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 ...

 

[noob2222]

Umm...no. Minimum FPS is the minimum FPS produced over a 1 second timespan. This can occur naturally due to a variety of reasons not related to CPU/GPU performance, and can give a REALLY bad impression if not understood properly. Its the frame latency that measures the time difference between two frames, which ideally would be equal to 16.67ms (or 60 FPS, over a 1 second timespan).

well, at least now I know all your trying to do is argue what the definition of minimum fps is.

second scatter plot isolates gaming performance by itself, with our latency-focused 99th percentile frame time results converted to FPS for easy readability.

STOP READING HERE AND TELL ME WHAT FPS IS IN THAT PLOT.

Exactly what fps is 99th percentile frame lag?
A: Maximum fps
B: Average fps
C: Minimum fps
D: some unkown variable that has never existed before

Spoiler

note the minimum FPS is frame 6 ... not frame 6-25, FRAME 6

you can have a minimum average fps, ie frames in any given second, but its still averaged in that particular second. Instantaneous minimum fps is just that. minimum is minimum, no matter what.

every one of those minimum fps is at frame 6. not second 6, not second 1, frame 6.

 

[truegenius]

since i am already in the queue for free crisis 3 😛
here is an opertunity for others too 😗
http://www.facebook.com/permalink.php?story_fbid=394562460616247&id=165710583501437

And no, nobody listens to me.
I'm a girl after all.
What would I know about computers?

 

[proffet]

sexism, racism, all kinds of prejudice suck big-time especially if your a good person and others around you are acting ignorant...

 

[gamerk316]

Exactly what fps is 99th percentile frame lag?
A: Maximum fps
B: Average fps
C: Minimum fps
D: some unkown variable that has never existed before

http://techreport.com/review/21516/inside-the-second-a-new-look-at-game-benchmarking/3

One way to address that question is to rip a page from the world of server benchmarking. In that world, we often measure performance for systems processing lots of transactions. Oftentimes the absolute transaction rate is less important than delivering consistently low transaction latencies. For instance, here is an example where the cheaper Xeons average more requests per second, but the pricey "big iron" Xeons maintain lower response times under load. We can quantify that reality by looking at the 99th percentile response time, which sounds like a lot of big words but is a fundamentally simple concept: for each system, 99% of all requests were processed within X milliseconds. The lower that number is, the quicker the system is overall. (Ruling out that last 1% allows us to filter out any weird outliers.)

 

[mayankleoboy1]

sexism, racism, all kinds of prejudice suck big-time especially if your a good person and others around you are acting ignorant...

Thats elitism 😛

 

[fazers_on_stun]

Baron(with some others) the one who convinced me Bulldozer was going to be some 50% faster, Its ok he can stay at Amdzone.

Think he wore out his welcome there as well 😛..

Somebody posted a link to his webpage (Christian Howell if you wanna google him), and apparently he is into some sort of anti-rap music nowadays..

 

[fazers_on_stun]

You OBVIOUSLY missed the reverse hyperthreading memo.

Heh, so I did, so I did.. 😛

That idea has been kicking around since I dunno - almost the entire last decade IIRC - and despite being debunked by Intel, AMD and I dunno who else, it still rears up in the rumor mill from time to time..

 

[gamerk316]

http://arstechnica.com/uncategorized/2006/07/7263-2/

And the points are STILL true.

 

[earl45]

Think he wore out his welcome there as well 😛..

Somebody posted a link to his webpage (Christian Howell if you wanna google him), and apparently he is into some sort of anti-rap music nowadays..

I warned him about his blind cheerleading for AMD, while not having a clue to what the performance
is on a unreleased CPU.
Each time looking like the dope on the ropes, and then he vanish only to return not learning from
past experiences.

 

[proffet]

Thats elitism 😛

heard of the phrase, "it's hard to soar like an eagle when surrounded by a bunch of turkey's..."
:lol:

 

[Dangi]

They do turn to "computer experts" at my school, those esteemed few who seem to know what they're talking about.

There's a variety of them:
1 - A certain someone who claims that he owns a 7Ghz Pentium 4 laptop and loves nothing more than claiming extreme overclocks as computers he runs 24/7 and loves doing the "3Ghz x 4 cores = 12GHz CPU" trick. It is no surprise that he buys AMD Bulldozer.
2 - A certain someone who goes about showing off his latest feat of tech expertise (but actually is about the easy as clicking a button) like building a Hackintosh or Minecraft server.
3 - A certain someone who goes about insulting AMD and claiming that his Intel Core i5-2500K is superior gaming computer to the computer of yours truly despite playing games with the IGP and a 19" 4:3 screen.
4 - A special someone who's never built a computer in his life but just likes to show off. He claims that he knows everything about software and hardware right after person 2 made his computer for him after your truly picked out the parts for him. Has nothing better to do than to ask people in they know what a kernel or etc. is and asking them to define it after memorising the definition off Wikipedia. His biggest claim to "understanding exactly how software works" is installing Windows. As I'm sure that's really difficult...

And no, nobody listens to me.

I'm a girl after all.

What would I know about computers?

I also have met people that does that 4 cores at 3Ghz = 12 Ghz

My old lab partner was unable to install linux using virtualbox, our teacher made a "special" distro for our subject

One companion bought a GTX 680 while having a 768p screen, kernel ?? that's the core of a matrix !!!! (lame joke)

 

[noob2222]

http://techreport.com/review/21516/inside-the-second-a-new-look-at-game-benchmarking/3

One way to address that question is to rip a page from the world of server benchmarking. In that world, we often measure performance for systems processing lots of transactions. Oftentimes the absolute transaction rate is less important than delivering consistently low transaction latencies. For instance, here is an example where the cheaper Xeons average more requests per second, but the pricey "big iron" Xeons maintain lower response times under load. We can quantify that reality by looking at the 99th percentile response time, which sounds like a lot of big words but is a fundamentally simple concept: for each system, 99% of all requests were processed within X milliseconds. The lower that number is, the quicker the system is overall. (Ruling out that last 1% allows us to filter out any weird outliers.)

nice try to take somethig out of context. Here is the question that you answered

but what we really need to understand is how well a video card maintains that steady illusion of motion.

Now, try to answer my question.

WHAT IS THE FPS IN THAT CHART? that chart isn't labeled frame lag, its labeled FPS, what is that fps?

 

[jdwii]

http://www.slashgear.com/sony-ps4-orbis-based-on-tweaked-amd-a10-tip-devs-01255214/

Since their all using Amd parts they should help keep Amd in business.

 

[-Fran-]

You know, today I noticed something really interesting.

While looking around in the directory of Dirt Showdown, they have another executable that has the "_avx" postfix.

Now... What are the FPS differences between Intel and AMD in Dirt Showdown? xD

Cheers!

 

[mayankleoboy1]

A thought: Anything that can be done by AVX, would perform better if done on the GPU.

 

[jimmysmitty]

My main problem is, real smart geeks are far and few between.

More posers than anything else.

You have no idea how many of those types I deal with. People who are "IT Consultants" who don't know jack.

It saddens me.

well, at least now I know all your trying to do is argue what the definition of minimum fps is.

http://techreport.com/r.x/amd-fx-8350/gaming-scatter.gif

second scatter plot isolates gaming performance by itself, with our latency-focused 99th percentile frame time results converted to FPS for easy readability.

STOP READING HERE AND TELL ME WHAT FPS IS IN THAT PLOT.

Exactly what fps is 99th percentile frame lag?
A: Maximum fps
B: Average fps
C: Minimum fps
D: some unkown variable that has never existed before

Spoiler

http://i145.photobucket.com/albums/r229/Lousysod/metrobench.jpg

note the minimum FPS is frame 6 ... not frame 6-25, FRAME 6

you can have a minimum average fps, ie frames in any given second, but its still averaged in that particular second. Instantaneous minimum fps is just that. minimum is minimum, no matter what.

http://media.bestofmicro.com/T/C/268032/original/Metro 2033 1680.png

every one of those minimum fps is at frame 6. not second 6, not second 1, frame 6.

I think that's just getting way to complicated TBH.

 

[sarinaide]

SteamRoller - Kaveri APU's

http://technewspedia.com/amd-and-arm-fusion-redefine-beyond-x86/

http://wccftech.com/amds-kaveri-based-28nm-richland-apu-features-steamroller-cores-compatibility-fm2-socket/

Interests: 3rd Party IP's/ARM CPU to accelarate HSA, Sea Islands integrated Graphics, Steamroller core improvements, Level 3 Cache, Native 2133mhz support.

Looking promising.

 

[amdfangirl]

^

On the GPU front, Richland would consist of the GCN based 8000 series IGP featuring the same core count as Trinity of 384 but a new architecture which would offer superior performance than a HD 7750

I doubt it. The HD 7750 is otherwise a 75W card!

 

[de5_Roy]

^ having an arm core in the apus is old news. amd is using arm's truszone tech because unlike intel, amd doesn't have dedicated security hardware in it's (especially mobile) chips. imo this doesn't have anything to do with hsa.
@wccffghijkltech article: i dunno where they got the idea of the apus having 4 mb l3 cache or just screwed up typing l2 cache. hopefully the apus will improve with l3 cache if they do get it.... as long as amd can properly use l3 cache.

 

[sarinaide]

More on Kaveri:

http://www.extremetech.com/computing/130939-the-future-of-amds-fusion-apus-kaveri-will-fully-share-memory-between-cpu-and-gpu

http://vr-zone.com/articles/amd-afds-seattle-trinity-follow-on-kaveri-to-have-true-shared-memory/16258.html

Again they talk of L3 cache and more importantly unified memory controller on both CPU and GPU componant along with other bells and whistles.

Its getting tedious that things regarding AMD are not plausible, already a A10 is 4x faster than a intel i7 in HSA bound applications, this is going to advance it further, this is now AMD's bread and butter, one would think that people can go beyond the archaic x86 which is fast becoming inefficient, one would think that AMD getting innovative with HSA partners to streamline traditional computing would in some little light be given credit. HSA vs X86 performance has clearly shown where computing can become more efficient.

I need to find the link where a 3970X was pitted against a HD7970 in pure compute performance, the 7970 showed up exactly the limitations of narrowmindedly sticking with x86 until its consumed and dry.

 

[mayankleoboy1]

When is kaveri expected to be released ?

 

[fazers_on_stun]

When is kaveri expected to be released ?

If we're taking bets 😛, I'll say 2H 2014. Otherwise, AMD's official position IIRC is Q4 of next year (i.e., about a year from now)..

 

[gamerk316]

Now, try to answer my question.

WHAT IS THE FPS IN THAT CHART? that chart isn't labeled frame lag, its labeled FPS, what is that fps?

Simple. If you know the average frame latency for 99% of the benchmark (again, 99% to remove the outliers), you can convert that to how many STEADY FPS you get by a very simple equation:

Steady FPS = 1000 ms / Average Frame Latency

So when you have the "ideal" 16.67ms latency, this equation comes out to 59.98 FPS, which matches the 59.98Hz refresh rate of the display. (Remember the NTSC refresh rate is technically just under 60Hz).

The assumption behind the FPS metric is the frame latency is equal to or under that 16.67ms limit, so 60 unique frames would be drawn to the screen, and any extra would just get dropped. If you have high amounts of latency however, the following can occur:

Frame 1 Created
Display requests image update
Frame 1 Sent to display
Display requests image update
Frame 1 Sent to display
Frame 2 Created
Frame 3 Created
Display requests image update
Frame 3 Sent to display

Note the fact that Frame 1 was sent to the display twice, and frame 2 gets dropped. That explains how you can get stuttering at high FPS rates (which occurs significantly more in SLI/CF configs, due to said latency). If this pattern continued though, FRAPS would say "60 FPS", even though you skip TWO frames every three image updates (only 30 unique frames would be drawn to the screen, with one repeating every 3 updates).

 

[sarinaide]

If we're taking bets 😛, I'll say 2H 2014. Otherwise, AMD's official position IIRC is Q4 of next year (i.e., about a year from now)..

Its simple really, product = money, no product = greater losses so from people at AMD and from experts in the Hardware and IT field AMD will release Steamroller even if only Kaveri, why its AMD's bread and butter and a very revolutionary product. I have been hypercritical of the need for FX but APU represents the horse AMD is choosing to back. While HSA is still early they will be far ahead by the time it becomes the norm.

To take stock of Kaveri;

1] Steamroller core architecture, cache improvements and latency reductions said to be around 30%, IPC gains, 35% less branch mispredicts, wider instruction execution and 15% power reduction.

2] Improved IGPU, expected SeaIslands GCN core, highest end iGPU expected to have around 850 radeon cores and over 1GFlops compute performance, will be the most powerful iGPU to date.

3] Unified memory, faster iGPU performance.

4] Significant HSA improvements, ARM x64 support.

5] Faster CPU, Faster iGPU, Faster memory performance all in the same TPD as existing Trinity parts.

So far so good methinks.

 

[de5_Roy]

Its simple really, product = money, no product = greater losses so from people at AMD and from experts in the Hardware and IT field AMD will release Steamroller even if only Kaveri, why its AMD's bread and butter and a very revolutionary product. I have been hypercritical of the need for FX but APU represents the horse AMD is choosing to back. While HSA is still early they will be far ahead by the time it becomes the norm.

To take stock of Kaveri;

1] Steamroller core architecture, cache improvements and latency reductions said to be around 30%, IPC gains, 35% less branch mispredicts, wider instruction execution and 15% power reduction.

2] Improved IGPU, expected SeaIslands GCN core, highest end iGPU expected to have around 850 radeon cores and over 1GFlops compute performance, will be the most powerful iGPU to date.

3] Unified memory, faster iGPU performance.

4] Significant HSA improvements, ARM x64 support.

5] Faster CPU, Faster iGPU, Faster memory performance all in the same TPD as existing Trinity parts.

So far so good methinks.

......
sea islands igpu? (kaveri's rumored to have southern islands (iirc)gcn cores. sea islands is hsa architecture.)
850 radeon cores in the igpu? (imo it's possible, but very very unlikely)
arm x64 ( i assume you mean 64bit arm) support? (are you sure you're not mistaking trustzone for 64bit arm instructions? those two are different. afaik if nothing drastically changes, amd won't be baking 64 bit arm cores into cpus until 2015.)
where do you get this stuff from? link, if you have any.
and no mention of the fab's capability to supply enough apus. so far this is the factor that has gone largely ignored. amd is totally at the mercy of glofo here.
all that igpu power must not be bottlenecked by the imc or the cpu. i've noticed that trinity's cpu bottlenecks the igpu at low res.