AMD Piledriver rumours ... and expert conjecture

[Reynod]

We have had several requests for a sticky on AMD's yet to be released Piledriver architecture ... so here it is.

I want to make a few things clear though.

Post a question relevant to the topic, or information about the topic, or it will be deleted.

Post any negative personal comments about another user ... and they will be deleted.

Post flame baiting comments about the blue, red and green team and they will be deleted.

Enjoy ...

 

[viridiancrystal]

They have to be getting something, otherwise they wouldn't agree to it. I would say maybe 15% of the profit goes to them. That is just a shot in that dark though. They are making something, that is for sure.

 

[fazers_on_stun]

Until fairly recently I allowed ads to run in my browser. One day I was annoyed by video ads that would pop up when I rolled my mouse over text. I could not move the video window or stop it until a certain amount of time has passed. From that point on I have been running Adblock Plus and see no ads from any website. I view it as self defense.

Well I have been getting by with IE's popup blocker, but lately it is pretty ineffective - guess the website devs have got it figured out, at least IE9's version 😛.

Will give adblock a try, thanks.

 

[fazers_on_stun]

Still the next year for AMD is crucial, they have to hit their marks and compete with Intel, not themselves in the CPU market be it DT or server. As well they need to push hard into the UMD space as Intel is already there and will only get better as time goes. Plus with Haswell coming up in about 6 months and it supposivley having much better power management it might be hard for AMD to compete again unless Intel screws up again (like with P4), which I doubt they will.

Now the latest rumors are that next week's press conference with Read, Su and a "mystery guest" involves another AMD purchase of some company, presumbably to make them more competitive in the mobile market.

If so, seems like AMD is going for broke - all or nothing - since unless their new purchase can generate tons of revenue in the next few quarters starting with this one, they may run out of cash & credit and be forced to go Chapt. 11.

 

[truegenius]

I agree, its getting to where I can't even go here on my non apple android phone.

another hint
try opera mini 😉

 

[jerry6]

http://www.newegg.com/Product/Product.aspx?Item=N82E16819113284

Sold out as of this post lolo

plenty at Tiger , but expensive .

 

[jerry6]

true, but the problem is no one actually even tries. As I said, a PII 560 can handle single player fine, but mp .. forget it. doesn't matter about lag, server, latency. It won't work, but according to toms, because they never even tried, it works fine.

The thing is building a machine based on one single aspect alone doesn't give a clear picture of the unkonwn. Case in point,



PD's problem, TEchreport posted numbers based on one game and solely on minimum fps (they called it 99th percentile frame lag, its minimum fps), but for apps, average it all together.

Lets face it, who buys a computer based on one program alone and only looks at the one single frame that lagged the longest. forget averages, forget highest, only look at the one slowest frame in only one game. After all, thats how they gave Intel an even greater advantage. Skew the results as much as possible.

Why not Batman? http://techreport.com/r.x/amd-fx-8350/arkham-99th.gifBF3http://techreport.com/r.x/amd-fx-8350/bf3-99th.gif
or Crysis 2http://techreport.com/r.x/amd-fx-8350/crysis-99th.gif

Answer: because skyrim is the only one that gives Intel a bigger lead.

Thats how you do an Anti-AMD review.

Good point , I have a 1090 and an 1100 , going to replace the 1090 with an 8350 , there are good enough improvements in the games my son plays to make it a good upgrade . I'll give it a shot with video encoding and C A D on his comp to see if it's a worth while upgrade on my 1100 machine . So far I'm happy with the 1100 performance , cut my work load times big time over the dual core it replaced .

 

[dattimr]

Yep, I've read that as well. My question is, exactly how much of a problem is this, if only 5% or so of all apps are affected due to using Intel's compiler? IOW, AMD complaining about this would be a red herring since it wouldn't really affect them much, plus they could always develop and release their own compiler..

They have their own compiler (Open64). But then, in trying to answer your question, we'd have to ask whether the sofware that is commonly used by reviewers doesn't get compiled with ICC. Cinema 4D (Cinebench's core, apparently) uses ICC, for instance. The problem is that a compiler is one of the most intricate pieces of software, and ICC seems to be so much ahead of the competition in several areas that, in fact, using it for AMD processors ends up being faster than the other options.

Off-topic question: Have you ever used the USB BIOS Flashback feature of your board or had any issues after flashing the BIOS with an ASUS board? My M5A99X EVO doesn't get past POST after I flashed it last night, but it was done within the BIOS (through EZ Flash 2). I suspect it might be due to my using an external HD with two partitions instead of a conventional pendrive with a single FAT32 partition to store the new BIOS file. Other than that, the board might be having a short because of something, but it doesn't seem very likely -- because I didn't swap or touch anything after the update. I have tried to boot the PC with a single RAM stick, use MemOK!, clear the CMOS through the proper jumper, remove the battery and try to use the CMOS jumper again, but nothing has worked so far. The CPU_LED is always on, and I am using a Phenom II X4 970. Perhaps I should try to boot it outside of the case and open a forum thread in the support section.

 

[cgner]

What i don't understand is you keep saying games don't support beyond 2-3 cores which is true for most games made today, but then you say it takes a 8 core Piledriver to beat a Phenom II and then you say the Phenom II is still a better choice Those extra cores are not even being used for gaming(as you just pointed out) just 4 or less 90% of the time

I'm not saying its worth upgrading to piledriver for gaming if you're on phenom II but a Phenom II by no means is still a better choice if you gave me a Phenom II x4 955 and a 6300 and a H100 or 212+ i could achieve higher FPS with a 6300 same thing with a 1100T vs a 8350.

What i'm trying to do is find all the things about steamroller and try to make a honest estimate on its performance improvements over Piledriver.

Cache is easily one thing that hurts BD/PD and it seems that issue isn't being touched on

The L2 cache accounts for a large chunk of AMD’s core leakage, so shutting half or more of it down can definitely help with battery life. The resized cache is no faster (same access latency); it just consumes less power.

Steamroller brings no significant reduction in L2/L3 cache latencies. According to AMD, they’ve isolated the reason for the unusually high L3 latency in the Bulldozer architecture

http://www.anandtech.com/show/6201/amd-details-its-3rd-gen-steamroller-architecture/2

Power consumption will definitely go down by the looks of it.

Clarification: its not worth upgrading for gaming since single core/thread performance of phenom ii is either on par or better. 8 cores to beat 6 cores- in programs, not games.

 

[cgner]

name some games made this year which don't.

Mass Effect 3 - 1
COD Blops - 3 (Assuming they did not change the engine since Marfare 3)
D3 - 2-3
Source games - 2.5
TW: Shogun 2 Fall of the samurai - 1

a few others, but too lazy search.

 

[jdwii]

Clarification: its not worth upgrading for gaming since single core/thread performance of phenom ii is either on par or better. 8 cores to beat 6 cores- in programs, not games.

For starters a 8 core PD has 4 FPU where a 6 core Phenom has 6 not to mention CMT is only 70-80% scaling not 93% which is what the Phenom ii is at.

here ya go..
http://www.newegg.com/Product/Product.aspx?Item=N82E16819113287
Fx-4300 for all the hardcore gamers..

??? Lol its still better than the I3(and a Phenom II in most games) if anyone ever thought a CMT processor would be best for gaming they need to learn a few things about latency and sharing resources.

 

[proffet]

AMD FX-4300 vs i3 3220
http://www.anandtech.com/bench/Product/700?vs=677

AMD FX-4300 vs 980BE
http://www.anandtech.com/bench/Product/700?vs=362

AMD FX-4300 vs i5-2400
http://www.anandtech.com/bench/Product/700?vs=363

(it's not all that and the 980BE beats it in both Cinebench..)

 

[esrever]

Mass Effect 3 - 1
COD Blops - 3 (Assuming they did not change the engine since Marfare 3)
D3 - 2-3
Source games - 2.5
TW: Shogun 2 Fall of the samurai - 1

a few others, but too lazy search.

ME 3 console port not relevant.
Cod console port not relevant.
D3 uses minimal of 3 cores
source games will use 4+ cores and have done this since L4D
TW: shogun 2 was made in 2011, this was just an expansion.

nothing you listed except for Diablo 3 was a high profile PC game made in 2012.

 

[esrever]

AMD FX-4300 vs i3 3220
http://www.anandtech.com/bench/Product/700?vs=677

AMD FX-4300 vs 980BE
http://www.anandtech.com/bench/Product/700?vs=362

AMD FX-4300 vs i5-2400
http://www.anandtech.com/bench/Product/700?vs=363

(it's not all that and the 980BE beats it in both Cinebench..)

beats the i3 in almost everything. Sysmark doesn't count as a useful benchmark.

 

[proffet]

and for fun:

4300 vs i5-760
http://www.anandtech.com/bench/Product/700?vs=191

 

[proffet]

beats the i3 in almost everything. Sysmark doesn't count as a useful benchmark.

of course you say that...

 

[jdwii]

Mass Effect 3 - 1
COD Blops - 3 (Assuming they did not change the engine since Marfare 3)
D3 - 2-3
Source games - 2.5
TW: Shogun 2 Fall of the samurai - 1

a few others, but too lazy search.

I was a little disappointed when sleeping dogs used only 2 cores. honestly i don't see this changing much in the future and i don't really think Amd is ever going to have the best single core performance ever again which means it won't have the best gaming performance that everyone wants.

 

[jdwii]

AMD FX-4300 vs i3 3220
http://www.anandtech.com/bench/Product/700?vs=677

AMD FX-4300 vs 980BE
http://www.anandtech.com/bench/Product/700?vs=362

AMD FX-4300 vs i5-2400
http://www.anandtech.com/bench/Product/700?vs=363

(it's not all that and the 980BE beats it in both Cinebench..)

Again with the 4300 vs the 980BE battle the 4300 beats it while being able to overclock much higher most likly 4.8Ghz were the 980 will only get to 4.2Ghz maybe on air. APP's don't count as again it only has 2 vs 4 FPU vs the Phenom its really a 80+80+80+80= 3.2 Core where the Phenom had 93% scaling(under my benchmarks).

As for the care less as its a retarded comparison as its not even close in its price range.

Some people

 

[esrever]

of course you say that...

Its a non real world benchmark which doesn't reflect the realistic performance of the chips with real software. Is there anything other conclusion from it?
http://www.brightsideofnews.com/news/2011/6/22/bapco-post-mortem-is-sysmark-actually-intelmark.aspx
there is a reason nvidia and via and AMD have called quits on it.

http://www.brightsideofnews.com/news/2012/4/19/opinion-are-benchmarks-worthless.aspx

We start off with probably the most controversial organization of them all. BAPCo used to be an industry body originally founded by Intel. In fact, for good amount of time, the allegedly independent organization was operating out of Intel's offices in Silicon Valley. As Intel began to focus more and more on optimizing, more and more companies left the organization. The companies that left (in alphabetical order) are AMD, Apple, Microsoft, NVIDIA, SanDisk, VIA and many more. A good explanation of how Intel controlled the benchmark body was an organization, ARCIntuition, which always voted in Intel's favor. As our confidential source and a former board member of BAPCo said "Intel has many ways in which it influenced the outcome of BAPCo votes. One glaring example is ARCintuition, a shell company that has no other purpose than providing services to BAPCo. The company is basically an Intel sock-puppet."

 

[esrever]

I was a little disappointed when sleeping dogs used only 2 cores. honestly i don't see this changing much in the future and i don't really think Amd is ever going to have the best single core performance ever again which means it won't have the best gaming performance that everyone wants.

http://www.dsogaming.com/pc-performance-analyses/sleeping-dogs-pc-performance-analysis/

thats totally only 2 cores right?

 

[viridiancrystal]

Mass Effect 3 - 1
COD Blops - 3 (Assuming they did not change the engine since Marfare 3)
D3 - 2-3
Source games - 2.5
TW: Shogun 2 Fall of the samurai - 1

a few others, but too lazy search.

:lol:

Games can scale to more than 4 cores. BF3 does it, Dragon Age: Origins did it, and, yes, Source games do it. It is possible, it just takes some real effort.

Edit: Had to spell out a name because it made an emoticon.

 

[cgner]

ME 3 console port not relevant.
Cod console port not relevant.
D3 uses minimal of 3 cores
source games will use 4+ cores and have done this since L4D
TW: shogun 2 was made in 2011, this was just an expansion.

nothing you listed except for Diablo 3 was a high profile PC game made in 2012.

well duh. Most games are console ports, thats why I said they dont use many cores. Thats the whole point. High profile games are coming in the next few weeks, not the first 8 months. Framing a question and nitpicking the answer is a weak way to prove your point.

I was a little disappointed when sleeping dogs used only 2 cores. honestly i don't see this changing much in the future and i don't really think Amd is ever going to have the best single core performance ever again which means it won't have the best gaming performance that everyone wants.

That.

 

[cgner]

:lol:

Games can scale to more than 4 cores. BF3 does it, Dragon Age: Origins did it, and, yes, Source games do it. It is possible, it just takes some real effort.

Edit: Had to spell out a name because it made an emoticon.

I know that....... problem is over half of them dont use 4+ cores because they are console ports. *facepalm*

 

[jdwii]

http://www.dsogaming.com/pc-performance-analyses/sleeping-dogs-pc-performance-analysis/
http://www.dsogaming.com/wp-content/uploads/2012/09/Untitled-Sleeping-Dogs.jpg
thats totally only 2 cores right?

I was like this better not be a picture showing the task messenger then boom wow people its 2012 and we still think this type of crap

If i play that game on my processor it shows all 6 cores are being used so does resident evil 4 not 5 and so does other games hell even a single threaded cincebench run will show that 😀

http://benchmark3d.com/sleeping-dogs-benchmark

Wait what numbers what does that show again by the way on my processor under the task messenger i limited that program to just 2 cores and the FPS did not go down.

 

[jdwii]

Here is my laptop doing a cincebench single core benchmark

 

[noob2222]

I was like this better not be a picture showing the task messenger then boom wow people its 2012 and we still think this type of crap

If i play that game on my processor it shows all 6 cores are being used so does resident evil 4 not 5 and so does other games hell even a single threaded cincebench run will show that 😀

http://benchmark3d.com/sleeping-dogs-benchmark

Wait what numbers what does that show again by the way on my processor under the task messenger i limited that program to just 2 cores and the FPS did not go down.

look at the cpu/gpu usage. all the charts have their gpu at 95-98%. conclusion : a 6950 1gb doesn't take much cpu to max it out.

drop in 2x 7970 or gtx 690 and watch that change.

its like my testing on metro 2033. single 6970 gpu stressed 2 cores ~60-85%. CF 6970 pushed 4 cores up to 60-75% ... does that mean its only a 2-core game since a single gpu ran fine with 2?

It just takes more gpu to find a cpu bottleneck. In order to properly test a game's cpu usage, a proper review should use the absolute maximum gpu setup, and definately not a 6950.

problem is no review site wants to spend that kind of money.