AMD Zen CPUs?

[Pieter3dnow]

Wrong. You are looking too hard at the surface and the inane posting of diehard Intel-ites that to this day still use Cinebench 11.5 to show IPC diff when each uses a different Instruction set.

If you look at the pipelines and the per cycle process AMD = 3 Intel = 4 then you quickly see that @4Ghz why AMD is behind. They constructed the architecture to scale to 8Ghz there-by nearly doubling that per cycle process. You cant think in terms of actual clocking results and see that they were basing on future prospects that NOW we know would not come to fruition.

Ranting and raving at short-sighted references proves nothing. Look at the white pages of the designs and for-go the benches as they tell you little.

It is such a big pie in the sky story that this almost brought me to tears. It is simply not true. If Bulldozer was close to this there would be a serious effort on getting their instead AMD dropped Bulldozer for past several years the only reason why it is still being made is prolly due to the wafer commitment with GF.

The white pages might as well had pigs can fly and the corresponding schematic attached to it does not hold _any_ value.

 

[MaddMutt]

K10.5 already did get shrinked to 32nm. It was called Llano. It didn't clock any better than Phenom II did on 45nm -- in fact it clocked worse, topping out at only 3.6ghz best case scenario. It had slightly higher IPC (6~7%) but this was offset by the inferior clocking ability. They had nowhere to take the K10.5 design, it was at its wall and at end of life status. Don't forget that K10 itself was basically still K8, itself a design that originated in the late 90's and initially released in 2003.

Yes I know that it was shrunk to 32nm but it also had a IGPU stuck in with it.
What I'm saying, is if they had stuck just with the CPU, they could have easily had an 8 core 10MB L3 @32nm.
The tweaks would of course have been SSE4a,FM3,FM4,and others. There was no need for the Dozer. The IPC would have improved as the chip moved to a 16-18 stage pipeline. This would have also increased the speed to over 4GHz+.

 

[MaddMutt]

We don't know how much of Llano problem was due to being first apu design.

First bulldozers had around 600Mhz advantage which wasn't enough to outperform PII
http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested

Also from that review - 6 real cores thuban - 900 milion transistors
4 module/8 "core" bulldozer - 1200 milions

True, AMD didnot work on the CPU part of Llano to improve it.
They already had a CPU handy, When they started looking for a place to put their NEW APU.

Like I posted earlier, AMD has the right idea but no R&D $$$$$. Intel will see that xx is the way to go and with their $$$$$$, will end up doing it better.

AMD had the right idea on the monolithic core but @ 65nm, it was not the right time. Intel waited until they had the uarch and @ 45nm did it better than AMD. You can do this when you control 80%+ of the market.

I Think Intel is paying AMD to screw up, BUT also paying them to stay in the x86 game.
As some one mentioned earlier, if AMD folds, no one gets to step in and claim their patents. This would leave Intel with over 90%+ market share!!!
Since they are based in the USA, that would qualify them, by the GOVERMENT as a monopoly!!!!
If you are old enough, you will remember the IBM, AT&T, Microsoft, ect, ect, monopolies and what happened to them.

 

[durquavian]

It is such a big pie in the sky story that this almost brought me to tears. It is simply not true. If Bulldozer was close to this there would be a serious effort on getting their instead AMD dropped Bulldozer for past several years the only reason why it is still being made is prolly due to the wafer commitment with GF.

The white pages might as well had pigs can fly and the corresponding schematic attached to it does not hold _any_ value.

I think you keep putting the cart before the horse. You assume they built bulldozer based on the existence of a 8Ghz possibility, when in fact they built it based on the future existence that never came to fruition. At that point they had no choice but to move forward regardless of the actual speed the chip would attain. You keep acting like we are saying they had an 8Ghz processor when what we said was they built around the possibility of one that never came to be.

And you realize the white pages document the architecture and the possibilities right, cause it doesn't sound like you do.

Fact: at Haswells release it barely performed 5% better than Ivy. Looking at the white pages one could discern a far better performance increase. Now as of today that 5% has grown to as much as 10% or more. How? Did the chip mature under use conditions? Did it miraculously learn to do better (that would be awesome)? No! What did happen is software began to better make use of the architectural advantages and instruction sets ie: AVX2 in Haswell.

So bear in mind that the 8150 bulldozer is performing far better today than it did at release = no changes in hardware. Reason = better use of the architecture. The reason windows was known as wintel. Today there are a number of users that speak to windows running near flawless with the same AMD CPUs that didn't so well earlier.

So no need for the doom and gloom approach as you attempt to convey. In reality the bulldozer model works very well albeit not as well as it was dreamt of so many years ago.

 

[Wascrash]

When you look at these reviews http://www.tweaktown.com/tweakipedi...80-sli-vs-gtx-980-sli-at-2560x1440/index.html and then compare to the H reviews that says the AMD chips don't scale with sli and are a bottleneck..confusing.

 

[durquavian]

Depends on targets and games being benched. As things have been progressing CPU performance is not quite as pivotal as it was. And @4K where the cap is 60fps, most CPUs can easily achieve.

 

[Ramses]

So bear in mind that the 8150 bulldozer is performing far better today than it did at release = no changes in hardware. Reason = better use of the architecture. The reason windows was known as wintel. Today there are a number of users that speak to windows running near flawless with the same AMD CPUs that didn't so well earlier.

So no need for the doom and gloom approach as you attempt to convey. In reality the bulldozer model works very well albeit not as well as it was dreamt of so many years ago.

I've noted this in all aspects of use. These chips are going to remain relative in the economy performance segment for awhile to come imo. I don't think I've ever seen hardware actually get better with age till now. I also hadn't till recently seen a new version of Windows run better/faster on a given piece of existing hardware than it's previous version. That happens too now, with both 7 and 8 in my experience.

 

[Ramses]

When you look at these reviews http://www.tweaktown.com/tweakipedi...80-sli-vs-gtx-980-sli-at-2560x1440/index.html and then compare to the H reviews that says the AMD chips don't scale with sli and are a bottleneck..confusing.

I think in many ways hardware reviews are getting carried away with minutiae.
It isn't that specifics aren't valid and interesting, but it wouldn't hurt to preface a lot of
such things for the less uber technically inclined with "they both run X fine, but this one runs it better and here's how so" or some such. Tweaktown did that toward the end of that review and I appreciate it, but it's not as common as I might hope. I understand professional impartiality too but it don't take a genius to know if something runs fine for less money. I don't see the practical quality and frankness in hardware reviews today I seem to remember seeing ten or more years ago. It feels like business now.

 

[durquavian]

I've noted this in all aspects of use. These chips are going to remain relative in the economy performance segment for awhile to come imo. I don't think I've ever seen hardware actually get better with age till now. I also hadn't till recently seen a new version of Windows run better/faster on a given piece of existing hardware than it's previous version. That happens too now, with both 7 and 8 in my experience.

When you consider that the architecture was brand new with no other x86 similar it is understandable and explains why we are seeing so much better performance.

 

[MaddMutt]

When you consider that the architecture was brand new with no other x86 similar it is understandable and explains why we are seeing so much better performance.

You can say the same about the Intel P IV, when they introduced SMT. At first, it was recommended to turn HT off because windows could not tell the difference. (Sound Familiar)
The software, at that time, was not coded on the desktop for multiple threads. Windows scheduler had to be reworked for SMT, or the whole system was inefficient. If you look back, even though Intel dropped SMT in their CORE CPU's, it still took years before the software was there. That was just for FULL Dual Core Cpu's, lets not forget the software lag for Quad Core Cpu's.
I have read and heard here in this forum, that current software is designed for Quad's.
Is this because Intel's #2 seller, is a Quad W/O HT??????
I know there is a HUGH increase, going from a Quad W/O HT to a Quad that does.
It's the same with AMD, going from a FX-4xxx to a FX-8xxx.

 

[defaultluser]

I've never seen the claim that Phenom II had hit it's limit on 45nm as a valid justification for releasing Bulldozer. This is the same bullshit excuse Intel made to themselves to justify the complete redesign that was the Pentium IV, which was late to release and under-performing. It was proven to be complete bullshit after a tweaked P3 core turned into the Pentium M, which was again revised into Core and Core 2.

Since the mistake that was the Pentium IV disaster, every new chip in the Intel lineup builds on the previous generation, so there is no "starting over" (and thus massively reduced risk). You see that in Nehalem, which was basically a Core 2 execution unit and tweaked front-end with an entirely new communications system (this includes the memory controller, ring bus, brand new cache), with HT and redesigned power states. They also increased the pipeline length slightly to enable higher clock speeds. REALITY: not a new core design, and as such it sometimes got outperformed by Core 2 Quad, but on the whole it was an improvement.

Since the introduction of Nehalem, Intel has added wider fetch/AVX and trace cache to Sandy, added new FMA and execution resources in Haswell, along with a complete remaster of the power states and voltage control. Through this, they have kept the basic architecture, tweaking performance while lowering power. They make this work even though there are two separate teams developing processors - they TALK TO EACH OTHER (what a concept)!

While you can sorta forgive Intel's stupid Netburst decision made in the mid-90s, it's a lot harder to understand AMD's similarly bad decison 10 years later, when so much more was known about process technology limits, and the risk inherent in developing/testing/optimizing a radical new core design SHOULD have been better understood by management. When AMD did not build on the success that was Phenom II (a successful core that just needed some power controls and core tweaks added to make it a competitive product in notebooks versus Nehalem), they were shooting themselves in the foot. There is nothing even remotely quantifiable as "reused" in Bulldozer, which was why it took so many years to get a core rev that did not suck (I'm including the delays before it was released in that total). They already took the time and effort to add full power states with Llano, so with just a little bit more effort they could have made a higher-clocking, better-performing revision.

 

[MaddMutt]

defaultluser [H]ardness Supreme, 9.0 Years

+1

Not sure how to give you a cookie.

That was well said.

 

[NaroonGTX]

It's not that K10 hit its limit on 45nm, it's that it pretty much hit its limit period. Yes, Intel went back to P3 and (very heavily) modified the piss out of it and built much better architectures out of it, but that doesn't necessarily mean AMD could've done the same with K8. AMD's brilliant management thought it was a good idea to get rid of their original ideology of having small teams of experienced, good engineers to having massive teams of inexperienced engineers working on their CPU's. That's why AMD seemingly got lazy around the mid-00's and all they did was just bolt on more cores to their pre-existing designs, and didn't make any huge strides after Core 2 came out. It's silly to think this AMD could do some major re-work with "K10". They did some small changes with Llano but it didn't amount to much, and that was on 32nm.

I see people keep saying "all they had to do was add two more cores (to Istanbul I guess) and put it on 32nm and they would've had a great competitor!" but this is obviously bunk for reasons I shouldn't even have to explain.

 

[Pieter3dnow]

So no need for the doom and gloom approach as you attempt to convey. In reality the bulldozer model works very well albeit not as well as it was dreamt of so many years ago.

It is not about doom and gloom it is about the feasibility of the Bulldozer line as much as you said with the Phenom II the Bulldozer has a problem. Even if the performance for me is not an issue it is not the core AMD sees fit building a future for itself.

The benchmarks from Oxide Games' Dan Baker shows that it is still very much a decent cpu at modest number of batches using Mantle. But if this number grows over 100K will it still be something that will work well. Not that we are going to see games that do that stuff but it might be within a year or so from now.

 

[Michaelius]

It's not that K10 hit its limit on 45nm, it's that it pretty much hit its limit period. Yes, Intel went back to P3 and (very heavily) modified the piss out of it and built much better architectures out of it, but that doesn't necessarily mean AMD could've done the same with K8. AMD's brilliant management thought it was a good idea to get rid of their original ideology of having small teams of experienced, good engineers to having massive teams of inexperienced engineers working on their CPU's. That's why AMD seemingly got lazy around the mid-00's and all they did was just bolt on more cores to their pre-existing designs, and didn't make any huge strides after Core 2 came out. It's silly to think this AMD could do some major re-work with "K10". They did some small changes with Llano but it didn't amount to much, and that was on 32nm.

I see people keep saying "all they had to do was add two more cores (to Istanbul I guess) and put it on 32nm and they would've had a great competitor!" but this is obviously bunk for reasons I shouldn't even have to explain.

All we are saying is that they would be in better position that they are now without designing whole new architecture from zero.

Also as I side note I think realistic target for Zen will be to match single threaded performance of Sandy Bridge and then with cheaper mobos and unlocked cpu <200$ they would at least be worth buying for enthusiasts on the budget.

 

[durquavian]

But isn't that the issue at heart? Remember my Haswell 3-5% at launch scenario that now demonstrates 10% or more over Ivy? The issue with bulldozer to a great extent was the initial snails pace to update drivers and get the software to be more efficient, in other words use Bulldozers strengths. Today, more-so with Piledriver, it performs quite well in the mid range and struggles to meet the extreme (an incredibly small market). But what has to be realized is that the software coding has more to do with performance than the physical architecture hence the increase in performance from launch.

 

[cageymaru]

14nm? Anyone ready for 14nm in 2016? And DDR4 of course.

Summit Ridge says, "Hi!"

http://www.sweclockers.com/nyhet/19954-amd-zen-forst-ut-i-summit-ridge-for-14-nanometer

 

[NaroonGTX]

That article didn't really reveal much new, but at least if they're on 14nm they'll be competitive with Intel on that front. A bit iffy on how they claim it's been "pushed back" to Q3 2016, I was expecting either a late Q1 or Q2 release. We'll see though. I'm looking forward to later this year (probably Computex in May or something) when they do a presentation on K12 and Zen.

 

[R3MF]

interesting that the suggestion is they will arrive on socket [F]M3.

 

[Pieter3dnow]

That article didn't really reveal much new, but at least if they're on 14nm they'll be competitive with Intel on that front. A bit iffy on how they claim it's been "pushed back" to Q3 2016, I was expecting either a late Q1 or Q2 release. We'll see though. I'm looking forward to later this year (probably Computex in May or something) when they do a presentation on K12 and Zen.

If that article is true then I would hope they start making samples before the end of the year if things needed to be improved then they still can make 2016 Q3.

interesting that the suggestion is they will arrive on socket [F]M3.

The thing is they outline power use also that would make it an APU by default.

 

[darkeen]

The thing is they outline power use also that would make it an APU by default.

I'm not sure 95W power usage implies it's an apu, but given amd's HSA push I would be surprised if it weren't an apu by default.

 

[Dangman]

But isn't that the issue at heart? Remember my Haswell 3-5% at launch scenario that now demonstrates 10% or more over Ivy?

Proof of this?

 

[NaroonGTX]

I'm not sure 95W power usage implies it's an apu, but given amd's HSA push I would be surprised if it weren't an apu by default.

I think just about every upcoming chip will be an APU. Even the "high-end" stuff.

Lisa Su stated that their partners have requested AMD's return to being competitive in terms of high-performance server/enterprise processors (very specific words), so it's possible we may see a CPU-only variant as well. I think in general, there will be a lot of different SKU's that can be dropped into the same singular socket.

 

[durquavian]

Proof of this?

You read enough forums you see plenty proof of this. And really it isn't that hard to understand, when software is written to the architecture and takes advantage of it it will run better and more efficient.

 

[Dangman]

You read enough forums you see plenty proof of this.

That's it? So no real proof then?

I've been on enough forums to know to discount most user/forum poster written performance tests/posts due to their lack of impartiality, placebo effects, data manipulation, improper methodologies for testing, use of synthetic benchmarks, etc. Not to mention that telling to someone to "read enough forums" isn't exactly a solid method of backing up your claims. It's just as lazy as telling someone to "google" for evidence for your arguments.

And really it isn't that hard to understand, when software is written to the architecture and takes advantage of it it will run better and more efficient.

It isn't hard to understand. However, I'm asking for actual legitimate proof showing that Haswell shows a 10%+ performance in real world apps and games and not some bullshit synthetic benchmarks.

 

[durquavian]

That's it? So no real proof then?

I've been on enough forums to know to discount most user/forum poster written performance tests/posts due to their lack of impartiality, placebo effects, data manipulation, improper methodologies for testing, use of synthetic benchmarks, etc. Not to mention that telling to someone to "read enough forums" isn't exactly a solid method of backing up your claims. It's just as lazy as telling someone to "google" for evidence for your arguments.


It isn't hard to understand. However, I'm asking for actual legitimate proof showing that Haswell shows a 10%+ performance in real world apps and games and not some bullshit synthetic benchmarks.

Oh my god you are gonna be the snippy one , ehh. Well it isn't too hard when looking at Video encoding. It definitely has an advantage there AVX2 which again requires the software to be coded to it. Gaming is not likely gonna see huge improvements as they generally stick with the more common code and rarely if never optimize it for any particular CPU brand or generation.

Seriously is that everyone likes being a pessimist or refuses to open there minds to truth? Just a slight bit of intelligence and rationale would lead to the conclusion I mentioned, especially after reading the white papers on the architecture. It is how it works and always has. Windows has definitely improved GREATLY for AMD over the last 3 years, to the point many now feel AMD is a far better/smoother experience. Again it leads to driver and software updates that better make use of the hardware. VIOLA!

 

[Dangman]

Oh my god you are gonna be the snippy one , ehh. Well it isn't too hard when looking at Video encoding. It definitely has an advantage there AVX2 which again requires the software to be coded to it. Gaming is not likely gonna see huge improvements as they generally stick with the more common code and rarely if never optimize it for any particular CPU brand or generation.

Seriously is that everyone likes being a pessimist or refuses to open there minds to truth? Just a slight bit of intelligence and rationale would lead to the conclusion I mentioned, especially after reading the white papers on the architecture. It is how it works and always has. Windows has definitely improved GREATLY for AMD over the last 3 years, to the point many now feel AMD is a far better/smoother experience. Again it leads to driver and software updates that better make use of the hardware. VIOLA!

How is it snippy to ask for legitimate proof of what you're talking about?

It's more of a case of most of us being extremely disappointed by AMD in the past with the release of the Zambezi CPU. If you were around then, you would have seen that many of us were extremely hopeful that AMD would finally have a serious contender against Intel. Many people used similar methods of rationalizing like you are doing right now to show that AMD will finally turn the tide against Intel. Zambezi turned out to be effectively a failure when actual data was shown.

As such, it's more of cautious wariness than any out-right pessimism or close-mindedness. This time around, more solid evidence is needed. Which you still haven't provided by the way. You have provided relatively solid hypotheses and predictions but you have not provided the data to support said hypotheses and therefore prove said hypotheses/predictions. That's part of having an open-mind: Looking out for actual data.

Also, in addition to the request for "proof showing that Haswell shows a 10%+ performance in real world apps and games and not some bullshit synthetic benchmarks", now I'm requesting that you provide the proof showing that AMD is now the far better/smoother experience with Windows.

 

[Michaelius]

I'm guessing it's more about software finally using more cores so whole FX line went from always being horribly outperformed by i5 (and i3 sometimes) to sometimes getting close

 

[Ramses]

Software really has improved, and I expect DX12 and Windows 10 to improve it further.
Back in the possibly not so good old days, a new version of windows generally meant
new cpu time and double your RAM. Since Windows 7 it's been almost the opposite.
My FX box is faster on 8 than 7 in general use, no doubt. Games come out taking advantage
of a handful of cores and run better than what was out a few years ago. Yeah, it's been nice.
Sucks for AMD/Intel I guess in a way, I don't see needing to upgrade for the next couple of years
unless something dramatic comes along and suddenly uses the 75% of my processing capacity that
sits around doing nothing most of the time lol. I'm not sure where you go from "instant" performance wise.
Maybe software that opens and does things before I want them to?

 

[durquavian]

How is it snippy to ask for legitimate proof of what you're talking about?

It's more of a case of most of us being extremely disappointed by AMD in the past with the release of the Zambezi CPU. If you were around then, you would have seen that many of us were extremely hopeful that AMD would finally have a serious contender against Intel. Many people used similar methods of rationalizing like you are doing right now to show that AMD will finally turn the tide against Intel. Zambezi turned out to be effectively a failure when actual data was shown.

As such, it's more of cautious wariness than any out-right pessimism or close-mindedness. This time around, more solid evidence is needed. Which you still haven't provided by the way. You have provided relatively solid hypotheses and predictions but you have not provided the data to support said hypotheses and therefore prove said hypotheses/predictions. That's part of having an open-mind: Looking out for actual data.

Also, in addition to the request for "proof showing that Haswell shows a 10%+ performance in real world apps and games and not some bullshit synthetic benchmarks", now I'm requesting that you provide the proof showing that AMD is now the far better/smoother experience with Windows.

Snippy as in negative, naysaying. And at no point was any of what I said about AMD being better than Intel or rationalizing its place. It is simple FACTS that software when coded to utilize the architecture/hardware will increase performance: should be a given though you seem hard-pressed to see it. And the 10% proof isn't hard to find when you look at video-encoding using AVX2, thought that made it more obvious. I have no desire to spend the time looking it back up and thumbing thru pages of posts for the actual proof.

And as far as AMD on windows, anyone will be hard pressed to find quantitative proof, benchmarks. It is a personal preference and I alluded to it as such when I said FEELS smoother, snappier.

I trust users far more than any tech site. Sure there are many who try to make it sound better than it is. But there are a ton of trustworthy users on both sides that give great accurate info. Tech sites do real crappy work with Overclocks and testing. You know they are bad when your 4.6Ghz CPU and slower ram with HDD is beating their 5.0Ghz and faster ram with SDD in benchmarks.

 

[maxius]

TOPIC ON BACK

unified socket and zen is looking to be a cpu only
http://wccftech.com/amd-bristol-ridge-apu-2016/

so essentially the consumer space will be this mobile then mainstream desktop and the server chip will live its own life

 

[lutjens]

Q3 2016 for the performance CPU? At that point, why even bother? It'll likely be obsolete when it's launched...it'll be going up against Skylake...

 

[Dangman]

Snippy as in negative, naysaying. And at no point was any of what I said about AMD being better than Intel or rationalizing its place. It is simple FACTS that software when coded to utilize the architecture/hardware will increase performance: should be a given though you seem hard-pressed to see it. And the 10% proof isn't hard to find when you look at video-encoding using AVX2, thought that made it more obvious. I have no desire to spend the time looking it back up and thumbing thru pages of posts for the actual proof.

Then don't make claims if you're not willing to back them up when people ask for proof of said claim. Simple as that.

Again, your hypothesis is solid. But you're missing the data that proves that said hypothesis is actually happening.

And as far as AMD on windows, anyone will be hard pressed to find quantitative proof, benchmarks. It is a personal preference and I alluded to it as such when I said FEELS smoother, snappier.

So basically placebo effects then. In other words, useless.

I trust users far more than any tech site. Sure there are many who try to make it sound better than it is. But there are a ton of trustworthy users on both sides that give great accurate info. Tech sites do real crappy work with Overclocks and testing. You know they are bad when your 4.6Ghz CPU and slower ram with HDD is beating their 5.0Ghz and faster ram with SDD in benchmarks.

While there are plenty of bad tech sites, users are not easily held accountable as a tech site is. Not to mention that in my experience, users are extremely prone to placebo effects and complete unwillingness to provide actual data for their claims. As such, I don't put a real heavy value on user performance results.

Q3 2016 for the performance CPU? At that point, why even bother? It'll likely be obsolete when it's launched...it'll be going up against Skylake...

If you're the extremely hopeful type, if AMD can get their single-threaded performance up to around Ivy Bridge or Haswell level AND have 6 or more cores, that could be a huge boon for AMD. Another factor that could help AMD is if Mantle is more widespread in 2016 thereby limiting the effects of a faster CPU.

 

[durquavian]

Then don't make claims if you're not willing to back them up when people ask for proof of said claim. Simple as that.

Again, your hypothesis is solid. But you're missing the data that proves that said hypothesis is actually happening.

So basically placebo effects then. In other words, useless.


While there are plenty of bad tech sites, users are not easily held accountable as a tech site is. Not to mention that in my experience, users are extremely prone to placebo effects and complete unwillingness to provide actual data for their claims. As such, I don't put a real heavy value on user performance results.

If you're the extremely hopeful type, if AMD can get their single-threaded performance up to around Ivy Bridge or Haswell level AND have 6 or more cores, that could be a huge boon for AMD. Another factor that could help AMD is if Mantle is more widespread in 2016 thereby limiting the effects of a faster CPU.

You cant just throw the term placebo out every time the result cant be quantified. There have been plenty of true double-blind tests for determining if one could tell the difference between Intel and AMD in gaming and most could not tell. As far as being smoother/fluid as some claim, I have seen it to, but there is no way to prove it with data/bench, it is just a feeling/observation. If you choose to not believe then it is your choice but to outright ignore or condemn does not make it any less real.

.

 

[Dangman]

You cant just throw the term placebo out every time the result cant be quantified. There have been plenty of true double-blind tests for determining if one could tell the difference between Intel and AMD in gaming and most could not tell. As far as being smoother/fluid as some claim, I have seen it to, but there is no way to prove it with data/bench, it is just a feeling/observation. If you choose to not believe then it is your choice but to outright ignore or condemn does not make it any less real.

In fact I can throw out the term placebo every time the result can't be quantified. I've seen plenty of situations in multiple forums where posters try to justify their poorly researched purchases (faster RAM, AMD CPUs over $170 price point, etc) by throwing out that they "feel" something different with said purchases. The problem with that it doesn't allow for any sort of conclusive debate since I can't see what you feel. In fact, it's a mere attempt to shut-down debate because you can't see what someone else feels. Not to mention, then those posters would try to present their "feels" as actual fact that applies to everyone. Which is a load of bullshit.

 

[durquavian]

In fact I can throw out the term placebo every time the result can't be quantified. I've seen plenty of situations in multiple forums where posters try to justify their poorly researched purchases (faster RAM, AMD CPUs over $170 price point, etc) by throwing out that they "feel" something different with said purchases. The problem with that it doesn't allow for any sort of conclusive debate since I can't see what you feel. In fact, it's a mere attempt to shut-down debate because you can't see what someone else feels. Not to mention, then those posters would try to present their "feels" as actual fact that applies to everyone. Which is a load of bullshit.

And this is where you fail to see fact. Placebo means a false positive. SO in a double-blind test when users can not tell the difference between two systems one AMD and the other Intel, both using the same GPU, then by your account it is a placebo effect because no benchmark review would prove this. Although there was one using the 7990(I think) and the 9590 for AMD and the 7990 and the Intel 3960x(the extreme one may have number wrong). The results showed that AMD in a title or 2 was behind the Intel but not by much. In the other titles the diff was non-existent or extremely hard to discern.

Sounds more to me like you just don't want to believe it more than it being real or not. I have seen it and so have others. Is it worth throwing away your Intel to get AMD? No. But it also shows that there is more to experience than any benchmark or review can show. Intel has the advantage, as I have already stated. In the basic market it isn't so easy to see ie: 60hz/fps 1080p. Move up to more expensive and extreme solutions and it is quite a bit bigger. Again if that is what you are after then get Intel. As I have advised many who have asked recently, right now the best option if you are desperate to get a new computer and have no preference then I say go i7 4770. AMD for good performance is going cost a bit even if the CPUs are cheaper. You will need more cooling and a high quality board which then offsets that price advantage. But if it is a small matter then go AMD and have fun OCing it to the moon, you wont regret it.

(Seriously I had hoped this negative AMD bashing would have remained at OCN. It gets old fast)

 

[Ramses]

This is a known phenomenon throughout human history.
Phrases like "more than the sum of it's parts" exist because of it.

There is a page in one of Richard Bach's books, A Gift Of Wings maybe, that speaks to why
an aircraft that is one of the most rigidly engineered and documented machines on earth, clearly
known and stated to take x feet to lift off, can be coaxed into doing so with much less.
This too speaks to such, and you either get it or you don't. I feel genuinely sorry for folks that do not.

 

[Dangman]

And this is where you fail to see fact. Placebo means a false positive. SO in a double-blind test when users can not tell the difference between two systems one AMD and the other Intel, both using the same GPU, then by your account it is a placebo effect because no benchmark review would prove this. Although there was one using the 7990(I think) and the 9590 for AMD and the 7990 and the Intel 3960x(the extreme one may have number wrong). The results showed that AMD in a title or 2 was behind the Intel but not by much. In the other titles the diff was non-existent or extremely hard to discern.

The placebo effect, according to Google, is:

A remarkable phenomenon in which a placebo -- a fake treatment, an inactive substance like sugar, distilled water, or saline solution -- can sometimes improve a patient's condition simply because the person has the expectation that it will be helpful.

So let's break it down. Earlier you made the claims that an AMD system is smoother/fluid. There's no empirical data supporting that AFAIK. What if someone actually bought an AMD system based on what you claim? They buy that AMD setup, use it, and because they were told that it was smoother, believes that AMD is smoother?

(Seriously I had hoped this negative AMD bashing would have remained at OCN. It gets old fast)

No it's more of me wanting more solid evidence. I've been a computer enthusiast long enough to know not to fully trust the words of other computer hardware enthusiasts. I've learned a long time ago to get more data and evidence. It's not bashing if we're asking for proof. If there's solid evidence showing AMD is the better solution, then I want to see it. I also don't like seeing unsubstantiated claims made on the forums as well. Stuff like that spreads before real information is found and therefore continues a cycle of misinformation.

Just to be clear: I have no problem with other people recommending AMD. I do have a problem with people recommending AMD based on poor or unsubstantiated claims or stances or information.

 

[Kai]

This is a known phenomenon throughout human history.
Phrases like "more than the sum of it's parts" exist because of it.

There is a page in one of Richard Bach's books, A Gift Of Wings maybe, that speaks to why
an aircraft that is one of the most rigidly engineered and documented machines on earth, clearly
known and stated to take x feet to lift off, can be coaxed into doing so with much less.
This too speaks to such, and you either get it or you don't. I feel genuinely sorry for folks that do not.

Wut.

That's because things are built with certain tolerances. For example, engine components are specifically engineered to be able to withstand a certain amount of stress for so long, but that doesn't mean that they can't be stressed further, just that they aren't guaranteed any further. That has nothing to do with being "more than the sum of its parts" at all. Nor does it have anything to do with this topic. That phrase insinuates a certain emotional element that we humans apply to machines. The machine will be able to perform up to when it fails no matter what we feel or think about it.

 

[Pieter3dnow]

I'm guessing it's more about software finally using more cores so whole FX line went from always being horribly outperformed by i5 (and i3 sometimes) to sometimes getting close

Older software on new cpu is something AMD does not seem to care for. And in way they are right, if you can do processing a lot faster with better code why make a cpu that runs only old code fast?

Q3 2016 for the performance CPU? At that point, why even bother? It'll likely be obsolete when it's launched...it'll be going up against Skylake...

I think you can find a lot of the same sentiment when AMD did not release on lower nm and people kept saying but Intel is going to launch 14nm "yesterday" 2 years ago already.

When people that praise Intel should learn one thing Intel does not care about is your expectations and what is on the roadmap the roadmap is not holy to them if there is no competing product on the market there not going to burn their inventory because you think they said they would introduce something different.
It is not that big of a deal since they control most of the x86 market it will not impact their financial status.