AMD CPU speculation... and expert conjecture

[8350rocks]

I have assurances that we have not seen the last "halo" product line from AMD. There have been allusions to the fact that it is not a great production environment for such a product right now. That being the main issue, we may not see another FX specifically, however, we will see something along the same vein. I don't have a crystal ball to say it will be AM4 or FM2+ or even FM3, though AMD is not done with HEDT. I have assurances that is not the case at all from a source that would know.

 

[etayorius]

The extra Decoder inside each Module by itself should be enough to gain at least a 10-15% increase, i would go as far as to to say the extra Decoder is by far the largest performance improvement from all of the changes in Steamroller Arch, i can understand the GPU Benchmarks since AMD has probably not released nor provided any Optimized Drivers for Kaveri, but the CPU benchmarks are just horrible if they are found to be true (i seriously doubt it), Steamroller main focus in CPU was to greatly improve IPC or Single Core performance, but now you`re suggesting that all the improvements AMD did to Steamroller in order to increase IPC are basically useless because as you claim "they are not driving performance", with this argument it would mean AMD are just clueless on what improves performance on their own Arch.

Different types of application have bottlenecks in different parts of the CPU. Taking your example, if the application's performance is not driven by the decoder module, then improvements to that area aren't going to lead to significant performance increases for that particular application.

Lets take games as an example: Do you really think the L1 cache is driving performance? No, the larger L2/L3 are more important. So improvements to the L1 aren't going to have significant effects to performance, even if the improvements to the L1 are quite large.

Combine that with a clockspeed decrease of 300MHz, and you could easily see why some applications would offer virtually the same performance, even if several parts of the CPU were improved up to 30%.

When I look at the improvements AMD made, I see improvements that will reduce some of the overhead to CMT, and that will make applications that scale across all the cores run better. For applications that don't scale as well, such as games, I really don't see much in the way of CPU performance increases. I'd suspect those apps are going to be driven by improvements to the iGPU, not CPU.

We know for a Fact that Piledriver is AT least 10-12% faster than Bulldozer in Single Core Exection and up to 20 Faster mostly in Multithreaded, AMD promised 20% more IPC by Ehnancing the Arch on Steamroller and another 10% out of pure higher Speed (30% target), take the 300Mhz less Speed compared to Richland and you still get the 20% more IPC with Kaveri, now how on earth is Richland overall faster than Kaveri in the Pudget Benchmarks?

"In some tasks". AMD's words, not mine. See my above point: If a particular part of the CPU is not driving performance for a particular app, that app should not expect to see any performance increases by increasing the performance of that part of the CPU.

For example: Say AMD made the cost overhead of CMT 5%, rather then the 20% PD has. So you'd say "15% more IPC", and thus, 15% more performance. But for tasks that never use CMT [one core per module being utilized], this 15% performance increase to CMT has no performance impact on that task. Combine that with a clock decrease, and you'd actually expect performance to drop due to lower clocks and no change to IPC.

Lets take your Best Case scenario of up to 10-15% more performance like you previously predicted several pages back, your predictions fell too generous compared to Pudget Benchmarks it would mean Kaveri is about 10% Slower than Richland, there for AMD is lying with their 20% more IPC compared to Richland using bad faith and false advertising for a product they may know it will fail, then inflated the price of 175USD compared to Richland 140USD, then to make all this short it means AMD wants to rip all of us with an inferior and more expensive Product.

Remember the BD launch? When the PII X4 965 BE was beating the FX-4100 head to head in the majority of benchmarks? Its happened before, just two years ago.

I don't think it will be that bad. There aren't many pure single-threaded applications, so the number of apps that lose performance is going to be vanishingly small. Also remember that games are driven more by GPU performance, and its possible driver side issues on the GPU side could be understating the results. Simply one data set, and we'll see where the others line up.

But now seriously, we have no reason to trust Pudget or the hand behind them, seeing as how small they are, they would have no issues getting rid of AMD products and having a nice Pay Check from Intel, they have nothing to lose from backstabbing AMD in the back.

And...we're back to conspiracy theories. Rather then try and explain the results [poor CPU design, shoddy drivers, issues with the OS], we fall back to bias instead. And people wonder why politics in this country is so bad right now...

Even if it is 10% best case scenario compared to Richland it will still be faster than my PhenomII 980 at Default Clock, i get a nice GPU to the levels of the HD5770 that will boost like crazy using MANTLE and HSA in the future and i will also get a Key for Battleshit4 on Origin bundled with my APU, which i can sell easily for 30-40 Bucks... not too bad at the end of the day.

Notice I've never been one of the people claiming that a PII CPU would beat Kaveri? That's just sad people are still claiming as such. For BD, ok, but that hasn't been true since PD came out. As for Mantle, I've made my opinions on the API very clear by now.

Fair enough, but i still doubt the good will of Pudget Benchmarks... good thing we are 4 days away from knowing the truth.

And 8350Rocks, i have the same feeling and i think we will see in 8-10 months a refresh to Kaveri, possibly running at 4Ghz or faster at default and with more GCN Cores, just a few months before Carrizo arrives.

 

[jimmysmitty]

I have assurances that we have not seen the last "halo" product line from AMD. There have been allusions to the fact that it is not a great production environment for such a product right now. That being the main issue, we may not see another FX specifically, however, we will see something along the same vein. I don't have a crystal ball to say it will be AM4 or FM2+ or even FM3, though AMD is not done with HEDT. I have assurances that is not the case at all from a source that would know.

AM4 will probably only arrive when they get DDR4 and since Intel is barely going to have it with Haswell-E this year I don't see AMD adopting it till at least 2015 as they normally are behind Intel in adoption by a year. Could be they are wanting to wait till it is cheaper to keep up the lower cost platform angle or it could be because they don't care to adopt it when it is new.

I don't think AMD is done I just don't think they are going to be pushing Intel and that's what worries me. Without competition the market gets stagnant and we suffer, not AMD or Intel. Both will continue to move on and make money in other markets but if the CPUs stop advancing fast enough we wont see the amazing performance boosts we saw. Haswell had a big boot to cache performance, especially L1 cache which doubled SB and almost doubled IB.

But without fierce competition, Intel can't afford to be labeled a monopoly so they will not push as hard to keep a decent lead over AMD without being too good that AMD makes too little sales.

 

[gamerk316]

And NVIDIA fires a major shot here:

http://www.anandtech.com/show/7622/nvidia-tegra-k1/3

NVIDIA's T1 GPU, the GPU being used in Project Denver, their ARM SoC, is basically just a cut down Kepler GPU. Farther, all their new designs, including GeForce, are going to start out as mobile designs.

This is significant. NVIDIA just got the most developer support for mobile, since they support the FULL OGL/OGL ES/DirectX/CUDA instruction sets. Makes it much more attractive for traditional PC gaming studios to make a low-cost mobile offering. Makes Mantle that much of a harder sell now, since it won't be a player in mobile.

I'm still waiting for AMD to make a mobile SoC. I'm stunned they haven't gone this route yet, especially since, via Trinity, they should be best positioned for it.

 

[UnrelatedTopic]

But they are under strict competition. The era of apus is dawning and, like x64, AMD is leading the charge. Like it or not (I was disappointed at first) the time of dCPU is coming to a close and like single core cpus it's going the way of the dodo. There will be no steamroller fx due in the next two years not because amd had given up on HEDT but there simply is no demand for 8 core cpus anymore and besides who needs 8 cores when you have 12 core apus right? Lol

 

[con635]

Mantle will succeed because of the consoles I think, I know they dont use mantle as such but thats because they dont need it, its the pc that does. amd will take big chunks of nvidia/intel market share in the next few years imo with the console wins, kaveri/huma and mantle.

 

[de5_Roy]

And NVIDIA fires a major shot here:

http://www.anandtech.com/show/7622/nvidia-tegra-k1/3

NVIDIA's T1 GPU, the GPU being used in Project Denver, their ARM SoC, is basically just a cut down Kepler GPU. Farther, all their new designs, including GeForce, are going to start out as mobile designs.

This is significant. NVIDIA just got the most developer support for mobile, since they support the FULL OGL/OGL ES/DirectX/CUDA instruction sets. Makes it much more attractive for traditional PC gaming studios to make a low-cost mobile offering. Makes Mantle that much of a harder sell now, since it won't be a player in mobile.

I'm still waiting for AMD to make a mobile SoC. I'm stunned they haven't gone this route yet, especially since, via Trinity, they should be best positioned for it.

i thought nvidia's demo was very weak and lacking. it rehashed the old stuff and made a big deal about project denver only to reveal it as a 64-bit dual core cpu and a 32bit version with a15 cores. wth, not even a53? did apple really node-block nvidia from 20nm?(rhetorical question, idc) the 64bit soc had a lot of cache but i don't think it'll help in 'wide' tasks.

i am a bit disappointed at only "45% better perf" information. i was expecting more on mantle (unless i've missed it). i blame call of .. i mean bf4.

afaik, amd's best shot at mobile soc will come around the time they launch seattle, prolly with a57+gcn 2.0 cores soc built on 20/28nm. jaguar/puma can't go lower power without sdp (intel's latest scam that amd joined), gcn 1.0, 1.1 don't seem too friendly towards low power. er.. assuming by mobile you mean smartphone soc.

 

[con635]

But they are under strict competition. The era of apus is dawning and, like x64, AMD is leading the charge. Like it or not (I was disappointed at first) the time of dCPU is coming to a close and like single core cpus it's going the way of the dodo. There will be no steamroller fx due in the next two years not because amd had given up on HEDT but there simply is no demand for 8 core cpus anymore and besides who needs 8 cores when you have 12 core apus right? Lol

lol technically we could argue that top kaveri is a 516 core apu, mooooooooar coooooars

 

[gamerk316]

But they are under strict competition. The era of apus is dawning and, like x64, AMD is leading the charge. Like it or not (I was disappointed at first) the time of dCPU is coming to a close and like single core cpus it's going the way of the dodo. There will be no steamroller fx due in the next two years not because amd had given up on HEDT but there simply is no demand for 8 core cpus anymore and besides who needs 8 cores when you have 12 core apus right? Lol

The APU basically replaces the older SoC concept. Same deal though. The issue is, with a CPU+GPU on the same die, you have limited space to put other enhancements to improve performance. You'll never get top CPU or top GPU performance as a result. Dedicated CPUs and dedicated GPUs aren't going anywhere.

Mantle will succeed because of the consoles I think, I know they dont use mantle as such but thats because they dont need it, its the pc that does

Not really. Aside from the poor logic, PCs offer more performance then next-gen consoles. So how do PCs need Mantle more?

 

[gamerk316]

And NVIDIA fires a major shot here:

http://www.anandtech.com/show/7622/nvidia-tegra-k1/3

NVIDIA's T1 GPU, the GPU being used in Project Denver, their ARM SoC, is basically just a cut down Kepler GPU. Farther, all their new designs, including GeForce, are going to start out as mobile designs.

This is significant. NVIDIA just got the most developer support for mobile, since they support the FULL OGL/OGL ES/DirectX/CUDA instruction sets. Makes it much more attractive for traditional PC gaming studios to make a low-cost mobile offering. Makes Mantle that much of a harder sell now, since it won't be a player in mobile.

I'm still waiting for AMD to make a mobile SoC. I'm stunned they haven't gone this route yet, especially since, via Trinity, they should be best positioned for it.

i thought nvidia's demo was very weak and lacking. it rehashed the old stuff and made a big deal about project denver only to reveal it as a 64-bit dual core cpu and a 32bit version with a15 cores. wth, not even a53? did apple really node-block nvidia from 20nm?(rhetorical question, idc) the 64bit soc had a lot of cache but i don't think it'll help in 'wide' tasks.

Frankly, I find the GPU more important here. If the slides speak true, NVIDIA is going all in on mobile. This means NVIDIA GPUs are going to get less power hungry, and more efficient. Could be interesting to see Maxwell performance as a result.

Development side, you now have the same GPU API support on mobile that you do on PCs. For development, that's HUGE, and a big feather in NVIDIAs cap, especially since AMD is absent from the mobile space [which still boggles me].

Tegra 5, honestly, is looking like the first good SoC out of NVIDIA. If it gains any traction, they could really make a move to take over mobile.

i am a bit disappointed at only "45% better perf" information. i was expecting more on mantle (unless i've missed it). i blame call of .. i mean bf4.

Time and money. And the fact they had to stop development to make the game work in the first place. And I still suspect, on higher tier hardware, the effects of Mantle are going to be a LOT smaller.

 

[de5_Roy]

Frankly, I find the GPU more important here. If the slides speak true, NVIDIA is going all in on mobile. This means NVIDIA GPUs are going to get less power hungry, and more efficient. Could be interesting to see Maxwell performance as a result.

Development side, you now have the same GPU API support on mobile that you do on PCs. For development, that's HUGE, and a big feather in NVIDIAs cap, especially since AMD is absent from the mobile space [which still boggles me].

ah, now i understand. it's an advantage for nvidia but from performance, they're still behind imagination and qualcomm. i read somewhere that nvidia's socs were too power hungry. i think same issue plagues amd's low power x86 cores - those may improve with 20nm.

Time and money. And the fact they had to stop development to make the game work in the first place. And I still suspect, on higher tier hardware, the effects of Mantle are going to be a LOT smaller.

may be. but in entry level where users will be willing to squeeze out every performance without paying more, mantle makes sense to me. by users i mean people who always play games at 768p/800p/900p tn display fed from a pc running jetway/ecs/biostar mobo, single dimm 4GB ram, powered by psus of questionable quality. unfortunately, amd neglects entry level products' software support, it either trickles down from high end, or it's neglected altogether. for those guys, mantle will be very welcome. also for the poor suckers (who are annoyed by framerate drops and lags) got tricked into buying amd cpus for online multiplayer/gaming - for those, reduction of cpu overhead may literally reduce headaches.

 

[etayorius]

MANTLE is a WIN, the only bad thing i see about it... well, i dont really see anything wrong with it as a Gamer and Final User, even Johan Andersson said the cost and time to port any game to MANTLE are minimum, it took them like a month to prepare MANTLE and up to 10% max of the total cost, as he said it really is worth it and affordable at the end of the day, since most code from the Consoles can be re used in MANTLE it is actually quite easy to sopport it.

The only hurted with MANTLE are nVidia and DirectX, i have no idea what the hell was nVidia thinking when announced GameWorks as MANTLE response... it seems no one is talking about it and the very people talking about GameWorks is mostly negative stuff.

MS is already working on DX12 and they said they were doing it it from scratch mostly, not long ago MS said they had no plans for a new version of DX but i guess seeing all the hype with MANTLE it meant horrible PR for MS, they will probably concentrate on a lot of performance with DX12 or some form of lower level calls (but not as low as MANTLE), at the end of the day MANTLE will force MS to evolve DX, which is great! i wish DX will go OpenSource and start supporting Close to Metal code and MANTLE disappears, if this happens there will be no need for MANTLE, but as of now MANTLE is awesome for all those people who got HD7000 and R7-R9 Series, even if stronger GPUs dont get the full 45% performance boost, any performance boost is greatly appreciated.

 

[UnrelatedTopic]

I agree dCPU still viable but theyll slowly fade away. 8 Core vishera is still better than quad core cranberries but that'll change over time as software increases support for HSA

 

[-Fran-]

You mean gamerk, you're wondering why we would want a fat free API to build applications (graphics in this case)? JEEZ, I WONDER WHY WE WOULD WANT THAT.

Glide in its time was actually quite famous. The problem was 3DFX not supporting new stuff from OGL and DX and being bought by nVidia, lol.

I do agree with you it will be a matter of cost, but I don't think, from the technical point of view, it will fail right away as a piece of technology. As a developer, and a nerd at heart, I do want new technologies that reduce the fat from my programming. I feel dirty when I have to use Hibernate (for example) instead of Springs JDBC Template. I'm PRETTY sure graphical veterans will rather use the close-to-metal API most of the time and newbies in programming will transition from DX/OGL to MANTLE with little effort (specially the ones coding for consoles).

So, again, if MANTLE fails, it won't be from a technical point of view. As is the norm for AMD, it will be from low adoption, which is not a technical thing. And so far, adoption is quite fine, for being a new tech IMO.

In regards to AMD not building SoCs. I think they do, but they're keeping them for Embedded systems for the time being IIRC. Also, remember when they got ATI and kind of like... You know... Sold their low power graphical patents/licenses to Qualcomm? They might be feeling the lose like, right now, hahaha.

Cheers!

 

[con635]

But they are under strict competition. The era of apus is dawning and, like x64, AMD is leading the charge. Like it or not (I was disappointed at first) the time of dCPU is coming to a close and like single core cpus it's going the way of the dodo. There will be no steamroller fx due in the next two years not because amd had given up on HEDT but there simply is no demand for 8 core cpus anymore and besides who needs 8 cores when you have 12 core apus right? Lol

The APU basically replaces the older SoC concept. Same deal though. The issue is, with a CPU+GPU on the same die, you have limited space to put other enhancements to improve performance. You'll never get top CPU or top GPU performance as a result. Dedicated CPUs and dedicated GPUs aren't going anywhere.

Mantle will succeed because of the consoles I think, I know they dont use mantle as such but thats because they dont need it, its the pc that does

Not really. Aside from the poor logic, PCs offer more performance then next-gen consoles. So how do PCs need Mantle more?

To take proper full advantage of amd hardware eg huma, dual graphics, xfire, remember theres going to be x86 amd gcn/huma apus in around 100 million homes across the world in a few years not even counting pc/laptop sales. I'm not sure whats in the wii u ibm cpu + gcn gpu??

 

[jdwii]

Juanrga facts are starting to fail, look he has even resorted to stating things such as " that main goal of AMD behind APUs is not cheap gaming" note that this is used to overcome his information previously about steamroller being 20% more powerful on average(which he used to say 30% but will deny it) and 30% more IGPU performance.

right before Bulldozer came out JF-AMD started to state that IPC and single core performance did not matter. Maybe he did not know that Intel did not give us 1 core for 220$ but 4 which most software used anyways and according to Amd's roadmaps will only be produced from Amd.
Like i said its about detecting BS and fanboyism, The older you get and the more your on fourms the more you will notice this.

Let me tell you the same thing I told the last guy ... stop trolling or there will be consequences.

Your deliberately trying to provoke another user and that will not be tolerated.

Not trying to start anything but this forum(from the mods) has allowed this guy to continue putting others down and its almost like he gets special treatment,

Again after he continued on and on for weeks and weeks when the time comes and he is wrong its called provoking the other user?

 

[etayorius]

Juanrga facts are starting to fail, look he has even resorted to stating things such as " that main goal of AMD behind APUs is not cheap gaming" note that this is used to overcome his information previously about steamroller being 20% more powerful on average(which he used to say 30% but will deny it) and 30% more IGPU performance.

right before Bulldozer came out JF-AMD started to state that IPC and single core performance did not matter. Maybe he did not know that Intel did not give us 1 core for 220$ but 4 which most software used anyways and according to Amd's roadmaps will only be produced from Amd.
Like i said its about detecting BS and fanboyism, The older you get and the more your on fourms the more you will notice this.

Let me tell you the same thing I told the last guy ... stop trolling or there will be consequences.

Your deliberately trying to provoke another user and that will not be tolerated.

Not trying to start anything but this forum(from the mods) has allowed this guy to continue putting others down and its almost like he gets special treatment,

Again after he continued on and on for weeks and weeks when the time comes and he is wrong its called provoking the other user?

I have not seen any offensive or trolling comments from Juanrga, the only thing i noticed is that he think to be right every single time... but there is no rule against that.

 

[jdwii]

" the only thing i noticed is that he think to be right every single time... but there is no rule against that."
Yes and he puts other's down when he thinks so.

 

[etayorius]

" the only thing i noticed is that he think to be right every single time... but there is no rule against that."
Yes and he puts other's down when he thinks so.

I don`t think that qualifies as an offense or as trolling.

Anyway i got a question for who ever can answer, i was wondering if any applications can take advantage of HUMA of that or if the applications need to be written specifically to support HUMA.

 

[juggernautxtr]

" the only thing i noticed is that he think to be right every single time... but there is no rule against that."
Yes and he puts other's down when he thinks so.

He is constant about always being right.
problem about dropping dcpu and dgpu, an APU is far from reaching high lvl performance requirements needed to produce what they use.
notice the server maps no 6-10 core server parts? APU no where near strong enough.
give up kraken super computer? worth millions in upgrades alone.
umm this is still a huge market that is still growing and giving up any chance to make money there is completely stupid.
kaveri in all actuallity would have to dump all over an I7 4770k to even think of taking those spots and i very much doubt it has that kinda power.

steamroller is either a complete flop or it is so stupidly ready at task that AMD is keeping mouth shut.
the APU is in no way ready to replace DCPU or DGPU it's going to take a few more years for these to kick in.

If AMD dropped dcpu amd would be losing even more server space,causing even higher loss in income. that 5% is still huge in terms of cash flow.

 

[de5_Roy]

Microsoft Surface 3 to feature Nvidia K1, Surface Mini launching this year
http://vr-zone.com/articles/microsoft-surface-3-feature-nvidia-k1-surface-mini-launching-year/69102.html

edit: just wanted to make sure: is this thread gonna go bye-bye on january 14th? since amd's dt cpu roadmap is stalled with pd and kaveri/carrizo are moving forward... and stuff.

 

[UnrelatedTopic]

Microsoft Surface 3 to feature Nvidia K1, Surface Mini launching this year
http://vr-zone.com/articles/microsoft-surface-3-feature-nvidia-k1-surface-mini-launching-year/69102.html

edit: just wanted to make sure: is this thread gonna go bye-bye on january 14th? since amd's dt cpu roadmap is stalled with pd and kaveri/carrizo are moving forward... and stuff.

well we could start

AMD Steamroller/Excavator speculation... and expert conjecture

or
Just plain and simple

AMD speculation... and expert conjecture Steamroller and beyond

 

[noob2222]

@etayorius

Try reading his comments belittling people instead of trying to pretend he is a nice person.

 

[Reynod]

Lets move on eh? If Juan gets annoying I'll waterboard him too ... Why not see what you can find on cache latency for this new part ... thats always an interestign discussion because Intel do it so well ... and AMD have some clear ground to make up, as it obviously detracts from individual core IPC.

 

[juanrga]

When I look at the improvements AMD made, I see improvements that will reduce some of the overhead to CMT, and that will make applications that scale across all the cores run better. For applications that don't scale as well, such as games, I really don't see much in the way of CPU performance increases. I'd suspect those apps are going to be driven by improvements to the iGPU, not CPU.

But we already showed you current games that scale well up to 8 cores. I also showed you a thread scheduler for killzone showing six-cores of the PS4 working fully. All forthcoming games developed for PS4/Xbox1 will be well-threaded.

Moreover, Kaveri also comes in dual-core configuration and will run faster the older dual-threaded games that you mention often.

I have assurances that we have not seen the last "halo" product line from AMD. There have been allusions to the fact that it is not a great production environment for such a product right now. That being the main issue, we may not see another FX specifically, however, we will see something along the same vein. I don't have a crystal ball to say it will be AM4 or FM2+ or even FM3, though AMD is not done with HEDT. I have assurances that is not the case at all from a source that would know.

Nobody here said that AMD is abandoning HEDT. At contrary it was said that AMD has plans for an ultra-high-performance APU since several years ago. I already mentioned the original doc from AMD reports stacked RAM (32-64GB) in the GPU, 10 TFLOP of performance, quad-channel memory, and 4 modules (8 cores), with 256-bit FMAC units per core in a 200-250W rated APU. Nvidia has a similar design with a 300W rated APU. Both are aimed to supercomputers/servers.

I will add that the original AMD design mentions cache coherent L3 shared between CPU and iGPU. I don't think we will see that design but an updated model. AMD has announced a pair of weeks ago that are collaborating with Hynix to release stacked RAM based in JEDEC standard.

Kaveri/Berlin and Carrizo/Toronto will use DDR memory and GlobalFoundries immature node. This is why the FX/Warsaw CPU line will be extended up to 2015. Then will come Basilisk APU (rumor). Depending of process technology available and the status of stacked RAM AMD could frop FX line then and move to an all APU strategy from tablets to supercomputers.