Leaked AMD Zen Engineering Sample Benchmarks?

Isn't that what was done? I mean pic shows all cpus were run at 3.2 GHz. Or was turbo enabled on the 5960x and the FX 8370 and not Zen? Or the Zen was run at 2.8 + turbo and not really 3.2 GHz?

its 2.8 + turbo. and I know for a damn fact that when all cores are engaged my i7 does not boost to its turbo speed. So in Ashes it is probably running at 2.8ghz since it will engage all cores.
 
Apparently the Oxide engine scales well even with 16 threads on a CPU ;), which it does seem to do that well.

Its one of the few engine/games that does this right now, most likely because the game takes advantage of the extra resources from the CPU.

If AMD matches Haswell with a 8 core Zen, err, they won't have an answer for Kaby Lake, Kaby Lake is likely not be the regular 8%-15% performance increases we have been seeing from Intel. From what I have heard it sounds to be more, more like the 10%-20% range.

But competition at the high end isn't everything, if AMD can get it power consumption down they will become a viable option for lower end laptop and desktops, cause OEM's, system builders can have an alternative to Intel with all other components staying roughly at the same price.

from everywhere I have read almost every tech site is saying don't expect miracles from kaby lake, your post seems to be the only one suggesting to expect more. There is not really much difference or upgrade reason for people to upgrade from haswell to skylake. I got a 6700k because I was building a new system but if I had haswell there wouldn't be much reason for me.
 
Thank god, my 3570k was the most overpriced pos hardware I've bought. If i was half the price back then, then I wouldn't be annoyed with its lack of multitasking oomph. $300 on amazon.ca is current price for anyone won
3570k is a good cpu even by todays standards i would hardly call it overpriced and you can get used ones on ebay for like 150
 
I didnt see temps or anything for zen but that is a low stock clock speed im wondering if it runs hot or something i would have expected atleast 3.5 ghz stock for this cpu.the only way ill buy one is if they can hit 4.2 ghz pretty easily otherwize ill stick with my 3930k
 
I didnt see temps or anything for zen but that is a low stock clock speed im wondering if it runs hot or something i would have expected atleast 3.5 ghz stock for this cpu.the only way ill buy one is if they can hit 4.2 ghz pretty easily otherwize ill stick with my 3930k

We've known for a VERY long time Zen is produced on Global Foundries 14nm LPP node, which is optimized for sub -3GHz clocks. Going beyond that will lead to excessive power draw and significant cooling issues. The clocks are pretty much what I've expected, and Zen will probably top out at about 3.2 GHz base. I also suspect it will be a very poor Overclocker for the same reasons.
 
Ran some math on the benchmark scores. I'm disregarding SMT performance effects to make the math simpler, and assuming the benchmark is running at Turbo speeds.

Performance = (IPC * Clock) * Number of Cores

i7 4790k:

65.4 = 4.0 * 8 * IPC
IPC = ~2.04

Zen:

58 = 3.2 * 16 * IPC
58 = 6.4 * IPC
IPC = ~1.13

FX-8350:

42 = 4.2 * 8 * IPC
IPC = 1.25

In the AoS benchmark, IPC per-core is higher in PD then it is in Zen. The reason is likely that AoS isn't stressing all 16 cores, leading to worse absolute numbers. Unless I have Core usage statistics, I can't correct and compute the "real" IPC. In any case, the one thing I can say is the majority of those CPU cores aren't doing any real work in AoS.

Honestly, AMD would likely have been far better served by cutting the core count in half to gain an extra 200MHz or so.
 
We've known for a VERY long time Zen is produced on Global Foundries 14nm LPP node, which is optimized for sub -3GHz clocks. Going beyond that will lead to excessive power draw and significant cooling issues. The clocks are pretty much what I've expected, and Zen will probably top out at about 3.2 GHz base. I also suspect it will be a very poor Overclocker for the same reasons.

I had not heard this before, and I certainly hope this is not true.

Of what benefit is raising the IPC by ~40%, if at the same time they are dropping the clocks by almost as much?

This was supposed to be their return to a competitive product. They can't possibly hope to compete with that strategy.

Is it possible that the GloFo 14nm LPP process is being used for APU's only? I understand that process is better for GPU performance, but produces lower clocked CPU parts. This would make sense for APU models. For the standalone FX parts they will hopefully be using a different process more suited to high clocking CPU's.
 
I had not heard this before, and I certainly hope this is not true.

Of what benefit is raising the IPC by ~40%, if at the same time they are dropping the clocks by almost as much?

This was supposed to be their return to a competitive product. They can't possibly hope to compete with that strategy.

Is it possible that the GloFo 14nm LPP process is being used for APU's only? I understand that process is better for GPU performance, but produces lower clocked CPU parts. This would make sense for APU models. For the standalone FX parts they will hopefully be using a different process more suited to high clocking CPU's.

Nope. Pretty much all of AMDs production is done at Global Foundries [formally AMDs fabrication unit, before they spun it off].

In any case, I've been betting on 3.2/3.4 Stock/Turbo clocks for the first stepping of Zen. Clocks will almost certainly be lower by some degree compared to Piledriver based CPUs, which will eat into some of the performance gain in IPC. That's why I've been predicting Ivy Bridge i7 performance levels. I'll be surprised if Zen matches Haswell based i7's.
 
Nope. Pretty much all of AMDs production is done at Global Foundries [formally AMDs fabrication unit, before they spun it off].

In any case, I've been betting on 3.2/3.4 Stock/Turbo clocks for the first stepping of Zen. Clocks will almost certainly be lower by some degree compared to Piledriver based CPUs, which will eat into some of the performance gain in IPC. That's why I've been predicting Ivy Bridge i7 performance levels. I'll be surprised if Zen matches Haswell based i7's.

I've been predicting Ivy Bridge performance levels as well, but according to my calculations if the IPC increase is 40% they will need high clocks in order to match a high end Ivy Bridge chip.

If they can't meet 3.8-4.0ghz, I fear they aren't even going to be able to reach Sandy Bridge levels.
 
I've been predicting Ivy Bridge performance levels as well, but according to my calculations if the IPC increase is 40% they will need high clocks in order to match a high end Ivy Bridge chip.

If they can't meet 3.8-4.0ghz, I fear they aren't even going to be able to reach Sandy Bridge levels.

To follow up on this. Lets assume 40% over current chips is accurate, IPC improvement is accurate, and we are only expecting a 2.8ghz base 3.2ghz turbo, compared to the current top chips with 4.1ghz base, 4.3ghz turbo, that's only a 4.2% performance increase at turbo clocks and a 4.4% performance DECREASE at base clocks...

Considering that in the most reliable single core benchmarks we have (Cinebench) AMD is trailing Intel by more than 50%, this is very very bad, and I am certainly hoping you are wrong.
 
I'm pretty sure they're expecting IB levels of IPC performance not single threaded performance. Looks like they did what intel did when they went from netburst to core architecture and scarified clock speed for IPC without intel's subsequent increase in clock speed. AMD's cop out this time will be "but you said you wanted higher IPC, we gave you higher IPC!" all while completely and utterly ignoring the real problem people were complaining about, that being single threaded performance. In the end single threaded performance from one generation to the next will be around 5% with a massive increase in IPC and a massive drop in clock speed to compensate which is about the same as an intel generational increase except with intel you get 5% every year and a half or so where with AMD it takes 3 or 4 years to get it, you'll still need an 8 core AMD chip to compete with intel 4 cores, which will still be single thread kings and AMD will only really compete with higher threaded applications, fundamentally nothing will have changed. My only concern is some of this "IPC" improvement might actually be from SMT where it's not actually doing more per clock but 5-10% or more of the 40% IPC boost might only be applicable to multiple threads which means Zen true single thread performance might be worse than their old stuff, which would be awful, misleading at best on AMDs part, and facepalm worthy.


And something I can totally see them do because we've seen them do all of that time and time again.
 
I'm pretty sure they're expecting IB levels of IPC performance not single threaded performance. Looks like they did what intel did when they went from netburst to core architecture and scarified clock speed for IPC without intel's subsequent increase in clock speed. AMD's cop out this time will be "but you said you wanted higher IPC, we gave you higher IPC!" all while completely and utterly ignoring the real problem people were complaining about, that being single threaded performance. In the end single threaded performance from one generation to the next will be around 5% with a massive increase in IPC and a massive drop in clock speed to compensate which is about the same as an intel generational increase except with intel you get 5% every year and a half or so where with AMD it takes 3 or 4 years to get it, you'll still need an 8 core AMD chip to compete with intel 4 cores, which will still be single thread kings and AMD will only really compete with higher threaded applications, fundamentally nothing will have changed. My only concern is some of this "IPC" improvement might actually be from SMT where it's not actually doing more per clock but 5-10% or more of the 40% IPC boost might only be applicable to multiple threads which means Zen true single thread performance might be worse than their old stuff, which would be awful, misleading at best on AMDs part, and facepalm worthy.


And something I can totally see them do because we've seen them do all of that time and time again.


If this is the case, they are dead to me.

They would be launching a product in 2017 that is completely and totally outclassed by my Intel CPU I bought in 2011 in everything but power consumption.

This might be OK consumer laptops, tablets and all in ones, but for our market, the performance oriented desktop user, if this is what they are doing it will be a massive fail of even greater than budozer proportions, IMHO.

From my perspective as a performance desktop hobbyist, the ONLY performance metric I care about is single threaded performance. I don't care how they get there, through clock, through IPC or through a balance of both, but they need to get there, or I'll just hold on to my aging relic of a CPU/Motherboard. The one I've had the longest by a VERY WIDE margin in all of my 25 years in this hobby.

I was hoping to buy a Zen to support them and refresh my system, but overclocked single threaded performance has to be able to meet what my 3930k does at 4.8ghz, or they won't even be a consideration. As I have said before, I'll do a performance side-grade to get more modern features and lower power consumption, but there is no way in hell I'm doing a performance downgrade.

This is what I'm talking about.

They need to be able to match a 5 year old (at time of launch) Ivy Bridge 3770K in this benchmark, or they will have lost the enthusiast market completely.
 
Last edited:
This is what I'm talking about.

They need to be able to match a 5 year old (at time of launch) Ivy Bridge 3770K in this benchmark, or they will have lost the enthusiast market completely.

For comparison, right now AMD's fastest entry in this table, The FX-9590 is slower than Intel's Pentium G2030.

Some linear interpolation suggests that assuming 40% IPC increase over Kaveri, but only 3.2Ghz turbo clocks, Zen would only score 1.16 in this benchmark, in other words between a Pentium G850 and a Pentium G2030.

If they can reach - say - a 3.8Ghz base, 4.0Ghz turbo like what I had expected, we are talking a 1.45 result, putting it in league with a Sandy Bridge i5-2500k

The latter is marginally passable for an enthusiast system (despite just being able to break even with a 6 year old chip at launch being a pretty bad embarrassment). If we are talking 3.2Ghz level turbo clocks, they are dead to me.
 
For comparison, right now AMD's fastest entry in this table, The FX-9590 is slower than Intel's Pentium G2030.
Some linear interpolation suggests that assuming 40% IPC increase over Kaveri, but only 3.2Ghz turbo clocks, Zen would only score 1.16 in this benchmark, in other words between a Pentium G850 and a Pentium G2030.
If they can reach - say - a 3.8Ghz base, 4.0Ghz turbo like what I had expected, we are talking a 1.45 result, putting it in league with a Sandy Bridge i5-2500k
The latter is marginally passable for an enthusiast system (despite just being able to break even with a 6 year old chip at launch being a pretty bad embarrassment). If we are talking 3.2Ghz level turbo clocks, they are dead to me.

There is a lot unclear still the numbers from the benchmark still can't be verified where it even did 3.2 ghz in another piece you can find the all cores turbo would be at 3.05 ghz so that is a little bit less then 3.2 ghz but regardless of these speeds it will still end up doing badly as you described at lower frequency.
 
To be honest in this day and age the enthusiast market can eat a dick as far as AMD and Intel are both concerned as it relates to their bottom line, since while profits are great revenue is low. What AMD really needs for a win financially is performance per watt so they can get themselves in high performance low power systems (surface systems and similar). If they can match intel or at least get closer for cheaper in the perf/W category it may not matter if we all hate them here. But if AMD's CPU perf/W relative to intel is the same as AMD's polaris perf/W is relative to nvidia's pascal, AMD is straight F'd in the A.
 
To be honest in this day and age the enthusiast market can eat a dick as far as AMD and Intel are both concerned as it relates to their bottom line, since while profits are great revenue is low. What AMD really needs for a win financially is performance per watt so they can get themselves in high performance low power systems (surface systems and similar). If they can match intel or at least get closer for cheaper in the perf/W category it may not matter if we all hate them here. But if AMD's CPU perf/W relative to intel is the same as AMD's polaris perf/W is relative to nvidia's pascal, AMD is straight F'd in the A.


As I mentioned above, they may be able to play in the laptop/tablet/convertible type market with these chips, and while it may be the most financially viable thing for them to do, it doesn't do anything for users like us, which is a disappointment.

If the sub 3ghz optimization for GloFo's processes is accurate, then I'm hoping that they might do some smaller runs on someone else's process for high performing non APU parts.
 
Has there even been any talk of a 2C/4T zen though? I always hear 4/6/8C variants but never a 2C one yet, and the tablet/convertible market gobbles up intel 2C/4T chips all day, don't see how AMD can get in the 10-15W range with an APU with 4 real cores.
 
I'm pretty sure they're expecting IB levels of IPC performance not single threaded performance. Looks like they did what intel did when they went from netburst to core architecture and scarified clock speed for IPC without intel's subsequent increase in clock speed. AMD's cop out this time will be "but you said you wanted higher IPC, we gave you higher IPC!" all while completely and utterly ignoring the real problem people were complaining about, that being single threaded performance. In the end single threaded performance from one generation to the next will be around 5% with a massive increase in IPC and a massive drop in clock speed to compensate which is about the same as an intel generational increase except with intel you get 5% every year and a half or so where with AMD it takes 3 or 4 years to get it, you'll still need an 8 core AMD chip to compete with intel 4 cores, which will still be single thread kings and AMD will only really compete with higher threaded applications, fundamentally nothing will have changed. My only concern is some of this "IPC" improvement might actually be from SMT where it's not actually doing more per clock but 5-10% or more of the 40% IPC boost might only be applicable to multiple threads which means Zen true single thread performance might be worse than their old stuff, which would be awful, misleading at best on AMDs part, and facepalm worthy.


And something I can totally see them do because we've seen them do all of that time and time again.

They will increase FLOATING POINT IPC by a significant amount, due to actually being able to properly feed the ALUs. I can easily envision a case where Floating Point workloads see a much larger performance gain then Integer ones, which weren't as starved under BD/PD. I always took the 40% IPC claim as "best case, likely FP Workloads", because AMD has a history of overstating performance.
 
Has there even been any talk of a 2C/4T zen though? I always hear 4/6/8C variants but never a 2C one yet, and the tablet/convertible market gobbles up intel 2C/4T chips all day, don't see how AMD can get in the 10-15W range with an APU with 4 real cores.

Zen modules seem to be made up of four cores, hence the 4/8. It's possible they could cut off half the die, but that wouldn't be terribly efficient. I suspect we'll only see 4/8 Core parts at launch.
 
Zen modules seem to be made up of four cores

I thought the ones that will be used in the server platform and the 8 core / 16 threaded CPU that is expected be released this year will be using 8 core modules in a single die. Although APUs will have 4C / 8T. I assume that will be a different die.
 
If AMD's pow power APUs are going to be 4C/8T to compete in the same market segment as intel's 2C/4T that has horrifying implications for their single thread performance. Bursty performance where a CPU gets a task done quickly and goes back to a low power mode is greatly preferred to something that plods along slower even if it uses a little less power while doing it, also it makes the device feel more responsive. I don't do a lot of encoding on my laptop or tablet nor do I intend to play a lot of vulcan or DX12 games on it so being a threaded powerhouse is fairly meaningless for 90%+ of my (and most peoples) mobile workload, so having 3 idle cores while one plods along won't really gain anybody anything in the mobile world, which is just a rehash of the situation AMD is in today, no products that can compete on the high end with performance and no products that can compete on the mobile arena with efficiency... Seems Zen might be another AMD product that is radically different from their old one that fundamentally doesn't change a thing.
 
If AMD's pow power APUs are going to be 4C/8T to compete in the same market segment as intel's 2C/4T that has horrifying implications for their single thread performance. Bursty performance where a CPU gets a task done quickly and goes back to a low power mode is greatly preferred to something that plods along slower even if it uses a little less power while doing it, also it makes the device feel more responsive. I don't do a lot of encoding on my laptop or tablet nor do I intend to play a lot of vulcan or DX12 games on it so being a threaded powerhouse is fairly meaningless for 90%+ of my (and most peoples) mobile workload, so having 3 idle cores while one plods along won't really gain anybody anything in the mobile world, which is just a rehash of the situation AMD is in today, no products that can compete on the high end with performance and no products that can compete on the mobile arena with efficiency... Seems Zen might be another AMD product that is radically different from their old one that fundamentally doesn't change a thing.

Agreed. It doesn't matter how well threaded your code is. Fewer strong cores will always be preferable to more weak cores. The only exceptions would be server applications and render/encoding machines where overall throughput matters more than a quick burst of activity.
 
Basing performance off one game is hardly a useful look at Zen. I also expect clock speed will be up by about 500 mhz over the current es samples. But I doubt it will convince anyone that already has a high end cpu to change over to Zen, but it will give anyone looking at a new computer a new choice.
 
  • Like
Reactions: NKD
like this
Basing performance off one game is hardly a useful look at Zen. I also expect clock speed will be up by about 500 mhz over the current es samples. But I doubt it will convince anyone that already has a high end cpu to change over to Zen, but it will give anyone looking at a new computer a new choice.

I agree. ES samples tend to be lower clocked than actual production hardware. its a test chip they don't have to have 4ghz to test they how it scales.
 
You can't tell anything about clock speeds from ES samples. There's been plenty of examples where ES chips were actually clocked higher than the final products but were lowered to either meet stability, cooling, or power envelope requirements. But this is AMD we're talking about, if you're expecting them to massively improve before launch you're just setting yourself up to be disappointed, just remember what AMD stands for, Always Manages Disappointment.
 
I had not heard this before, and I certainly hope this is not true.

Of what benefit is raising the IPC by ~40%, if at the same time they are dropping the clocks by almost as much?

This was supposed to be their return to a competitive product. They can't possibly hope to compete with that strategy.

Is it possible that the GloFo 14nm LPP process is being used for APU's only? I understand that process is better for GPU performance, but produces lower clocked CPU parts. This would make sense for APU models. For the standalone FX parts they will hopefully be using a different process more suited to high clocking CPU's.

ES sample. I haven't really seen es samples clocked up to max like consumer samples. These are just test chips probably designed to test IPC they can scale them up if they wanted so I don't see any reason to stress about clock speeds.
 
Here's a question, and it might be a dumb one.

Apart from price and brand loyalty, why would anyone purchase an AMD chip that is competitive only with 5-year old hardware from a competitor?

I'm assuming that the primary object of most posts in here has been regarding games, so I get that gamers without a tight budget are unlikely to buy something that isn't competitive. What about everyone else?

Apart from being cheaper, what advantage does Zen have, if any, over current Intel offerings?

Taking price somewhat into account, are they competitive at any particular price point? Say at $250-300 for a processor. Is there any reason anyone would choose AMD over Intel if that was a hard budget?
 
Here's a question, and it might be a dumb one.
Apart from price and brand loyalty, why would anyone purchase an AMD chip that is competitive only with 5-year old hardware from a competitor?
I'm assuming that the primary object of most posts in here has been regarding games, so I get that gamers without a tight budget are unlikely to buy something that isn't competitive. What about everyone else?
Apart from being cheaper, what advantage does Zen have, if any, over current Intel offerings?
Taking price somewhat into account, are they competitive at any particular price point? Say at $250-300 for a processor. Is there any reason anyone would choose AMD over Intel if that was a hard budget?
You would choose Zen if you have a need for cores 8 core Zen will be cheaper then Intel 8 core.
The chain of thought is that even if you are serious about 5 year old Intel hardware it is about competition as well. Modern API (Mantle DX12 Vulkan) are less dependant on IPC then their predecessors so it makes it less important to "buy" Intel for gaming.

In the end you need to take AMD seriously not just for the market where you can choose things but also because of the past where Intel was charging premium prices for mid range cpu this alone warrens that you consider AMD as an option.

Some people think that you can get away with crapping on AMD but the reality of things are that Intel got caught several times of anti competitive behaviour they did that because of AMD that says enough. If you look at their R&D budget it is a very nice chip regardless of hitting the mark or not.

You and many others make the mistake of portraying people that buy AMD as brand loyalists should those people keep portraying Intel users as accessory to a corrupt corporate entity ?
Would like to see some people not make the mistake of calling people names for the sake of crapping on AMD, if you have a point make it , no name calling...
 
You would choose Zen if you have a need for cores 8 core Zen will be cheaper then Intel 8 core.
Which brings up the question of who is it that needs 8 cores? I get that 8 cores is great, but who has a pressing need for 8 cores instead of 4 w/ HT? Or 6?

People who run home servers? Do serious video editing (but not casual youtube vids)? What usage scenario makes a typical home user with 1-3 computers choose Zen over Skylake?

You and many others make the mistake of portraying people that buy AMD as brand loyalists should those people keep portraying Intel users as accessory to a corrupt corporate entity ?
Uhm, in what way did I portray people as such? I'm running AMD in this rig right now and generally find myself relatively pleased with AMD's offerings.

There are some people who buy AMD because of brand. I'd guess it's very few. There are people who buy AMD because they are on a budget, there are more of these. In my original post I simply mentioned those two categories to get them 'out of the way' to ask why other people not on a budget might prefer Zen.

I know that people buy AMD for other reasons, and I wanted to know what those reasons were. 8 Cores. Great - but for what?

Would like to see some people not make the mistake of calling people names for the sake of crapping on AMD, if you have a point make it , no name calling...
I was asking a question not remotely trying to make a point. I did not call anyone a name at any time. And even if I had been trying to paint with a wide brush using my "loyalists" comment (I wasn't), I still wouldn't consider that an insult.

Thank you for somewhat answering the question, but you might want to get that chip on your shoulder looked at.
 
Which brings up the question of who is it that needs 8 cores? I get that 8 cores is great, but who has a pressing need for 8 cores instead of 4 w/ HT? Or 6?

People who run home servers? Do serious video editing (but not casual youtube vids)? What usage scenario makes a typical home user with 1-3 computers choose Zen over Skylake?

Gamers on a budget is your usage scenario.

As time goes on more cores is becoming more relevant. Doom addresses up to 12 threads. BF4 addresses up to 8 threads. BF3 addressed up to 6. Vulkan, Mantle and DX12 are making these ancient high core AMD chips relevant. Now people also want modern platforms with PCI-E 2, m.2 etc. That being said, folks with 8350s are now playing Doom as just well as folks with Intel Quads. That being said, yes Intels are more powerful but AMD still has the bang for the buck crown. Not everyone gives a shit about brands. Some folks just care about getting the most for their money. AMD still delivers that in the latest titles thanks to these new APIs.

Intel isn't the answer in every scenario.
 
Here's a question, and it might be a dumb one.

Apart from price and brand loyalty, why would anyone purchase an AMD chip that is competitive only with 5-year old hardware from a competitor?

My rationale WAS that in most cases the CPU doesn't matter anymore. I haven't been CPU limited in anything for way more than 5 years. With current AMD CPU's there is potential for being CPU limited, but with Zen - if that goes away, and they are now fast enough to not be CPU limited anymore - it doesn't matter which CPU you buy. Intel may be a littel faster, but both would be fast enough.

It doesn't matter if you have the fastest PC on the block. What matters is if it is fast enough for what you do. Bragging rights are for morons.

If it doesn't matter, I'd altruistically like to support AMD, because the consequences of them disappearing from the marketplace are so horrifically bad for our hobby, that I can't let that happen, if I ahve any say what so ever.

I'm assuming that the primary object of most posts in here has been regarding games, so I get that gamers without a tight budget are unlikely to buy something that isn't competitive. What about everyone else?

If a CPU is fast enough to not be CPU limited in any of their games, why would a gamer care, and spend extra money on a CPU they don't need?

Apart from being cheaper, what advantage does Zen have, if any, over current Intel offerings?

None what so ever. Unless they have some ridiculously well hidden ace up their sleeve, which sounds highly unlikely to me, the best we can hope for is per-core performance equivalent to a 6 year old (at launch time) sandy bridge Core i5-2500k at stock clocks. Sure, they will have more cores than a Core i5, but that doesn't impress me much. I'll take fewer more powerful cores over more less powerful cores 100% of the time in client applications. This is th ebest we can hope for, if they are able to reach higher clocks.

According to comments above (if they are accurate), the process they are manufacturing on won't be able to go much above 3Ghz though, which means they won't even be able to catch up to where Intel was 6 years ago. if true, this is a huge bummer, and I won't be buying.

I could be convinced to buy if they hit my best case scenario above, but if the clocks are low then there isn't a chance in hell.

Taking price somewhat into account, are they competitive at any particular price point? Say at $250-300 for a processor. Is there any reason anyone would choose AMD over Intel if that was a hard budget?

Too much speculation at this point. No one knows how final silicon will perform or how much it will cost, and we probably won't know until December the earliest. These analyses will have to wait until then.

As for educated guesses, if you want many cores, and don't care about low per-core performance, they will probably be the best buy.
 
Not CPU limited for 5 years? Try reach 60FPS in Fallout 4 in a heavy area for once. That can pretty much only be done on a 6700K with OC and fast memory. There is no such thing as "fast enough".
 
Not CPU limited for 5 years? Try reach 60FPS in Fallout 4 in a heavy area for once. That can pretty much only be done on a 6700K with OC and fast memory. There is no such thing as "fast enough".

That's interesting. I haven't gotten around to buying it yet. (I'm still working my way through New Vegas, but I'll get there eventually)

I wonder what could possibly be driving up the CPU use so high? It's not a particularly advanced game...
 
That's interesting. I haven't gotten around to buying it yet. (I'm still working my way through New Vegas, but I'll get there eventually)

I wonder what could possibly be driving up the CPU use so high? It's not a particularly advanced game...

Warhammer is another game you are massively CPU limited in, DX12 or not. And the list could go on.
 
That's interesting. I haven't gotten around to buying it yet. (I'm still working my way through New Vegas, but I'll get there eventually)

I wonder what could possibly be driving up the CPU use so high? It's not a particularly advanced game...

Typical bad coding is my guess.
 
  • Like
Reactions: Ruoh
like this
Gamers on a budget is your usage scenario.

As time goes on more cores is becoming more relevant. Doom addresses up to 12 threads. BF4 addresses up to 8 threads. BF3 addressed up to 6. Vulkan, Mantle and DX12 are making these ancient high core AMD chips relevant. Now people also want modern platforms with PCI-E 2, m.2 etc. That being said, folks with 8350s are now playing Doom as just well as folks with Intel Quads. That being said, yes Intels are more powerful but AMD still has the bang for the buck crown. Not everyone gives a shit about brands. Some folks just care about getting the most for their money. AMD still delivers that in the latest titles thanks to these new APIs.

Intel isn't the answer in every scenario.

The ability to use more cores doesn't matter if those cores aren't stressed. That's why, despite DOOM using 12+ threads, the i3 outperforms the FX-8350. It's the same basic dynamic: If no INDIVIDUAL CPU core gets overworked, the faster per-core performance wins the day. The advantage you gain by threading is you spread out the workload among cores, preventing any individual one from bringing the entire process to a halt. But for games, where 90% of the work is handled by two VERY large threads (The main executable and the primary rendering thread), a dual-core with hyperthreading is enough to run the application without running into CPU side bottlenecks, regardless of how many threads the program is capable of using.

I wonder what could possibly be driving up the CPU use so high? It's not a particularly advanced game...

Bethdesia has done a LOT of shortcuts over the years. The most famous example was Skyrim, which uses a hacked 32-bit executable, rather then properly compiling a 64-bit one. The result was a "win-64" executable that invoked Win32 .dll's. That's part of the reason their titles run terribly. Not sure if this is the case with FO4, but I've seen no evidence they've upgraded their engine to actually use Win64 properly.
 
I was asking a question not remotely trying to make a point. I did not call anyone a name at any time. And even if I had been trying to paint with a wide brush using my "loyalists" comment (I wasn't), I still wouldn't consider that an insult.

Thank you for somewhat answering the question, but you might want to get that chip on your shoulder looked at.
Yeah this does not work for me when you type this:
Apart from price and brand loyalty, why would anyone purchase an AMD chip that is competitive only with 5-year old hardware from a competitor?

No one would buy AMD stuff that is outdated by 5 years as hardware upgrade , this suggests that you have an opinion of people that do want to buy this "instantly outdated" Zen ..
Also very good insight as to what the Zen retail performance would be.
 
I like having a mix of different hardware: Intel, AMD, Nvidia, components from different OEMs, just to play with and experiment.

Typically I try to switch between Intel/Nvidia and AMD/AMD for every other rig. Kind of broke that trend recently (since I needed a VR rig and AMD's current CPU offerings aren't enough) but I plan my next rig to be all AMD.

For me, it's just fun to test out different types of hardware and different brands.
 
Back
Top