AMD Piledriver rumours ... and expert conjecture

[Reynod]

We have had several requests for a sticky on AMD's yet to be released Piledriver architecture ... so here it is.

I want to make a few things clear though.

Post a question relevant to the topic, or information about the topic, or it will be deleted.

Post any negative personal comments about another user ... and they will be deleted.

Post flame baiting comments about the blue, red and green team and they will be deleted.

Enjoy ...

 

[jimmysmitty]

They will sell and that is the main thing,but certainly they will have a server presence and they say Vishera by years end,by then Trinity will have sold
and I read that they may have Steamroller apu by sept. and excavator apu full HSA by 2013 Q2 .
Now that's following their new Tick Tock schedule.

As far as I can tell I see AMD shares hitting 12$ to15$ by Q4 2012 by Q4 2013 they may well be back to 25$ or more.
They wiil own low and medium PC format as well as make inroads into high end with APU desk top models .
pricing today is king Fancy like my Intel 2500K is out priced by the APU that price ratio is the difference that the Billions of people on this planet will see.
People from China ,India ,Russia,etc...sees price,cores,Freq. The fanatics with money to blow may buy Intel and businesses will probably stay Intel.
But the burgeoning ,yearning billions of people of India, China know value not fancy.The 40 million Americans on food stamps now know value too.
all the consoles will most likely be APU .
Beyond 2014 Intel may have solutions as good but they can't be expected to snap their fingers and produce as good igp until then.
By Haswell 2014 long after release. The Apu is a moving target until at least 2014.Price wise Intel has to be competitive.
Not even Intel can hold back evolution,the HSA revolution will continue and will those who don't swim behind in it's wake.

A mass amount of people in other countries are one thing but you forget, the majority in China are lower middle class and cannot afford the same luxurious CPUs/GPUs we do. Add to that that China is thinking of creating their own Goverment approved CPU and that will kill all hope of Intel or AMD taking that market.

As for Russia, mostly the same except its more about money. There a lot of things are luxuries, even orange juice that we here in the states complain about paying a couple bucks for, they don't even get that kind of stuff or if they do its like an expensive bottle of wine. Trust me, my wife is Russian and her family tells her this all the time.

Either way, the business for CPUs is businesses and servers. Consumer desktops don't make nearly the same amount of money. Thats where AMD needs to focus again instead of low end desktop.

I agree AMD and INTEL both want to bundle their own chipset with the mobo ... expecially for APU's.

For AMD squeezing NVidia like that must be some recompense for the way NVidia have worked to compete in the discrete graphics market and do so well ... despite yeild issues and heat over the last two ticks ... they still produce a more refined graphics product on average ... drivers / support / etc.

It still leaves NVidia without an X86 license and with APU's touted take a bigger slice of the market as users want portability ... thats a tough area to be in at the moment.

I find it hard as an AMD fan to dislike NVidia ... they keep producing the goods.

I find it much easier to dislike Intel ... though my i7 2600 makes it hard to dislike them ... hard indeed.

Oh the pain ... the pain of it all jimmy !!!

Hopefully trinity won't be one great flash and then its over ... < two little puzzles there for the older folk>

nVidia does do good. I may not buy them, but they make good GPUs. And Trinity will be fine for a while. While Intel is making great strides on their GPUs, even a Intel hater has to admit their IGP is now viable, I don't see them blowing AMD out of the water just yet. Maybe if they find a way to stuff Larrabee into it. Or is that Knights Corner now......

you sure about that?

http://developer.amd.com/sdks/AMDAPPSDK/Pages/default.aspx

The problem is AMD developed it and didn't push it at all, until recently. Hopefully this is a sign that Read will push more development into AMD's techs instead of saying "this is what it can do, now use it".

http://blogs.amd.com/fusion/2012/04/24/adobe-and-amd-enable-brilliant-experiences/

In fact, a mainstream notebook PC based on the AMD A8-3530MX APU is up to 672% faster when accelerated by the horsepower of the AMD Radeon™ graphics technology in the APU.*

627% faster than what? Than Quick Synce gen 1, 2 or than the APU without it? I would assume such a large number is against the APU without it not QS since even nVidias version isn't quite a QS killer yet. And it may stay that way since Intel does have the one up on them in terms of time. I imagine Haswell will give a QS 3.0.

Granted VCE is more geared directly to video encoding and should be faster, ATI Stream is also usable for Video encoding.

Look familiar yet?

Most enthusiasts here have learned the hard way to never count AMD chickens before they are fully hatched & independently benchmarked 😀..

It was Kenstfield wont beat Barcelona. Still same arch but it was more thaat people were spouting that the FSB would be a major bottleneck for desktops and that Barcelonas monolithic die and IMC with HyperTransport would cream Kenstfield MCM FSB approach.

Still people wont learn. Never say anything cannot happen. People also said Intels IGP couldn't play graphic intensive games but SB plays Portal 2 very well and thats a pretty nice looking game. Of course it looks even better maxed out but still, anything could happen.

 

[sarinaide]

More doom and gloom prophacy, at the end it all looks like "AMD will never do anything in the CPU market again, it is over and we must all use Intel because somehow AMD doesn't give me ample enough performance to run winamp and I feel word is not quick enough". It also appears to be saying never doubt intel but doubt AMD, nothwithstanding the financial differences between the two manufacturers.


I got so bored of the synthetic arguements and too and fro's, that I decided to run game synthetics for all my older CPU's which included E6400, 6700, 8300, Q6600, 9450, QX, some Pentium D samples, Athlon II X2, 4, Phenom II's, first gen i3/i5's, i7 950 and a 2500K, I am without a Bulldozer right now to compare, but at standard playable frame rates the chips by era rather stack up well with each other and at the same price point AMD's performance is not as desernable at times as what 3rd party reviews and the thumbsuck synthetics make out to be.

I happened to browse a topic with people arguing that SB was at least 12% improvement on first gen comparitive products, and yet they say that first gen is significantly faster than FX, yet there is not even a 10% difference at the same price point between FX and SB bar Dirt 3 which seems very much against AMD based CPU's and GPU's. Its all much of a muchness, maybe I am pedantic, maybe I am just easy to please, but I really don't see how AMD is not giving me enough performance to be happy.

As for Intel IGP, that is one area I will not agree with anyone on when they claim it to be useful, it is pityful at best, with plenty titles dumping to desktop on splash with failure to initialize errors. Having seen the highest end Llano with asymetrical crossfire run BF3 med-high at 1900x1080, Intel HD comes nowhere close.

 

[sarinaide]

All fine and well if you are after Directx 9 titles or want to play on DX9, but when a Llano can play on DX11 at equal or higher FPS than HD on DX9 that is enough to make a budget builder buy llano(trinity) based chips.

 

[palladin9479]

As for Intel IGP, that is one area I will not agree with anyone on when they claim it to be useful, it is pityful at best, with plenty titles dumping to desktop on splash with failure to initialize errors. Having seen the highest end Llano with asymetrical crossfire run BF3 med-high at 1900x1080, Intel HD comes nowhere close.

This is where everyone screws up. Building a high performing SIMD array processor is one thing (a GPU), making the drivers for it are a completely different matter. Having personally lived through the "growing pains" of S3, Nvidia and ATI, I can say with completely authority that a GPUs drivers account for 70~80% of its worth. This is why I say Intel is years behind ATI and NVidia, making a piece of hardware that calculates pixels is not enough, you need to create the software support base that goes along with it. Thus assuming their GPU is 100% HW perfect right ~NOW~ (something on manufacture has ever done, ever), they still have years of software work ahead of them.

Of course the blue painted people won't hear a word of this.

 

[noob2222]

Blue paint reflects any negative blue comments, but absorbs all good things. and Vice versa, the blue paint reflects any positive green comments but absorbs all the bad.

Avengers paint however is awesome.

 

[sarinaide]

As for Intel IGP, that is one area I will not agree with anyone on when they claim it to be useful, it is pityful at best, with plenty titles dumping to desktop on splash with failure to initialize errors. Having seen the highest end Llano with asymetrical crossfire run BF3 med-high at 1900x1080, Intel HD comes nowhere close.

This is where everyone screws up. Building a high performing SIMD array processor is one thing (a GPU), making the drivers for it are a completely different matter. Having personally lived through the "growing pains" of S3, Nvidia and ATI, I can say with completely authority that a GPUs drivers account for 70~80% of its worth. This is why I say Intel is years behind ATI and NVidia, making a piece of hardware that calculates pixels is not enough, you need to create the software support base that goes along with it. Thus assuming their GPU is 100% HW perfect right ~NOW~ (something on manufacture has ever done, ever), they still have years of software work ahead of them.

Of course the blue painted people won't hear a word of this.

I had the HD 4000 is 45% better than HD 3000 arguement, the real issue is that it is rather pathetic even at budget orientated markets, if you consider GPU budget builds, even then Intel don't have anything that really is appealing to a budget builder, often the TH build offs have a 2400 on some H61 or Z75 chipset now I guess but all of those said builds lack upgradability from the motherboard standpoint and from a PSU stand point, resale value on the last two $600 builds will end up in massive losses at best. Having seen a FX 8120 + Sabertooth 990fx it rather mutes the budget market. Intel as I have held and still do target specific consumers and enthusiasts.

Blue paint reflects any negative blue comments, but absorbs all good things. and Vice versa, the blue paint reflects any positive green comments but absorbs all the bad.

Avengers paint however is awesome.

I would regard myself as a psuedo enthusiast or someone with a profound interest in hardware and increasing knowledge from those that know more, I have Intel setups new and old, AMD setups new and old, I have Nvidia, 3DFx, ATI/AMD cards and by and large have gone through the wows and the lows of all. Personally I get a lot more joy out of the AMD setups, so it is with that I really hope that PD gets AMD's architecture back on to the merry road, it is difficult to say what the appeal is but I find the AMD setups appeals more to a person, Intel appeals more to a consumer.

 

[triny]

Actually it was Nov. 23, 2011 when AMD closed at $5.05. Today it closed at $7.18.

Back on Sept. 15th of last year, it was trading at $7.34 and about 2 weeks later, dropped to $4.53 or nearly three bucks a share. So gambling on stock prices can be risky.

Risks are taken to earn big bucks fast, my next 2 computers are free

 

[triny]

Ok got done doing preliminary benchmarks on my 3550MX this weekend, and wow is all I gotta say. This chip has plenty of headroom, just needs to be tweaked, also I'm completely bamboozled on why AMD didn't push this further.

3550MX stock, no modification

B0: 2700 @1.3500
P0: 2000 @1.1125
P1: 1700 @1.0875
P2: 1600 @1.0625
P3: 1400 @1.0250
P4: 1200 @1.0000
P5: 1000 @0.9625
P6: 0800 @0.9375

CB 11.5

CPU:2.57
Single Thread:0.64
MP Ratio: 4.03

Watching HWinfo64 I could see all four cores at 2.0Ghz the entire time, temps were 60~65c

3550MX Overclock / Undervolt modification with K10
B0: 3000 @1.3500
P0: 2700 @1.3500
P1: 2000 @1.1125
P2: 1600 @0.9250
P3: 1400 @0.9125
P4: 1200 @0.9000
P5: 1000 @0.8875
P6: 0800 @0.8750

Up: 200ms @ 60%
Down: 2000ms @ 20%


CB 11.5
CPU:3.22
Single Thread: 0.75
MP Ratio:4.29

This is allowing Windows to dynamically move the thread around. This resulting in the four cores constantly clocking at 2.7Ghz, they never made it to 3.0 due to the thermal headroom being hit. Temps were 89~94c but holding steady and not rising. When I set the affinity mask to core 3 I got the following.

Single Thread: 0.86

HWinfo64 had Core 3 pegged at 3.0Ghz and temps at ~64c and holding. The other cores clocked down to their 800 idle speed.

Finially I clock locked (K10 has the ability to force a core to a set speed) core 3 at 3.0ghz and reran the single threaded test.

Single Thread: 0.87

Ran it over and over on both and got a consistent minor improvement. I can attest this to core not having to spool up from 800 to 3.0 during the start of the test.

K10 allows you to adjust the spool up / down times. The way it works is if a core is being hit hard (60%) for a certain period of time (200ms) it will force it to the next P state, if the core is under low utilization (20%) for a certain period of time (2000ms) it will force it to the next lower P state. With HWInfo64 up you can actually see the cores going up / down.

I really hope trinity keeps this level of control over the CPU. Using my own manual tweaking I was able to get a very large performance boost of 25.7% on multi-core tests and a solid performance boost of 15~17% on single threaded tests.

If you know the type of program your going to run, you can further gain performance by manipulating the affinity flag to force it to run on the same core(s) and thus gain from the higher clocking overhead. Window's scheduler is under the assumption that every core is the same speed which is why nobody can get any benefit from boosting technology. You have to manually manipulate the settings yourself to get the most out of it.

Very nice, I believe Lano was a trial balloon to test support before putting a major effort forth .

 

[triny]

More doom and gloom prophacy, at the end it all looks like "AMD will never do anything in the CPU market again, it is over and we must all use Intel because somehow AMD doesn't give me ample enough performance to run winamp and I feel word is not quick enough". It also appears to be saying never doubt intel but doubt AMD, nothwithstanding the financial differences between the two manufacturers.


I got so bored of the synthetic arguements and too and fro's, that I decided to run game synthetics for all my older CPU's which included E6400, 6700, 8300, Q6600, 9450, QX, some Pentium D samples, Athlon II X2, 4, Phenom II's, first gen i3/i5's, i7 950 and a 2500K, I am without a Bulldozer right now to compare, but at standard playable frame rates the chips by era rather stack up well with each other and at the same price point AMD's performance is not as desernable at times as what 3rd party reviews and the thumbsuck synthetics make out to be.

I happened to browse a topic with people arguing that SB was at least 12% improvement on first gen comparitive products, and yet they say that first gen is significantly faster than FX, yet there is not even a 10% difference at the same price point between FX and SB bar Dirt 3 which seems very much against AMD based CPU's and GPU's. Its all much of a muchness, maybe I am pedantic, maybe I am just easy to please, but I really don't see how AMD is not giving me enough performance to be happy.

As for Intel IGP, that is one area I will not agree with anyone on when they claim it to be useful, it is pityful at best, with plenty titles dumping to desktop on splash with failure to initialize errors. Having seen the highest end Llano with asymetrical crossfire run BF3 med-high at 1900x1080, Intel HD comes nowhere close.

The difference between fx and SB is mostly one of semantics

 

[gamerk316]

I'm still waiting for new revisited and re-compiled versions of windows programs that support BD. We should see a lot of performance gains if what GCC showed is easily "ported" to windows compilers (don't remember how this works with compilers).

You're going to be waiting for quite some time them.

What incentive is there for coorporations to change their compiler of choice, port all their programs over to it, debug everything again [race conditions that got through testing become a MAJOR problem now], all just to "maybe" gain a little bit of performance for a specific CPU arch that the majority of users won't be using?

You typically don't see business change their infrastructure without a good reason.

 

[gamerk316]

More doom and gloom prophacy, at the end it all looks like "AMD will never do anything in the CPU market again, it is over and we must all use Intel because somehow AMD doesn't give me ample enough performance to run winamp and I feel word is not quick enough". It also appears to be saying never doubt intel but doubt AMD, nothwithstanding the financial differences between the two manufacturers.


I got so bored of the synthetic arguements and too and fro's, that I decided to run game synthetics for all my older CPU's which included E6400, 6700, 8300, Q6600, 9450, QX, some Pentium D samples, Athlon II X2, 4, Phenom II's, first gen i3/i5's, i7 950 and a 2500K, I am without a Bulldozer right now to compare, but at standard playable frame rates the chips by era rather stack up well with each other and at the same price point AMD's performance is not as desernable at times as what 3rd party reviews and the thumbsuck synthetics make out to be.

I happened to browse a topic with people arguing that SB was at least 12% improvement on first gen comparitive products, and yet they say that first gen is significantly faster than FX, yet there is not even a 10% difference at the same price point between FX and SB bar Dirt 3 which seems very much against AMD based CPU's and GPU's. Its all much of a muchness, maybe I am pedantic, maybe I am just easy to please, but I really don't see how AMD is not giving me enough performance to be happy.

As for Intel IGP, that is one area I will not agree with anyone on when they claim it to be useful, it is pityful at best, with plenty titles dumping to desktop on splash with failure to initialize errors. Having seen the highest end Llano with asymetrical crossfire run BF3 med-high at 1900x1080, Intel HD comes nowhere close.

Beware the GPU bottleneck when benching games. Turn down every graphical setting, and bench again, and you see which CPU arch is the more powerful. The point is, even though SB and BD may give the same FPS in a specific game when settings are maxed, its due to the GPU being more important. 9 time out of 10, SB has significantly more headroom before it becomes a CPU bottleneck in games.

If you benched a Q9650, I wonder how competitive it would still be in games compared to BD...

 

[de5_Roy]

The difference between fx and SB is mostly one of semantics

hahaha.. you say funny things, triny

 

[sarinaide]

I have no Zambezi's to test, and certainly no FX 8XXX series chips to test, but what I did do on the Zambezi's prior all got good frame rates, but very irratic rates too.

Athlon II's stacked well against earlier Core2 chips, Phenom II's stack up with later Core2 and first gen core I excluding the higher end i7's, Bulldozer from what I assume will sometimes match SB, and sometimes first gen, and most of the time will perform like the Phenom II's.

In the IGP performance, the APU's destroy everything before them, a heavily oc'ed 2700K with HD3000 is absolutely no match in IGP performance, which is a moral victory.

Crysis 2
Skyrim
Oblivion
Crysis Warhead
F1 2011
Battlefield 3
Battlefield 2 Bad Company
Shogun 2
Far Cry 2
Metro 2033
Stalker Call of the Prypiyat.

Overall on core grunt, Intel has more often than not, the Thubans still game very well compared with Sandybridge in terms of FPS difference. As to the GPU's used HD 6970, HD 6990, HD 7870, HD 5970, HD 5870, GTX 460 1GB, GTX 480, GTX 285, GTX 560ti 448, GTX 580, 8800 Ultra

 

[gamerk316]

I have no Zambezi's to test, and certainly no FX 8XXX series chips to test, but what I did do on the Zambezi's prior all got good frame rates, but very irratic rates too.

Athlon II's stacked well against earlier Core2 chips, Phenom II's stack up with later Core2 and first gen core I excluding the higher end i7's, Bulldozer from what I assume will sometimes match SB, and sometimes first gen, and most of the time will perform like the Phenom II's.

In the IGP performance, the APU's destroy everything before them, a heavily oc'ed 2700K with HD3000 is absolutely no match in IGP performance, which is a moral victory.

Crysis 2
Skyrim
Oblivion
Crysis Warhead
F1 2011
Battlefield 3
Battlefield 2 Bad Company
Shogun 2
Far Cry 2
Metro 2033
Stalker Call of the Prypiyat.

Overall on core grunt, Intel has more often than not, the Thubans still game very well compared with Sandybridge in terms of FPS difference. As to the GPU's used HD 6970, HD 6990, HD 7870, HD 5970, HD 5870, GTX 460 1GB, GTX 480, GTX 285, GTX 560ti 448, GTX 580, 8800 Ultra

Again, due to a GPU bottleneck. Games are horrible for benchmarking for that reason alone.

 

[fazers_on_stun]

More doom and gloom prophacy, at the end it all looks like "AMD will never do anything in the CPU market again, it is over and we must all use Intel because somehow AMD doesn't give me ample enough performance to run winamp and I feel word is not quick enough". It also appears to be saying never doubt intel but doubt AMD, nothwithstanding the financial differences between the two manufacturers.

?? There's a gigantic difference between "AMD will never do anything in the CPU market again" and predicting that AMD will dominate the high end while ARM takes over the low end, presumably leaving Intel on the brink of bankruptcy, as Triny was posting.. However he says he owns some AMD stock, so he has a vested interest in propagandizing AMD's upcoming CPUs.

I got so bored of the synthetic arguements and too and fro's, that I decided to run game synthetics for all my older CPU's which included E6400, 6700, 8300, Q6600, 9450, QX, some Pentium D samples, Athlon II X2, 4, Phenom II's, first gen i3/i5's, i7 950 and a 2500K, I am without a Bulldozer right now to compare, but at standard playable frame rates the chips by era rather stack up well with each other and at the same price point AMD's performance is not as desernable at times as what 3rd party reviews and the thumbsuck synthetics make out to be.

Hmm, well "good enough" is OK but remember this is an enthusiast site, and I specifically referred to enthusiasts in my post. I don't think anybody here doubts AMD CPUs are adequate for whatever tasks you want to throw at them, or a good value for the $$. But, most of us are going to require hard & fast proof of AMD having taken the halo end of the performance spectrum, as Triny would have us believe is imminent. And while your post belittles the benchmarks and reviews, there's really nothing else for us to rely on, short of buying every new CPU and testing them for whatever purpose we have in mind.

 

[sarinaide]

?? There's a gigantic difference between "AMD will never do anything in the CPU market again" and predicting that AMD will dominate the high end while ARM takes over the low end, presumably leaving Intel on the brink of bankruptcy, as Triny was posting.. However he says he owns some AMD stock, so he has a vested interest in propagandizing AMD's upcoming CPUs.

I got so bored of the synthetic arguements and too and fro's, that I decided to run game synthetics for all my older CPU's which included E6400, 6700, 8300, Q6600, 9450, QX, some Pentium D samples, Athlon II X2, 4, Phenom II's, first gen i3/i5's, i7 950 and a 2500K, I am without a Bulldozer right now to compare, but at standard playable frame rates the chips by era rather stack up well with each other and at the same price point AMD's performance is not as desernable at times as what 3rd party reviews and the thumbsuck synthetics make out to be.

Hmm, well "good enough" is OK but remember this is an enthusiast site, and I specifically referred to enthusiasts in my post. I don't think anybody here doubts AMD CPUs are adequate for whatever tasks you want to throw at them, or a good value for the $$. But, most of us are going to require hard & fast proof of AMD having taken the halo end of the performance spectrum, as Triny would have us believe is imminent. And while your post belittles the benchmarks and reviews, there's really nothing else for us to rely on, short of buying every new CPU and testing them for whatever purpose we have in mind.

All is fair.

I will not say I am belittling the reviews, I am just against synthetics as every review site has largely differing synthetics and that leads to my obvious assumption that it is largely setup related in so much as it is the CPU alone. I treat them as something to read rather than the gospel due to bad experiences relying on benchmarks. I will also say that I do forward a lot of the reviews from TH to third party members as I find the reads to be interesting and pass on content. I am sorry that you feel it belittles Toms, but that is not the intention.

I will also like to clarify that I in no way made any inferance that AMD was going to take the performance mantle from Intel, at least not in the short term with the state of finances, the divorce from Global Foundries and the ATI situation has rather limited AMD's hand in regards to R&D, but they do have APU's and the GPU segment which is winning a few positives.

 

[fazers_on_stun]

All is fair.

I will not say I am belittling the reviews, I am just against synthetics as every review site has largely differing synthetics and that leads to my obvious assumption that it is largely setup related in so much as it is the CPU alone. I treat them as something to read rather than the gospel due to bad experiences relying on benchmarks. I will also say that I do forward a lot of the reviews from TH to third party members as I find the reads to be interesting and pass on content. I am sorry that you feel it belittles Toms, but that is not the intention.

Guess I took your "thumbsuck synthetics" comment the wrong way then .

I will also like to clarify that I in no way made any inferance that AMD was going to take the performance mantle from Intel, at least not in the short term with the state of finances, the divorce from Global Foundries and the ATI situation has rather limited AMD's hand in regards to R&D, but they do have APU's and the GPU segment which is winning a few positives.

Understood, and most people here agree with that assessment. Except those who need AMD's stock to climb to $25 a share by 2014 😛..

@ GamerK316 - wasn't it you who bought AMD stock near its low of $1.63 a few years back, then sold when it hit near $10? Now that's the way to speculate! 😀

 

[gamerk316]

@ GamerK316 - wasn't it you who bought AMD stock near its low of $1.63 a few years back, then sold when it hit near $10? Now that's the way to speculate! 😀

Buy low, sell high. If the company isn't going to go bankrupt, then you will always recoup, and then some.

 

[de5_Roy]

"desktop trinity to debut in august"
http://www.fudzilla.com/home/item/27062-desktop-trinity-coming-in-august

 

[-Fran-]

Buy low, sell high. If the company isn't going to go bankrupt, then you will always recoup, and then some.

Well, once Intel has a mean (or way) to justify how they won't be a monopoly if no other Company owns part of the x86 license, AMD will go down in one blow, maybe VIA as well.

And I understand why doing a recompile is costly (time wise at least), but why companies don't give different executables for different CPUs. Is not that hard to run the compile 2 or 3 times for palpable increases in performance. It's just a matter of compiling flags being passed, nothing more difficult than that. I know the hassle of the libraries used, but oh well... Certainly beating a dead horse...

Cheers!

 

[noob2222]

627% faster than what? Than Quick Synce gen 1, 2 or than the APU without it? I would assume such a large number is against the APU without it not QS since even nVidias version isn't quite a QS killer yet. And it may stay that way since Intel does have the one up on them in terms of time. I imagine Haswell will give a QS 3.0.

I missed this since you worte it in the middle of quoting what I said. I take it you didn't read the article http://blogs.amd.com/fusion/2012/04/24/adobe-and-amd-enable-brilliant-experiences/

OpenCL increases the speed of the A8 3530MX vs Not OpenCL. Granted this tech would still work with an AMD video card, but in an APU (Trinity, steamroller?), it will still work with an Nvidia gpu installed, off the APU. May even work with Intel cpu +Amd video card, unless AMD took notes from Intel (compiler) and Nvidia (physx) about making things not work when competition is detected.

Like I said, AMD needs to push their tech, not just invent it and say "why arent you using it?" Its a step in the right direction, get the software support as the hardware is being developed instead of after.

 

[gamerk316]

Well, once Intel has a mean (or way) to justify how they won't be a monopoly if no other Company owns part of the x86 license, AMD will go down in one blow, maybe VIA as well.

Not really. AMD owns the X86-64 license; what is Intel going to do when AMD retaliates by getting rid of that license agreement, go back to Intel64?

And I understand why doing a recompile is costly (time wise at least), but why companies don't give different executables for different CPUs.

Because then you need to invest in multiple test benches, debug and maintain several different variants of code, have multiple help support teams, update/maintain multiple builds, and last but not least, have two separate media installs [DVD's]. And this is before the end user installs the wrong version.

Compiler optimizers do all this ALREADY. Granted, you get a bigger executable, but it solves the multiple CPU features problem.

This is the same reason why we don't have X64 games yet; you think a game like Skyrim would benefit from being able to use more the ~3.5GB of RAM when running on X64?

You have to understand, even if 95% of the code is common to both builds, you WILL run into different bugs in both versions. That greatly increases testing, development, and at the end of the day, more buggy code.

 

[noob2222]

Here is an interesting article, kind of puts into perspective just how much power a massive server utilizes.

http://www.theregister.co.uk/2012/05/08/amd_eda_cloud_buildout/

The Austin, Markham, Sunnyvale, and Kuala Lampur data centers together have around 80,000 cores of the total 120,000 cores in the design cloud. The Austin data center is rated at 3.8 megawatts and is not only maxxed out in terms of how much power it can use, but additionally is not able to be extended without the power company having to route new power feeds out to the facility - something that was very expensive to do.

3.8 megawatts ... ouch.

At the moment, the Swamee data center has 2 megawatts of power activated and the 40,000 cores are drawing about 1.4 megawatts, down from the 2 megawatts that 40,000 cores using the two prior generations of Opteron processors would draw if AMD was using them in the new facility, according to Dana's math.

Goes to show why huge server farms need lower power cpus. Whole different ballgame when you say 40,000 cores vs the 16-32 that reviews use.

 

[gamerk316]

OpenCL increases the speed of the A8 3530MX vs Not OpenCL. Granted this tech would still work with an AMD video card, but in an APU (Trinity, steamroller?), it will still work with an Nvidia gpu installed, off the APU. May even work with Intel cpu +Amd video card, unless AMD took notes from Intel (compiler) and Nvidia (physx) about making things not work when competition is detected.

Theres a difference between "disabling a feature" versus "not optimising". If Intels proprietary compiler does not optimize for AMD, thats their business. If they actually DISABLED manually programmed SSE calls however, that would be blatently illegal. Theres a difference between the two.

As for PhysX, its a proprietary standard owned by NVIDIA. They implemented it through CUDA, another proprietary standard owned by NVIDIA. I fail to see why they should offer it to AMD for free. And PhysX is the best thing out there, at least until someone writes up an OpenCL based physics engine...

 

[noob2222]

As for PhysX, its a proprietary standard owned by NVIDIA. They implemented it through CUDA, another proprietary standard owned by NVIDIA. I fail to see why they should offer it to AMD for free. And PhysX is the best thing out there, at least until someone writes up an OpenCL based physics engine...

Never said that. Remember when people were using an old Nvidia card with AMD gpus to get physX support? What was Nvidia's solution? Nvidia did that to their own customers and had very little to do with AMD other than to drive people back to buying Nvidia only. Somehow I don't see AMD stooping to their level.