AMD CPU speculation... and expert conjecture

[esrever]

UMA => single access time (uniform access).
NUMA => multiple access times (non-uniform access).

http://en.wikipedia.org/wiki/Non-Uniform_Memory_Access
http://en.wikipedia.org/wiki/Uniform_Memory_Access

Under NUMA, a processor can access its own local memory faster than non-local memory (memory local to another processor or memory shared between processors).

Not only HSA is using the correct term unified, but Nvidia also uses the correct term unified for its next gen dGPU

Nvidia Maxwell GPUs will be able to access its own local memory (GDDR5) and also the remote memory (system DDR3). Two physical memory pools => two access times => the memory model is non-uniform but unified.

====================================================

More stuff. Charlie speculation about the XboxOne CPU being clocked at ~1.9GHz was not right:

We recently just went into full production, so we're now producing en masse Xbox One consoles. We've had real good progress on the system. In fact, we just updated the CPU performance to 1.75 GHz on top of the graphics performance improvement, so the system is really going to shine [and] the games look pretty incredible.

http://arstechnica.com/gaming/2013/09/xbox-one-gets-a-cpu-speed-boost-to-go-with-its-faster-gpu/

Maxwell was cancelled entirely per S|A...

http://news.softpedia.com/news/NVIDIA-Maxwell-Graphics-Cards-Confirmed-for-Q1-2014-377838.shtml

that says maxwell is 28nm.

 

[8350rocks]

UMA => single access time (uniform access).
NUMA => multiple access times (non-uniform access).

http://en.wikipedia.org/wiki/Non-Uniform_Memory_Access
http://en.wikipedia.org/wiki/Uniform_Memory_Access

Under NUMA, a processor can access its own local memory faster than non-local memory (memory local to another processor or memory shared between processors).

Not only HSA is using the correct term unified, but Nvidia also uses the correct term unified for its next gen dGPU

Nvidia Maxwell GPUs will be able to access its own local memory (GDDR5) and also the remote memory (system DDR3). Two physical memory pools => two access times => the memory model is non-uniform but unified.

====================================================

More stuff. Charlie speculation about the XboxOne CPU being clocked at ~1.9GHz was not right:

We recently just went into full production, so we're now producing en masse Xbox One consoles. We've had real good progress on the system. In fact, we just updated the CPU performance to 1.75 GHz on top of the graphics performance improvement, so the system is really going to shine [and] the games look pretty incredible.

http://arstechnica.com/gaming/2013/09/xbox-one-gets-a-cpu-speed-boost-to-go-with-its-faster-gpu/

Maxwell was cancelled entirely per S|A...

http://news.softpedia.com/news/NVIDIA-Maxwell-Graphics-Cards-Confirmed-for-Q1-2014-377838.shtml

Ok, so you think Charlie is wrong, and the massive staff reduction and technology cancellation was something other than Maxwell?

You really should read the article @ S|A

 

[jdwii]

Although this article is older i still feel like posting it, http://www.theverge.com/2013/6/21/4452488/amd-sparks-x86-transition-for-next-gen-game-consoles

 

[Cazalan]

Apparently this part of this recent news didn't receive the attention that deserves:

AMD will sell the 64-bit ARM processors alongside x86 server processors, but the company expects ARM to outsell x86 in the long run.

Everyone and their mother is making ARM 64 bit chips next year. Except you can expect Apple and Qualcomm to have custom and faster cores like they have for a while. Apple, Applied Micro, AMD, Broadcom, Calxeda, HiSilicon, NVidia, Samsung, ST Microelectronics, Qualcomm, X-Gene, etc. Only one has actually showed live hardware running, and that's X-Gene.

The ARM market is actually more competitive than x86 so it's not a walk in the park for AMD.

 

[Cazalan]

Ok, so you think Charlie is wrong, and the massive staff reduction and technology cancellation was something other than Maxwell?

You really should read the article @ S|A

20nm has come later than both NVidia and AMD expected. As fabless companies they don't get much choice in the matter.

When did he say there was massive layoffs? The company has 3 billion in cash and no debt. He even said they accelerated Logan to be ahead of schedule.

 

[de5_Roy]

Everyone and their mother is making ARM 64 bit chips next year. Except you can expect Apple and Qualcomm to have custom and faster cores like they have for a while. Apple, Applied Micro, AMD, Broadcom, Calxeda, HiSilicon, NVidia, Samsung, ST Microelectronics, Qualcomm, X-Gene, etc. Only one has actually showed live hardware running, and that's X-Gene.

The ARM market is actually more competitive than x86 so it's not a walk in the park for AMD.

i consider qualcomm to be intel's arm-rival and mediatek as amd's arm-rival, along with nvidia. i agree that arm market is more competitive than x86 (almost) duopoly.
i think nvidia, calxeda and samsung will specially try to push into server markets at different target sectors. this is as far as my limited knowledge does.

edit: samsung and mediatek both have 'octo-core' arm socs, so amd's 'moar cores' hype work work as effectively as it did in pcs.

 

[esrever]

Apparently this part of this recent news didn't receive the attention that deserves:

AMD will sell the 64-bit ARM processors alongside x86 server processors, but the company expects ARM to outsell x86 in the long run.

Everyone and their mother is making ARM 64 bit chips next year. Except you can expect Apple and Qualcomm to have custom and faster cores like they have for a while. Apple, Applied Micro, AMD, Broadcom, Calxeda, HiSilicon, NVidia, Samsung, ST Microelectronics, Qualcomm, X-Gene, etc. Only one has actually showed live hardware running, and that's X-Gene.

The ARM market is actually more competitive than x86 so it's not a walk in the park for AMD.

thats why AMD isn't competing directly with most of the ARM players.

 

[juanrga]

Apparently this part of this recent news didn't receive the attention that deserves:

AMD will sell the 64-bit ARM processors alongside x86 server processors, but the company expects ARM to outsell x86 in the long run.

Everyone and their mother is making ARM 64 bit chips next year. Except you can expect Apple and Qualcomm to have custom and faster cores like they have for a while. Apple, Applied Micro, AMD, Broadcom, Calxeda, HiSilicon, NVidia, Samsung, ST Microelectronics, Qualcomm, X-Gene, etc. Only one has actually showed live hardware running, and that's X-Gene.

The ARM market is actually more competitive than x86 so it's not a walk in the park for AMD.

thats why AMD isn't competing directly with most of the ARM players.

I suspect the HSA foundation includes some kind of hidden clause/agreement. Traditional ARM vendors will focus in mobile, whereas AMD will focus in servers and probably HPC, but with ARM instead of x86.

In any case, the facts are that AMD is replacing x86 jaguar based servers by ARM based servers, and has said that they expect ARM sales to outperform x86 sales in the long run (2016?). This indicates, at least to me, that AMD will be not releasing high-end x86 server chips in the future. And I suspect that AMD assault to the HPC market will be made with HSA ARM designs.

In fact, I am almost sure that I will choose HSA ARM in the poll about what architecture will dominate the world. Apparently Nvidia and others think something similar.

 

[de5_Roy]

I suspect the HSA foundation includes some kind of hidden clause/agreement. Traditional ARM vendors will focus in mobile, whereas AMD will focus in servers and probably HPC, but with ARM instead of x86.

doesn't make sense to me, from a business standpoint.
they're businesses, not secret hobby club. why would anyone in the right mind give up a lucrative sector so that amd can make money off of selling arm-based server solutions?

In any case, the facts are that AMD is replacing x86 jaguar based servers by ARM based servers, and has said that they expect ARM sales to outperform x86 sales in the long run (2016?). This indicates, at least to me, that AMD will be not releasing high-end x86 server chips in the future. And I suspect that AMD assault to the HPC market will be made with HSA ARM designs.

In fact, I am almost sure that I will choose HSA ARM in the poll about what architecture will dominate the world. Apparently Nvidia and others think something similar.

what facts about amd replacing x86 jaguar for arm? i'd like to see proof, something official. no logical(!) leaps or 'out there' speculation, please.
by the time arm gains traction, it'll compete with jaguar's successor. unless arm v8(or whatever 64bit is called) is so awesome that it launches without ever needing any tweaks or revisions.

 

[blackkstar]

So now we are going from AMD leaving desktop enthusiast market to AMD leaving all x86 markets?

They are going to give up a lot of things by abandoning these markets. Currently 30% of their sales are FX. Professional market, gone because no one wants a workstation chip that can't run Adobe and Autodesk products. Gaming market, gone unless you think AMD wants to sell $500 dGPUs for Android games.

AMD needs to shift focus to mobile, APUs, and HSA, there is no doubt about it. But they will still need big cores and they will still need big dGPU.

AMD is gonna dig themselves into a hole. If they push HSA too hard and ignore traditional workloads, they're going to force everyone else who has applications that aren't HSA ready or that don't scale to tons of cores very well onto other platforms.

AMD's biggest problem seems to me that their x86 big cores are designed for multi-threading and they're pushing for HSA, which will do multi-thread fantastically. Far better than a traditional CPU. Which means that their CPUs are going to be made obsolete by their own new products instead of complimenting them well by being strong at what HSA/GPGPU is bad at, which is single thread.

AMD big x86 cores should be single thread oriented to compliment HSA GPGPU instead of trying to compete with it.

However I still do not buy HSA replacing traditional computing entirely. ignoring traditional CPU workloads means that AMD is going to have a product line that only works for HSA enabled applications, and if you want to use it for traditional workloads, you're going to have an awful experience. The trade off will be huge, and I would imagine that the amount of traditional software will vastly out-number HSA software for a very long time, if forever.

Hardly anyone is going to choose "significantly slower in existing software but very fast in specific workloads" over "pretty good at everything and there's only a few applications where things are slow."

So overall, no, I'm not buying the HSA hype like everyone is. Yes, it'll be absolutely amazing for specific, niche roles that grow overtime. However, APU can't replace traditional CPUs or dGPUs and it never will be able to as long as it's a single socket solution with a die size that isn't ridiculous.

If you want my opinion on where HSA is going to sit in regards to marketshare and usage, it'll sit somewhere between CUDA/OpenCL and traditional CPU workloads. It will suppliment it, perhaps greatly in specific workloads, but it can't replace it because software will take forever to catch up and it will be nearly impossible to get software developers to jump on something with such small market share.

The only shining light is gaming. Where, if AMD goes APU only, their strongest CPU will be around FX 4300 level, FX 6300 if you're optimistic about Steamroller. That's the end of AMD's gaming platform. It means that the only reason to buy a high end AMD dGPU will be to pair it with an Intel CPU, because something along the lines of FX 4300 isn't going to push anything extreme. And then you won't have HSA. Meaning that a large portion of gamers won't care about HSA at all. So then you have consoles pushing for HSA and then Intel CPU PC gaming rigs without HSA being purchase while you can buy HSA PC gaming rigs.

And look how bad AMD wants to have a high end gaming CPU. They released FX 9590 with 220w TDP and $900 price tag just so they could try and have something competitive in the high end gaming segment.

Going HSA only is going to push a lot of people away from AMD products. Also, I don't recall ever reading that HSA is going to completely replace traditional computing. I have seen numbers saying it can do great things and that it is good (I do think HSA is awesome), but never anything regarding "traditional computing is dead, RIP in peace!!!!"

That seems to come from forum conjecture that people assume HSA is going to replace everything.

APU with HSA, if that's AMD's best performing product, will get mauled by Intel + Phi and Power8 + CUDA. There is absolutely no way a single AMD APU is going to come remotely close to the levels of performance both of those offer.

Now, 16 core Steamroller or Excavator Opteron with 7970 replacements, that'd be a scary thought.

I just wanna know, that if this market is useless and dying, why are Nvidia, IBM, Intel, and even AMD (with FirePro line) pushing towards it?

Like I said before, AMD needs HSA for desktop with CPU + dGPU. Not having that will create an awkward problem in their product line where CPU + dGPU is faster in most workloads, but loses horribly in HSA workloads while APU loses horribly in majority of workloads while it wins hard in the specific HSA ones.

http://www.amazon.com/s/ref=sr_st?keywords=desktop+cpu&qid=1378573411&rh=n%3A172282%2Cn%3A541966%2Cn%3A193870011%2Cn%3A229189%2Ck%3Adesktop+cpu&sort=popularity-rank

Sorted by popularity. Notice where FX 8350 is? Now where is desktop APU?

Popularity of FX 8350 is really contradicting a lot of things that are getting thrown around in this thread. I find it very hard to believe that AMD is going to give up their top 3 selling products on Amazon and leave that market entirely to Intel.

FX 8350 is more popular than 3770k. Amazingly i3 is more popular than FX 6300. People like FX 8000 series a lot more than people in this thread are realizing, probably because they hang out in forums where people masturbate to a product being superior because it scores 120fps in a 720p gaming benchmark with settings on low as compared to one that scores 110fps.

AMD knows this, I'm assuming they thought that "moar coars and moar jigahurtz" was working so they released FX 9000 series, plus it would function as a halo product and it looks better for AMD if people who don't know better look at products on Amazon and see AMD HAS products that can be priced that high.

FX 9590 is on the same page as Intel 3960x. Serious under-estimation of what FX is to AMD going on in this thread, and serious over-estimation of what HSA is going to do in the short term (next 3 years or so).

EDIT: Just for fun, newegg results of desktop CPUs sorted by number of reviews:
http://www.newegg.com/Processors-Desktops/SubCategory/ID-343?Order=REVIEWS

I'm making a leap and assuming all CPUs have the same percentage of users review the product, but if that is true, AMD has sold a lot of FX chips. Much more FX than Intel has sold i5 to folks that shop for CPUs only at Newegg.

 

[juanrga]

I suspect the HSA foundation includes some kind of hidden clause/agreement. Traditional ARM vendors will focus in mobile, whereas AMD will focus in servers and probably HPC, but with ARM instead of x86.

doesn't make sense to me, from a business standpoint.
they're businesses, not secret hobby club. why would anyone in the right mind give up a lucrative sector so that amd can make money off of selling arm-based server solutions?

In any case, the facts are that AMD is replacing x86 jaguar based servers by ARM based servers, and has said that they expect ARM sales to outperform x86 sales in the long run (2016?). This indicates, at least to me, that AMD will be not releasing high-end x86 server chips in the future. And I suspect that AMD assault to the HPC market will be made with HSA ARM designs.

In fact, I am almost sure that I will choose HSA ARM in the poll about what architecture will dominate the world. Apparently Nvidia and others think something similar.

what facts about amd replacing x86 jaguar for arm? i'd like to see proof, something official. no logical(!) leaps or 'out there' speculation, please.
by the time arm gains traction, it'll compete with jaguar's successor. unless arm v8(or whatever 64bit is called) is so awesome that it launches without ever needing any tweaks or revisions.

Because AMD is licensing the fusion technology to the HSA foundation. Do you really believe AMD would say hey guys take all this technology for almost free, which will improve your products a lot of, and now compete against me please. It seems evident that AMD is licensing to mobile players, because those will be competing with Intel in the mobile market. Whereas AMD has already shared its intention that they want take Intel servers market using ARM servers. Everyone in the HSA foundation wins each one in a market.

The facts about AMD replacing x86 jaguar for ARM were given before. AMD official server roadmap and the presentation explaining why are replacing jaguar servers by arm servers. Also AMD quote, given above, claiming that they expect the sales of ARM servers to surpase the sales of x86 servers. Do you believe AMD will invest lots of resources in future x86 servers when is predicting that sales will drop?

So now we are going from AMD leaving desktop enthusiast market to AMD leaving all x86 markets?

They are going to give up a lot of things by abandoning these markets. Currently 30% of their sales are FX. Professional market, gone because no one wants a workstation chip that can't run Adobe and Autodesk products. Gaming market, gone unless you think AMD wants to sell $500 dGPUs for Android games.

AMD needs to shift focus to mobile, APUs, and HSA, there is no doubt about it. But they will still need big cores and they will still need big dGPU.

AMD is gonna dig themselves into a hole. If they push HSA too hard and ignore traditional workloads, they're going to force everyone else who has applications that aren't HSA ready or that don't scale to tons of cores very well onto other platforms.

AMD's biggest problem seems to me that their x86 big cores are designed for multi-threading and they're pushing for HSA, which will do multi-thread fantastically. Far better than a traditional CPU. Which means that their CPUs are going to be made obsolete by their own new products instead of complimenting them well by being strong at what HSA/GPGPU is bad at, which is single thread.

AMD big x86 cores should be single thread oriented to compliment HSA GPGPU instead of trying to compete with it.

However I still do not buy HSA replacing traditional computing entirely. ignoring traditional CPU workloads means that AMD is going to have a product line that only works for HSA enabled applications, and if you want to use it for traditional workloads, you're going to have an awful experience. The trade off will be huge, and I would imagine that the amount of traditional software will vastly out-number HSA software for a very long time, if forever.

Hardly anyone is going to choose "significantly slower in existing software but very fast in specific workloads" over "pretty good at everything and there's only a few applications where things are slow."

So overall, no, I'm not buying the HSA hype like everyone is. Yes, it'll be absolutely amazing for specific, niche roles that grow overtime. However, APU can't replace traditional CPUs or dGPUs and it never will be able to as long as it's a single socket solution with a die size that isn't ridiculous.

If you want my opinion on where HSA is going to sit in regards to marketshare and usage, it'll sit somewhere between CUDA/OpenCL and traditional CPU workloads. It will suppliment it, perhaps greatly in specific workloads, but it can't replace it because software will take forever to catch up and it will be nearly impossible to get software developers to jump on something with such small market share.

The only shining light is gaming. Where, if AMD goes APU only, their strongest CPU will be around FX 4300 level, FX 6300 if you're optimistic about Steamroller. That's the end of AMD's gaming platform. It means that the only reason to buy a high end AMD dGPU will be to pair it with an Intel CPU, because something along the lines of FX 4300 isn't going to push anything extreme. And then you won't have HSA. Meaning that a large portion of gamers won't care about HSA at all. So then you have consoles pushing for HSA and then Intel CPU PC gaming rigs without HSA being purchase while you can buy HSA PC gaming rigs.

And look how bad AMD wants to have a high end gaming CPU. They released FX 9590 with 220w TDP and $900 price tag just so they could try and have something competitive in the high end gaming segment.

Going HSA only is going to push a lot of people away from AMD products. Also, I don't recall ever reading that HSA is going to completely replace traditional computing. I have seen numbers saying it can do great things and that it is good (I do think HSA is awesome), but never anything regarding "traditional computing is dead, RIP in peace!!!!"

That seems to come from forum conjecture that people assume HSA is going to replace everything.

APU with HSA, if that's AMD's best performing product, will get mauled by Intel + Phi and Power8 + CUDA. There is absolutely no way a single AMD APU is going to come remotely close to the levels of performance both of those offer.

Now, 16 core Steamroller or Excavator Opteron with 7970 replacements, that'd be a scary thought.

I just wanna know, that if this market is useless and dying, why are Nvidia, IBM, Intel, and even AMD (with FirePro line) pushing towards it?

Like I said before, AMD needs HSA for desktop with CPU + dGPU. Not having that will create an awkward problem in their product line where CPU + dGPU is faster in most workloads, but loses horribly in HSA workloads while APU loses horribly in majority of workloads while it wins hard in the specific HSA ones.

http://www.amazon.com/s/ref=sr_st?keywords=desktop+cpu&qid=1378573411&rh=n%3A172282%2Cn%3A541966%2Cn%3A193870011%2Cn%3A229189%2Ck%3Adesktop+cpu&sort=popularity-rank

Sorted by popularity. Notice where FX 8350 is? Now where is desktop APU?

Popularity of FX 8350 is really contradicting a lot of things that are getting thrown around in this thread. I find it very hard to believe that AMD is going to give up their top 3 selling products on Amazon and leave that market entirely to Intel.

FX 8350 is more popular than 3770k. Amazingly i3 is more popular than FX 6300. People like FX 8000 series a lot more than people in this thread are realizing, probably because they hang out in forums where people masturbate to a product being superior because it scores 120fps in a 720p gaming benchmark with settings on low as compared to one that scores 110fps.

AMD knows this, I'm assuming they thought that "moar coars and moar jigahurtz" was working so they released FX 9000 series, plus it would function as a halo product and it looks better for AMD if people who don't know better look at products on Amazon and see AMD HAS products that can be priced that high.

FX 9590 is on the same page as Intel 3960x. Serious under-estimation of what FX is to AMD going on in this thread, and serious over-estimation of what HSA is going to do in the short term (next 3 years or so).

EDIT: Just for fun, newegg results of desktop CPUs sorted by number of reviews:
http://www.newegg.com/Processors-Desktops/SubCategory/ID-343?Order=REVIEWS

I'm making a leap and assuming all CPUs have the same percentage of users review the product, but if that is true, AMD has sold a lot of FX chips. Much more FX than Intel has sold i5 to folks that shop for CPUs only at Newegg.

AMD will be not focusing in releasing products for a very tiny minory. Neither Intel is, look to reviews of recent IB-E, most people (even Anadtech) is dissapointed. I think that what really said AMD was they expect 30% of sales to come from _CPUs_ alone. I think they didn't mention FX.

Traditional workloads are done. ___All___ players (Intel, Nvidia, Qualcom, IBM, AMD...) are focusing in heterogeneous computing. Tradittional workloads will become legacy.

Not only HSA in the desktop is of first importance for AMD. Take a look to servers. They are replacing Opteron CPUs by Berlin HSA APUs.

AMD big x86 cores are not competing with GPGPU. Take a look to AMD slides: CPU for serial workloads, GPU for paralell workloads.

Also, as I said before Kaveri CPU will be at i5/ FX-6 level approx. That is enough for inmense mayority of buyers.

AMD has already shared its firm intention to go to an all APU strategy.

You don't need to pair a high end AMD dGPU with Intel, you can use FX-8/FX-9 CPUs before AMD releases the top high-end APUs in 2015.

The FX-9590 is not at $900, AMD has forced a huge drop in the price. Last time I checked it was at Haswell level price.

"Going HSA only is going to push a lot of people away from AMD products." Market predictions shows otherwise.

AMD will be not competing in HPC with HSA APUs alone, but with HSA APUs + HSA dGPUs.

I don't buy amazon 'popularity' list. An i4-4770k the most popular, even above the i5-3570k? Really? LOL Look to number of voters also 50(i7) vs 345 (i5). Does amazon want to convince me that the i7 Haswell is the more popular, but only 1 of each 7 buyers vote for it? Moreover, that list ignores pre-built computers that people purchase. Steam stastistics show that the FX-8350 is not even a 1% of gaming systems.

 

[de5_Roy]

Because AMD is licensing the fusion technology to the HSA foundation. Do you really believe AMD would say hey guys take all this technology for almost free, which will improve your products a lot of, and now compete against me please. It seems evident that AMD is licensing to mobile players, because those will be competing with Intel in the mobile market. Whereas AMD has already shared its intention that they want take Intel servers market using ARM servers. Everyone in the HSA foundation wins each one in a market.

amd is licensing the fusion technology? this is the first time i've heard of a 'fusion technology'. please explain what it means. from what i know, 'fusion' used to be x86 cores with radeon gpu. amd has not associated 'fusion' with arm, so far. prolly because they were sued...

afaik, they're developing an open hardware standard. it will be up to vendors how they implement it in hardware and software, then get certified for hsa compliance. that's how the competition will happen. amd can't help if someone else outdoes them. you know, amd sometimes beat intel with amd's own x86 cpus. my memory's a bit hazy, but i think those used to be called athlon fx.

there is no 'mobile players'. they're 'players' aiming for wherever they can get their foothold. mobile (i.e. smartphones and tablets) just happens to be the widely known one.

you have it backwards. arm has such commanding presence in mobile that amd and intel will be the ones competing against arm 'armada'.

your 'everybody wins' is an illusion. i woulda said delusion but i'd rather wait a few years to see how amd actually does in arm-server sector against vendors that already provide (32 bit, calxeda et al) machines. before you name-drop 64-bit arm v8, keep in mind that a commercially available product based on arm v8 is currently non-existent. well, i searched and didn't find any.

The facts about AMD replacing x86 jaguar for ARM were given before. AMD official server roadmap and the presentation explaining why are replacing jaguar servers by arm servers. Also AMD quote, given above, claiming that they expect the sales of ARM servers to surpase the sales of x86 servers. Do you believe AMD will invest lots of resources in future x86 servers when is predicting that sales will drop?

facts given before where? i am not going back searching through walls of text. if you have official proof of amd stating they're replacing jaguar with arm, please provide. pretty please.
what server roadmap? the last one i saw showed coexistence. if there's a different one, i'd like to see that.
amd expects that server arm soc will surpass x86 but they gave no size or scale of the market itself. expectation is not the same as certainty.
i don't believe, i know that amd invested a lot of r&d into fabbing and selling zambezi cpus in the pc market and they predicted that zambezi will outsell everything. in the end those got humiliated by amd's own apus, let alone intel's cpus. that's not a prediction nor expectation.

parrotting promo slides have it's downsides. people who parrot promo slides often overlook the facts stated in the very last slide where vendors disclaim all responsibilities to protect themselves from any bad outcomes.

 

[The Q6660 Inside]

I bought two haswell cpus on ebay an i5 4430 and an i5 4670k. Intel cpus are so cheap now I will have to decide which to keep. I am leaning towards keeping the i5 4430 as its good enough. AMD are uncompetitive as an all round cpu.

Competiton keeps the 4670K cheapish. Heck, I found an in box 8320 for 120$ on ebay. If AMD was to fall off the face of the earth, Intel will price jack your i5 to $400, same for vice versa. $TFU about you "all round" (deprecated, biased and or single threaded apps soon to be irrelavent that have little to no increase in realistic performance) The FX line has plenty of customers, and in the end that is all that matters.

 

[noob2222]

I suspect the HSA foundation includes some kind of hidden clause/agreement. Traditional ARM vendors will focus in mobile, whereas AMD will focus in servers and probably HPC, but with ARM instead of x86.

doesn't make sense to me, from a business standpoint.
they're businesses, not secret hobby club. why would anyone in the right mind give up a lucrative sector so that amd can make money off of selling arm-based server solutions?

In any case, the facts are that AMD is replacing x86 jaguar based servers by ARM based servers, and has said that they expect ARM sales to outperform x86 sales in the long run (2016?). This indicates, at least to me, that AMD will be not releasing high-end x86 server chips in the future. And I suspect that AMD assault to the HPC market will be made with HSA ARM designs.

In fact, I am almost sure that I will choose HSA ARM in the poll about what architecture will dominate the world. Apparently Nvidia and others think something similar.

what facts about amd replacing x86 jaguar for arm? i'd like to see proof, something official. no logical(!) leaps or 'out there' speculation, please.
by the time arm gains traction, it'll compete with jaguar's successor. unless arm v8(or whatever 64bit is called) is so awesome that it launches without ever needing any tweaks or revisions.

Because AMD is licensing the fusion technology to the HSA foundation. Do you really believe AMD would say hey guys take all this technology for almost free, which will improve your products a lot of, and now compete against me please. It seems evident that AMD is licensing to mobile players, because those will be competing with Intel in the mobile market. Whereas AMD has already shared its intention that they want take Intel servers market using ARM servers. Everyone in the HSA foundation wins each one in a market.

The facts about AMD replacing x86 jaguar for ARM were given before. AMD official server roadmap and the presentation explaining why are replacing jaguar servers by arm servers. Also AMD quote, given above, claiming that they expect the sales of ARM servers to surpase the sales of x86 servers. Do you believe AMD will invest lots of resources in future x86 servers when is predicting that sales will drop?

So these facts are your own speculation? nice ... I wonder what "facts" I can make up.

Unless the a57 is 85% faster than the a15 ... replacing jaguar would fail.

Aside from that, your "facts" are based on AMD scrapping the Puma+ core architecture wich is the ACTUAL jaguar replacement. Beema, Mullins, and Nolan after that according to your made up "facts" don't exist.

Kinda funny considering amd's latest "roadmap update" specifically talks about beema.

 

[noob2222]

I bought my i5 4430 for £97 and a new i5 4670k with grid 2 for £150.

This is an AMD discussion not what crap you bought on ebay. I hope you get them and they are both defunct. GO AWAY.

Aside from that, the fact that you got them that cheap should tell you how crappy they are.

 

[noob2222]

Better then anything amd have, then again we all hear amd are cheap when they are to below average to sell for any more. AMD ain't stupid, they have to sell at there prices they do otherwise they would be bankrupt as everyone would buy intel instead. intel baytrail easily beats the jaguar cpus, looks like steamroller cores are only amd hope as it seems one sided at the moment in intels favour.

easily ... ROFL

This is on Intel's own "cripple AMD" benchmark. easily ... rofl, can't even beat jaguar at that. Go back to your Intel-blue painted apartment.

Atoms are junk, always have been, always will be.

 

[noob2222]

^^ thats 2.4 ghz turbo blue boy and 1.5 ghz isn't the fastest jaguar available. Bay trail will probably show up in time to compete with puma+

 

[The Q6660 Inside]

^^ thats 2.4 ghz turbo blue boy and 1.5 ghz isn't the fastest jaguar available. Bay trail will probably show up in time to compete with puma+

+1 Intel will have to sharpen Atom's spear to hell and back to even get close to jaguar and lets not forget our friend ARM. Hafijur needs to leave honestly, he has ruined the thread. Nobody cares about his Pentium M being "ungodly", how a 4700MQ beats an 8350 in a biased benchmark, how he got a GX9200 for dirt cheap, his unedcated rants about how Nehalem was Intel's worst product since Netburst (smells Itanic and Atom) soley due to its high TDP and how a decently clocked i7 1st gen will bottleneck a 670.. and finally his "ermehgawd AMD faildozer mawdlurr shawks" crap all night long.

 

[hcl123]

As I said before an uniform memory access model provides a single access time to physical memory. A non-uniform memory access provides two times: one for local memory and other for remote memory. This is standard stuff.

In the slide, AMD clearly represent a single physical memory pool. This works for kaveri APU, but will not work for dGPUs. Once again, the HSA specification clearly states that its memory model is unified. There is not a single mention to uniform or hUMA in it.

If we want now reinvent the words and give them other meanings then the confusion is guaranteed.

The confusion is yours, there isn't such ting as single access time

UMA => single access time (uniform access).
NUMA => multiple access times (non-uniform access).

http://en.wikipedia.org/wiki/Non-Uniform_Memory_Access
http://en.wikipedia.org/wiki/Uniform_Memory_Access

Under NUMA, a processor can access its own local memory faster than non-local memory (memory local to another processor or memory shared between processors).

All very correct, but is like that by the "arbitration" scheme, which is local then remote, which invariability leads to mem "hops" and different access times for NUMA. But what defines NUMA is address spaces not access time.. meaning different "physical" address spaces in an hierarchical view (kind of "single" view but not linear unified) , which invariably leads to different access times.

Even in Numa access times can vary substantially according to the interconnect of the different mem pools... but is not because of that that there must be a different nomenclature or different scheme based on the "amount of time" .

Accept it, the distinctive feature is not access times, those are consequences, not the determinant factor. NUMA and more ccNUMA are coherent physical memory in an hierarchical view... and in here virtual memory becomes orthogonal, cache coherency is not dependent on VM.

hUMA is different for not saying the opposite... the coherent unified view is "Virtual Memory" and the cache coherency is somehow dependent of this view. (first make VM an unified view, even for "devices", then attach cache coherency on top of that).

 

[hcl123]

Apparently this part of this recent news didn't receive the attention that deserves:

AMD will sell the 64-bit ARM processors alongside x86 server processors, but the company expects ARM to outsell x86 in the long run.

Everyone and their mother is making ARM 64 bit chips next year. Except you can expect Apple and Qualcomm to have custom and faster cores like they have for a while. Apple, Applied Micro, AMD, Broadcom, Calxeda, HiSilicon, NVidia, Samsung, ST Microelectronics, Qualcomm, X-Gene, etc. Only one has actually showed live hardware running, and that's X-Gene.

The ARM market is actually more competitive than x86 so it's not a walk in the park for AMD.

thats why AMD isn't competing directly with most of the ARM players.

I suspect the HSA foundation includes some kind of hidden clause/agreement. Traditional ARM vendors will focus in mobile, whereas AMD will focus in servers and probably HPC, but with ARM instead of x86.

In any case, the facts are that AMD is replacing x86 jaguar based servers by ARM based servers, and has said that they expect ARM sales to outperform x86 sales in the long run (2016?). This indicates, at least to me, that AMD will be not releasing high-end x86 server chips in the future. And I suspect that AMD assault to the HPC market will be made with HSA ARM designs.

In fact, I am almost sure that I will choose HSA ARM in the poll about what architecture will dominate the world. Apparently Nvidia and others think something similar.

I can't find any "hidden clause" not even slightly suggested on HSA.

HSA is not about ARM + x86 or something like that... HSA is about an Virtual Machine with its own kind of ISA, which is called HSAIL. HSA programs even have its own "binary format" called BRIG...

http://www.slideshare.net/hsafoundation/hsail-final-11junepptx

HSA should work on x86, ARM, MIPS, Power etc... only needs is the runtime definitions ported to those CPU or LU (latency units) ISAs. See it like Java... if the runtime/JVM (java virtual machine) itself is ported to those CPU ISAs, the same Java program can run "unchanged" in a x86, ARM, MIPS or Power system without issues... The same with HSA BRIG program, only HSA is more low level, it can even have JVM on top of it, or like in the "Sumatra enable JVM" the HSA runtime is embedded in the JVM itself.

HSA is a SIMT( single instruction multi-thread) mostly JIT paradigm... to go for the GPU (though those could be accessed directly by other languages/runtimes) it needs a runtime that is dependent on the OS for the Virtual Memory management... it can have "(Cacheable) coherent system memory" mixed with "(non-cacheable) non coherent system memory" in a seamless way, dependent on the MMU of those system processing elements... an in here is where IOMMU enters, is the definition of those MMUs for HSA...

The "cacheable coherent" factor is on top of that, and it MUST be a broad definition, different vendors can have different "cache coherency" schemes (as long it is memory coherent by the "virtual memory") and i'm convinced that hUMA will be only an AMD "name", other vendors can have exactly similar functionality but call it different... and it doesn't matter, the same "programs will run unchanged" in any system (x86, ARM etc + different accels), because it depends on a runtime/driver close tight with the OS virtual memory management.

http://hsafoundation.com/hsa-developer-tools/

HSA is a Low Level Virtual Machine... very flexible, allowing direct interaction with hardware (example Sumatra JVM), and with innovating features AQL (access queue language) approach with Hardware Queue Buffers (and that will be how much of hardware definition there will be, Queues, IOMMU MMUs, and *your own* cache coherency on top if you like)..

[ Matter of fact HSA as a LLVM, can even work on the most varied discrete implementation... matter of fact "Berlin" is HSA (probably those MMUs, Queues and CC on top)... and Berlin is Seamicro clusters, so we will see a Seamicro cluster with Freedom Fabric running HSA programs out of the box ...all the way from SoC and ultraportable, to big cluster machines LOL (most probably the ARM Seattle will be the same, its not an APU, but could have discrete GPGPU, DSP, FPGA etc) ... ]

 

[The Q6660 Inside]

intel baytrail is miles ahead when released of jaguar and current arm cpus. The 4700mq is actually faster at most things when turbo'd compared to the fx8350. Intels generally have faster laptop cpus then desktop cpus, perfect
example intel fastest laptop dual cores are faster then any intel
desktop i3. These are just facts.
Nehalem was embarrassing, my
brother university computer room
with nehalem i7's and sandy
bridge i7's and the sandy bridge
room was very cool while the
nehalem computer room was like
a boiling room with similar
cooling. Its not hard to know why,
sandy bridge machines had
there own igpu so it disables the
high end gpus doing normal
tasks and it takes half the power
for more performance.

I got it now, amd fans like you dismiss facts and like to live in
fantasy land and not accept the
truth.

You sir, live in fantasyland. Your beloved GX9200 would do the same. Itanic and Atom was a disaster. The i7 9XX was like the GTX 480 of CPUs sure it will draw a sh*t ton of power, but give it a nice OC with water or a custom fan + repaste and you are still at 760 levels, which is more than enough for 1080p. In the case of the i7 9XX just BCLK and voltage boost to victory.

 

[hcl123]

So now we are going from AMD leaving desktop enthusiast market to AMD leaving all x86 markets?

They are going to give up a lot of things by abandoning these markets. Currently 30% of their sales are FX. Professional market, gone because no one wants a workstation chip that can't run Adobe and Autodesk products. Gaming market, gone unless you think AMD wants to sell $500 dGPUs for Android games.

AMD needs to shift focus to mobile, APUs, and HSA, there is no doubt about it. But they will still need big cores and they will still need big dGPU.

AMD is gonna dig themselves into a hole. If they push HSA too hard and ignore traditional workloads, they're going to force everyone else who has applications that aren't HSA ready or that don't scale to tons of cores very well onto other platforms.

It depends mostly on AMD than anyone else on ARM... even ARM itself...

ARM 64bit is clearly a better ISA(uarch) than x86_64 for a CPU... now the all it needs is a definition like the chinese of Godson are doing with MIPS for vectors, a wide vector definition, and Godson will have 512bit vectors even before Intel...

AMD could "port" (a subsidy for language) XOP 256bit or even 512bt to ARM etc (at least try)... that is, have a superior Vector definition (XOP is better than AVX), and an execute pipe definition for x86_64 that doesn't depend on Intel for licensing... in ARM and others... like Godson that has similar pipe (BUT VERY LIMITED) to help/accelerate run x86 on MIPs by "on-the-fly binary translation" (like Transmeta)... AMD could even negotiate "royalties" with ARM Holdings about that, and not only with ARM but also MIPS etc to.

**IF** that happens, honestly it starts to smell like rotten for the sides of x86... x86 which no matter its dominance on the PC and traditional small server markets, seems clearly now obsolete...

So... if one is long sighted, and clever sighted... the future of Intel is on AMD hands ... that is why they want to kill it the fast as they can.

Me if i were AMD CEO i'll start doing it tomorrow... and its late... Intel is a vicious entity, enemy of every body, even its own partners and costumers( sorry, but the long track of fines all over the world and law suits, seems to confirm this)... and deserves to die.

First no more patent agreements, what is done is done... then try to license XOP and x86_64 (what possible) to ARM and MIPS and why not even OpenPower... all joined by a good push for royalty free HSA and royalty free HyperTransport 4.0... it would be a slow transition, but for one thing the x86 dominating allover factor will be dead and buried( intel vision will be dead). If Intel dies or not is a totally different issue, but i would have the upper hand, and i think in the end Intel would be forced to "license" ARM or MIPS (other) to stay relevant( they could take back to life Alpha)... it would be slow but inexorable (my view).

XOP and x86_64 on ARM (etc) is most relevant because *the* principal factor for the success of a new BETTER ISA, was and is exactly the "software ecosystem", not the technical prowess of a ISA (that is why x86 won so far)... no "software ecosystem" and you can even have the upper duper frincking best ISA/uarch in the world, that it goes nowhere...

In the end i think something in those lines is unavoidable... even for x86... it will be Intel on one side, with its very peculiar views about heterogeneous and else, and it will be everybody else on the other side.

Make no mistake, its NOT a technical approach, its a "political" approach... follow the leader and you'll always be a second choice no matter what, and no matter advertising and publicity or propaganda, so the *wise political choice* is exactly NOT follow the leader, and more and "free" technical approaches are better than alone, and if you break then you could be dead on the leader command... like happened to nvidia chipset business, and could happen to nvidia discrete adapters (1.5B has can have an ensnaring irresistible appeal, but nvidia is dancing with dead...)

 

[The Q6660 Inside]

So now we are going from AMD leaving desktop enthusiast market to AMD leaving all x86 markets?

They are going to give up a lot of things by abandoning these markets. Currently 30% of their sales are FX. Professional market, gone because no one wants a
workstation chip that can't run Adobe and Autodesk
products. Gaming market, gone unless you think AMD
wants to sell $500 dGPUs for Android games.

AMD needs to shift focus to mobile, APUs, and HSA, there is
no doubt about it. But they will still need big cores and they will still need big dGPU.

AMD is gonna dig themselves into a hole. If they push
HSA too hard and ignore
traditional workloads, they're going to force everyone else who has applications that aren't HSA ready or that don't scale to tons of cores very well onto other platforms.

It depends mostly on AMD than anyone else on ARM... even ARM itself...

ARM 64bit is clearly a better ISA(uarch) than x86_64 for a CPU... now the all it needs is a
definition like the chinese of Godson are doing with MIPS for vectors, a wide vector definition, and Godson will have 512bit vectors even before Intel...

AMD could "port" (a subsidy for language) XOP 256bit or even 512bt to ARM etc (at least try)... that is, have a superior Vector definition (XOP is better than AVX), and an execute pipe definition for x86_64 that doesn't depend on Intel for licensing... in ARM and others... like Godson
that has similar pipe (BUT VERY LIMITED) to help/accelerate run x86 on MIPs by "on-the-fly binary translation" (like Transmeta)... AMD could even negotiate "royalties" with ARM Holdings
about that, and not only with ARM but also MIPS etc to.

**IF** that happens, honestly it starts to smell like rotten for the
sides of x86... x86 which no matter its dominance on the PC and traditional small server markets, seems clearly now obsolete..

So... if one is long sighted, and clever sighted... the future of Intel is on AMD hands ... that is why they want to kill it the fast as they can.

Me if i were AMD CEO i'll start doing it tomorrow... and its late... Intel is a vicious entity, enemy of every body, even its own partners
and costumers( sorry, but the long track of fines all over the world and law suits, seems to confirm this)... and deserves to die.
First no more patent agreements, what is done is done... then try to license XOP and x86_64 (what possible) to ARM and MIPS and

why not even OpenPower... all joined by a good push for royalty free HSA and royalty free HyperTransport 4.0... it would be a slow transition, but for one thing the x86 dominating allover factor will be dead and buried( intel vision will be dead). If Intel dies or not is a totally different issue, but i would have the upper hand, and i think in the end Intel would be forced to "license" ARM or MIPS (other) to stay relevant( they could take back to life Alpha)... it would be slow but inexorable (my view).

XOP and x86_64 on ARM (etc) is most relevant because *the* principal factor for the success of a new BETTER ISA, was and is exactly the "software ecosystem", not the technical





prowess of a ISA (that is why x86 won so far)... no "software ecosystem" and you can even have the upper duper frincking best ISA/uarch in the world, that it goes nowhere...

In the end i think something in





those lines is unavoidable... even for x86... it will be Intel on one side, with its very peculiar views about heterogeneous and else, and it will be everybody else on the other side.

Make no mistake, its NOT a

technical approach, its a "political" approach... follow the leader and you'll always be a second choice no matter what, and no matter advertising and publicity or propaganda, so the *wise political choice* is exactly NOT follow the leader, and more and "free" technical approaches are better than alone, and if you break then you could be dead on the leader command... like happened to nvidia chipset business, and

could happen to nvidia discrete adapters (1.5B has can have an ensnaring irresistible appeal, but nvidia is dancing with dead...)

+1 In the end, ARM is going to spread and dominate the markets, eventually that is.

 

[noob2222]

Maybe once arm gets past 486 speeds and has a stable os. ARM is energy efficient, although its anything but "fast".

 

[juanrga]

Because AMD is licensing the fusion technology to the HSA foundation. Do you really believe AMD would say hey guys take all this technology for almost free, which will improve your products a lot of, and now compete against me please. It seems evident that AMD is licensing to mobile players, because those will be competing with Intel in the mobile market. Whereas AMD has already shared its intention that they want take Intel servers market using ARM servers. Everyone in the HSA foundation wins each one in a market.

amd is licensing the fusion technology? this is the first time i've heard of a 'fusion technology'. please explain what it means. from what i know, 'fusion' used to be x86 cores with radeon gpu. amd has not associated 'fusion' with arm, so far. prolly because they were sued...

Therefore, you have been missing the reviews, the news, and the marketing of last years... LOL

AMD fusion plans started to be developed many years ago. Why do you believe AMD bought ATI? And those plans were made public in many occasions:

The Bulldozer design has been influenced by AMD's long-term beliefs about the way processors should be built. First, the company believes that workloads will become increasingly multithreaded; processors should be optimized for multithreaded throughput—more concurrent threads—rather than single-threaded performance.

Second, it believes that heavy floating point tasks shouldn't be done on the CPU at all. They should execute on GPUs. This belief underscores AMD's Fusion strategy: the integration of CPU cores and GPU cores into accelerated processing units (APUs) so that mathematical tasks can use the GPU cores.

[...]

Eight concurrent threads provides high throughput for highly multithreaded applications. The belief that floating point-heavy workloads should use the GPU justifies the separate integer/shared floating point design. With floating point heavy lifting performed by the GPU, it no longer matters that two threads have to share access to the floating point unit.

http://arstechnica.com/gadgets/2011/10/can-amd-survive-bulldozers-disappointing-debut/

Therefore, the modern HSA idea of using the GPU to compute tasks, offloading the CPU, was invented many years ago by AMD, much before the HSA foundation was born. If you believed that AMD idea of APUs was something with cheap integrated graphics, then you were very wrong.

AMD is promoting the HSA foundation for spreading this new heterogeneous technology, because they did learn of the mistakes when they tried to push the (then new) 64bits technology alone.