop cache issues on the A0 CPC tested or SMT and uop cache issues on the newer chips, remember they were using an old ES for testing. We also have no idea what issue was, it might not be a issue at all but rather a not at the time of A0 100% implement feature (eg micro code engine or something not handling SMT correctly)

You think Zen would have same IPC/power for blender/x264 as BW-E with either a non working SMT or non working uop cache? I get the feeling so much "fake news" is going around at the moment, just look at the available data and ocam's razor that S***.....

 

[Dresdenboy]

Anyone know if Zen will support transactional memory?

https://en.wikipedia.org/wiki/Advanced_Synchronization_Facility

I haven't seen a related ISA extension patch yet.

 

[DrMrLordX]

SMT and Uop cache issues on the A0 CPC tested or SMT and uop cache issues on the newer chips, remember they were using an old ES for testing. We also have no idea what issue was, it might not be a issue at all but rather a not at the time of A0 100% implement feature (eg micro code engine or something not handling SMT correctly)

You think Zen would have same IPC/power for blender/x264 as BW-E with either a non working SMT or non working uop cache? I get the feeling so much "fake news" is going around at the moment, just look at the available data and ocam's razor that S***.....

I was sort of thinking the same thing, it's just no fun watching digitimes call for a March release for everything, including the high-end chip. I still think they may be a bit mistaken there and that we may see X370 + BE Summit Ridge in January, with B350 + everything else in March.

That A0 ES was supposed to be from June, surely they've done another spin since then.

To me it isn't about FUD or not FUD, it's about whether or not I can buy the thing in January or March. How much longer do I have to wait AMD?

 

[Abwx]

I haven't seen a related ISA extension patch yet.

At the beginning of the article, chapter :

Le jeu d'instruction

http://www.hardware.fr/news/14758/amd-detaille-architecture-zen.html

google translate should be good enough.

 

[Arachnotronic]

I was sort of thinking the same thing, it's just no fun watching digitimes call for a March release for everything, including the high-end chip. I still think they may be a bit mistaken there and that we may see X370 + BE Summit Ridge in January, with B350 + everything else in March.

That A0 ES was supposed to be from June, surely they've done another spin since then.

To me it isn't about FUD or not FUD, it's about whether or not I can buy the thing in January or March. How much longer do I have to wait AMD?

DigiTimes says March 2017. Only a few more months now, not the end of the world.

 

[DrMrLordX]

DigiTimes says March 2017. Only a few more months now, not the end of the world.

Think of it this way: had AMD launched diy Bristol Ridge in October as they very well could have, I'd have bought one of those instead. So now I'm off-track waiting for Summit Ridge instead. It's that or wait for Raven Ridge.

Christmas in January is one thing, but March? Bleargh.

 

[farosis]

In the meantime Skylake-E is not far away after all these delays. They have to deal with Skylake-E for the most part during the lifetime of Zen v1, the window where it could compete against Broadwell-E becomes smaller and smaller.

Actually AMD got a lot of time to make up a good plan for HEDT too, don't be naive.

 

[biostud]

So probably around Broadwell-E performance. Then pricing will be the key selling point.

 

[krumme]

So probably around Broadwell-E performance. Then pricing will be the key selling point.

I dont think its that important.
If amd for once would like to build brand what they want to avoid is beeing seen as the cheap alternative. Compared to hedt line that could easily happen.
But it they price their 4 6 8c zen line aprox 50-100 usd higher than the core i3 i5 i7 They would be in a situation where they are not seen as cheap. Because they are faster and more expensive. They are the premium product.
The product looks so fine it seems you can use a little short term profit in exchange for a lot much needed brand value.

 

[DrMrLordX]

But it they price their 4 6 8c zen line aprox 50-100 usd higher than the core i3 i5 i7 They would be in a situation where they are not seen as cheap.

They can get away with that, yes. If they charge $100 more for their top-of-the-line 8c/16t product than the i7-7700k for example, that would put them at ~$500, which is reasonable.

Still nowhere near $1k+ for a 6900k. I do not think they'll do very well selling Summit Ridge for Broadwell-E prices.

 

[krumme]

Going along the lines of building brand; What they want to combat intel hedt line is something that have same functionality and is just more premium. That could be just a less efficient server bin of 4 ccx zen giving 16 cores on quad memory. Then price it 200 usd above top end hedt.
Then make some oem bundles with some huge prof vega. Its all about bragging rights selling the 350 usd segment cpu.

 

[krumme]

They can get away with that, yes. If they charge $100 more for their top-of-the-line 8c/16t product than the i7-7700k for example, that would put them at ~$500, which is reasonable.

Still nowhere near $1k+ for a 6900k. I do not think they'll do very well selling Summit Ridge for Broadwell-E prices.

Yep. If their naming sr3 sr5 sr7 is true its pretty obvious what they are trying to do segment wise. It cant be anymore obvious. And looking at perf it seems to me they can price it higher than the eqivalent core line.
Pretty smart imo. But its a good deal easier to be smart if you have the right products

 

[biostud]

But if they perform like broadwell and cost more, why would anyone buy them?

I'm not saying they should be super cheap but if the 8c zen is priced like the 6800k and the 6c like the 7700k, then it would be a viable alternative to intel.

 

[KTE]

I always dislike how reviews have to recommend a new product/or not mention a recommendation at all, even tho you may be able to buy older products that are now far cheaper and hence a better performance/value proposition.

For instance, I would rather pay 500 for Core i7-5960x today rather than any of the 7x00/6xx0 series chips (I've just bought one on sale for slightly less).

They can get away with that, yes. If they charge $100 more for their top-of-the-line 8c/16t product than the i7-7700k for example, that would put them at ~$500, which is reasonable.

Still nowhere near $1k+ for a 6900k. I do not think they'll do very well selling Summit Ridge for Broadwell-E prices.

Sent from HTC 10
(Opinions are own)

 

[Sweepr]

www.3dcenter.org/news/blockdiagramm-zu-amds-zen-prozessorenkernen-aufgetaucht

 

[zinfamous]

That's pretty disappointing, if true. We had been told before that it was the lower-end chips shipping in March, not the entire set. I guess if they have SMT bugs they haven't been able to sort out, they need more delays . . .



I sincerely wish they would keep to their earlier timeline of selling their high-end chip in January! How are they going to have a Ryzen giveaway on Jan 18th when the chips don't even ship until March? Do you get a voucher for a future chip or what?

How long did we hold onto those free "Half Life 2!" vouchers with GPU bundles before the game was actually released? 1, 1.5 years? I know it's not the same thing, but there have been such delays with free vouchers for hardware in the past, right?

 

[Doom2pro]

Really does anyone here think that Broadwell is not over priced? Does anyone really believe AMD is going to price Ryzen near overpriced Broadwell simply because it's in the same performance class?

The only reason Intel got away with over pricing Broadwell (as well as others) is due to lack of competition... AMD doesn't have that situation, and seems to me neither does Intel anymore. If AMD prices Ryzen near Broadwell's current MSRP, they are just asking for Intel to slash prices dramatically leaving AMD holding the bag.

Seems the only way this comes out good for AMD is if they price Ryzen where Broadwell SHOULD have been priced had it actually had competition, possibly a bit lower to protect against Intel price cuts..

With the new WSA and Samsung's Fabs, AMD is going to be baking quite a lot of these babies so production shouldn't be an issue.

 

[Doom2pro]

Btw, does anyone have any idea if Node knowledge learned at GF's 14LPP flows to Samsung and vice versa? Do the minor differences between the two prevent this? If not, could this possibly be in the licensing deal between them? Thanks.

 

[jpiniero]

Really does anyone here think that Broadwell is not over priced? Does anyone really believe AMD is going to price Ryzen near overpriced Broadwell simply because it's in the same performance class?

May as well get what you can from the early adopters before Skylake-X hits.

Edit: Still think the 8 core is going to be $700-$800. The 6 and 4 depend on clocks more but I could see $400 and $300.

 

[sm625]

Really does anyone here think that Broadwell is not over priced? Does anyone really believe AMD is going to price Ryzen near overpriced Broadwell simply because it's in the same performance class?

There is very little evidence that it is in the same performance class. Blender by itself isnt enough. What we do have is a gaming benchmark from CPC that shows a 6700K beating Zen by an utterly unreachable 20% across 8 games. That is a complete stomp. If that is how this plays out then AMD has themselves a $400 chip at most. Nobody is going to give up 20% of their fps just to transcode and render faster. Otherwise, there wouldnt be thousands of 12 core systems on ebay for $500 each. What sells will be what produces top gaming fps. AMD has to win at gaming. AMD has to be, AMD MUST be at the very top of the gaming charts in order to break its way back into the HEDT and server markets. We can argue that the workloads are different and yada yada bla bla but its just the way it is. HEDT wouldnt exist if everyone had to lose 10-20% fps just to game on them. And AMD isnt producing any data to suggest they have captured the gaming crown.

 

[William Gaatjes]

I have a strong hunch that the i3 competitors from AMD will be a quad core ryzen with both smt enabled and disabled. The i5 competitor will be a six core ryzen with smt enabled. And the i7 competitor will be the full eight core model with smt enabled.

 

[Doom2pro]

There is very little evidence that it is in the same performance class. Blender by itself isnt enough. What we do have is a gaming benchmark from CPC that shows a 6700K beating Zen by an utterly unreachable 20% across 8 games. That is a complete stomp. If that is how this plays out then AMD has themselves a $400 chip at most. Nobody is going to give up 20% of their fps just to transcode and render faster. Otherwise, there wouldnt be thousands of 12 core systems on ebay for $500 each. What sells will be what produces top gaming fps. AMD has to win at gaming. AMD has to be, AMD MUST be at the very top of the gaming charts in order to break its way back into the HEDT and server markets. We can argue that the workloads are different and yada yada bla bla but its just the way it is. HEDT wouldnt exist if everyone had to lose 10-20% fps just to game on them. And AMD isnt producing any data to suggest they have captured the gaming crown.

6700K is a higher clocked Quad Core, if Zen is at Broadwell E class, who's to say a Quad Core version wouldn't clock higher and make up this "gap"? Besides you are ignoring the lower clocks and bugs in the early silicon this benchmarked 20% gap you are going off of.

The problem with large core count SKUs are their shared package TDP, all those cores have to fit in the same package TDP, so clocking them all at a high base clock without having an insane TDP is very hard.

This is why Intel quad cores are kings at gaming, they have around half the cores of 8 and 10c SKUs and they have the same TDP so they clock much higher and with most games having issues utilizing more than 4 threads, it only increases this effect.

You can't compare an 8c/16t ENGINEERING sample to a Production Ready QUAD CORE 6700K. Wait for the 4c/8t Ryzen benchmarks to make any determination.

 

[zinfamous]

There is very little evidence that it is in the same performance class. Blender by itself isnt enough. What we do have is a gaming benchmark from CPC that shows a 6700K beating Zen by an utterly unreachable 20% across 8 games. That is a complete stomp. If that is how this plays out then AMD has themselves a $400 chip at most. Nobody is going to give up 20% of their fps just to transcode and render faster. Otherwise, there wouldnt be thousands of 12 core systems on ebay for $500 each. What sells will be what produces top gaming fps. AMD has to win at gaming. AMD has to be, AMD MUST be at the very top of the gaming charts in order to break its way back into the HEDT and server markets. We can argue that the workloads are different and yada yada bla bla but its just the way it is. HEDT wouldnt exist if everyone had to lose 10-20% fps just to game on them. And AMD isnt producing any data to suggest they have captured the gaming crown.

I haven't yet seen any benchmarks of the 4C or 6C Zen chips, so I'm not sure what you can assume comparing the 8C Ryzen and its clocks to the 4C 6700K and game performance? Or am I missing something?

 

[Toettoetdaan]

There is very little evidence that it is in the same performance class. Blender by itself isnt enough. What we do have is a gaming benchmark from CPC that shows a 6700K beating Zen by an utterly unreachable 20% across 8 games. That is a complete stomp. If that is how this plays out then AMD has themselves a $400 chip at most. Nobody is going to give up 20% of their fps just to transcode and render faster. Otherwise, there wouldnt be thousands of 12 core systems on ebay for $500 each. What sells will be what produces top gaming fps. AMD has to win at gaming. AMD has to be, AMD MUST be at the very top of the gaming charts in order to break its way back into the HEDT and server markets. We can argue that the workloads are different and yada yada bla bla but its just the way it is. HEDT wouldnt exist if everyone had to lose 10-20% fps just to game on them. And AMD isnt producing any data to suggest they have captured the gaming crown.

That is indeed what people will see, that an octa-core Zen isn't a match for a quad i7. But if you look at the 6900, at the same clocks, the 6900 would be just 4% faster in that collection of games.

I guess switching to AMD with just a 4% difference can be justified with a nice discount.

 

[leoneazzurro]

N

There is very little evidence that it is in the same performance class. Blender by itself isnt enough. What we do have is a gaming benchmark from CPC that shows a 6700K beating Zen by an utterly unreachable 20% across 8 games. That is a complete stomp. If that is how this plays out then AMD has themselves a $400 chip at most. Nobody is going to give up 20% of their fps just to transcode and render faster. Otherwise, there wouldnt be thousands of 12 core systems on ebay for $500 each. What sells will be what produces top gaming fps. AMD has to win at gaming. AMD has to be, AMD MUST be at the very top of the gaming charts in order to break its way back into the HEDT and server markets. We can argue that the workloads are different and yada yada bla bla but its just the way it is. HEDT wouldnt exist if everyone had to lose 10-20% fps just to game on them. And AMD isnt producing any data to suggest they have captured the gaming crown.

I think you have understood very little of the benchmarks published. Now, take a look again at the sample frequencies, at the fact that it is an early sample and that the announced frequencies of Zen are mre than a bit higher, with a turbo going surely more than 3.4 GHz. Then take a look at the Broadwell-E results, because you cannot compare the number of cores with a 6700K or 7700K. And finally, don't think everyone has you priorities, many people would give 20% of low resolution CPU power (that could translate in less than 10% in more GPU limited situations, which are the "real life" situations).