AMD CPU speculation... and expert conjecture

[truegenius]

Spoiler

dude
did you even read
didn't i told you to repost the specs as i missed your posts where you posed links and specs

and to mods
can i report post of this troll as he never posted any links to reply me in previous 3 pages and saying that he posted the same
he never posted tech specs even after asking him many times and saying that he already did
isn't he abusing the spirit of a proper discussion and every other poster is getting frustrated of his trolls
i would like to report his posts for trolls, can i do that ?

I gave die size, process node, memory size, gflops, TDPs, number of latency cores, bandwidths, interconnects... and I did it five or six times in this same thread.

You ignore what I say about hardware X, then you make silly claims "hardware X cannot Y", and latter you ask me to give you again the specs of hardware X. LOL!

Second, when contacting them, don't forget to mention that you started this charade with a direct personal attack against me

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-261#13260783

somehow i missed your posts
so post the specs, year, die size, consumer cost of your apu

 

[juanrga]

What four cores they have to develop? A57 is ready. Puma+ is ready and only needs to be tweaked a bit to adapt it to ARMv8. The only two cores that remain are K12 and its x86 sister. But the x86 sister is being developed from ideas used for the K12 core, thus it is not the same as designing an entire core from scratch.

I expect them to continue making improvements to all the platforms each year or they will stagnate. None of AMD's competitors are going to slow down on the ARM side and Intel certainly isn't slowing down on the x86 side.

This is not answering my question. What four cores? So far as I know A57 and Puma+ will be replaced by K12 and its "sister core" in 2016.

 

[truegenius]

Spoiler

Spoiler

did you saw that i had written something else there
"so tell me what does it mean"
it is in english so i don't think it will be that hard for you to understand that you just ignored

It did mean that I wrote just before the part that you bolded. it is in english

yeah, current gen consoles are making pc gamers laugh on them

It is funny that you still believe that exascale APUs are for gaming consoles :lol:

let me remind you my reply again
i din't saw any post where paladin admitted that dgpus will not be able to provide 4-5x more performance than an apu
so where did you debunked it juan ? imaginary land !
give me answer juan

The answer to the same question has not changed

and now will you enlighten me by giving me the specs of your apu, year and cost ( keep in mind that 390x is supposed to have 4224 gcn cores (2x core count in 3 years so you can imagine the performance of single die dgpu by 2020 let alone dual gpu cards and multi cards config which is your apu going to take-on) and it is only 2015 so your apu should be well ahead of 390x)
and how will it cope with future high resolution gaming, what resolution you are expecting this apu to be able to handle for lag free gaming

In short: you started ignoring everything I have said about those APUs currently under development by Nvidia, AMD, and Intel; ignoring what the engineers say; what gaming experts say, ignoring quotes, links, and technical details (die size, process node, memory, flops, TDPs...); and after ignoring all that you started spreading FUD about the APUS, pretending that those APUs don't have enough RAM to "store game data" lol, but when pressed down you are unable to say how many RAM those APUs have.

here are all the replies of yours to me
so tell me where is the link you are talking about

unsurprisingly there is no link
because you have posted no links and making foolish baseless comments just to show how imaginary are your replies
you just proving your self as a big big troll
stop posting links in imaginary land and post links here

Spoiler

i didn't thought that english will be so hard language , seems like its time for me to join some english classes
here it is ( EnglishClassesJoinedTogether )
now i think i studied enough english to understand your question
so you want me to predict the amount of ram that you hypothetical apus will offer
if yes then considering cpu cores, igpu (better than "hawai") and iram on same die then we won't see ram size in double digits till 2020 even if we use hbm v2 or 3
all these things on same die without burning a hole in your cabinet and pocket/wallet/debit/credit card

Wrong ram size. This show again that you did spread FUD about hardware that you don't understand.

i don't know if roy is pulling your legs , if not then you can use difference of qdr and ddr to explain gddr5 and ddr3's difference

Not a good idea, because GDDR5 is not quad data rate. Both DDR3 and GDDR5 are... double data rate.

Wrong ram size. This show again that you did spread FUD about hardware that you don't understand.

am i supposed to give right answer for something which is wrong

You are supposed to know some basic facts about the hardware on which you are commenting on.

You have shown you have no idea of which is the memory size of that hardware, but this didn't stop you from spreading FUD about the hardware being unable (in your imagination) to store game data for 8k. :lol:

do you know hardware you are commenting on or are you just copy pasting the things your friend and crystal ball told you
your first claim that your apu will annihilate dgpus ( even the extreme ones ) was enough to come to conclude that you don't know some basic facts and ignoring physics

I already explained why "some basic facts" and physics show that GPUs will be killed.

lets suppose we are in 2020 and your hypothetical apus are real then what would you see in tom's system section
you will see a question like this
"what apu do i need to play gta8 @8k @60fps , i will need apu with more ram to create ramdisk as i will record the game too, budget is 200$, should i go for 9900k"
and answer should be like this
"9900k is overkill, you will need a20-9800k because it have 16 cpu cores, 8192 gpu cores and 100GB of on die ram which is suitable for your needs and it consumes only 200w"

This proves again that you don't have any idea about the hardware that was being discussed. You are shutting random numbers from your...

You have shown you have no idea of which is the memory size of that hardware, but this didn't stop you from spreading FUD about the hardware being unable (in your imagination) to store game data for 8k. :lol:

it is your apu so you tell what is the configuration of your apu

Are you admitting that your 8K comment was FUD and that you dont have any idea of the hardware that you are commenting on?

A genius doesn't make claims about hardware and latter ask which is the hardware configuration.

and to back this imagination of your you said that it will overcome the bandwidth problem by using ram on apus thus kills ram too
next what ?
will it kill hdd too ?

This is all in your imagination. I never said that kill ram or HDD.

i am just saying that it will have these things ( like current apus have ) but you will have dedicated things which will provide you 5x the performance and upgradability with countless possible combinations of specs, everybody agrees on this thing that you can get 4-5x more performance than what an apu of that time offers

The 5x myth was debunked before. The upgradability issue was also replied before...

you are just creating FUD like bulldozer team did by saying that bulldozer will kill kill i7 and now you are saying that apus will kill gaming dgpus

Except that you confound bulldozer marketing with an academic claim based in the laws of physics. I am not reproducing the claim of particular company who want sell its products.

Today we can extract heat from 220-300W CPUs using air cooling. The main problem for high-performance APUs has been the memory bottleneck of PCs. AMD can push a 2TFLOP APU for the PS4 because the console is using a fast system memory beyond the slow DDR memory used in PCs.

Memory bottleneck is solved by HBM/HMC memory. AMD is already integrating HBM on the next APUs. Nvidia is doing the same (APUs with up to 1.6TB/s BW). Intel will start selling 'CPU' with packaged MCDRAM (500GB/s BW) the next year.

For the sake of comparison, a GTX680 has 192 GB/s BW available.

so tell me what does it mean
different memory for cpu and igpu that is not for hsa
or killing ram

I suppose that you are asking for the bold part. Well, it doesn't means your "different memory for cpu and igpu that is not for hsa" nor your "killing ram". Those two are only in your imagination. This show again that you continue pretending to criticize hardware that you don't understand.

for apu which is aimed for hsa the first option was not an option anyway so what i left with is "killing ram"
or is it doing something like xboxone to give a gaming experience of 720p @30fps

i din't saw any post where paladin admitted that dgpus will not be able to provide 4-5x more performance than an apu
indeed i always saw something like this

Juan has a very poor understanding of what "locality" is, and I believe you do to. All it means is that the latency of a particular circuit is incredibly low due to it being on the same die. In the case of dGPU's vs iGPU's we get into a particular quandary. Anything you can do on the iGPU you can do 4~5x of on a dGPU, this relates to everything from vector processing power to memory access speeds / bandwidth. In order for a iGPU to be "better" then a dGPU the workload would have to be small enough that it would be complected in less time then it takes the dGPU to receive it.

Spoiler

Ex
100 instruction sets, 3 cycles each (load, execute, store) = 300 total instructions
iGPU: 2 per cycle with 1 cycle latency per batch
dGPU: 10 per cycle with 5 cycle latency per batch (500% more latency then iGPU)

iGPU = 50 batch's (50 total cycles used in latency) = (50 * 3) +50 = 200 cycle total execution time
dGPU = 10 batch's (50 total cycles used in latency) = (10 * 3) +50 = 80 cycle execution time

2 instruction sets, 3 cycles each (load, execute, store) = 6 total instructions
iGPU: 1 batch (1 total cycle used in latency) = (1 * 3) +1 = 4 cycle total execution time
dGPU: 1 batch (5 total cycle used in latency) = (1 * 3) +5 = 8 cycle total execution time

In the first instance the instruction load was big enough to fill the entire dGPU and take advantage of it's larger resources. The 500% increased latency is a big factor in it's total performance (30 execution cycles vs 50 latency) yet due to it's 5x higher throughput it still easily beats the integrated option. In the second instance there was a much smaller set of instructions because we need the results back before we can branch to the next segment of code. In that scenario we are unable to take advantage of the higher performance of the dGPU and the iGPU becomes the better option.

This is what we are talking about when we mention that dGPU's won't be replaced. In larger workloads they are significantly superior to the iGPU's due to them having larger heat and space constraints. The iGPU's only pull ahead in small quick workloads when you need the results of the calculation ASAP.

so where did you debunked it juan ? imaginary land !

And that post from him was adequately replied. Of course, you missed the answer, how unsurprising!

 

[truegenius]

Spoiler

did you saw that i had written something else there
"so tell me what does it mean"
it is in english so i don't think it will be that hard for you to understand that you just ignored

It did mean that I wrote just before the part that you bolded. it is in english

and this shows how you ignore main part of the posts
how you ignored the part in which i was asking you to explain your comment """ AMD is already integrating HBM on the next APUs"""
and after ignoring 2-3 times you just say that you gave answer to this 5 to 6 times, this shows your lack of knowledge and your ignorance behaviour

for apu which is aimed for hsa the first option was not an option anyway so what i left with is "killing ram"
or is it doing something like xboxone to give a gaming experience of 720p @30fps

yeah, current gen consoles are making pc gamers laugh on them

It is funny that you still believe that exascale APUs are for gaming consoles :lol:

[/quotemsg]

and it is even funnier that you believe that i was talking about apus of consoles, as per you your apu is supposed to be on pc and i was talking about that and most powerfull igpus are on consoles which are unable to match pc gaming thus i was passing another humorous comment on that, this shows your either lack of IQ or lack of humor or lack of awareness about difference between pc and consoles , also unsurprisingly how unaware you are of the current topic of discussion, so get out of imaginary land , or take this as an advice and seek a doctor (and you will think that i am personally attacking you)

let me remind you my reply again
i din't saw any post where paladin admitted that dgpus will not be able to provide 4-5x more performance than an apu
so where did you debunked it juan ? imaginary land !
give me answer juan

The answer to the same question has not changed

the answer to the question was never given
will you ever get out of the imaginary land
paladin never ever admitted to your claim that apus igpu will have better performance than dgpu
so unsurprisingly you just avoided the same question second time, how big of a troll you are

and let me over simplify all the discussion going on
here i accept that i don't know abcd of computers ( and admit that Playing the flute to a buffalo (you) is a waste )
so take my question as as fresh question ( and think that i missed all the previous 263 pages of discussion so don't talk crap like "i said it in my earlier posts")
and now will you enlighten me by giving me the specs of your apu, year and cost ( keep in mind that 390x is supposed to have 4224 gcn cores (2x core count in 3 years so you can imagine the performance of single die dgpu by 2020 let alone dual gpu cards and multi cards config which is your apu going to take-on) and it is only 2015 so your apu should be well ahead of 390x)
and how will it cope with future high resolution gaming, what resolution you are expecting this apu to be able to handle for lag free gaming

In short: you started ignoring everything I have said about those APUs currently under development by Nvidia, AMD, and Intel; ignoring what the engineers say; what gaming experts say, ignoring quotes, links, and technical details (die size, process node, memory, flops, TDPs...); and after ignoring all that you started spreading FUD about the APUS, pretending that those APUs don't have enough RAM to "store game data" lol, but when pressed down you are unable to say how many RAM those APUs have.

here i asked a fresh question and said you to give a separate answer, and unsurprisingly instead of giving answer to me you just said that you have already given (which you never gave) the answer, because you never built any system so you don't even know the abc of how computer system works which de5_roy have already proved

and about that medicine thing, i posted that just like some friendly joke, as you are a Physical chemist so i thought that you will understand it more but unsurprisingly you seriously lack sense of humor

and again i am asking you the same question whose answer you never gave

so take my question as as fresh question ( and think that i missed all the previous 263 pages of discussion so don't talk crap like "i said it in my earlier posts")
and now will you enlighten me by giving me the specs of your apu, year and cost ( keep in mind that 390x is supposed to have 4224 gcn cores (2x core count in 3 years so you can imagine the performance of single die dgpu by 2020 let alone dual gpu cards and multi cards config which is your apu going to take-on) and it is only 2015 so your apu should be well ahead of 390x)
and how will it cope with future high resolution gaming, what resolution you are expecting this apu to be able to handle for lag free gaming

 

[colinp]

I like coming here to read about speculation and leaks but lately there has been too much arguing between folks who can't agree to disagree.

Agreed. Can someone just move all that nonsense into a "juanrga vs. the world" thread and leave this one clean for good natured, reasoned discourse about AMD?

 

[palladin9479]

Everyone, cease fighting with Juan and clean up your posts with spoiler tags to hide the massive quote pyramids that are taking up the entire page.

 

[cdrkf]

AMD's new stratergy looks sound to me.

Being able to mix and match X86, GCN and ARM logic onto processors is fantastically flexible.

But remember, that's a lot of extra die that won't be used in native applications. Its fantastically inefficient.

At the end of the day, the chip will run EITHER X86 code, or ARM code. Not both. So that extra ARM logic goes usused when running X86 tasks, at which point its just eating die that could be used for something else to improve performance (L3 cache, for example).

The ONLY benefit is you can support two markets with one chip, which for a fabless company is attractve. But you have to sacrifice performance from both parts to make it work. AMD is not beating Qualcomm in ARM, and not beating Intel in X86, so this will backfire in spectacular fashion.

That wasn't quite what I meant- I agree if they take and x86 cpu + graphics and bung a load of arm cores on there as well that would be inefficient in most cases, although given AMD's goals with HSA the idea is eventually that the OS would be unaware of what resources are being used and the processor can divvy up work as required (meaning everything gets used for what it's best at).

The point of the flexibility though is to be able to offer a tailored solution to a customers needs- it doesn't really help us (enthusiasts / gamers) but I think it's a sound strategy to get the company it's 'bread and butter' work. The prime example at the moment is the console chips- whatever people say about them or AMD the fact is AMD is *the only* company that could have fulfilled the requirements. Now being able to drop ARM cores onto a chip as well- even more flexible. I can even see server boxes being built with a mix of ARM and x86 cpus as server software is somewhat more flexible than desktop stuff it should be able to use everything available (hence the new pin compatible x86 and arm chips).

I doubt the ARM side of things is going to impact desktop parts much (they have included the Cortext A5 core as a security processor but the die space that takes up is minuscule). The new arm chips might be more interesting in tablets and thin laptops though...

 

[juanrga]

The point of the flexibility though is to be able to offer a tailored solution to a customers needs- it doesn't really help us (enthusiasts / gamers) but I think it's a sound strategy to get the company it's 'bread and butter' work. The prime example at the moment is the console chips- whatever people say about them or AMD the fact is AMD is *the only* company that could have fulfilled the requirements. Now being able to drop ARM cores onto a chip as well- even more flexible. I can even see server boxes being built with a mix of ARM and x86 cpus as server software is somewhat more flexible than desktop stuff it should be able to use everything available (hence the new pin compatible x86 and arm chips).

I doubt the ARM side of things is going to impact desktop parts much (they have included the Cortext A5 core as a security processor but the die space that takes up is minuscule). The new arm chips might be more interesting in tablets and thin laptops though...

From last AMD conference:

Since we're talking about a replacement for Bulldozer, this is an entirely different class of beast from the "Puma+" and Cortex-A57 cores in the first SkyBridge parts. AMD's execs noted that these are high-frequency, high-performance CPU cores that will span the range from laptops to desktops to servers—and not just "microservers."

I agree on that the x86 cores will be massively preferred by Windows desktop users, but the non-Windows users could use the ARM version. There is an interesting discussion at other forums, with some people speculating about the possibility of Apple using the K12 in a future Mac Pro.

 

[colinp]

I agree on that the x86 cores will be massively preferred by Windows desktop users, but the non-Windows users could use the ARM version. There is an interesting discussion at other forums, with some people speculating about the possibility of Apple using the K12 in a future Mac Pro.

People have been "speculating" on forums about Apple using (or buying!) AMD since Jesus was a boy. Yawn.

 

[juanrga]

Keller gives some details about the new cores:

They will be closer in frequency to the 4 GHz of AMD's latest Kaveri x86 SoC than today's 2 GHz ARM chips.

For Skybridge, AMD tweaked its new Puma x86 cores to look more like an ARM V8 core.

AMD upgraded its current on-chip fabric, I/O and memory-management units so they could accommodate either the Puma+ or standard ARM cores in Skybridge. The fabric gets an upgrade for the custom ARM cores coming in 2016.

Keller suggested AMD will consolidate its work on separate Bulldozer and Jaguar x86 cores into one processor line in the future.

Keller said about the 16/14nm FinFET process "looks pretty good, we're happy with it."

http://www.eetimes.com/document.asp?doc_id=1322237&page_number=2

Considering K12 IPC and that it will hits about 4GHz then an 8-core K12 @4GHz will be faster than FX-9000, whereas consuming much less power.

K12 IPC >= A57 IPC
A57 IPC ~ 10/7 (1.9/2) Jaguar IPC
Jaguar IPC ~ 1.075 Piledriver IPC
---------------------------------------
K12 IPC ~ 1.46 Piledriver IPC

The sister x86 core would perform about the same than the K12.

There is some people in other forums speculating about the new core outperforming skylake (they believe AMD Skybridge is a pun to Intel Skylake). I think a more reasonable target for the new core is single thread performance near Haswell (maybe 90%).

 

[cdrkf]

I agree on that the x86 cores will be massively preferred by Windows desktop users, but the non-Windows users could use the ARM version. There is an interesting discussion at other forums, with some people speculating about the possibility of Apple using the K12 in a future Mac Pro.

People have been "speculating" on forums about Apple using (or buying!) AMD since Jesus was a boy. Yawn.

Apple were actually interested in going with AMD APUs- they were looking at the original A8 but unfortunately AMD were about a year too late (it was delayed repeatedly).

Apple do use some AMD products as well- namely their graphics cards in a number of different things. I do think however that Apple are looking to use more of their own processor designs again (I think the move to 64bit on the A7 was future looking, I'm almost certain we'll see a Macbook Air based around some derivative of A7 in the near future).

 

[colinp]

Apple were actually interested in going with AMD APUs- they were looking at the original A8 but unfortunately AMD were about a year too late (it was delayed repeatedly).

Apple do use some AMD products as well- namely their graphics cards in a number of different things. I do think however that Apple are looking to use more of their own processor designs again (I think the move to 64bit on the A7 was future looking, I'm almost certain we'll see a Macbook Air based around some derivative of A7 in the near future).

I remember that story about Apple and Llano. I wonder where the story originated from, as I doubt it's veracity.

But you're right of course on the dGPU side of things. And I agree with you that Apple are more likely to move to their own core designs.

 

[Cazalan]

This is not answering my question. What four cores? So far as I know A57 and Puma+ will be replaced by K12 and its "sister core" in 2016.

From how I read it they are planning to maintain a low end core and high end core in both x86/ARM. Similar to how there is Jaguar/Steamroller now. If they can merge those to a single design then great but a single core for all market segments is unlikely.

 

[de5_Roy]

s/a slaps together a cheap kabini htpc based on AM1 platform
http://semiaccurate.com/2014/05/15/lets-build-kabini-htpc/
boxes like these would be perfect for kaveri and carrizo. not that one, but similar.

Nope.

it is sad to see your argument reduced to a single, negative word. i assumed you had reasonable amount of data and credible research.

You used an argument about cost, which was irrelevant to discussing throughput.

did i really. i surely remember using sram as reference for bandwidth. i guess you must have forgotten... 'cuz otherwise, you'd be lying. so, to refresh your memory:

Spoiler

are you pointing to the bandwidth numbers? you have never explained where they come from or how they were measured/calculated or how they work. comparing GB/s numbers are near-meaningless without other factors explained. for example (and example only, don't try pulling an argument out of this), sram can go into TB/s bandwitdh, yet we don't see sram displacing dram.

First you asked "what" and I replied. Then you asked "why" and I replied again. Now you are asking me the "how". You change the question with each new answer.

Your argument about SRAM is irrelevant to the facts discussed here: GDDR5 has replaced the slow DDR3 in any high-end graphics card; Sony uses GDDR5 instead of slow DDR3 as system memory for the PS4; I already explained "why".

and you did end up using it for arguing, lol.
in the same post, you offered the only statement as explanation to why gddr5 is beyond the slow ddr3 in pcs:

Spoiler

DDR3 only handles an input or output but not both on the same cycle. GDDR handles input and output on the same cycle.

this is prolly the 4/5th time i've asked what this means after you dodged it repeatedly.

Are not forum threads. I gave you again the link in the last post. Not only you ignored it once again, but you delete it.
....
It was a link about latencies of DDR3 and GDDR5.

if you're refering to the redgamingtech link, the guy is on a personal blog, comparing consoles excluding pcs. even he doesn't outright calculate gddr5 latencies, only posts results while admitting that latencies vary by make and models. amusingly, i haven't seen You calculate gddr5 latencies and post the results and calculations.

and another factor everyone seems to ignore is the prefetch buffer and it's affect on latencies. the rgt writer and other gddr5 proponents also ignore the prefetch buffer and full random access, 8Gbit ddr3 chip vs 2Gbit gddr5, cost - from what i've read so far on gddr5 and ddr3. there is a lot of unkown factors left unexplained.

Spoiler

yea i didn't even bother reading.
It seems the rule in this thread.

you said it!

Beautiful example of how you take a phrase original from you "yea i didn't even bother reading that link", cut it to "yea i didn't even bother reading", and then attribute it to me. Nice manipulation!

ahaha, i had a feeling it wouldn't go over the way i intended. feel free to ignore it.

 

[noob2222]

Spoiler

Take this highly efficient design and build me a 10 second car.

Someone is already working on that one.

http://www.carcraft.com/featuredvehicles/1404_big_block_chevy_smart_car_at_drag_strip/

so essentially the solution to the problem is to forgo efficiency for raw power.

In the same sense, taking a V8 and trying to catch the efficiency of a 3 cylinder is just as futile.

http://www.computingcompendium.com/p/arm-vs-intel-benchmarks.html

Trying to say Intel's Atom = haswell is the biggest flawed statement in this whole ARM vs x86 discussion. Atom is Intel's attempt to make an efficient x86 (aka V8) by neutering it.

ARM has a long way to go to catch that 16,774 score, at best its currently less than 3000. Even single core @ 1.3 ghz each, the infamous apple A7 is less than half of what haswell is and Atom being 1/3 as fast.

But hey, its faster than Atom, lets keep pretending that Atom = high end.

 

[juanrga]

This is not answering my question. What four cores? So far as I know A57 and Puma+ will be replaced by K12 and its "sister core" in 2016.

From how I read it they are planning to maintain a low end core and high end core in both x86/ARM. Similar to how there is Jaguar/Steamroller now. If they can merge those to a single design then great but a single core for all market segments is unlikely.

From the EETIMES article I mentioned before:

Keller suggested AMD will consolidate its work on separate Bulldozer and Jaguar x86 cores into one processor line in the future.

 

[juanrga]

http://www.computingcompendium.com/p/arm-vs-intel-benchmarks.html

Trying to say Intel's Atom = haswell is the biggest flawed statement in this whole ARM vs x86 discussion. Atom is Intel's attempt to make an efficient x86 (aka V8) by neutering it.

ARM has a long way to go to catch that 16,774 score, at best its currently less than 3000. Even single core @ 1.3 ghz each, the infamous apple A7 is less than half of what haswell is and Atom being 1/3 as fast.

But hey, its faster than Atom, lets keep pretending that Atom = high end.

I don't see anyone in this thread pretending that Atom is high end. I don't see anyone in this thread comparing K12 to Atom. I see comparisons of K12 to Piledriver and Haswell desktops.

You comparing the performance of a 1.3GHz phone chip (~2W) against a 3.5GHz desktop chip (80W) and claiming "ARM has a long way to go to catch" is so useless and wrong as comparing the performance of a 1.3GHz i3-4010Y (11.5W) against an 3.5GHz i7-4960x (130W) and claiming that "Haswell has a long way to go to catch Sandy Bridge".

You are confounding architectures and ISAs, with specific implementations at particular products.

A serious analysis of cyclone microarchitecture compared to Haswell was made by Anand

http://www.anandtech.com/show/7910/apples-cyclone-microarchitecture-detailed

An exhaustive analysis of armv8 ISA is given by Kanter. Keller has also commented which are the strong points of the ISA.

Finally, it is worth mention AMD Seattle is already a 25W chip. Evidently Seattle is not designed for phones but for microservers.

 

[Cazalan]

From the EETIMES article I mentioned before:

Keller suggested AMD will consolidate its work on separate Bulldozer and Jaguar x86 cores into one processor line in the future.

Kind of scarey if you think about it. ARM can't do it, Intel can't do it, but somehow AMD will do it. A one size fits all core. The holy grail of computing.

There sure were some harsh comments regarding Keller on there.

 

[gamerk316]

*sigh*

Its the BD Architecture Reveal all over again. Some people just never learn...

 

[Cazalan]

https://www.semiwiki.com/forum/content/3460-samsung-endorses-fd-soi.html

Samsung resurrecting 28nm FDSOI to deal with the high prices of 20nm bulk. Will GF follow suit?

 

[juanrga]

From the EETIMES article I mentioned before:

Keller suggested AMD will consolidate its work on separate Bulldozer and Jaguar x86 cores into one processor line in the future.

Kind of scarey if you think about it. ARM can't do it, Intel can't do it, but somehow AMD will do it. A one size fits all core. The holy grail of computing.

There sure were some harsh comments regarding Keller on there.

I will reserve my opinion about what I think about AMD strategy.

 

[anxiousinfusion]

AMD held APU14 today in China. According to this talk, AMD plans to not move away from desktop. Rather, talks even about growth. Who here was saying that AMD plans to abandon desktop?

 

[de5_Roy]

http://img1.mydrivers.com/img/20140516/s_a3847e06fde04444bb6d668ec312caef.jpg
AMD held APU14 today in China. According to this talk, AMD plans to not move away from desktop. Rather, talks even about growth. Who here was saying that AMD plans to abandon desktop?

there's a bit of a marketing hype there, but they're not wrong. in countries like china, south korea, malayasia and indonesia afaik (but can't confirm atm) pcs have a very good base. china only recently lifted the console ban while possessing a well-established competitive pc gaming culture. in u.s., pc markets are saturated, so amd and intel are looking elsewhere.

@cazalan: does this mean the previous alliance of st-ericsson is dissolved and samsung will use st's booster tech in their 28nm process? it seemed like samsung hit issues with their own 20nm process... and this makes me question glofo and samsung's 14nm bulk's viability.

edit: who else here thinks that keller is being lined up and rory reed's successor and koduri being the cto? i mean, if keller becomes the ceo, and him being a former insider, the cpu core/ghz wars may heat up again.... just saying...

 

[noob2222]

http://www.computingcompendium.com/p/arm-vs-intel-benchmarks.html

Trying to say Intel's Atom = haswell is the biggest flawed statement in this whole ARM vs x86 discussion. Atom is Intel's attempt to make an efficient x86 (aka V8) by neutering it.

ARM has a long way to go to catch that 16,774 score, at best its currently less than 3000. Even single core @ 1.3 ghz each, the infamous apple A7 is less than half of what haswell is and Atom being 1/3 as fast.

But hey, its faster than Atom, lets keep pretending that Atom = high end.

I don't see anyone in this thread pretending that Atom is high end. I don't see anyone in this thread comparing K12 to Atom. I see comparisons of K12 to Piledriver and Haswell desktops.

You comparing the performance of a 1.3GHz phone chip (~2W) against a 3.5GHz desktop chip (80W) and claiming "ARM has a long way to go to catch" is so useless and wrong as comparing the performance of a 1.3GHz i3-4010Y (11.5W) against an 3.5GHz i7-4960x (130W) and claiming that "Haswell has a long way to go to catch Sandy Bridge".

You are confounding architectures and ISAs, with specific implementations at particular products.

A serious analysis of cyclone microarchitecture compared to Haswell was made by Anand

http://www.anandtech.com/show/7910/apples-cyclone-microarchitecture-detailed

An exhaustive analysis of armv8 ISA is given by Kanter. Keller has also commented which are the strong points of the ISA.

Finally, it is worth mention AMD Seattle is already a 25W chip. Evidently Seattle is not designed for phones but for microservers.

ya ... ok ..

The performance of the new Opteron is about 3x the performance of that which replaces, which fall in the "between two and four times" range mentioned. You don't need to estimate the performance of each core, it is already in my original link: the new A57 ARM core falls at about 136% of AMDs X86 core, IPC wise

your the one comparing ARM to x86, claiming that ARM is far superior.

It also appears that AMD's cpus are about as fast as Intel's Atom ... since that 136% falls right where atom is

 

[noob2222]

From the EETIMES article I mentioned before:

Keller suggested AMD will consolidate its work on separate Bulldozer and Jaguar x86 cores into one processor line in the future.

Kind of scarey if you think about it. ARM can't do it, Intel can't do it, but somehow AMD will do it. A one size fits all core. The holy grail of computing.

There sure were some harsh comments regarding Keller on there.

pretty much sais it all right there.