AMD CPU speculation... and expert conjecture

[de5_Roy]

NOOOOOO.......
.....
.....
okay i expected this....
No GDDR5 Support on Kaveri
http://www.fudzilla.com/home/item/33018-no-gddr5-support-on-kaveri
i do hope that amd finds a way to keep that hungry hungry igpu well-fed. or i'll cry injustice.

google-translated from the original link:

It is reported that early stage of development (18-24 months ago) AMD there is indeed going to let Kaveri APU supports GDDR5 memory, is similar to an example is the Sony PS4 host, the latter used as a unified 8GB GDDR5 memory, after leaked the news show Kaveri also integrates 128bit GDDR5 128bit DDR3 memory controller and memory controller, AMD had expectations Kaveri and the corresponding product came out, GDDR5 memory prices dropped to an acceptable range that people can become mainstream, but now look at to AMD's idea is a little too far ahead.

Not to mention GDDR5, the current DDR3 memory prices have remained high, really do 4GB or 8GB GDDR5 memory, which will cost 80,150 dollars, so the price is not too mainstream.

Benefits of using GDDR5 can achieve higher bandwidth, dual-channel DDR3-2133 memory bandwidth of about 35GB / s, replace the words GDDR5 bandwidth of up to 70GB / s. Now, with the bursting of this vision, we still continue to use DDR3 memory bar, memory manufacturers have considered a good day, and now there is the trend of rising prices.

btw, if the kaveri apus follow 7xxx numbering then the igpus might have 9xxx e.g. 9660D, 9760D numbering.

Should I sell my graphics card and get an APU ?

depends on what you want to do with your pc. why not open a new thread in the systems section?

 

[anxiousinfusion]

NOOOOOO.......
.....
.....
okay i expected this....
No GDDR5 Support on Kaveri
http://www.fudzilla.com/home/item/33018-no-gddr5-support-on-kaveri
i do hope that amd finds a way to keep that hungry hungry igpu well-fed. or i'll cry injustice.

google-translated from the original link:

But isn't Kaveri's memory controller supposedly reworked? They're trying to do more with less.

 

[jdwii]

looks like xbone is going to lose

http://www.theguardian.com/technology/2013/oct/31/ps4-playstation-4-xbox-one-consoles-call-of-duty-ghostss

60 fps target, cut off graphics quality to achieve it, ps4 should look better, but still no PC quality.

Something that i've been saying for quite some time now

 

[de5_Roy]

But isn't Kaveri's memory controller supposedly reworked? They're trying to do more with less.

it's a new uarch, new features, so the imc should be improved as well. but improved how much and where? more of what with less of what? how will amd counteract the rising ddr3 prices without eradicating imc bottleneck? i wanna see amd provide answers to these in kaveri's performance.
imo gddr5 was a good alternative to high speed EDRAM acting as a big last level cache. i guess now it'll be up to the new buses.

 

[kansasboy001]

I agree, unless they put the pcie controller back on the north bridge for their enthusiast APUs. I think that AMD will be releasing some high end APU to replace the fx line. Or still possible, phenom III, they are reusing Athlon, so...

detaching pcie controller for 'enthusiast apus' may have adverse affect. due to market positioning, apus benefit from more integration.

mobile version of athlon core was used in the first mainstream apu - llano. those didn't clock higher than 3 ghz. bd uarch is designed for easier integration with other components.

amd is reusing the athlon brand only, like they reused the fx brand for bd cpus. the underlying 'cpu's are apus with igpus lasered off.

Thats what I meant, reuse the
Phenom branding for the enthusist cpus agian, to distance themselver from the failded FX line.

 

[de5_Roy]

Thats what I meant, reuse the
Phenom branding for the enthusist cpus agian, to distance themselver from the failded FX line.

ah. that could create customer confusion. some of the phenom ii cpus are still on sale. chances are people might associate phenom iii with am3 platform instead of am3+. it's too soon, barely 2 years since amd re-introduced fx branding.
it really depends on amd's marketing division. if they think phenom iii is a credible brand, they'll likely do it.

 

[tracker45]

So any news on when it is out ?

 

[8350rocks]

Spoiler

1. Your confusion about the word CISC was funny, but I was really referring to all the further nonsense that you posted against RISC/ARM.

2. The chip runs background processes to the game such as social stuff, recording, checking for online updates... it doesn't run the OS.

3. You said that that you "saw" PS4 benchmarks @2.6 Ghz, that you waited the "actual final specs will likely fall closer to the 2.4 mark, with possibly a turbo core feature to allow a higher clock under some instances that use fewer cores." Then you strongly misinterpreted the PS4 patents doc and said to everyone here that the CPU run at 2.75 GHz. Even hapidupi had to correct your nonsense. Hapidupi! LOL!

4. You claimed that bulk was impossible, that the people who said bulk was no working for AMD anymore, you insinuated that he didn't know the tech and pretended that the delay was caused by the migration to FD-SOI. When people complained. You did search and post a talk by Glofo and tried to convince us that FD-SOI was cheaper than bulk and AMD was going all SOI. "It is evident" you claimed...

5. Therefore you still insist on saying every novice that a SR FX 8-core CPU is coming for AM3+ and that AM3+ is a good platform for upgrade?

6. Easy. That is an ancient mobo with AM4 in the model code. Someone at the database did a mistake and indexed it as AM4 socket mobo. The rest is only in your imagination. It is not a new mobo with a new socket for a forthcoming 5GHz SR 8-core FX CPU...

I don't think you understand the points I was making about RISC/CISC, however, you don't have the architecture knowledge to understand it anyway. So, I am going to write this off as you simply don't understand. In gamer terms "Learn 2 Research".

My understanding is enough to detect your nonsense.

You did not read a word I said:

2. Sony has confirmed an additional chip on the PS4 to run background processes, and the design lead for the entire project stated the OS would be run primarily off the GPU leaving 8 cores available for developing games. Do you think Sony's project lead was lying? Why would he? That information will be verifiable easily...

Can you read it better now?
--------------------------------

This is the second time that I read your nonsense. The question is: do you read? I don't know what is more funny, that you write "GPU" or that you don't understand that the OS runs on the jaguar cores in the APU and that the secondary chip in the PS4 runs background tasks to the game, such as decoding, downloading, social stuff...

There were ES benchmarks run @ 2.6 GHz, Jaguar cores will actually clock quite a bit higher than that, you know that right? They were likely testing to see what the thermal limits and power consumption would be. I did say that I suspected it would end up closer to ~2.2-2.4 GHz, than the 1.6 GHz many were claiming. By the way, what is the clockspeed on the PS4 APU? Just curious...oh...that's right...it's 2.0 GHz. Who was right?

Hum, let us see you claimed the PS4 runs at 2.75GHz, but that was plain nonsense. Sony did claim 1.6Ghz; therefore, you are off by more than 1GHz or about 72% of error.

I never said bulk was impossible...I did say it was improbable, and that there would be massive clockspeed and thermal penalties to going bulk. I was not wrong on any of those counts. Your poor mastery of the English language is showing.

No. You said us it was SOI. Someone wrote:

I believe that the 28nm process used for SR and Kaveri will be bulk but I'm not sure.

And your answer was:

No, it will be a PD-SOI process...from GloFo.

One thing is that you are wrong, but another is that you pretend now to hide what you wrote during months.

I tell them a SR FX replacement is coming, I also tell them I have no idea what socket, or in that regard, much of any information about it at all.

No. You said us that Steamroller FX is AM3+ socket. Someone asked "8350 or wait for Steamroller?" and your answer was:

The good news is though, if you buy an AM3+ board and a new CPU, you won't need to buy a new board for steamroller, it will be AM3+.

I wonder how many newbies followed your advice, believing that you know what you are writing.

It could very easily be the case, *very* easily; however, Foxconn should not be advertising AM4 socket MB drivers when no such product clearly exists. I was a bit curious when I saw that...and pointed it out to you and a few others to confer. The consensus, that it was not what it looked like, was reached before you ever wrote back to me.

There is no AM4 sockect mobo on Foxconn. I already explained there is a mistake in their indexing of an ancient mobo, but please ignore facts and continue fantasizing. It is much more funny.

i will just leave it here :

http://amdfx.blogspot.com/2013/10/amd-kaveri-benchmarks-more-steamrolling.html

If legit, it confirms that SR is ~30% faster than BD. Precisely in my estimation of kaveri performance I assumed a 30% faster than BD (or ~20% faster than PD) to show that Kaveri CPU would be at the i5 level of performance.

The regression in FP could be related to SR using a simplified (streamlined) FPU. This is not a serious problem because SR comes in HSA APUs where the GPU will be used as a giant FPU.

They're probably tweaking it to prepare for 256 bit FMAC FPUs (AVX2 instructions) in excavator and running into some issues. That would be my guess...

At least now you make clear that you are guessing.... lol

256-bit FMAC units don't make any sense for me, but of course AMD can prove me wrong.

1.

•Thirdly, said Cerny, "The original AMD GCN architecture allowed for one source of graphics commands, and two sources of compute commands. For PS4, we’ve worked with AMD to increase the limit to 64 sources of compute commands -- the idea is if you have some asynchronous compute you want to perform, you put commands in one of these 64 queues, and then there are multiple levels of arbitration in the hardware to determine what runs, how it runs, and when it runs, alongside the graphics that's in the system."

"The reason so many sources of compute work are needed is that it isn’t just game systems that will be using compute -- middleware will have a need for compute as well. And the middleware requests for work on the GPU will need to be properly blended with game requests, and then finally properly prioritized relative to the graphics on a moment-by-moment basis."

http://www.gamasutra.com/view/feature/191007/inside_the_playstation_4_with_mark_.php?page=2

Maybe now you understand what GPU means...because the GPU is not an additional ARM core...

Unsurprisingly that is unrelated to what you wrote before about GPU and unrelated to your nonsense that the OS is run in a secondary chip and 8 jaguar cores are for games exclusively.

2.

RISC runs differently, and yes x86 CPUs have had underlying RISC architecture characteristics for a long time. Since the Pentium days really; however, as was pointed out by Palladin earlier, x86 allows MUCH larger instructions to be converted into bytecode to be run on the CPU itself.

The added complexity of x86 was my point. That is easily it's greatest strength while also being an inherent weakness.

ARM's strength and weakness both lie in it's simplicity. That's why it excels for low power devices and simple things like Micro Servers. It's also the primary reason it's not a terribly valid Desktop uarch option.

In order to make ARM a serious x86 competitor, you would have to add several layers of complexity. That complexity would drive up transistor counts and die complexity, requiring quad channel memory controllers, and many other things that draw more power. Once you've done that, you have an x86 competitor...that no longer consumes power like a mobile/low power solution. Because the added complexity draws more power, your consumption numbers spike upward dramatically.

x86 ISA has been dealing with this in it's architecture since x586 (K6-2 days). ARM has not had the benefit of time spent tweaking their architecture for such added complexities, and it takes a *LONG* time to get that stuff right. Which is part of what we see in the current AMD and Intel uarch's, Intel has a far more *REFINED* uarch, because it hasn't changed dramatically since P4 days. Where as AMD's uarch is only 2-3 years old at this point, and not nearly as well refined and tuned.

So, what you're talking about with ARM taking over DT, or even a large share of notebooks, etc., will not come about for quite some time.

The reasons for this are simple:

1.) Most of the consumer DT world runs on Windows, like it or not, M$ still has some clout. They don't want to redesign their OS entirely for ARM, and WART is a terrible execution.

2.) With no major OS player taking ARM seriously any time soon, hardware advances will come slowly because it's not in high demand. Only way ARM becomes a big player is if some large player in the PC world backs it and pushes hard. AMD making Micro Servers using ARM is not that push into the consumer sector you expect. It's a gimmick to say "see, we can do low power better than Intel", nothing more.

3.) Without a Major OS player backing ARM, the development of consumer software will be slow. Open source groups may do something, but how has that worked out for Linux so far? Outside of Android, it's still not terribly popular as an OS in the PC world considering roughly 3% of the world is likely running it on a DT PC. I think Linux should see more use than it does, but in the consumer space, M$ is still king.

4.) As ARM adds complexity, it will add power consumption, and the more you need to be able to ask the ISA to do, the more convoluted the hardware, middleware, software becomes to do those things. So, as the hardware adds complexity, the power draw increases. Once you get ARM running at a 50W+ TDP, x86 becomes a clear winner. Which is what would happen with a billion transistor ARM chip running at 4 GHz.

So, you may not like what I am saying, and you may disagree entirely. However, your declaration that ARM will rule desktop any time soon is a mere pipe dream...much like Acorn was back when it first started. That's why it's a non-profit organization that designs cores and licenses them, not an organization making chips and selling PCs.

That was my CISC vs. RISC rant...If you don't understand what I am talking about then that is your own fault for not knowing the difference between the 2.

That only covers a fraction of the nonsense that you have said about RISC and about ARM. Both that fraction and what you don't quote was corrected before.

3. PS4 runs @ ~2.0 GHz, which is 25% higher than the 1.6 GHz initially claimed. I also stated that there were ES's of the custom chip run @ 2.6 GHz, and there were.

(6) PS4 APU ES's have been benchmarked in the 2.4-2.6 GHz range already. I don't anticipate they will necessarily end up at 2.6 GHz, though I think 2.2-2.4 GHz with a turbo core for less threaded situations is entirely feasible.

http://www.tomshardware.com/forum/id-1721986/amd-cpu-performance-increase-nextgen-consoles-release.html

Sony said 1.6Ghz for the PS4. The 2.0Ghz is only a rumor. Moreover both are far from your previous pretension that it runs a 2.75Ghz. The funny part is that hapidupi was right and you wrong. LOL

4. Look at how old that thread is...seriously? We had less information then, than we do now...and we don't know much now. That was purely speculation based on expectations of current trends continuing from FX. You had some posts from 6 months ago saying things that we now know to be different from reality as well. I am not going to waste time digging them up...but they are there for sure.

LOL. A pair of posts ago you affirmed that you never said such thing. Now the excuse change to ups I said but that is old

5. Are you dense? Read what I wrote...it was determined that it was not, in fact, AM4 MB before you ever wrote me back. Blackkstar and I came to that conclusion before you ever replied to me.

Seeing as you pretended above to hide what you wrote, do you really pretend I to trust what you claim you were thinking one day? LOL

You clearly did not read what I wrote...how long are you going to continue this blatant ignorance?

Does anyone else want to explain to juanrga what I said?

What is your native language? It's clearly not English because you don't understand it very well. Where did I say 2.75 GHz in the post? You reference topics from May, and claim that it's relevant since the information we had at the time changed. You also said exactly what I did at that time, because that's what we knew then. Now, you would deny that you were prepared for AM3+ steamroller on FD-SOI????

You can try that tactic, you can pretend that you never said so or when caught with a quote, you can pretend that you said it in "May", but not latter. However. it is not a wise thing to do, because I can link to your posts.

Look at this one, you said that the PS4 is clocked at 2.75Ghz, someone corrects you and explain you that 2.75 is the freq. of the memory and that the CPU is clocked at 1.6GHz. Your typical ignorant answer was

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-138#11425669

Now pay attention to the next message after yours, where the same person corrects you again and explain you again that 2.75 is the memory freq. Your reply to him was again "NO".

You were plain wrong for days. I had to confirm here

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-141#11457221

that 2.75Ghz is the WCK of the memory, and that the freq. of the CPU is 1.6GHz. My post was followed by a series of funny replies.

Did you learn anything? Nope. Because days latter you wrote:

It has already been all but confirmed that the APUs will be running closer to ~2.0-2.2 GHz for the XBone and PS4. If you think that number won't increase over time, you're far more naïve than I thought. Just because the memory is clocked at one frequency has nothing to do with CPU clockspeed.

Microsoft has confirmed that Xbox1 is clocked at 1.75GHz. But you continue posting misleading info.

About SOI. So late as September I wrote that Kaveri/SR is bulk:

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-142#11461096

This generate a series of answers from several experts (including you) trying to convince me that Kaveri was being made on SOI process. Some quotes:

My thoughts run to doing it on SOI, as JK was a big proponent of the K7/K8 going to SOI originally.

Other sources would even say their 22nm FinFET on bulk was actually lackluster compared to the projections...though that would mean I am right about bulk vs. SOI and you'd be wrong about the delay from AMD for Kaveri. Thus, I don't anticipate you'll acknowledge that.

Until i asked you "If you have one source that says that Kaveri is being manufactured on SOI process, please share it with us." Your answers:

You can't follow the link to the SOI conference anymore, however, shortly after AMD announced going bulk over a year ago, they then fired a bunch of people (those that spoke of bulk included) and then they were talking to GF a month later about FD-SOI @ 28nm at the global SOI conference...you do the math.

You insisted with your it is SOI, SOI is ready, AMD is talking about SOI... My answer:

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-142#11467899

and during page 143 you continued claiming your confidence in that was SOI, because SOI was ready, trololo

You can now pretend that you never said that Steamroller is AM3+, that you never said that Kaveri/SR is SOI, that you never said that PS4 clock is 2.75Ghz, that you never said that Xbone ~2.0--2.2GHz, that you never said that PhysX is not in PS4, that you never said...

You can pretend that you said those in "May", but not latter. The problem is that your posts show otherwise.

I said, and I quote:

Per the patent applications, the maximum clockspeed listed for PS4 is 2.75 GHz

That statement is entirely accurate. I didn't say specifically that it would be clocked at those speeds...in fact, I said that I expected it would come in quite a bit closer to ~2.2-2.4 GHz multiple times in this thread and others.

The complain is not about that statement, but about others quoted above where you pretended that the CPU was clocked at 2.75GHz. You can continue negating it...

I did say Steamroller would be AM3+, at the time AMD had announced one more generation of processors for AM3+ and the Centurion line had not been announced...what was the logical conclusion to make?

Funny how you changed your words, from your initial I never said AM3+ to your I said AM3+. Any bet that if I had not liked to a post with your exact words you would be still negating what you said?

I did say Kaveri would likely be SOI...I never denied that. In fact, I am not entirely convinced to this point that it will still be on bulk as you claim.

Spoiler

http://www.advancedsubstratenews.com/2013/07/globalfoundries-on-cost-vs-performance-for-fd-soi-bulk-and-finfet/

There it shows GF is ready to produce 28nm SHP (FD-SOI), and that the costs are less than bulk process. This is because the design phase doesn't require all the masking layers that bulk does. AMD has said many things, and then done something different before...we will wait to see what the product actually is to confirm bulk or SOI. However, if it's bulk, I expect Kaveri will not be the knight in shining armor they need for APUs. Performance will end up similar to Richland because of the lower clocks and higher heat generation thresholds.

http://www.globalfoundries.com/dac2013/seminardetails.aspx

http://www.electronicsnews.com.au/news/globalfoundries-to-take-on-stmicroelectronics%E2%80%99-fd

http://www.globalfoundries.com/technology/tech_elements.aspx

http://www.globalfoundries.com/technology/advanced_tech.aspx

There is no reason to suggest that GF is not ramping FD-SOI. Especially after they announced at the consortium this year that production would ramp up Q4 2013. I think this is credible because they have been trying to ramp it up since end of 2012.

Charlie @ S|A dissected the PS4 APU, and said it would run right at ~2.0 GHz. He dissected the XBone APU and said it would run at ~1.8 GHz. So, I was right, it was closer to 2.2-2.4 than 1.6 GHz.

I fail to see what you're talking about where I was wrong...where do you have proof that AMD has said anything about bulk since 2012? You don't...considering the plethora of things that have changed since Q1 2012...I would say it's fair to think that bulk may entirely be wrong, just as much as it may entirely be right. We will see...

Effectively, you said it.

I didn't specifically say "The CPU will be 2.75 GHz"...I said, it could potentially be 2.75 GHz...whether that was in turbo or otherwise. Your comprehension is low on this, clearly. Additionally, I have not denied that was my position in light of the information available at the time.

I said it would be AM3+ back when the word from AMD was that another generation of processors was coming to AM3+ and it was supported through 2015. I was not the only one who said that either.

Since new information has come out, I am no longer sure that it would be AM3+, it could be an entirely new platform, for all we know, that is, as of yet, unannounced.

Now, could you stop trying to be a martyr and discuss the topic of the thread.

 

[8350rocks]

http://www.tomshardware.com/forum/id-1863987/official-amd-representatives.html

Maybe we can get word straight from the horse's mouth...?

Though I would expect some sort of PR spin to some degree...we will see what happens.

 

[noob2222]

If that's your point then you need a different benchmark. The quad-core FX is markedly slower than the octo-core FX, despite running at the same clockspeed. More cores does matter for AMD.

Emphasis mine. More cores matter more for AMD because, with only four cores, they are getting totally overloaded due to their poor IPC. As a result, if even ONE core gets loaded, performance tanks. Hence why AMD's performance tends to scale linearly as you OC. Intel quads, by contrast, don't have that problem (for the most part) at base clocks, hence why OCing tends to offer less performance increase (no bottleneck to remove).

Hence why games like BF4, where AMD gains significant performance as more cores are added, yet Intel performance is still sorted largely by IPC + Clock performance, and also why Intel tends to maintain its gaming lead.

See:
http://www.bf4blog.com/battlefield-4-retail-gpu-cpu-benchmarks/
http://pclab.pl/art55318-3.html

FX-8350 ~ i5 2500k in BF4, despite how well threaded it is. That's how superior IPC is better for overall performance in a nutshell. For Intel, more cores essentially just lowers overall core loading, since its high tier i5's aren't bottlenecked.

This was my exact argument against BD from the very beginning, and three years in, AMD still can't beat SB i7's. Even in games using over a dozen threads, like BF4, AMD still looses, or at best, matches Intel, head to head.

Lets examine this shall we?

Using PClabs since it favors intel anyway.

I7 4770k @3.5 ghz = 78.7 fps
I7 4770k@ 4.5 ghz = 91.4 fps ... bottleneck ... no

lets check clock speed scaling

4.5/3.5 = 1.285 or 28.5% increase in clock speed
91./78.7 = 1.166 or 16.6% faster.
Ratio of clock speed to performance = 58%

8350, @4.0 ghz = 61.8 fps
8350 @ 5 ghz = 73.6
same math
skipping to ratio = 76%

sure, AMD scales faster because its slightly more bottlenecked, but that doesn't mean that Intel isn't bottlenecked

It could also mean that if none of them are close to being bottlenecked, that AMD will still scale better (8 semi-real cores vs 4+4 ht cores)

Ahhh, but you forget to realize that once you escape the effects of a CPU bottleneck, performance increases tend to be very minimal. You could make the argument that Intel's scaling is worse because the 4770k eventually escapes the effects of being a bottleneck, where the 8350 is still not at that point.

Isn't it interesting there that the 8350 @ 4.48 outperforms the 3960x @ 5.10

Some tasks naturally scale, and for those tasks, the FX-8350 would likely outperform most Intel CPU's. I've never argued against that particular point. [Go back to the BD thread; I predicted it would stink at gaming, but be a benchmark monster.]

For games, where most parallel processing is done on the GPU (Rendering + Physics), you are left doing mainly serial workloads, which favors a CPU with higher clock and higher IPC. THATS the point I'm arguing: For gaming, AMD does not have a particularly strong architecture. But for a subset of tasks, AMD's arch is better.

But for a gaming rig? I'd recommend Intel every time. The only attractive CPU AMD has for gaming is the FX-8350 [I do not think the 6300 is going to age well at all.]

its not exactly weak either.

more often than not this is the case. AMD and Intel aren't that far apart. you have to dig for cases to support "dual core is good enough"

Intel's strength is their memory controller. When does main memory speed/latency affect gpu performance? meaning does memory affect when the gpu is doing very little rendering or when the gpu is being stressed?

Put another way. How does intel and AMD compare when you faceplant a wall so that the cpu isn't doing anything but feeding the gpu? How does each cpu compare when full on destruction is happening?

Wich one is more important, faceplanting a wall to boost average fps, or minimum fps?

 

[de5_Roy]

http://www.tomshardware.com/forum/id-1863987/official-amd-representatives.html

Maybe we can get word straight from the horse's mouth...?

Though I would expect some sort of PR spin to some degree...we will see what happens.

none from the cpu/apu dept. only the discreet gfx and marketing guys.
i wish i knew what they were trying to promote!!!

 

[esrever]

http://www.tomshardware.com/forum/id-1863987/official-amd-representatives.html

Maybe we can get word straight from the horse's mouth...?

Though I would expect some sort of PR spin to some degree...we will see what happens.

none from the cpu/apu dept. only the discreet gfx and marketing guys.
i wish i knew what they were trying to promote!!!

their new GPU they just launched maybe

 

[juanrga]

Spoiler

Spoiler

1. Your confusion about the word CISC was funny, but I was really referring to all the further nonsense that you posted against RISC/ARM.

2. The chip runs background processes to the game such as social stuff, recording, checking for online updates... it doesn't run the OS.

3. You said that that you "saw" PS4 benchmarks @2.6 Ghz, that you waited the "actual final specs will likely fall closer to the 2.4 mark, with possibly a turbo core feature to allow a higher clock under some instances that use fewer cores." Then you strongly misinterpreted the PS4 patents doc and said to everyone here that the CPU run at 2.75 GHz. Even hapidupi had to correct your nonsense. Hapidupi! LOL!

4. You claimed that bulk was impossible, that the people who said bulk was no working for AMD anymore, you insinuated that he didn't know the tech and pretended that the delay was caused by the migration to FD-SOI. When people complained. You did search and post a talk by Glofo and tried to convince us that FD-SOI was cheaper than bulk and AMD was going all SOI. "It is evident" you claimed...

5. Therefore you still insist on saying every novice that a SR FX 8-core CPU is coming for AM3+ and that AM3+ is a good platform for upgrade?

6. Easy. That is an ancient mobo with AM4 in the model code. Someone at the database did a mistake and indexed it as AM4 socket mobo. The rest is only in your imagination. It is not a new mobo with a new socket for a forthcoming 5GHz SR 8-core FX CPU...

I don't think you understand the points I was making about RISC/CISC, however, you don't have the architecture knowledge to understand it anyway. So, I am going to write this off as you simply don't understand. In gamer terms "Learn 2 Research".

My understanding is enough to detect your nonsense.

You did not read a word I said:

2. Sony has confirmed an additional chip on the PS4 to run background processes, and the design lead for the entire project stated the OS would be run primarily off the GPU leaving 8 cores available for developing games. Do you think Sony's project lead was lying? Why would he? That information will be verifiable easily...

Can you read it better now?
--------------------------------

This is the second time that I read your nonsense. The question is: do you read? I don't know what is more funny, that you write "GPU" or that you don't understand that the OS runs on the jaguar cores in the APU and that the secondary chip in the PS4 runs background tasks to the game, such as decoding, downloading, social stuff...

There were ES benchmarks run @ 2.6 GHz, Jaguar cores will actually clock quite a bit higher than that, you know that right? They were likely testing to see what the thermal limits and power consumption would be. I did say that I suspected it would end up closer to ~2.2-2.4 GHz, than the 1.6 GHz many were claiming. By the way, what is the clockspeed on the PS4 APU? Just curious...oh...that's right...it's 2.0 GHz. Who was right?

Hum, let us see you claimed the PS4 runs at 2.75GHz, but that was plain nonsense. Sony did claim 1.6Ghz; therefore, you are off by more than 1GHz or about 72% of error.

I never said bulk was impossible...I did say it was improbable, and that there would be massive clockspeed and thermal penalties to going bulk. I was not wrong on any of those counts. Your poor mastery of the English language is showing.

No. You said us it was SOI. Someone wrote:

I believe that the 28nm process used for SR and Kaveri will be bulk but I'm not sure.

And your answer was:

No, it will be a PD-SOI process...from GloFo.

One thing is that you are wrong, but another is that you pretend now to hide what you wrote during months.

I tell them a SR FX replacement is coming, I also tell them I have no idea what socket, or in that regard, much of any information about it at all.

No. You said us that Steamroller FX is AM3+ socket. Someone asked "8350 or wait for Steamroller?" and your answer was:

The good news is though, if you buy an AM3+ board and a new CPU, you won't need to buy a new board for steamroller, it will be AM3+.

I wonder how many newbies followed your advice, believing that you know what you are writing.

It could very easily be the case, *very* easily; however, Foxconn should not be advertising AM4 socket MB drivers when no such product clearly exists. I was a bit curious when I saw that...and pointed it out to you and a few others to confer. The consensus, that it was not what it looked like, was reached before you ever wrote back to me.

There is no AM4 sockect mobo on Foxconn. I already explained there is a mistake in their indexing of an ancient mobo, but please ignore facts and continue fantasizing. It is much more funny.

i will just leave it here :

http://amdfx.blogspot.com/2013/10/amd-kaveri-benchmarks-more-steamrolling.html

If legit, it confirms that SR is ~30% faster than BD. Precisely in my estimation of kaveri performance I assumed a 30% faster than BD (or ~20% faster than PD) to show that Kaveri CPU would be at the i5 level of performance.

The regression in FP could be related to SR using a simplified (streamlined) FPU. This is not a serious problem because SR comes in HSA APUs where the GPU will be used as a giant FPU.

They're probably tweaking it to prepare for 256 bit FMAC FPUs (AVX2 instructions) in excavator and running into some issues. That would be my guess...

At least now you make clear that you are guessing.... lol

256-bit FMAC units don't make any sense for me, but of course AMD can prove me wrong.

1.

•Thirdly, said Cerny, "The original AMD GCN architecture allowed for one source of graphics commands, and two sources of compute commands. For PS4, we’ve worked with AMD to increase the limit to 64 sources of compute commands -- the idea is if you have some asynchronous compute you want to perform, you put commands in one of these 64 queues, and then there are multiple levels of arbitration in the hardware to determine what runs, how it runs, and when it runs, alongside the graphics that's in the system."

"The reason so many sources of compute work are needed is that it isn’t just game systems that will be using compute -- middleware will have a need for compute as well. And the middleware requests for work on the GPU will need to be properly blended with game requests, and then finally properly prioritized relative to the graphics on a moment-by-moment basis."

http://www.gamasutra.com/view/feature/191007/inside_the_playstation_4_with_mark_.php?page=2

Maybe now you understand what GPU means...because the GPU is not an additional ARM core...

Unsurprisingly that is unrelated to what you wrote before about GPU and unrelated to your nonsense that the OS is run in a secondary chip and 8 jaguar cores are for games exclusively.

2.

RISC runs differently, and yes x86 CPUs have had underlying RISC architecture characteristics for a long time. Since the Pentium days really; however, as was pointed out by Palladin earlier, x86 allows MUCH larger instructions to be converted into bytecode to be run on the CPU itself.

The added complexity of x86 was my point. That is easily it's greatest strength while also being an inherent weakness.

ARM's strength and weakness both lie in it's simplicity. That's why it excels for low power devices and simple things like Micro Servers. It's also the primary reason it's not a terribly valid Desktop uarch option.

In order to make ARM a serious x86 competitor, you would have to add several layers of complexity. That complexity would drive up transistor counts and die complexity, requiring quad channel memory controllers, and many other things that draw more power. Once you've done that, you have an x86 competitor...that no longer consumes power like a mobile/low power solution. Because the added complexity draws more power, your consumption numbers spike upward dramatically.

x86 ISA has been dealing with this in it's architecture since x586 (K6-2 days). ARM has not had the benefit of time spent tweaking their architecture for such added complexities, and it takes a *LONG* time to get that stuff right. Which is part of what we see in the current AMD and Intel uarch's, Intel has a far more *REFINED* uarch, because it hasn't changed dramatically since P4 days. Where as AMD's uarch is only 2-3 years old at this point, and not nearly as well refined and tuned.

So, what you're talking about with ARM taking over DT, or even a large share of notebooks, etc., will not come about for quite some time.

The reasons for this are simple:

1.) Most of the consumer DT world runs on Windows, like it or not, M$ still has some clout. They don't want to redesign their OS entirely for ARM, and WART is a terrible execution.

2.) With no major OS player taking ARM seriously any time soon, hardware advances will come slowly because it's not in high demand. Only way ARM becomes a big player is if some large player in the PC world backs it and pushes hard. AMD making Micro Servers using ARM is not that push into the consumer sector you expect. It's a gimmick to say "see, we can do low power better than Intel", nothing more.

3.) Without a Major OS player backing ARM, the development of consumer software will be slow. Open source groups may do something, but how has that worked out for Linux so far? Outside of Android, it's still not terribly popular as an OS in the PC world considering roughly 3% of the world is likely running it on a DT PC. I think Linux should see more use than it does, but in the consumer space, M$ is still king.

4.) As ARM adds complexity, it will add power consumption, and the more you need to be able to ask the ISA to do, the more convoluted the hardware, middleware, software becomes to do those things. So, as the hardware adds complexity, the power draw increases. Once you get ARM running at a 50W+ TDP, x86 becomes a clear winner. Which is what would happen with a billion transistor ARM chip running at 4 GHz.

So, you may not like what I am saying, and you may disagree entirely. However, your declaration that ARM will rule desktop any time soon is a mere pipe dream...much like Acorn was back when it first started. That's why it's a non-profit organization that designs cores and licenses them, not an organization making chips and selling PCs.

That was my CISC vs. RISC rant...If you don't understand what I am talking about then that is your own fault for not knowing the difference between the 2.

That only covers a fraction of the nonsense that you have said about RISC and about ARM. Both that fraction and what you don't quote was corrected before.

3. PS4 runs @ ~2.0 GHz, which is 25% higher than the 1.6 GHz initially claimed. I also stated that there were ES's of the custom chip run @ 2.6 GHz, and there were.

(6) PS4 APU ES's have been benchmarked in the 2.4-2.6 GHz range already. I don't anticipate they will necessarily end up at 2.6 GHz, though I think 2.2-2.4 GHz with a turbo core for less threaded situations is entirely feasible.

http://www.tomshardware.com/forum/id-1721986/amd-cpu-performance-increase-nextgen-consoles-release.html

Sony said 1.6Ghz for the PS4. The 2.0Ghz is only a rumor. Moreover both are far from your previous pretension that it runs a 2.75Ghz. The funny part is that hapidupi was right and you wrong. LOL

4. Look at how old that thread is...seriously? We had less information then, than we do now...and we don't know much now. That was purely speculation based on expectations of current trends continuing from FX. You had some posts from 6 months ago saying things that we now know to be different from reality as well. I am not going to waste time digging them up...but they are there for sure.

LOL. A pair of posts ago you affirmed that you never said such thing. Now the excuse change to ups I said but that is old

5. Are you dense? Read what I wrote...it was determined that it was not, in fact, AM4 MB before you ever wrote me back. Blackkstar and I came to that conclusion before you ever replied to me.

Seeing as you pretended above to hide what you wrote, do you really pretend I to trust what you claim you were thinking one day? LOL

You clearly did not read what I wrote...how long are you going to continue this blatant ignorance?

Does anyone else want to explain to juanrga what I said?

What is your native language? It's clearly not English because you don't understand it very well. Where did I say 2.75 GHz in the post? You reference topics from May, and claim that it's relevant since the information we had at the time changed. You also said exactly what I did at that time, because that's what we knew then. Now, you would deny that you were prepared for AM3+ steamroller on FD-SOI????

You can try that tactic, you can pretend that you never said so or when caught with a quote, you can pretend that you said it in "May", but not latter. However. it is not a wise thing to do, because I can link to your posts.

Look at this one, you said that the PS4 is clocked at 2.75Ghz, someone corrects you and explain you that 2.75 is the freq. of the memory and that the CPU is clocked at 1.6GHz. Your typical ignorant answer was

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-138#11425669

Now pay attention to the next message after yours, where the same person corrects you again and explain you again that 2.75 is the memory freq. Your reply to him was again "NO".

You were plain wrong for days. I had to confirm here

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-141#11457221

that 2.75Ghz is the WCK of the memory, and that the freq. of the CPU is 1.6GHz. My post was followed by a series of funny replies.

Did you learn anything? Nope. Because days latter you wrote:

It has already been all but confirmed that the APUs will be running closer to ~2.0-2.2 GHz for the XBone and PS4. If you think that number won't increase over time, you're far more naïve than I thought. Just because the memory is clocked at one frequency has nothing to do with CPU clockspeed.

Microsoft has confirmed that Xbox1 is clocked at 1.75GHz. But you continue posting misleading info.

About SOI. So late as September I wrote that Kaveri/SR is bulk:

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-142#11461096

This generate a series of answers from several experts (including you) trying to convince me that Kaveri was being made on SOI process. Some quotes:

My thoughts run to doing it on SOI, as JK was a big proponent of the K7/K8 going to SOI originally.

Other sources would even say their 22nm FinFET on bulk was actually lackluster compared to the projections...though that would mean I am right about bulk vs. SOI and you'd be wrong about the delay from AMD for Kaveri. Thus, I don't anticipate you'll acknowledge that.

Until i asked you "If you have one source that says that Kaveri is being manufactured on SOI process, please share it with us." Your answers:

You can't follow the link to the SOI conference anymore, however, shortly after AMD announced going bulk over a year ago, they then fired a bunch of people (those that spoke of bulk included) and then they were talking to GF a month later about FD-SOI @ 28nm at the global SOI conference...you do the math.

You insisted with your it is SOI, SOI is ready, AMD is talking about SOI... My answer:

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-142#11467899

and during page 143 you continued claiming your confidence in that was SOI, because SOI was ready, trololo

You can now pretend that you never said that Steamroller is AM3+, that you never said that Kaveri/SR is SOI, that you never said that PS4 clock is 2.75Ghz, that you never said that Xbone ~2.0--2.2GHz, that you never said that PhysX is not in PS4, that you never said...

You can pretend that you said those in "May", but not latter. The problem is that your posts show otherwise.

I said, and I quote:

Per the patent applications, the maximum clockspeed listed for PS4 is 2.75 GHz

That statement is entirely accurate. I didn't say specifically that it would be clocked at those speeds...in fact, I said that I expected it would come in quite a bit closer to ~2.2-2.4 GHz multiple times in this thread and others.

The complain is not about that statement, but about others quoted above where you pretended that the CPU was clocked at 2.75GHz. You can continue negating it...

I did say Steamroller would be AM3+, at the time AMD had announced one more generation of processors for AM3+ and the Centurion line had not been announced...what was the logical conclusion to make?

Funny how you changed your words, from your initial I never said AM3+ to your I said AM3+. Any bet that if I had not liked to a post with your exact words you would be still negating what you said?

I did say Kaveri would likely be SOI...I never denied that. In fact, I am not entirely convinced to this point that it will still be on bulk as you claim.

Spoiler

http://www.advancedsubstratenews.com/2013/07/globalfoundries-on-cost-vs-performance-for-fd-soi-bulk-and-finfet/

There it shows GF is ready to produce 28nm SHP (FD-SOI), and that the costs are less than bulk process. This is because the design phase doesn't require all the masking layers that bulk does. AMD has said many things, and then done something different before...we will wait to see what the product actually is to confirm bulk or SOI. However, if it's bulk, I expect Kaveri will not be the knight in shining armor they need for APUs. Performance will end up similar to Richland because of the lower clocks and higher heat generation thresholds.

http://www.globalfoundries.com/dac2013/seminardetails.aspx

http://www.electronicsnews.com.au/news/globalfoundries-to-take-on-stmicroelectronics%E2%80%99-fd

http://www.globalfoundries.com/technology/tech_elements.aspx

http://www.globalfoundries.com/technology/advanced_tech.aspx

There is no reason to suggest that GF is not ramping FD-SOI. Especially after they announced at the consortium this year that production would ramp up Q4 2013. I think this is credible because they have been trying to ramp it up since end of 2012.

Charlie @ S|A dissected the PS4 APU, and said it would run right at ~2.0 GHz. He dissected the XBone APU and said it would run at ~1.8 GHz. So, I was right, it was closer to 2.2-2.4 than 1.6 GHz.

I fail to see what you're talking about where I was wrong...where do you have proof that AMD has said anything about bulk since 2012? You don't...considering the plethora of things that have changed since Q1 2012...I would say it's fair to think that bulk may entirely be wrong, just as much as it may entirely be right. We will see...

Effectively, you said it.

I didn't specifically say "The CPU will be 2.75 GHz"...I said, it could potentially be 2.75 GHz.

You started saying was 2.2Ghz, then claimed 2.75Ghz by quoting the FCC doc. Someone explained you that 2.75GHz is the memory, the CPU is clocked at 1.6GHz and your answer was "NO", pretending that 2.75GHz was the freq. of the CPU. Instead learning you pretended to scorn him. You can now try to hide that episode, but it is not going to happen.

I said it would be AM3+ back when the word from AMD was that another generation of processors was coming to AM3+ and it was supported through 2015. I was not the only one who said that either.

Since new information has come out, I am no longer sure that it would be AM3+, it could be an entirely new platform, for all we know, that is, as of yet, unannounced.

Again (this is the second time that I repeat this) the problem is not if you were in mistake or not. The problem is that you pretended that you never said so.

Now, I will stop trying to be a martyr and discuss the topic of the thread.

^^ Fixed it for you.

Now you could maybe comment in the slide #13 of the recent talk given by AMD.

I interpret it as FX-4000 series is abandoned and AMD considers it a good idea to pair high-end R9 280X with an APU. In that slide, AMD claims that an APU+dGPU provides a gaming experience similar to using a 8350 with the same dGPU.

No what I agree/disagree, merely emphasizing that slide #13 seems to delineate AMD plans for the desktop.

 

[tracker45]

Is Steamroller even necessary ?

The 8350 is still good.

 

[Sigmanick]

AMD ASK ME ANYTHING LINK

8350, Thank you very much for posting this here. As a new member to THG, I have not seen or participated in an AMA thread and did not know they existed. I would have missed this opportunity. Again, Thank you.

 

[blackkstar]

I agree, unless they put the pcie controller back on the north bridge for their enthusiast APUs. I think that AMD will be releasing some high end APU to replace the fx line. Or still possible, phenom III, they are reusing Athlon, so...

detaching pcie controller for 'enthusiast apus' may have adverse affect. due to market positioning, apus benefit from more integration.

mobile version of athlon core was used in the first mainstream apu - llano. those didn't clock higher than 3 ghz. bd uarch is designed for easier integration with other components.

amd is reusing the athlon brand only, like they reused the fx brand for bd cpus. the underlying 'cpu's are apus with igpus lasered off.

Thats what I meant, reuse the
Phenom branding for the enthusist cpus agian, to distance themselver from the failded FX line.

Remember that Baeca leak? It was marketed as a Phenom IV.

Perhaps we will see something new in Phenom 4 branding instead?

AMD has been very quiet about their next generation HEDT platform. They were also very quiet with 290x release.

Imagine AMD drops a new chipset for HEDT that has some sort of technology that lets 290x crossfire really haul.

I really wouldn't put it past them at this point to drop something huge out of thin air. AMD did such a good job keeping Hawaii under wraps that Nvidia didn't even have an answer at all and got completely blindsided by it.

The Phenom 4/Baeca rumor was shot down really quickly because the only 25nm process that exists is only really used for memory and it's a collaboration between Micron and Intel. However perhaps there's at least a little bit of truth behind this.

I am just kind of thinking of how companies intentionally "leak" things like this to gauge public perceptions. Perhaps this was some sort of fake leak to see how people would respond to Phenom IV branding on a new socket with a lot of cores? Maybe it's all just completely fake? I don't know, a little digging on the internet leaves me reading a lot of people saying that it has to be fake because there's no way a 6ghz chip is 75w and about the 25nm, but this thing could be fake yet still hinting at things.

Examples of this happening in the past:

Rumor: Nvidia working on "Titan LE"
Later: Nvidia releases GTX 780 as a lightweight Gk110 product

Hate to bring up something that's extremely sketchy but it might at least provide some clues for us.

 

[Embra]

I still think AMD needs to come out with high end "FXish" CPUs. I know they market for them is quite small, but the publicity of having such a chip trickles down there entire line. Intel has benefited greatly from such press. It works with GPUs s well.

 

[8350rocks]

Spoiler

1. Your confusion about the word CISC was funny, but I was really referring to all the further nonsense that you posted against RISC/ARM.

2. The chip runs background processes to the game such as social stuff, recording, checking for online updates... it doesn't run the OS.

3. You said that that you "saw" PS4 benchmarks @2.6 Ghz, that you waited the "actual final specs will likely fall closer to the 2.4 mark, with possibly a turbo core feature to allow a higher clock under some instances that use fewer cores." Then you strongly misinterpreted the PS4 patents doc and said to everyone here that the CPU run at 2.75 GHz. Even hapidupi had to correct your nonsense. Hapidupi! LOL!

4. You claimed that bulk was impossible, that the people who said bulk was no working for AMD anymore, you insinuated that he didn't know the tech and pretended that the delay was caused by the migration to FD-SOI. When people complained. You did search and post a talk by Glofo and tried to convince us that FD-SOI was cheaper than bulk and AMD was going all SOI. "It is evident" you claimed...

5. Therefore you still insist on saying every novice that a SR FX 8-core CPU is coming for AM3+ and that AM3+ is a good platform for upgrade?

6. Easy. That is an ancient mobo with AM4 in the model code. Someone at the database did a mistake and indexed it as AM4 socket mobo. The rest is only in your imagination. It is not a new mobo with a new socket for a forthcoming 5GHz SR 8-core FX CPU...

I don't think you understand the points I was making about RISC/CISC, however, you don't have the architecture knowledge to understand it anyway. So, I am going to write this off as you simply don't understand. In gamer terms "Learn 2 Research".

My understanding is enough to detect your nonsense.

You did not read a word I said:

2. Sony has confirmed an additional chip on the PS4 to run background processes, and the design lead for the entire project stated the OS would be run primarily off the GPU leaving 8 cores available for developing games. Do you think Sony's project lead was lying? Why would he? That information will be verifiable easily...

Can you read it better now?
--------------------------------

This is the second time that I read your nonsense. The question is: do you read? I don't know what is more funny, that you write "GPU" or that you don't understand that the OS runs on the jaguar cores in the APU and that the secondary chip in the PS4 runs background tasks to the game, such as decoding, downloading, social stuff...

There were ES benchmarks run @ 2.6 GHz, Jaguar cores will actually clock quite a bit higher than that, you know that right? They were likely testing to see what the thermal limits and power consumption would be. I did say that I suspected it would end up closer to ~2.2-2.4 GHz, than the 1.6 GHz many were claiming. By the way, what is the clockspeed on the PS4 APU? Just curious...oh...that's right...it's 2.0 GHz. Who was right?

Hum, let us see you claimed the PS4 runs at 2.75GHz, but that was plain nonsense. Sony did claim 1.6Ghz; therefore, you are off by more than 1GHz or about 72% of error.

I never said bulk was impossible...I did say it was improbable, and that there would be massive clockspeed and thermal penalties to going bulk. I was not wrong on any of those counts. Your poor mastery of the English language is showing.

No. You said us it was SOI. Someone wrote:

I believe that the 28nm process used for SR and Kaveri will be bulk but I'm not sure.

And your answer was:

No, it will be a PD-SOI process...from GloFo.

One thing is that you are wrong, but another is that you pretend now to hide what you wrote during months.

I tell them a SR FX replacement is coming, I also tell them I have no idea what socket, or in that regard, much of any information about it at all.

No. You said us that Steamroller FX is AM3+ socket. Someone asked "8350 or wait for Steamroller?" and your answer was:

The good news is though, if you buy an AM3+ board and a new CPU, you won't need to buy a new board for steamroller, it will be AM3+.

I wonder how many newbies followed your advice, believing that you know what you are writing.

It could very easily be the case, *very* easily; however, Foxconn should not be advertising AM4 socket MB drivers when no such product clearly exists. I was a bit curious when I saw that...and pointed it out to you and a few others to confer. The consensus, that it was not what it looked like, was reached before you ever wrote back to me.

There is no AM4 sockect mobo on Foxconn. I already explained there is a mistake in their indexing of an ancient mobo, but please ignore facts and continue fantasizing. It is much more funny.

i will just leave it here :

http://amdfx.blogspot.com/2013/10/amd-kaveri-benchmarks-more-steamrolling.html

If legit, it confirms that SR is ~30% faster than BD. Precisely in my estimation of kaveri performance I assumed a 30% faster than BD (or ~20% faster than PD) to show that Kaveri CPU would be at the i5 level of performance.

The regression in FP could be related to SR using a simplified (streamlined) FPU. This is not a serious problem because SR comes in HSA APUs where the GPU will be used as a giant FPU.

They're probably tweaking it to prepare for 256 bit FMAC FPUs (AVX2 instructions) in excavator and running into some issues. That would be my guess...

At least now you make clear that you are guessing.... lol

256-bit FMAC units don't make any sense for me, but of course AMD can prove me wrong.

1.

•Thirdly, said Cerny, "The original AMD GCN architecture allowed for one source of graphics commands, and two sources of compute commands. For PS4, we’ve worked with AMD to increase the limit to 64 sources of compute commands -- the idea is if you have some asynchronous compute you want to perform, you put commands in one of these 64 queues, and then there are multiple levels of arbitration in the hardware to determine what runs, how it runs, and when it runs, alongside the graphics that's in the system."

"The reason so many sources of compute work are needed is that it isn’t just game systems that will be using compute -- middleware will have a need for compute as well. And the middleware requests for work on the GPU will need to be properly blended with game requests, and then finally properly prioritized relative to the graphics on a moment-by-moment basis."

http://www.gamasutra.com/view/feature/191007/inside_the_playstation_4_with_mark_.php?page=2

Maybe now you understand what GPU means...because the GPU is not an additional ARM core...

Unsurprisingly that is unrelated to what you wrote before about GPU and unrelated to your nonsense that the OS is run in a secondary chip and 8 jaguar cores are for games exclusively.

2.

RISC runs differently, and yes x86 CPUs have had underlying RISC architecture characteristics for a long time. Since the Pentium days really; however, as was pointed out by Palladin earlier, x86 allows MUCH larger instructions to be converted into bytecode to be run on the CPU itself.

The added complexity of x86 was my point. That is easily it's greatest strength while also being an inherent weakness.

ARM's strength and weakness both lie in it's simplicity. That's why it excels for low power devices and simple things like Micro Servers. It's also the primary reason it's not a terribly valid Desktop uarch option.

In order to make ARM a serious x86 competitor, you would have to add several layers of complexity. That complexity would drive up transistor counts and die complexity, requiring quad channel memory controllers, and many other things that draw more power. Once you've done that, you have an x86 competitor...that no longer consumes power like a mobile/low power solution. Because the added complexity draws more power, your consumption numbers spike upward dramatically.

x86 ISA has been dealing with this in it's architecture since x586 (K6-2 days). ARM has not had the benefit of time spent tweaking their architecture for such added complexities, and it takes a *LONG* time to get that stuff right. Which is part of what we see in the current AMD and Intel uarch's, Intel has a far more *REFINED* uarch, because it hasn't changed dramatically since P4 days. Where as AMD's uarch is only 2-3 years old at this point, and not nearly as well refined and tuned.

So, what you're talking about with ARM taking over DT, or even a large share of notebooks, etc., will not come about for quite some time.

The reasons for this are simple:

1.) Most of the consumer DT world runs on Windows, like it or not, M$ still has some clout. They don't want to redesign their OS entirely for ARM, and WART is a terrible execution.

2.) With no major OS player taking ARM seriously any time soon, hardware advances will come slowly because it's not in high demand. Only way ARM becomes a big player is if some large player in the PC world backs it and pushes hard. AMD making Micro Servers using ARM is not that push into the consumer sector you expect. It's a gimmick to say "see, we can do low power better than Intel", nothing more.

3.) Without a Major OS player backing ARM, the development of consumer software will be slow. Open source groups may do something, but how has that worked out for Linux so far? Outside of Android, it's still not terribly popular as an OS in the PC world considering roughly 3% of the world is likely running it on a DT PC. I think Linux should see more use than it does, but in the consumer space, M$ is still king.

4.) As ARM adds complexity, it will add power consumption, and the more you need to be able to ask the ISA to do, the more convoluted the hardware, middleware, software becomes to do those things. So, as the hardware adds complexity, the power draw increases. Once you get ARM running at a 50W+ TDP, x86 becomes a clear winner. Which is what would happen with a billion transistor ARM chip running at 4 GHz.

So, you may not like what I am saying, and you may disagree entirely. However, your declaration that ARM will rule desktop any time soon is a mere pipe dream...much like Acorn was back when it first started. That's why it's a non-profit organization that designs cores and licenses them, not an organization making chips and selling PCs.

That was my CISC vs. RISC rant...If you don't understand what I am talking about then that is your own fault for not knowing the difference between the 2.

That only covers a fraction of the nonsense that you have said about RISC and about ARM. Both that fraction and what you don't quote was corrected before.

3. PS4 runs @ ~2.0 GHz, which is 25% higher than the 1.6 GHz initially claimed. I also stated that there were ES's of the custom chip run @ 2.6 GHz, and there were.

(6) PS4 APU ES's have been benchmarked in the 2.4-2.6 GHz range already. I don't anticipate they will necessarily end up at 2.6 GHz, though I think 2.2-2.4 GHz with a turbo core for less threaded situations is entirely feasible.

http://www.tomshardware.com/forum/id-1721986/amd-cpu-performance-increase-nextgen-consoles-release.html

Sony said 1.6Ghz for the PS4. The 2.0Ghz is only a rumor. Moreover both are far from your previous pretension that it runs a 2.75Ghz. The funny part is that hapidupi was right and you wrong. LOL

4. Look at how old that thread is...seriously? We had less information then, than we do now...and we don't know much now. That was purely speculation based on expectations of current trends continuing from FX. You had some posts from 6 months ago saying things that we now know to be different from reality as well. I am not going to waste time digging them up...but they are there for sure.

LOL. A pair of posts ago you affirmed that you never said such thing. Now the excuse change to ups I said but that is old

5. Are you dense? Read what I wrote...it was determined that it was not, in fact, AM4 MB before you ever wrote me back. Blackkstar and I came to that conclusion before you ever replied to me.

Seeing as you pretended above to hide what you wrote, do you really pretend I to trust what you claim you were thinking one day? LOL

You clearly did not read what I wrote...how long are you going to continue this blatant ignorance?

Does anyone else want to explain to juanrga what I said?

What is your native language? It's clearly not English because you don't understand it very well. Where did I say 2.75 GHz in the post? You reference topics from May, and claim that it's relevant since the information we had at the time changed. You also said exactly what I did at that time, because that's what we knew then. Now, you would deny that you were prepared for AM3+ steamroller on FD-SOI????

You can try that tactic, you can pretend that you never said so or when caught with a quote, you can pretend that you said it in "May", but not latter. However. it is not a wise thing to do, because I can link to your posts.

Look at this one, you said that the PS4 is clocked at 2.75Ghz, someone corrects you and explain you that 2.75 is the freq. of the memory and that the CPU is clocked at 1.6GHz. Your typical ignorant answer was

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-138#11425669

Now pay attention to the next message after yours, where the same person corrects you again and explain you again that 2.75 is the memory freq. Your reply to him was again "NO".

You were plain wrong for days. I had to confirm here

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-141#11457221

that 2.75Ghz is the WCK of the memory, and that the freq. of the CPU is 1.6GHz. My post was followed by a series of funny replies.

Did you learn anything? Nope. Because days latter you wrote:

It has already been all but confirmed that the APUs will be running closer to ~2.0-2.2 GHz for the XBone and PS4. If you think that number won't increase over time, you're far more naïve than I thought. Just because the memory is clocked at one frequency has nothing to do with CPU clockspeed.

Microsoft has confirmed that Xbox1 is clocked at 1.75GHz. But you continue posting misleading info.

About SOI. So late as September I wrote that Kaveri/SR is bulk:

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-142#11461096

This generate a series of answers from several experts (including you) trying to convince me that Kaveri was being made on SOI process. Some quotes:

My thoughts run to doing it on SOI, as JK was a big proponent of the K7/K8 going to SOI originally.

Other sources would even say their 22nm FinFET on bulk was actually lackluster compared to the projections...though that would mean I am right about bulk vs. SOI and you'd be wrong about the delay from AMD for Kaveri. Thus, I don't anticipate you'll acknowledge that.

Until i asked you "If you have one source that says that Kaveri is being manufactured on SOI process, please share it with us." Your answers:

You can't follow the link to the SOI conference anymore, however, shortly after AMD announced going bulk over a year ago, they then fired a bunch of people (those that spoke of bulk included) and then they were talking to GF a month later about FD-SOI @ 28nm at the global SOI conference...you do the math.

You insisted with your it is SOI, SOI is ready, AMD is talking about SOI... My answer:

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-142#11467899

and during page 143 you continued claiming your confidence in that was SOI, because SOI was ready, trololo

You can now pretend that you never said that Steamroller is AM3+, that you never said that Kaveri/SR is SOI, that you never said that PS4 clock is 2.75Ghz, that you never said that Xbone ~2.0--2.2GHz, that you never said that PhysX is not in PS4, that you never said...

You can pretend that you said those in "May", but not latter. The problem is that your posts show otherwise.

I said, and I quote:

Per the patent applications, the maximum clockspeed listed for PS4 is 2.75 GHz

That statement is entirely accurate. I didn't say specifically that it would be clocked at those speeds...in fact, I said that I expected it would come in quite a bit closer to ~2.2-2.4 GHz multiple times in this thread and others.

The complain is not about that statement, but about others quoted above where you pretended that the CPU was clocked at 2.75GHz. You can continue negating it...

I did say Steamroller would be AM3+, at the time AMD had announced one more generation of processors for AM3+ and the Centurion line had not been announced...what was the logical conclusion to make?

Funny how you changed your words, from your initial I never said AM3+ to your I said AM3+. Any bet that if I had not liked to a post with your exact words you would be still negating what you said?

I did say Kaveri would likely be SOI...I never denied that. In fact, I am not entirely convinced to this point that it will still be on bulk as you claim.

Spoiler

http://www.advancedsubstratenews.com/2013/07/globalfoundries-on-cost-vs-performance-for-fd-soi-bulk-and-finfet/

There it shows GF is ready to produce 28nm SHP (FD-SOI), and that the costs are less than bulk process. This is because the design phase doesn't require all the masking layers that bulk does. AMD has said many things, and then done something different before...we will wait to see what the product actually is to confirm bulk or SOI. However, if it's bulk, I expect Kaveri will not be the knight in shining armor they need for APUs. Performance will end up similar to Richland because of the lower clocks and higher heat generation thresholds.

http://www.globalfoundries.com/dac2013/seminardetails.aspx

http://www.electronicsnews.com.au/news/globalfoundries-to-take-on-stmicroelectronics%E2%80%99-fd

http://www.globalfoundries.com/technology/tech_elements.aspx

http://www.globalfoundries.com/technology/advanced_tech.aspx

There is no reason to suggest that GF is not ramping FD-SOI. Especially after they announced at the consortium this year that production would ramp up Q4 2013. I think this is credible because they have been trying to ramp it up since end of 2012.

Charlie @ S|A dissected the PS4 APU, and said it would run right at ~2.0 GHz. He dissected the XBone APU and said it would run at ~1.8 GHz. So, I was right, it was closer to 2.2-2.4 than 1.6 GHz.

I fail to see what you're talking about where I was wrong...where do you have proof that AMD has said anything about bulk since 2012? You don't...considering the plethora of things that have changed since Q1 2012...I would say it's fair to think that bulk may entirely be wrong, just as much as it may entirely be right. We will see...

Effectively, you said it.

I didn't specifically say "The CPU will be 2.75 GHz"...I said, it could potentially be 2.75 GHz.

You started saying was 2.2Ghz, then claimed 2.75Ghz by quoting the FCC doc. Someone explained you that 2.75GHz is the memory, the CPU is clocked at 1.6GHz and your answer was "NO", pretending that 2.75GHz was the freq. of the CPU. Instead learning you pretended to scorn him. You can now try to hide that episode, but it is not going to happen.

I said it would be AM3+ back when the word from AMD was that another generation of processors was coming to AM3+ and it was supported through 2015. I was not the only one who said that either.

Since new information has come out, I am no longer sure that it would be AM3+, it could be an entirely new platform, for all we know, that is, as of yet, unannounced.

Again (this is the second time that I repeat this) the problem is not if you were in mistake or not. The problem is that you pretended that you never said so.

Now, I will stop trying to be a martyr and discuss the topic of the thread.

^^ Fixed it for you.

Now you could maybe comment in the slide #13 of the recent talk given by AMD.

I interpret it as FX-4000 series is abandoned and AMD considers it a good idea to pair high-end R9 280X with an APU. In that slide, AMD claims that an APU+dGPU provides a gaming experience similar to using a 8350 with the same dGPU.

No what I agree/disagree, merely emphasizing that slide #13 seems to delineate AMD plans for the desktop.

Well, considering the PS4 APU is actually 2.0+ GHz and not 1.6...you'd need to increase the numbers by roughly 25% then double them.

Looks like you're still not there...

Stop trying already...you're using bad/old information to attempt to create a false equivalency.

http://www.tomshardware.co.uk/forum/352312-28-steamroller-speculation-expert-conjecture/page-138

That is the only post I have seen where I discuss the clockspeed of the APU with any certainty...

Where did I say that it was going to be 2.75 GHz??? You say that I said that...however, there is no proof.

Stop the shenanigans. If you want to post that I said something...from this point forward, I will not acknowledge you unless you quote my original post and link me to it.

You are butchering nearly everything I say, and what you're not butchering, you're misreading, or completely not understanding. This is like talking to hajidurp....except you're on about ARM the way he was about Intel.

Are you 2 at all related? Just curious....

 

[blackkstar]

I still think AMD needs to come out with high end "FXish" CPUs. I know they market for them is quite small, but the publicity of having such a chip trickles down there entire line. Intel has benefited greatly from such press. It works with GPUs s well.

http://ir.amd.com/phoenix.zhtml?c=74093&p=irol-newsArticle&ID=1865755&highlight=

Read the quarterly report.

Basically, notebook mobile was the worst performing part of AMD last quarter.

Desktop class CPUs did second best and GPU growth is out of control (in a good way).

I don't see AMD not using their GPU wins to levitate their HEDT CPU platform.

People are swarming to AMD GPUs. Growth is near 100% and this time last year AMD was doing very well with 7970.

I think we are about to see things surface at APU13 which show that AMD is focusing on APUs because they have to as there is no HPP node for AMD to launch a HEDT class CPU.

Meaning I am feeling that AMD Is so quiet in regards to HEDT (FX successor or new FX chips) because they simply don't have a place to make a chip that runs at 4ghz and is around 300mm^2 right now. AMD also stated that they're not doing anything besides bulk at 28nm and that they're not commenting on what they will do after 28nm. Meaning that AMD is either going to wait for 14nm FD-SOI or release new chips on 32nm.

The best thing AMD could do right now is refresh the Vishera platform. Update PD to what they did with Richland where it overclocks to ridiculous levels, add a new chipset, and add some sort of thing that works with the new platform chipset that unlocks potential of R9 series GPUs.

That would have been the smartest thing to do, at least. It's probably too late for that though.

But keep your eyes open at APU 13 for lots of talk about 28nm bulk and i am pretty much expecting there to be absolutely no sign of anything SOI or any sort of part that would really love SOI.

 

[kansasboy001]

I agree, AMD has been very sneaky with releases lately, with their additional funds form MS and Sony, they have the funding to put in to a big release. Amd also hired two of the best CPU Architects in the market, (I forgot the names). If I remember right they did something similar when The released Athlon 64 and beat the crap out of Intel's CPUs.

 

[lilcinw]

http://www.tomshardware.com/forum/id-1863987/official-amd-representatives.html

Maybe we can get word straight from the horse's mouth...?

Though I would expect some sort of PR spin to some degree...we will see what happens.

none from the cpu/apu dept. only the discreet gfx and marketing guys.
i wish i knew what they were trying to promote!!!

I sent a question to jpishgar asking why the thread was in the CPU forum when there didn't seem to be any CPU personnel on the panel. The thread has been moved to the graphics forum.

I wouldn't expect them to discuss FX much.

Edit: I just noticed the titled changed to Official AMD Radeon Representatives

 

[8350rocks]

I agree, AMD has been very sneaky with releases lately, with their additional funds form MS and Sony, they have the funding to put in to a big release. Amd also hired two of the best CPU Architects in the market, (I forgot the names). If I remember right they did something similar when The released Athlon 64 and beat the crap out of Intel's CPUs.

They have hired a plethora of talent:

John Gustafson - Father of "Gustafson's Law", former Intel employee, GPU compute expert. (Chief Product Architect GPUs at AMD)

Raja Koduri - Ex Apple employee for GPUs, ex ex AMD employee for GPUs/Drivers. (VP of Graphical Solutions at AMD)

Jim Keller - Ex Apple employee for A5-A7, ex ex AMD employee, architect of K7/K8 and HTX. (Chief Product Architect CPUs at AMD)

With those guys working on this, I don't foresee an issue with them putting out phenomenal product designs. What I do see, right now, is a fab issue. 28nm FD-SOI @ GF via STMicro's designs is only just beginning to ramp. Yields won't be good for a while yet, until it's mature, and AMD has had to eat costly low yields @ GF before. I think if they do a 28nm FX type CPU, then it will not be until the process is mature enough to support a ~280-300mm^2 die size with moderately good yields. That may be 6-12 months out, depending on how the yields for the process go with the IBM processors that GF is producing. If the yields work out well for the POWER8 stuff, then AMD may dive into the pool.

This is particularly true of the position they took using bulk. It is far easier to port from bulk to SOI than it is to go the opposite direction. This is because the process for bulk requires more masking layers to perfect, and SOI removes a great many of those, so it is essentially "easier" to produce. The designs don't translate across 100%, though the portability is probably close to the 80% mark.

I also feel that they are holding something back, hopefully I can get a good, solid answer from AMD in the AMA. Though, I somehow think that it will be sidestepped by NDA's and other methods. At least if we prod them enough about what the FX successor will be, they will see the massive interest in the product.

 

[The Q6660 Inside]

I agree, AMD has been very sneaky with releases lately, with their additional funds form MS and Sony, they have the funding to put in to a big release. Amd also hired two of the best CPU Architects in the market, (I forgot the names). If I remember right they did something similar when The released Athlon 64 and beat the crap out of Intel's CPUs.

They have hired a plethora of talent:

John Gustafson - Father of "Gustafson's Law", former Intel employee, GPU compute expert. (Chief Product Architect GPUs at AMD)

Raja Koduri - Ex Apple employee for GPUs, ex ex AMD employee for GPUs/Drivers. (VP of Graphical Solutions at AMD)

Jim Keller - Ex Apple employee for A5-A7, ex ex AMD employee, architect of K7/K8 and HTX. (Chief Product Architect CPUs at AMD)

With those guys working on this, I don't foresee an issue with them putting out phenomenal product designs. What I do see, right now, is a fab issue. 28nm FD-SOI @ GF via STMicro's designs is only just beginning to ramp. Yields won't be good for a while yet, until it's mature, and AMD has had to eat costly low yields @ GF before. I think if they do a 28nm FX type CPU, then it will not be until the process is mature enough to support a ~280-300mm^2 die size with moderately good yields. That may be 6-12 months out, depending on how the yields for the process go with the IBM processors that GF is producing. If the yields work out well for the POWER8 stuff, then AMD may dive into the pool.

This is particularly true of the position they took using bulk. It is far easier to port from bulk to SOI than it is to go the opposite direction. This is because the process for bulk requires more masking layers to perfect, and SOI removes a great many of those, so it is essentially "easier" to produce. The designs don't translate across 100%, though the portability is probably close to the 80% mark.

I also feel that they are holding something back, hopefully I can get a good, solid answer from AMD in the AMA. Though, I somehow think that it will be sidestepped by NDA's and other methods. At least if we prod them enough about what the FX successor will be, they will see the massive interest in the product.

Perhaps they will launch new 3M/6C HEDT products on ye olde 32nm FD-SOI and then when 28nm FD-SOI matures, they could launch Excavator or SR HEDT on it...

 

[8350rocks]

http://www.tomshardware.com/forum/id-1863987/official-amd-representatives.html

Maybe we can get word straight from the horse's mouth...?

Though I would expect some sort of PR spin to some degree...we will see what happens.

none from the cpu/apu dept. only the discreet gfx and marketing guys.
i wish i knew what they were trying to promote!!!

I sent a question to jpishgar asking why the thread was in the CPU forum when there didn't seem to be any CPU personnel on the panel. The thread has been moved to the graphics forum.

I wouldn't expect them to discuss FX much.

Edit: I just noticed the titled changed to Official AMD Radeon Representatives

I still intend to prod them about FX replacements, either way. Hopefully they can get back to me about it, or give at least a simple answer regarding the possibilities.

 

[kansasboy001]

I agree, AMD has been very sneaky with releases lately, with their additional funds form MS and Sony, they have the funding to put in to a big release. Amd also hired two of the best CPU Architects in the market, (I forgot the names). If I remember right they did something similar when The released Athlon 64 and beat the crap out of Intel's CPUs.

They have hired a plethora of talent:

John Gustafson - Father of "Gustafson's Law", former Intel employee, GPU compute expert. (Chief Product Architect GPUs at AMD)

Raja Koduri - Ex Apple employee for GPUs, ex ex AMD employee for GPUs/Drivers. (VP of Graphical Solutions at AMD)

Jim Keller - Ex Apple employee for A5-A7, ex ex AMD employee, architect of K7/K8 and HTX. (Chief Product Architect CPUs at AMD)

With those guys working on this, I don't foresee an issue with them putting out phenomenal product designs. What I do see, right now, is a fab issue. 28nm FD-SOI @ GF via STMicro's designs is only just beginning to ramp. Yields won't be good for a while yet, until it's mature, and AMD has had to eat costly low yields @ GF before. I think if they do a 28nm FX type CPU, then it will not be until the process is mature enough to support a ~280-300mm^2 die size with moderately good yields. That may be 6-12 months out, depending on how the yields for the process go with the IBM processors that GF is producing. If the yields work out well for the POWER8 stuff, then AMD may dive into the pool.

This is particularly true of the position they took using bulk. It is far easier to port from bulk to SOI than it is to go the opposite direction. This is because the process for bulk requires more masking layers to perfect, and SOI removes a great many of those, so it is essentially "easier" to produce. The designs don't translate across 100%, though the portability is probably close to the 80% mark.

I also feel that they are holding something back, hopefully I can get a good, solid answer from AMD in the AMA. Though, I somehow think that it will be sidestepped by NDA's and other methods. At least if we prod them enough about what the FX successor will be, they will see the massive interest in the product.

AMD also has TSMC to fall back on