AMD CPU speculation... and expert conjecture

palladin9479 · Oct 6, 2013

juanrga · Oct 6, 2013

noob2222 :

It is very ironic coming from someone, you, who is the only here who has posted real insults in your previous posts.

noob2222 :

You because that quote doesn't say what you pretend. A pair of posts ago, I discussed:

- the possibility of AMD releasing refreshes of Piledriver for FX Vishera and/or Centurion

- the possibility of future high-performance APUs (6C/8C)

- how x86 would be maintained for legacy workloads

- how high-end ARM will be at the performance level of fastest x86 desktop chips

- Etcetera.

noob2222 :

What is stupid for you is a wise decision for others. This is what was said in the link that I provided in my above post (link you don't mention):

AMD has historically had a hard time getting its low-power chips into tablets. The old Brazos-based Hondo APU barely barely made it into any slates, and the newer Temash and Kabini chips don't have many design wins. The lowest-power Temash variant still has a relatively high 3.9W TDP, though; AMD can surely do better with ARM-based chips. Radeon graphics technology should make those offerings unique, and it could help AMD gain a foothold in the tablet market.

For me AMD is doing the correct moves and taking the correct decisions. I would be very worried if AMD followed your suggestions! :ange:

noob2222 :

Apple, AMD, Nvidia, Qualcomm, Dell, HP,... think otherwise.

juanrga · Oct 6, 2013

de5_Roy :

The benchmarks that I provided were obtained with an OC 9370. No real 9590 was benchmarked... Another example of how some people here don't read I suppose.

de5_Roy · Oct 6, 2013

juanrga :

i was discussing pricing of the cpu, not the bf4 gaming benchmark. your reading ability needs improvement.

The Q6660 Inside · Oct 6, 2013

griptwister :

Everything you need, screw the 1440p, get real!

PCPartPicker part list / Price breakdown by merchant / Benchmarks

CPU: AMD FX-8320 3.5GHz 8-Core Processor ($153.61 @ Amazon)
Motherboard: Asus Crosshair V Formula-Z ATX AM3+ Motherboard ($214.99 @ Amazon)
Case: Corsair 750D ATX Full Tower Case ($175.98 @ Newegg)
Other: AX750 ($199.99)
Other: AX750 Sleeving ($40.00)
Total: $784.57
(Prices include shipping, taxes, and discounts when available.)
(Generated by PCPartPicker 2013-10-06 10:59 EDT-0400)

The Q6660 Inside · Oct 6, 2013

de5_Roy :

Well, you could just get an RX240 Kit, slam it in your case and voila, 5GHz is in easy reach.

@palladin, tis cable management time! http://www.corsair.com/en/professional-series-gold-ax1200-individually-sleeved-modular-cables-blue.html

My build and also potentially terrible cable management (Yes that is the top of an original CM690 screwed on to the 500R top 120mm mounts and hidden under the mesh so I could mount my triple aging 80mm Violet LED fan setup http://imgur.com/xd4VKoX)
Build: http://imgur.com/a/3bhFO

anxiousinfusion · Oct 6, 2013

palladin9479 :

I love your Richland machine. The desktop of the future indeed.

jdwii · Oct 6, 2013

Not a mod but the continues name calling here is getting a bit too much and the so called reading disability that people feel entitled to point out is just childish, For example when i was correcting hafier(which i'm sure he didn't even read what i posted) i did not put him down in any way all i did was give him some info and told him he needs learn it before commenting on some things. Now i understand how that can get a bit annoying if someone keeps going on and on with nonsense but that is their right their own post make them lose credibility name calling makes you lose credibility and is one of the worst fallacies in a Fallacious Argument.

8350rocks · Oct 6, 2013

juanrga :

I have seen the Centurion benchmarks.

Also, I am not trolling, nor am I incapable of reading.

The issue lies in the fact that your statements have no bearing or foundation in real world scenarios.

The only 64 bit ARM chip is the A7 in the iphone 5S, and it isn't much of a competitor at all for any desktop chip.

Trying to extrapolate any real world useable metrics from it in terms of desktop performance is a fool's errand.

Also, AMD can say what they want, ARM will be good at mobile and micro servers. I won't dispute that, because it's true; however, I interpret what they're saying as though comparing x86 to ARM in workloads favorable to ARM.

Will ARM win? In mobile? They already have essentially... In micro servers? Maybe, the jury is out still. In DT? Not likely, see the chromebooks out running an OS that doesn't even remotely resemble a DT OS, and doesn't have half the capability of something like Ubuntu or Windows, or even OSX.

My point here is, the only ARM penetration into the world of PCs, at a consumer level, is a couple of tiny laptops that can run a browser and word processor. Ask them to do anything more than that, and they can't. Additionally...I haven't heard anyone refer to them as "blazing fast" or "snappy" at all!

blackkstar · Oct 6, 2013

tourist :

The latest Longsoon MIPS CPU (from China) could emulate something like 90% of x86 instructions for something like a 5% increase in die size.

So yeah, you could make an ARM chip that could handle x86 instructions. Whether AMD will do it or not or if they even can is another thing.

Although I do recall Cyrix making x86 CPUs without a license because they reverse engineered x86 chips. I think that's kind of what Longsoon is but I'm not sure if AMD would be in legal trouble over this or not.

8350rocks · Oct 6, 2013

tourist :

Then you would lose the performance advantage that juanrga wants to sing about all the time. So why do it? If ARM runs x86 code at 90% efficiency, and x86 runs it at 100% efficiency (as a comparative baseline...not in theoretical terms). Then why make ARM run x86 when it's inferior to x86 at doing such things?

EDIT: Crusoe was an interesting diversion into making a GPGPU (essentially) compute x86 instructions, and is ultimately the basis for GPU compute philosophies at this time. However, Crusoe was not very good at serial applications, only things that could run massively parallel. It ultimately lost out to x86 because the niche was too small in consumer space, and they were not capable of promoting the technology for the industrial use it could have served. They knew what they had, but did not know how to put it into the marketplace in such a manner as to leverage it's strengths, like fluid dynamics and protein synthesis, etc.

noob2222 · Oct 6, 2013

@juan

answer one question with proof honestly if thats even possible. What is the TDP of an ARM A7, A9, or A15 cpu?

I just want to see how horrible x86 power consumption is compared to ARM in actual terms.

Try if you can, to put some proof into your theories other than marketing gimmicks or your own personal opinion.

palladin9479 · Oct 6, 2013

The arm discussion has been nice and all, especially as AMD is experimenting with it, but lets not fall into name calling and insult slinging. Since this is a speculation thread, speculating in theoretical performance is great, lets just not attack each other over it. Sometimes the best answer is to agree to disagree.

GOM3RPLY3R · Oct 6, 2013

griptwister :

I know how everyone here Loooooooves Single-Threaded benchmarks. lol. I did super pi (really to do a quick temp and clock read). But here's the results:

*As a side note, My GPU was able to idle at 25ºC Today! :3*

juanrga · Oct 7, 2013

de5_Roy :

I am not discussing benchmarks either, I am saying you that there was not 9590 and that he was not mentioning any 9590. Your attempt to correct him on pricing was wrong, as is your reply to me now

juanrga · Oct 7, 2013

8350rocks :

Companies like AMD, Nvidia, Apple, Qualcomm... think otherwise.

8350rocks :

And once again you repeat the same nonsensical argument. The Apple A7 is a 2W dual-core phone chip, it is _NOT_ aimed to desktops neither to supercomputers or servers. The chips used for ARM desktops, ARM servers and ARM supercomputers are _NOT_ 2W dual-core phone chips.

This has been explained to you many times before. I feel as if I am talking to hapidupi clone: correcting the _same_ mistakes again and again and again.

I am going to bet that I will read the same crappy argument against ARM using a 2W dual-core phone chip.

gamerk316 · Oct 7, 2013

blackkstar :

juanrga :

Woah, 5ghz OCed $200 FX 8350 beating stock $1000+ Intel EE edition.

In before "MUH POEWR CONSUMPOOOTTTIONZZZ!!"

4ghz to 5ghz is a 25% increase in clock speed. If AMD released FX SR with 20% IPC increase over PD, STOCK FX 4m/8c part would do that to Intel EE in BF4.

You guys still think AMD is just gonna walk away from this market when they can be this competitive in a next gen game? AMD's high end gaming platform is literally right here and they have the stuff to make it happen right now, even without SR.

This is why I say AMD isn't walking away from high end x86 CPUs. They have the software under their control and they're getting software developers to write for their hardware and not Intel's. It'd be completely idiotic to squander this. BF4 is just the tip of the iceberg of what we're going to see in the future.

Oh, but no, AMD just released an expensive, high end CPU as a last hurrah and not to establish some sort of halo product.

May I ask how well an OCd 2500k would do instead? Intel can OC too you know...

gamerk316 · Oct 7, 2013

EDIT: Crusoe was an interesting diversion into making a GPGPU (essentially) compute x86 instructions, and is ultimately the basis for GPU compute philosophies at this time. However, Crusoe was not very good at serial applications, only things that could run massively parallel. It ultimately lost out to x86 because the niche was too small in consumer space, and they were not capable of promoting the technology for the industrial use it could have served. They knew what they had, but did not know how to put it into the marketplace in such a manner as to leverage it's strengths, like fluid dynamics and protein synthesis, etc.

Emphasis mine, and this is exactly what you would expect. A single GPU execution unit it weak; the strength of a GPU si when you feed it parallel tasks that scale to hundreds of execution units.

juanrga · Oct 7, 2013

noob2222 :

If you were really interested in this topic, you would take a look to all the TDP and power consumption figures mentioned before, instead asking me for the same data.

You would see power consumption and efficiency for older A9 and A15, you could see how dual-core phone-TDP-rated Apple A7 SoC competes with quad-core tablet-TDP-rated x86 Silvermont or with a quad-core laptop-TDP-rated x86 Jaguar.

You would see how the estimated TDP for Seattle was quoted and how its efficiency is superior to best-class Xeons.

You would see the plans to build a ARM supercomputer about 1000 more powerful than current x86 top supercomputers, but using less power than today.

And so on, links included.

But asking as if this data was not given before... and then pretending that all this is the "personal opinion" of someone clearly shown what is the real goal. :lol:

juanrga · Oct 7, 2013

gamerk316 :

It is not reduced to a question of if an execution unit is "weak" or not. If weak/strong was everything involved here, then a massively number of strong CPUs would execute parallel tasks better than GPGPUs. It is not the case.

de5_Roy · Oct 7, 2013

juanrga :

i... have no idea how you made this about you and how you speak for someone else...

i know there was no fx9590 in the bf4 benchmark, but the specs (from the chart) in the post and the chart directly corresponded to an fx9590 with incorrectly mentioned price. there is no other amd fx cpu with stock 5.0 ghz clockrate (max. turbo, still counts

). it's price still counts even if the cpu performance was simulated.

you are free to think whatever you want... prices and specs don't lie.

i'll just point out the fact that you're arguing semantics (again) and wrong one (again) at that.

even if you claim that an overclocked 5.0 ghz fx8350 was mentioned - you have to be lucky to stably overclock (as every cpu overclocks differently) an fx8350 even if the cpu cost $200. accompanying costs would still go against it's favor.

juanrga · Oct 7, 2013

gamerk316 :

blackkstar :

May I ask how well an OCd 2500k would do instead? Intel can OC too you know...

It could be behind the 8350 stock, but then the 8350 can be OC as well increasing again the gap.

8350rocks · Oct 7, 2013

juanrga :

His point was: your entire argument about this is based on marketing slides and theoretical numbers. There is no real world "proof" that this will even prove accurate. Much less is there any system in the world running ARM CPUs for DT grade OS's.

In other words, you're predicting a meteor will hit the earth in 2047 and claiming you know this because the stars told you. Even though there is currently no evidence to support the trajectory you're giving as legitimate.

See the parallel with the analogy?

juanrga · Oct 7, 2013

8350rocks :

The data about the A9, A15, Apple SoC... is _not_ based in marketing slides _neither_ in theoretical numbers.

The estimation of the TDP of Seattle was partially confirmed by Feldman (AMD) during an interview. AMD statements about the expected performance of A57-based products match very well with info and benchmarks that I have from ARM.

The predicted efficiency of the ARM supercomputer is based in theoretical numbers extrapolated from the _prototype_. That they are theoretical numbers is _evident_ because the supercomputer has not been still built :lol: However, those numbers are made by experts in the topic (e.g. the same supercomputer center that helped IBM to develop their supercomputer CPUs) and have been reviewed for _funding_ by a scientific European program panel of experts. Even if everyone involved is wrong by 100%, the final efficiency will be much higher than current x86 supercomputers.

As I already mentioned you at least _twice before_, there are ARM CPUs running "DT grade OS". There is no essential difficulty with coding desktop OSes and desktop applications for ARM.

You don't have any argument, benchmark, theoretical argument, link or anything rocks; you only have (i) elementary misconceptions, (ii) trivial statements, and (iii) continuous negation of the data and links given to _you_ again and again.

What I find very interesting is that _you_ very often buy rumours about Steamroller, Excavator, Radeon, MANTLE... and _you_ are often over-enthusiast about those rumours, even when AMD doesn't confirm them. However, now, suddenly, anything said by AMD about ARM either is wrong or has to be put under 90-days quarantine. This double standard is very funny.

Now feel free to reply with some rant or insult :bounce:

de5_Roy · Oct 7, 2013

juanrga :

this is the type of provocation we all (me included) should avoid.
argue the idea with other ideas instead.

AMD CPU speculation... and expert conjecture

Splendid

Distinguished

Distinguished

Splendid

Honorable

Honorable

Distinguished

Splendid

Distinguished

Honorable

Distinguished

Distinguished

Splendid

Honorable

Distinguished

Distinguished

Glorious

Glorious

Distinguished

Distinguished

Splendid

Distinguished

Distinguished

Distinguished

Splendid

Share this page