AMD Zen Performance Preview

[FrozenSun]

So its no faster than a Skylake I5 in games. Not sure if I believe that yet.

It doesn't seem that surprising. Most games don't use a ton of threads and Skylake almost certainly is faster in lightly threaded loads due to much higher clocks and likely higher IPC.

 

[FrozenSun]

Take with a grain of salt for now. I dont think the article is fake, but as always. More tests, more reviews etc.

Seems like pretty decent results. Was a bit better than the 6800k at heavily threaded task and a bit worse at lightly threaded tasks. Fits pretty well in line with the previous leaks of a 350-500 price tag. I almost wonder if that was the lower of the two 8c SKUs, given the clock speeds. I know AMD is known for smoke and mirrors, but they flat out said the top end one would be 3.4 baseline and higher.

At any rate, it seems likely at this point AMD finally has a desktop processor they can sell for 300+ for the first time in a long, long, long time.

 

[Rvenger]

Why is it only running at 3.15 GHz? If the base clock without boost is supposely 3.4? I doubt this is retail. Either way, looks pretty impressive for 3.15 GHz against higher clocked Intel parts, especially on games without much multi threading.

I keep forgetting that the games that only require 4 threads, the mainstream CPUs still have around the same Single threaded IPC as the enthusiast ones. It's certainly putting a whomping on the FX-8370 so there's that. Its a massive upgrade for AMD users with Phenom and BD/PD!

 

[SighTurtle]

I wonder if Global Foundries is ready to take on the orders without falling behind like on Polaris.

 

[OrangeKhrush]

It looks like an early ES chip, running 3.1Ghz is extremely low clocks yet it does as expected. with higher base clocks and XFR variable turbo, it should really deliver comparible performance to an i7 5960X. I think Zen is very comparible to Haswell on the performance front and that is a very solid showing.

Pricing will matter. I THink the top end SKU should be sold around 500USD with a very punchy entry level 4/8 around or below 200USD.

 

[razor1]

It looks like an early ES chip, running 3.1Ghz is extremely low clocks yet it does as expected. with higher base clocks and XFR variable turbo, it should really deliver comparible performance to an i7 5960X. I think Zen is very comparible to Haswell on the performance front and that is a very solid showing.

Pricing will matter. I THink the top end SKU should be sold around 500USD with a very punchy entry level 4/8 around or below 200USD.

Pricing will be very telling, no way will AMD price it lower unless they have to pretty simple, being the value seller has never worked in the CPU or GPU Markets.

And yeah the article stated that and much more. So try to keep it in a bit more perspective then your comments in the other threads.

 

[OrangeKhrush]

Pricing will be very telling, no way will AMD price it lower unless they have to pretty simple, being the value seller has never worked in the CPU or GPU Markets.

And yeah the article stated that and much more. So try to keep it in a bit more perspective then your comments in the other threads.

I can't read it as it is in french. I am not sure exactly what comments? I proved using my 4960X that the scores on anandtech for CB15 was using turbo clock, the math added up because it is a very simple equation. All I am waiting for now is a proper Zen CB15 as I feel the 1188 was actually score at 3.1ghz, a very similar 3.1Ghz score to a i7 5960X

 

[razor1]

I can't read it as it is in french. I am not sure exactly what comments? I proved using my 4960X that the scores on anandtech for CB15 was using turbo clock, the math added up because it is a very simple equation. All I am waiting for now is a proper Zen CB15 as I feel the 1188 was actually score at 3.1ghz, a very similar 3.1Ghz score to a i7 5960X

I translated parts of what were important just look above, they pretty much stated if Ryzen doesn't come out to 3.8-4.2, ghz, its not going to be competitive with Kaby Lake. And this is where the discrepancy is the IPC is not where Kaby Lake, Sky Lake, Haswell are at. and its showing up in gaming performance where IPC is much more of a factor then going with more cores.

 

[OrangeKhrush]

I translated parts of what were important just look above, they pretty much stated if Ryzen doesn't come out to 3.8-4.2, ghz, its not going to be competitive with Kaby Lake.

1) AMD's slide showed 3.4Ghz + emphasis on plus

2) we know AMD is using XFR technology with SenseMI, which the long an short is that your turbo is determined by your cooling quality

Kaby Lake is Skylake in the real world of perfomance, Kaby barely pushed 2% gains, in fact most gains were higher clock rates. Refined the node a bit as it is Intels' third 14nm iteration, they should have it down to a T by now, so it has moderate power gains but higher clocks.

 

[razor1]

1) AMD's slide showed 3.4Ghz + emphasis on plus

2) we know AMD is using XFR technology with SenseMI, which the long an short is that your turbo is determined by your cooling quality

Kaby Lake is Skylake in the real world of perfomance, Kaby barely pushed 2% gains, in fact most gains were higher clock rates. Refined the node a bit as it is Intels' third 14nm iteration, they should have it down to a T by now, so it has moderate power gains but higher clocks.

Do you really believe everything AMD has stated? Come on, ok 3.4 or more, but still that is far away from 3.8 to 4.2

And the rest of the stuff, doesn't matter, keep things a bit more realistic you will be a bit happier if it comes out and beats realistic expectations, then the other way around.

Yeah sense what ever, doesn't matter what they call it man. Trust me giving something a new marketing name that has been done for years now doesn't change the fact it old.

 

[OrangeKhrush]

Do you really believe everything AMD has stated? Come on, ok 3.4 or more, but still that is far away from 3.8 to 4.2

And the rest of the stuff, doesn't matter, keep things a bit more realistic you will be a bit happier if it comes out and beats realistic expectations, then the other way around.

Yeah sense what ever, doesn't matter what they call it man. Trust me giving something a new marketing name that has been done for years now doesn't change the fact it old.

Your responses are terribly close to that "F" word, you seem to think you know more about AMD technology than they do, SenseMI is not a rebrand of anything, it is built in sensory bolt ons that determines which state the CPU will deliver optimal performance. As stated in the New Horison demo, the premise of SenseMI with XFR is that better cooling results in higher frequency ranges. Can AMD do this, yes they can, in fact AMD and Nvidia have been doing this for years with Radeons and GTX GPU's, using firmware to make GPU's more efficient and cooler. This technology has been used in the automotive performance indusstry for over 20 years in the form of engine mappings.

What is your phobia with AMD having what appears to be a representative product? You keep talking about this expectations, nobody is expecting Zen to beat a Skylake/Broadwell EP, the "expectations" seem to be this phobia driven hostility from team blue googled broskies. Yes I run an intel/nvidia setup, I just choose not to be so damn defensive or devisive.

 

[razor1]

Your responses are terribly close to that "F" word, you seem to think you know more about AMD technology than they do, SenseMI is not a rebrand of anything, it is built in sensory bolt ons that determines which state the CPU will deliver optimal performance. As stated in the New Horison demo, the premise of SenseMI with XFR is that better cooling results in higher frequency ranges. Can AMD do this, yes they can, in fact AMD and Nvidia have been doing this for years with Radeons and GTX GPU's, using firmware to make GPU's more efficient and cooler. This technology has been used in the automotive performance indusstry for over 20 years in the form of engine mappings.

What is your phobia with AMD having what appears to be a representative product? You keep talking about this expectations, nobody is expecting Zen to beat a Skylake/Broadwell EP, the "expectations" seem to be this phobia driven hostility from team blue googled broskies. Yes I run an intel/nvidia setup, I just choose not to be so damn defensive or devisive.

Dude, I suggest you look up, what perceptrons are and then see what SenseMI is, then you will know its been used in many many chips prior to Ryzen, damn its been used for more then 10 years now in some way or form from CPU's to GPU's, to SOC's lol.

OK?

Don't be a sheep for marketing man......

 

[OrangeKhrush]

Dude, I suggest you look up, what perceptrons are and then see what SenseMI is, then you will know its been used in many many chips prior to Ryzen, damn its been used for more then 10 years now in some way or form from CPU's to GPU's, to SOC's lol.

OK?

Don't be a sheep for marketing man......

I bet 16 years ago you were a sucker for Intel spinning that fancy term called Hyper Threading and how it made a single core pentium 4 act like a dual core, everyone ate that BS like no tomorrow, heck SMT is about 20% efficiency of a real physical core, some apps better use it but it was anything but a extra like core.

Other crap that has had people fall hook line and sinker is this adaptive V sync, Mantle, Load Line Calibration, Ram boost, Lucid Logix, Quad channel, PCI e 3.0, SATA express, thunderbolt. All that is just a flavour added but in most circumstances is just a load of ball bags.

 

[razor1]

I bet 16 years ago you were a sucker for Intel spinning that fancy term called Hyper Threading and how it made a single core pentium 4 act like a dual core, everyone ate that BS like no tomorrow, heck SMT is about 20% efficiency of a real physical core, some apps better use it but it was anything but a extra like core.

Other crap that has had people fall hook line and sinker is this adaptive V sync, Mantle, Load Line Calibration, Ram boost, Lucid Logix, Quad channel, PCI e 3.0, SATA express, thunderbolt. All that is just a flavour added but in most circumstances is just a load of ball bags.

No never believed in Intel's hyperthreading because being an experienced programmer from way back when and 3d artist, I knew that it was great only if programs were made to take advantage of it..... lol, if you go back in my post history, even IPC isn't that important for all tasks, depends what you are doing, more cores might be more beneficial to certain people. But if its gaming, IPC is still king because of API's and programming techniques matter more.

 

[OrangeKhrush]

No never believed in Intel's hyperthreading because being an experienced programmer from way back when and 3d artist, I knew that it was great only if programs were made to take advantage of it..... lol, if you go back in my post history, even IPC isn't that important for all tasks, depends what you are doing, more cores might be more beneficial to certain people. But if its gaming, IPC is still king because of API's and programming techniques matter more.

I like the sound of SensMI and will give it a chance along with XFR before I declare it FUD or not, I mean Zen hasn't even passed beyond BETA ES and people are getting bent out of shape.

 

[razor1]

I like the sound of SensMI and will give it a chance along with XFR before I declare it FUD or not, I mean Zen hasn't even passed beyond BETA ES and people are getting bent out of shape.

Its not fud but its nothing new, things like this aren't going to be what saves AMD either. yeah its beneficial, but that is all it is.

 

[SighTurtle]

A full translation: https://forums.anandtech.com/threads/zen-es-benchmark-from-french-hardware-magazine.2495505/

 

[Silverbolth]

Guys, don't forget that this is the REAL 8 cores CPU, no hyper threading. So I can see a 4 cores or 6 cores able to get higher frequencies....

 

[razor1]

Guys, don't forget that this is the REAL 8 cores CPU, no hyper threading. So I can see a 4 cores or 6 cores able to get higher frequencies....

Depends have to wait and see, frequency is a process of architecture more so than in the past, so we don't know if less cores are going to give more mhz or even if it does how much more....

Theoretically to sustain a certain power envelope yes but without knowing the architecture limits on power (pretty much voltage needs), its not absolute at this point.

 

[SighTurtle]

Clamp down on the hype, at the end of the day, until final chips start getting to reviewers, you shouldn't start believing the sky is the limit (and have it fall on you, trust me it hurts). Its really great that CanardPC (the french tech magazine), managed to get a in-depth look at a Zen ES (I know I was bouncing everyday for more info), and give us more information than the questionable stuff AMD provided on performance, but as the author of the article noted, its not out of the woods, especially if you believe that it can compete at the very top with Intel, there are still lots of stuff that can mean that it won't be the best, or even compete with lower generations well, and especially there are plenty of obstacles ahead for AMD before anyone can hope for a true viable competitor in the market.

 

[Shintai]

A full translation: https://forums.anandtech.com/threads/zen-es-benchmark-from-french-hardware-magazine.2495505/

I love how the author calls one company lazy but not the one that sold the same product for over 4 years

 

[noko]

A full translation: https://forums.anandtech.com/threads/zen-es-benchmark-from-french-hardware-magazine.2495505/

Frankly that looks good for me on the multi-core runs. Also for games, I look at if it can maintain 60fps+, anything else above that for me is pointless. Also consider for games, more cores being effectively being used will become more and more the norm. Power also looks good and this is just a ES, early bios etc.

Talking about games, I find worthless the low resolution/setting benchmarks that would never reflect any type of game play that most would be using or would expect to use.

 

[noko]

I love how the author calls one company lazy but not the one that sold the same product for over 4 years

AMD just spent the money more wisely and developed an Intel killer instead .

 

[Shintai]

AMD just spent the money more wisely and developed an Intel killer instead .

So I keep hearing, but I haven't seen it yet

But if that's the case, congrats on AMDs new 1000$+ CPUs. They missed that since the K8.

 

[drescherjm]

AMD just spent the money more wisely

AMD has been grossly mismanaged for a long time. Maybe the new leader ship is fixing that but I can't agree that AMD did anything more wisely in the last decade.

 

[OrangeKhrush]

Its not fud but its nothing new, things like this aren't going to be what saves AMD either. yeah its beneficial, but that is all it is.

I will give time for AMD to lay it out for us to play it out. Why have AMD ensured that SensMI has been disabled on all ES, I think there is something fun about this that may end up a nice little goodie treat.

 

[OrangeKhrush]

Big Zen seems like a direct performer against a 5960X, been on that a while. I can see these chips being sold around 500-600USD for the top end flagship and sister, and a sub 200USD for the 4C/8T SKU.....I like this, I am just so curious for more.

 

[razor1]

Big Zen seems like a direct performer against a 5960X, been on that a while. I can see these chips being sold around 500-600USD for the top end flagship and sister, and a sub 200USD for the 4C/8T SKU.....I like this, I am just so curious for more.

I can't see that, cause there is no benefit to under cut to that degree. These markets don't function on price unless there is a reason for them to, Intel can compete on price, this isn't like back in the day where AMD could under cut Intel and maintain margins that were acceptable, and Intel couldn't because of their overhead in technology development.

Curiosity is great, but ignoring the fundamentals of how markets work and just expecting AMD to cut price to gain marketshare, has shown it never works, its harder for them even with Performance parity to gain marketshare just because of what has happened in the past decade with their CPU's, so how would going into a price war with Intel *which Intel can do too* be beneficial to anyone outside of us consumers?

These are corporations, they aren't there to make us happy they are there to make their board members and investors happy.

This is why I stated pricing will tell us A LOT about where Ryzen ends up.

 

[SighTurtle]

I can't see that, cause there is no benefit to under cut to that degree. These markets don't function on price unless there is a reason for them to, Intel can compete on price, this isn't like back in the day where AMD could under cut Intel and maintain margins that were acceptable, and Intel couldn't because of their overhead in technology development.

Curiosity is great, but ignoring the fundamentals of how markets work and just expecting AMD to cut price to gain marketshare, has shown it never works, its harder for them even with Performance parity to gain marketshare just because of what has happened in the past decade with their CPU's, so how would going into a price war with Intel *which Intel can do too* be beneficial to anyone outside of us consumers?

These are corporations, they aren't there to make us happy they are there to make their board members and investors happy.

This is why I stated pricing will tell us A LOT about where Ryzen ends up.

There is something to be said for AMD shedding its bargain bin status i think. For regular people and a good portion of enthusiasts (especially in the market AMD tried to appeal at New Horizon), I think AMD is gonna want to show itself as a competitor that can jostle for top-dog spot, and do it just priced enough to beat Intel. Those Intel stickers have appeared virtually across the board for regular consumers, and AMD does not need to be considered second-rate, or second-best, cause that will mean consumers will turn away at the first sign of "AMD". Solve that, price it a bit under and people will choose it. Plus, like you stated before, AMD needs money, like a lot and fast. They got Zen up to near Intel's best but if they can't compete with Intel on support, software, etc, no one will choose a dying company for their long term needs.

I think it was Shintai who made fun out of AMD seeking a 10% marketshare in servers. Assuming his statement on shit performance does not come true, AMD is still gonna have to show it can provide stability with their products to a marketplace under virtually Intel's complete proliferation for quite a bit. (I don't know if thats true, do they use ARM in servers yet?) Besides, 10% in marketshare is a shitload of money.

Edit: AMD, excuse me, Radeon, faces the same problem in the GPU market. Nvidia funds PR in such enormous amounts, I would not be surprised for people to have never heard of AMD, Radeon, or ATI (especially the esport market, see any red there recently? Its green all the way Oh, and getting that esports guy was a good idea, the key being 421K followers, seeing a AMD logo right as they check twitter. Should have paid for some color on it through. ), and aside from having to claw back with their own PR (what money? oh look, a market where AMD can't go lower than the virtual 0% it already has.), saying they have the best graphics card on the market is pretty good marketing itself.

Doesn't matter if its not exactly true, (cause Nvidia won't let that happen ever), AMD just needs to be able to say it. Especially with Polaris being so...."low-end".

Speaking not as a consumer, but as a something? AMD did what it needed to do with Polaris, bullshit its way through, despite enthusiasts being less than happy, Polaris did what it needed to do, not derail the train for Zen. Their immediate focus should/is be on Zen, thats where it will get the money to survive, push Vega back, does anyone actually think Nvidia isn't gonna make sure that AMD gets as little money as possible?

(This may be bullshit. Sorry for wasting your time.)

 

[razor1]

There is something to be said for AMD shedding its bargain bin status i think. For regular people and a good portion of enthusiasts (especially in the market AMD tried to appeal at New Horizon), I think AMD is gonna want to show itself as a competitor that can jostle for top-dog spot, and do it just priced enough to beat Intel. Those Intel stickers have appeared virtually across the board for regular consumers, and AMD does not need to be considered second-rate, or second-best, cause that will mean consumers will turn away at the first sign of "AMD". Solve that, price it a bit under and people will choose it. Plus, like you stated before, AMD needs money, like a lot and fast. They got Zen up to near Intel's best but if they can't compete with Intel on support, software, etc, no one will choose a dying company for their long term needs.

I think it was Shintai who made fun out of AMD seeking a 10% marketshare in servers. Assuming his statement on shit performance does not come true, AMD is still gonna have to show it can provide stability with their products to a marketplace under virtually Intel's complete proliferation for quite a bit. (I don't know if thats true, do they use ARM in servers yet?) Besides, 10% in marketshare is a shitload of money.

Well to gain 10% of server marketshare they have to do above and beyond Intel, just look at when they had their Athlon advantage, it took them 2 years to gain 25% marketshare with huge performance benefit and power advantage. To gain marketshare having equivalent products isn't enough when there is a virtual monopoly.

AMD's perception as bargain bin products were their doing, "smart" consumers understand this, "sheeple" don't lol, they won't see it, they buy by marketing and Intel is a bit too strong with their mind share right now. Arm server CPU from AMD isn't ready nor do I think they even have anything planned in the short term, cause of the way it kinda dropped off the radar 2 q's ago.

I'm going to wait on judgement on its performance, but impressed with what AMD showed, but if this magazine (which seems to be reputable) comes true, well, 500 bucks put Ryzen 8 core at 6 core performance of Intel overall (higher in some tests, lower in others), and the 500 bones is justified.

Edit: and AMD getting that 25% marketshare was when Intel didn't do any rebates in the server market too.

 

[FrozenSun]

I translated parts of what were important just look above, they pretty much stated if Ryzen doesn't come out to 3.8-4.2, ghz, its not going to be competitive with Kaby Lake. And this is where the discrepancy is the IPC is not where Kaby Lake, Sky Lake, Haswell are at. and its showing up in gaming performance where IPC is much more of a factor then going with more cores.

Looking at benches, that's more an issue with clock speed than IPC. For the clock listed on the bench (3.15/3.4), the IPC seems pretty good. Yeah, the 6700k beats it soundly in gaming, but that's kind of expected at 4GHz. The 6700k also beats the 6800k and 6900k easily as well for the same reason.

The big question will be if the 4 and 6 core Zens will be able to reach higher clocks to compete. It looks like the 8 core zen should compete somewhere around the 6800k level, which isn't a bad result at all. But it's worth pointing out that the 6800k/6850k aren't exactly volume processors. They don't seem to sell a whole lot of those. The enthusiast market is pretty niche as it is, but the 6800 line is the niche of the niche. They're great chips, but they don't seem to sell a lot of them, just looking at what people use in their rigs, number of reviews on reseller sites, etc.

The 6600k and 6700k are by far the most popular enthusiast chips, so it's important AMD competes there as well and that will be reliant on them getting the 4 and 6 core chips up close the 4GHz, which is a pretty big if.

Honestly, I'm just excited Zen doesn't seem to be a dud. I don't see it topping Intel in any regard, but it does look like it should compete with some of their lower and mid ranged i7s, which will hopefully lead to some slightly better prices. I still expect my next rig to be a 7700k, but we'll see how it plays out.

 

[razor1]

Looking at benches, that's more an issue with clock speed than IPC. For the clock listed on the bench (3.15/3.4), the IPC seems pretty good. Yeah, the 6700k beats it soundly in gaming, but that's kind of expected at 4GHz. The 6700k also beats the 6800k and 6900k easily as well for the same reason.

The big question will be if the 4 and 6 core Zens will be able to reach higher clocks to compete. It looks like the 8 core zen should compete somewhere around the 6800k level, which isn't a bad result at all. But it's worth pointing out that the 6800k/6850k aren't exactly volume processors. They don't seem to sell a whole lot of those. The enthusiast market is pretty niche as it is, but the 6800 line is the niche of the niche. They're great chips, but they don't seem to sell a lot of them, just looking at what people use in their rigs, number of reviews on reseller sites, etc.

The 6600k and 6700k are by far the most popular enthusiast chips, so it's important AMD competes there as well and that will be reliant on them getting the 4 and 6 core chips up close the 4GHz, which is a pretty big if.

Honestly, I'm just excited Zen doesn't seem to be a dud. I don't see it topping Intel in any regard, but it does look like it should compete with some of their lower and mid ranged i7s, which will hopefully lead to some slightly better prices. I still expect my next rig to be a 7700k, but we'll see how it plays out.

Most games don't scale past 4 cores, and that is what we are seeing, the frequency is part of it too, so yeah it could be more things, but what ever it is, we have to wait and see. I'm betting its both.

Any case, I don't see this being a market sifter by any means, it does force Intel to innovate which is good, but consumer side, I can't see AMD really pushing Intel to do much other then even out their line up regarding price with Zen if there will be any discrepancy and this is temporary, cause its all in Intel's hands if they have the capability to innovate more in the short to mid term (planned for this) or we have to wait a bit longer like 3 years.

 

[Gideon]

The older the game the more it will rely on the frequency of ZEN, some of the newest games will likely do better with the extra cores. The gaming market is slowly shifting to more threads. Looks pretty good since it was ES chip with low clocks, but much will hinge on what AMD releases the chip at.

 

[KazeoHin]

AMD just spent the money more wisely and developed an Intel killer instead .

The word "AMD" and the phrase "Spending money wisely" should never ever never




Ever







Never











Ever















EVER be used in the same sentence.

 

[pendragon1]

The word "AMD" and the phrase "Spending money wisely" should never ever never
.....
EVER be used in the same sentence.

as much as I root for them, this is pretty true

 

[razor1]

well at this point they have to, not much choice there, now if they start making money and they start doing crazy parties lol, I will be like WTF lol but right now they don't have money to play around with anything.

 

[primetime]

looks like the new AMD's faster than my current chip...I cant remember the last time that happened lol....

 

[OrangeKhrush]

I love how the author calls one company lazy but not the one that sold the same product for over 4 years

http://www.anandtech.com/show/9483/intel-skylake-review-6700k-6600k-ddr4-ddr3-ipc-6th-generation/16

A 2011 i7 2600K is still inside 5% of a Skylake 6700K's gaming performance, there are people like myself that have no reason to upgrade our CPU's because Intel has been dolling out the same crap for 5 years. I have a 4960X, 4790 and 4460, neither need to be replaced by a Skylake or Kabylake especially not for gaming.

i7 4790K is still the best bang for buck i7 on the market, it OC's like a boss and it games like a boss.

 

[Reality]

remember when everybody thought AMD knew how to properly evaluate product pricing, then released the Fury X and Nano at the exact same MSRP as the 980 TI? keep thinking AMD will undercut Intel by 50% lol

 

[Shintai]

Looking at benches, that's more an issue with clock speed than IPC. For the clock listed on the bench (3.15/3.4), the IPC seems pretty good. Yeah, the 6700k beats it soundly in gaming, but that's kind of expected at 4GHz. The 6700k also beats the 6800k and 6900k easily as well for the same reason.

The big question will be if the 4 and 6 core Zens will be able to reach higher clocks to compete. It looks like the 8 core zen should compete somewhere around the 6800k level, which isn't a bad result at all. But it's worth pointing out that the 6800k/6850k aren't exactly volume processors. They don't seem to sell a whole lot of those. The enthusiast market is pretty niche as it is, but the 6800 line is the niche of the niche. They're great chips, but they don't seem to sell a lot of them, just looking at what people use in their rigs, number of reviews on reseller sites, etc.

The 6600k and 6700k are by far the most popular enthusiast chips, so it's important AMD competes there as well and that will be reliant on them getting the 4 and 6 core chips up close the 4GHz, which is a pretty big if.

Honestly, I'm just excited Zen doesn't seem to be a dud. I don't see it topping Intel in any regard, but it does look like it should compete with some of their lower and mid ranged i7s, which will hopefully lead to some slightly better prices. I still expect my next rig to be a 7700k, but we'll see how it plays out.

We haven't seen any Zen parts with high clocks. And the reason may be that getting high IPC isn't hard. High clocks isn't hard either. But high clocks and high IPC is where it gets really hard.

Dont expect better prices, AMD wants better margins. Lisa Su even publicly said she wants more money. Also its suicide to start a price war, it benefits none of the companies.