Zen gets even with Broadwell-E in Blender

[gamerk2]

There's also another way of looking at it:

Broadwell and Zen scored about the same score when running at the same clock. Which is fine and dandy, until you remember that Zen has double the cores and Blender actually scales across all of them. So basic math dictates that in Blender, Zen has roughly HALF the IPC as Broadwell, given both finished in about the same time. Which doesn't bode well for gaming performance at all.

Now, if we take things farther and assume AMD got "creative" with some of the specific settings within Blender to bias performance...

 

[Chimpee]

There's also another way of looking at it:

Broadwell and Zen scored about the same score when running at the same clock. Which is fine and dandy, until you remember that Zen has double the cores and Blender actually scales across all of them. So basic math dictates that in Blender, Zen has roughly HALF the IPC as Broadwell, given both finished in about the same time. Which doesn't bode well for gaming performance at all.

Now, if we take things farther and assume AMD got "creative" with some of the specific settings within Blender to bias performance...

I thought both ran 8 cores 16 threads?

 

[Shadohh]

There's also another way of looking at it:

Broadwell and Zen scored about the same score when running at the same clock. Which is fine and dandy, until you remember that Zen has double the cores and Blender actually scales across all of them. So basic math dictates that in Blender, Zen has roughly HALF the IPC as Broadwell, given both finished in about the same time. Which doesn't bode well for gaming performance at all.

Now, if we take things farther and assume AMD got "creative" with some of the specific settings within Blender to bias performance...

I am not following this logic.

I am pretty sure both platforms are 8c/16t chips.

 

[gamerk2]

Leaked Geekbench Results: AMD Zen Performance Extrapolated | PC Perspective

Specifically:

AES numbers make sense, given AMD has dedicated resources on the CPU die. For all other tasks, you see performance 10-30% slower then IB at the same clocks.

If true (And that's a BIG if), Zen likely doesn't even match Sandy Bridge IPC levels.

 

[ccityinstaller]

Leaked Geekbench Results: AMD Zen Performance Extrapolated | PC Perspective

Specifically:

AES numbers make sense, given AMD has dedicated resources on the CPU die. For all other tasks, you see performance 10-30% slower then IB at the same clocks.

If true (And that's a BIG if), Zen likely doesn't even match Sandy Bridge IPC levels.

I just can't see this being true for release SKUs...AMD knows they cannot release a CPU that slow and have any chance of staying afloat..I just cannot believe that Lisa SU and company have their head buried that deep in the sand..I guess we will see in January...

 

[Marstg]

Broadwell and Zen scored about the same score when running at the same clock. Which is fine and dandy, until you remember that Zen has double the cores and Blender actually scales across all of them. So basic math dictates that in Blender, Zen has roughly HALF the IPC as Broadwell, given both finished in about the same time. Which doesn't bode well for gaming performance at all.

If you read the article I linked, it is clearly stated that the Zen chip ran against a 6900k which is Broadwell-E! Which, last time I checked was a 16t CPU Intel® Core™ i7-6900K Processor (20M Cache, up to 3.70 GHz) Specifications

 

[viscountalpha]

I just can't see this being true for release SKUs...AMD knows they cannot release a CPU that slow and have any chance of staying afloat..I just cannot believe that Lisa SU and company have their head buried that deep in the sand..I guess we will see in January...

After hearing her speak about the r460 and r470? I can totally believe she has her head in the sand that deep.

 

[noko]

While the test showing Zen beating Intel at the same speed would appear good - too much was left out to really conclude anything. The negative is why did they not show Zen running at 3.2ghz running the same speed Intel runs out of the box? Makes me think Zen is going to be around 3ghz and not even 3.2ghz! Still if running at 3ghz and those numbers roughly indicate threaded applications - it should blow away an I7-6700K which has half the number of cores of Zen. That does not mean it will perform well in other applications and we will just have to see how this turns out.

If AMD has a decent price and the performance above reflects some sort of broad truth/fact Zen should do well.

Now what I do not like are the very near sighted folks that will just concentrate on NEGATIVES and are blind to anything showing a success. In other words let the facts lay as they lay, financials, past shit I say has zero basis for determining if Zen is the processor for you or not. If I can get close to Broadwell rendering ability in Zen at half the cost - you can bet you can count me in having a few Zen's around.

Gaming wise if my FX systems so far have not had any significant negative effect on my games then I am sure Zen would be the same and most likely much better.

 

[NKD]

While the test showing Zen beating Intel at the same speed would appear good - too much was left out to really conclude anything. The negative is why did they not show Zen running at 3.2ghz running the same speed Intel runs out of the box? Makes me think Zen is going to be around 3ghz and not even 3.2ghz! Still if running at 3ghz and those numbers roughly indicate threaded applications - it should blow away an I7-6700K which has half the number of cores of Zen. That does not mean it will perform well in other applications and we will just have to see how this turns out.

If AMD has a decent price and the performance above reflects some sort of broad truth/fact Zen should do well.

Now what I do not like are the very near sighted folks that will just concentrate on NEGATIVES and are blind to anything showing a success. In other words let the facts lay as they lay, financials, past shit I say has zero basis for determining if Zen is the processor for you or not. If I can get close to Broadwell rendering ability in Zen at half the cost - you can bet you can count me in having a few Zen's around.

Gaming wise if my FX systems so far have not had any significant negative effect on my games then I am sure Zen would be the same and most likely much better.

Most people don't think this way. I am saying the same. If 8 core zen is decent at 3ghz. It can only get better from that point on. Further clock bumps and further IPC improvements over time. It is a good base to run on. I personally think that if its fricking 500 bucks and can trade blows with broadwell-e part few times and come with 20% of it overall. That is still a good processor for the price compared to competition going above $1000.

 

[Shintai]

I just can't see this being true for release SKUs...AMD knows they cannot release a CPU that slow and have any chance of staying afloat..I just cannot believe that Lisa SU and company have their head buried that deep in the sand..I guess we will see in January...

By that argument Bulldozer should never have been released.

Zen seems to build on the same principle. Moar cores to try and sell vastly inferior core performance.

Some people got extremely unrealistic expectations to AMD in both CPU and GPUs. Its easy to see the difference in R&D budgets. You just dont get something you didn't pay for. You may not get what you paid for either tho.

 

[JustReason]

By that argument Bulldozer should never have been released.

Zen seems to build on the same principle. Moar cores to try and sell vastly inferior core performance.

Some people got extremely unrealistic expectations to AMD in both CPU and GPUs. Its easy to see the difference in R&D budgets. You just dont get something you didn't pay for. You may not get what you paid for either tho.

Do you have any constructive criticism ever when you speak of AMD. Hell if AMD beat the living crap out of Intel you would still never buy them. On the positive side if they did at least I would never have to waste my time reading your incessant drivel anymore as you would have no choice but to run and hide after your Hate-everything-AMD-Crusade.

AND WHAT ARE THESE UNREALISTIC VIEWS POSTED HERE ON THIS FORUM? I haven't seen anyone posting anything unreasonable other than you.

 

[lolfail9001]

I haven't seen anyone posting anything unreasonable other than you.

Well, you should read more of your own posts, even if i agree that Shintai sometimes goes over the top.

 

[Shintai]

Do you have any constructive criticism ever when you speak of AMD. Hell if AMD beat the living crap out of Intel you would still never buy them. On the positive side if they did at least I would never have to waste my time reading your incessant drivel anymore as you would have no choice but to run and hide after your Hate-everything-AMD-Crusade.

AND WHAT ARE THESE UNREALISTIC VIEWS POSTED HERE ON THIS FORUM? I haven't seen anyone posting anything unreasonable other than you.

Its unrealistic when someone with a much smaller R&D budget is expected to compete in the top. Its like saying the next VIA CPU will rival the best from AMD and Intel. And if you dont agree, you are naturally a VIA hater and an AMD/Intel fanboy on a crusade, right? Because that is in short what you tell me. And we have seen the first couple of neutral benches.

What do you think R&D pay for? It not without reason that Nvidia beats AMD in the graphics segment as well.

Apples SoC performance isn't coming out of thin air either.

(Note quarter vs year with Apple.)

 

[JustReason]

Well, you should read more of your own posts, even if i agree that Shintai sometimes goes over the top.

I have stated for the record, if you will, Zen will not beat Intel. To have posters as asinine as some post that we think it will shows just how ignorant and intellectually limited they are.

As far as being competitive that requires criteria which most never give as a way to make their post seem rational.

 

[Pieter3dnow]

I have stated for the record, if you will, Zen will not beat Intel. To have posters as asinine as some post that we think it will shows just how ignorant and intellectually limited they are.

As far as being competitive that requires criteria which most never give as a way to make their post seem rational.

Most of "us" have stated that Zen would not beat Intel, the interesting thing is how close. And from there what is Zen+ doing that should be the one that might be able to "beat" Intel.

 

[lolfail9001]

Most of "us" have stated that Zen would not beat Intel, the interesting thing is how close. And from there what is Zen+ doing that should be the one that might be able to "beat" Intel.

Strictly speaking "how close" is unreasonable question, given AMD's position in recent years.

 

[Pieter3dnow]

Strictly speaking "how close" is unreasonable question, given AMD's position in recent years.

Were talking about Zen here not previous material because that would not reflect on realistic expectations. The rest is rather off topic and done to death already.

So the closer the better on retail products ....

 

[lolfail9001]

Were talking about Zen here not previous material because that would not reflect on realistic expectations. The rest is rather off topic and done to death already.

So the closer the better on retail products ....

Realistic expectations can only be based on reality, that's how it works.

Reality is... when was the last time AMD had competitive high end CPU, again?

Another reality: when was the last time relatively low RnD budget allowed to create high-end product in this sphere?

 

[JustReason]

Realistic expectations can only be based on reality, that's how it works.

Reality is... when was the last time AMD had competitive high end CPU, again?

Another reality: when was the last time relatively low RnD budget allowed to create high-end product in this sphere?

be careful. What you are stating is there has never been nor ever will be. Absolutes are dangerous. Also we have already stated it won't beat intel. I believe it will beat the construction cores and that is still good news. The interesting part is the growth potential, if it has one. Piledriver was definitely an improvement on bulldozer. So what matters is the initial landing of Zen.

 

[Chimpee]

I dont think any of us here are arguing AMD will beat Intel, just questioning how close they match Intel offering, whether it really matches Broadwell-E or Sandy Bridge depends on ones expectations. While it is fun to speculate, I rather just wait for actual review before jumping to conclusion.

 

[lolfail9001]

be careful. What you are stating is there has never been nor ever will be. Absolutes are dangerous. Also we have already stated it won't beat intel. I believe it will beat the construction cores and that is still good news. The interesting part is the growth potential, if it has one. Piledriver was definitely an improvement on bulldozer. So what matters is the initial landing of Zen.

Absolutes are fine when we are talking only about expectations. And right now, there is no realistic reason to expect anything than some improvement over Construction cores performance-wise and related metrics-wise. And that's about it.

 

[lolfail9001]

Which is based off those abstract Geekbench results that were interpolated by a select individual (or few individuals)? That doesn't make any sense considering that we aren't discussing CMT but a SMT based architecture.

No, it is based on entirely 0 leaks. It is realistic expectation, only dealing with known cold facts one cannot dispute. Based on those, expecting anything more than Westmere->Sandy Bridge jump, but with Excavator as baseline is overhyping, AMD simply ain't got resources.

 

[Gideon]

No one expected the Athlon to do as well as it did and scale as well as it did. I still miss that good old slot processor. Shintai the simple fact is Intel has to spend way more on R&D as they have many different things they are developing while AMD has a more narrow focus. Plus you have no idea on the breakdown on how Intel chose to spend the cash, how much went to Fab tech, gpu, cpu, memory, ssd, motherboard and so on and so on. Simple fact is all were doing is speculating on what little data we have scene and so far it looks good and since we dont work for Intel or AMD we have no clue much past that. In a few more months I think we will have a better picture on how Zen will fare.

 

[DrLobotomy]

If this chip isn't faster than my OC'ed i7-920 then no Zen for me. If it is 2x faster then I might think about it. Intel hasn't even done that so..........

 

[ManofGod]

Looking forward to Zen kicking some butt and meeting the realistic expectations that AMD fans have in this thread. Now, those Intel fans how are the AMD can do no good people, guess I will just do the Nelson point and laugh. Either way, I will probably not upgrade right away even though I know I would like too.

 

[tangoseal]

Zen will kick butt .... ftw!

 

[lolfail9001]

However it isn't based on any cold facts or evidence/arguments. It is based on conjecture which assumes that AMD will only make a 8% average IPC performance improvement if your example is to be believed. AMD doesn't need to beat Broadwell-E or Haswell-E if they can deliver 80-90% of the performance at lower prices. Intel can't expect to sell a 6950X for $1500+ while a 8 core Summit Ridge processor may be priced at 5820k-5930k price range.

It is based on cold facts in AMD's history ever since Penryn release and it's RnD budget.
Further, Westmere to Sandy was more like 20% jump, 30% in some cases even.
And btw, Intel can entirely expect to sell a 6950X for $1500+ if 8 core Summit Ridge is only as fast as 6800-6850k.

 

[darkeen]

It is based on cold facts in AMD's history ever since Penryn release and it's RnD budget.
Further, Westmere to Sandy was more like 20% jump, 30% in some cases even.
And btw, Intel can entirely expect to sell a 6950X for $1500+ if 8 core Summit Ridge is only as fast as 6800-6850k.

History does tend to repeat itself.... maybe we'll get a repeat of the athlon 64 days.

 

[Shadohh]

We should all hope for AMD to have a good product, competition drives cost/tech wars.

It is good the for the consumer, the companies participating and humankind.

 

[drescherjm]

maybe we'll get a repeat of the athlon 64 days

I don't see this happening soon. Remember when that happened when AMD had both better silicon and better design. Zen will certainly not have better silicon and I do not expect Zen to have a clearly better design either.

 

[drescherjm]

It doesn't need to be superior if it can compete in performance while offering similar or lower price.

Fully agreed.

 

[NKD]

It is based on cold facts in AMD's history ever since Penryn release and it's RnD budget.
Further, Westmere to Sandy was more like 20% jump, 30% in some cases even.
And btw, Intel can entirely expect to sell a 6950X for $1500+ if 8 core Summit Ridge is only as fast as 6800-6850k.

So you are basing zen on past amd history? Give me a 100% proof and guarantee or wait for reviews. Sounds like you are trying to hard. Too hard to wait for real benchmarks or you are going to go by cold hard facts in amd recent history? Yea I am not expecting to be overwhelmed but you are too concentrated on proving a product to be a failure when you have no real hard benchmarks of an actual product running at whatever retail spees will be. Its not hard for me to say oh AMD might suck, sure they could given past history. But I am not going to say about a product I have seen no real benchmarks. If AMD gets with 20% of Broadwell-E on average and gives it to you for 500 bucks vs 1500? You really going to compalin? I know I am not.

 

[NKD]

We should all hope for AMD to have a good product, competition drives cost/tech wars.

It is good the for the consumer, the companies participating and humankind.

Don't tell that to a fanboy it won't make sense. Some people are hell bent on hoping zen fails.

 

[lolfail9001]

So you are basing zen on past amd history? Give me a 100% proof and guarantee or wait for reviews. Sounds like you are trying to hard. Too hard to wait for real benchmarks or you are going to go by cold hard facts in amd recent history? Yea I am not expecting to be overwhelmed but you are too concentrated on proving a product to be a failure when you have no real hard benchmarks of an actual product running at whatever retail spees will be. Its not hard for me to say oh AMD might suck, sure they could given past history. But I am not going to say about a product I have seen no real benchmarks. If AMD gets with 20% of Broadwell-E on average and gives it to you for 500 bucks vs 1500? You really going to compalin? I know I am not.

Considering that my post was exclusively about setting expectations for Zen, your ramblings are dismissed. I have seen this picture with Polaris already, and i was never optimistic, so go figure. Btw, do you remember that Broadwell-E has $400 (okay, $450 ~_~) SKU? That's the target Zen needs to beat by a noticeable margin to compete with Broadwell-E.

 

[Shintai]

Don't tell that to a fanboy it won't make sense. Some people are hell bent on hoping zen fails.

You can turn it around as well. Some people are hell bent on hoping Zen succeeds.

But it comes down to economics. And the R&D budgets are what they are. You could just as well play the lottery and hope to win the big prize.

Polaris was also going to e so amazing. And when Polaris 10 failed, Raja on video interview said that Polaris 11 was the real one. A chip that funny enough ended up even worse.

AMD also played the stock card. They dont have high hopes either.

 

[Spirit_Retro]

You can turn it around as well. Some people are hell bent on hoping Zen succeeds.

But it comes down to economics. And the R&D budgets are what they are. You could just as well play the lottery and hope to win the big prize.

Polaris was also going to e so amazing. And when Polaris 10 failed, Raja on video interview said that Polaris 11 was the real one. A chip that funny enough ended up even worse.

AMD also played the stock card. They dont have high hopes either.

Phhht... I'm hell bent on pizza.

 

[Pieter3dnow]

You can turn it around as well. Some people are hell bent on hoping Zen succeeds.
But it comes down to economics. And the R&D budgets are what they are. You could just as well play the lottery and hope to win the big prize.Polaris was also going to e so amazing. And when Polaris 10 failed, Raja on video interview said that Polaris 11 was the real one. A chip that funny enough ended up even worse.AMD also played the stock card. They dont have high hopes either.

You have a relevant point for Zen or you just keep swinging ?

As it is now Zen is not a retail product until then you can come up with more crazy stuff then usual.

 

[NKD]

You can turn it around as well. Some people are hell bent on hoping Zen succeeds.

But it comes down to economics. And the R&D budgets are what they are. You could just as well play the lottery and hope to win the big prize.

Polaris was also going to e so amazing. And when Polaris 10 failed, Raja on video interview said that Polaris 11 was the real one. A chip that funny enough ended up even worse.

AMD also played the stock card. They dont have high hopes either.

When it's out and it sucks then it's reality. Until then whatever makes you feel better and sleep good at night cuz zen is not in retail. Polaris failed? Seems like a fine product to me for the price. Like I said you may be right when zen is out just cuz you hate amd you can't predict a products failure without having a retail sample. Could it fail? Yea! But I don't know cuz I haven't seen any reviews yet.

Funny you quoted me when I hadn't even quoted you lol. I guess you may be one of those people hell bent on hoping zen fails lol.

 

[Shintai]

When it's out and it sucks then it's reality. Until then whatever makes you feel better and sleep good at night cuz zen is not in retail. Polaris failed? Seems like a fine product to me for the price. Like I said you may be right when zen is out just cuz you hate amd you can't predict a products failure without having a retail sample. Could it fail? Yea! But I don't know cuz I haven't seen any reviews yet.

Funny you quoted me when I hadn't even quoted you lol.

Why dont you tell me why Zen will be good instead of playing the hate card for the nth time and act like the victim? What miracles will AMD do on their shoehorn R&D budget to make it better? They even got less R&D for it than Bulldozer had. Why did AMD issue stocks now, if Zen is going to be this great competitive product that will just reap in revenue in half a year? Why not show Zen left and right if it is so good when there is nothing to lose but everything to win from it? How is 1 AMD engineer going to compete with ~10 Intel engineers that uses a much better node and got the IDM benefit?

There is a reality to stick to, even when you root for the underdog with all your desires. Just as well as VIA isn't going to make a competitive higher end x86 product anymore either.

 

[razor1]

When it's out and it sucks then it's reality. Until then whatever makes you feel better and sleep good at night cuz zen is not in retail. Polaris failed? Seems like a fine product to me for the price. Like I said you may be right when zen is out just cuz you hate amd you can't predict a products failure without having a retail sample. Could it fail? Yea! But I don't know cuz I haven't seen any reviews yet.

Funny you quoted me when I hadn't even quoted you lol. I guess you may be one of those people hell bent on hoping zen fails lol.

We will know if Polaris did well when the Q3 numbers come out, I don't think Polaris is selling that well (it might be due to supply issues, but there is too much circumstantial evidence from newegg, amazon and the euro retailer that says otherwise. Its not these online retailers are colluding to show nV is out selling AMD .

Why dont you tell me why Zen will be good instead of playing the hate card for the nth time and act like the victim? What miracles will AMD do on their shoehorn R&D budget to make it better? They even got less R&D for it than Bulldozer had. Why did AMD issue stocks now, if Zen is going to be this great competitive product that will just reap in revenue in half a year?

There is a reality to stick to. Just as well as VIA isn't going to make a competitive higher end x86 product anymore either.

Zen doesn't need to be as fast as top end Xeons to be successful, as long as it gets to Ivy Bridge levels margin improvements will pull AMD into the black. The foundation of Zen if its good enough to do that, then they will stabilize future versions as well. Of course we still don't know what Intel will do, but that is too far ahead to really get any clear picture of that.

And NKD, this is all speculation so far, but the problem with AMD is from what I have seen, they themselves never made a great processor, they had to buy out other companies and those other companies created their next generation "great" processor, the likes of Cyrix and Nexgen, those two companies were he building blocks of AMD's great successes. Those engineers have long left AMD, one of them was Jim Keller (what about all the others before Jim, he wasn't the lead for most of AMD CPU's, just K8 and for that only 1 year in the beginning. For all that work for 1 year he gets that much hype? Hmm sounds a bit weird don't you think?

All the while Intel just took a bad direction with Pentium 4, they had the ground works for core architecture with PIII, they just took the wrong step and miscalculated that Moore's law would stop working for them. So by No means will I count Intel out, they have the resources and the talent to easily match AMD when it comes to CPU.

Zen if it reaches Ivy bridge levels or Haswell, AMD will have a 3 year gap where they can make profits, but that is only if Intel sits idle!, AMD has already shown their cards when it comes to future projections of Zen so Intel can plan ahead for their next CPU's, Cannon Lake and make sure AMD doesn't get anywhere. So that cuts that three years down to a year and half. This is all with only margins increases and slight marketshare uptake in desktop and notebook.

We can't really expect much more from AMD in the next year and half, unless they come out with a Kaby Lake killer, which is highly unlikely.