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 ...

 

[truegenius]

looks like an unlocked i3 may help out intel here 😉

 

[fazers_on_stun]

Yeah AMD is far behind Intel and it's only shoot is HSA, the only thing AMD is better than Intel right now is in iGPU and GPGPU performance, but Intel with his huge wallet can match AMD in 5 years or even catch up earlier with some kind of deal with Nvidia, as Nvidia was looking for a deal with Intel to use theirs fabs.

But AMD did a good thing believing in GPGPU performance, yeah the results have been subpar so far but I don't think Intel has more room left to improve single CPU performance, and moars cores is not the right answer, at least for general purpose.


But amdfanboys and intelfanboys must have high hopes in AMD, if AMD ends up closing Intel will be the only one left, meaning monopoly that is not beneficial for customers 'cause I still can see ARM in the mainstream PC no matter how much they improve their Cortex AXX

Yeah, HSA certainly looks like a promising idea, and hopefully Intel will jump on that bandwagon as well, although I doubt it due to Intel's hubris. But there's a long road between a promising idea and a kick-arse CPU. If Steamy rolls over and only delivers another 5% boost to PD then I guess we'll be told to "just wait for Excavator" 😛..

And if Excavator digs itself into a hole then I dunno what we'll be told - AFAIK AMD doesn't have anything past that on a roadmap..

 

[-Fran-]

Intel doesn't need "help". With current computing needs, an unlocked i3 would cannibalize i5 and even i7 sales at some point.

We already went through this, I think.

And regarding Trinity, not much difference in desktop as in laptops. They're keeping the same distance to Intel's price points regarding performance (around same P/P ratio). In the Tom's article, it was very clear: now they offer different experiences, for about the same performance per dollar.

Now, for the regular joe, I'd say, given the same P/P ratio around the block, Trinity is the better buy.

Cheers!

 

[jdwii]

Lol power.....what is it good for, absolutely nothing.......now what is it good for absolutely nothing.

I am sure there are people that actually give a damn about power numbers but I think most here are power users so its rather irrelevant.

It looks like even a Antec 380 watt for 35-45$ can handle a Trinity overclocked just fine. and just like Noob said Most PC's are on idle(and this is a fact) and when it is it uses LESS power then the Intel solution.

http://semiaccurate.com/2012/10/01/amds-trinity-faces-off-with-intels-ivy-bridge/

i7 3770k @ 3.5GHz (3.9GHz Turbo) vs A10-5800k @ 3.8GHz (4.2GHz Turbo) in CPU only workloads:

http://semiaccurate.com/assets/uploads/2012/09/Ivy-Bridge-versus-Trinity.png

http://semiaccurate.com/assets/uploads/2012/09/Trinity-Performance-Deficit.png

Hence why I'm not expecting much out of PD.

That is the worst thing you ever posted here at toms(i might have you beat though) we both know its not going to "TOP" Intel but with the AGGRESSIVE pricing of Trinity maybe Amd won't price Piledriver to high this time i can see after today a 8 core piledriver beating a I7 4 core with HT in most Multithreaded apps at a lower cost somthing a 8150 could do at best with the I7 2600K but it was so rare. Amd fans don't disagree because if it did i would have one since i do video encoding a LOT and Video rendering most of the time it lost to the 1100T and others it barely beat a I5 when all 6-8 cores were being used on Bulldozer.


And to others who keep looking at some benchmarks showing Trinity having less multithreaded performance over Llano while having better Single threaded performance that is because of CMT which means Piledriver will only scale 80% core to core while the K10 scaled around 93%(in my case). Also to some one who said L3 cache makes a 20% difference you need to look at some benchmarks because its really around 5%.

My conclusion of Trinity is its priced very well and it competes heavily towards the I3 and it also had a good enough video card for mainstream games as well as 1080P video with multiple screens without being overclock it seems to use less power than Llano and while overclocked it uses more but still nothing a Antec 380 watt for 35-45$ can't handle and if you overclock the A10 will be faster then a I3 in most cases and if you also overclock the video card you can get 6670 like performance which is a lot of bang per buck(i wish this was out when i started my first computer on a HUGE budget). I'd say once you throw in the L3 cache in Piledrivers improved on Bulldozer by around 6-8% in IPC while clocks improved around 9% which means in "Some" cases a 15% improvement over bulldozer could happen in some benchmarks(let me see handbrake then i'll know for sure).

 

[esrever]

looks like an unlocked i3 may help out intel here 😉

I doubt intel will ever do this. Im surprised they even make the 3225 tbh. Intel is about making you buy the more expensive product if you need it. That is why a lot of things are disabled as you go down their line.

 

[gamerk316]

1: I only posted S/A's conclusions, and my thoughts (about what I expected).

2: I said CPU performance is close to i3. GPU wise, not even close.

Again, these chips make sense for laptops and really cheap gaming rigs (for the computer illiterate, etc). But I really can't see any use for them outside that niche.

 

[viridiancrystal]

http://images.anandtech.com/graphs/graph6347/50398.png
thats as per my guesstimations 😗
that
pd is better than k10 in per core (but almost same performance per ghz)
and is almost same to 955/a8 in multithreaded 😀

I think that Cinebench does a terrible job of showing the performance of AMD's Module approach.
I would try this instead.
Compare this http://www.tomshardware.com/reviews/a10-5800k-a8-5600k-a6-5400k,3224-9.html
and this http://www.tomshardware.com/reviews/fx-8150-zambezi-bulldozer-990fx,3043-14.html
Look through the benches that are still somewhat close, and compare Trinity to the PII 980.

http://semiaccurate.com/2012/10/01/amds-trinity-faces-off-with-intels-ivy-bridge/
i7 3770k @ 3.5GHz (3.9GHz Turbo) vs A10-5800k @ 3.8GHz (4.2GHz Turbo) in CPU only workloads:

http://semiaccurate.com/assets/uploads/2012/09/Ivy-Bridge-versus-Trinity.png

http://semiaccurate.com/assets/uploads/2012/09/Trinity-Performance-Deficit.png

Hence why I'm not expecting much out of PD.

I bet a Radeon 7950 outperforms the Radeon 7660D. It's fair to say Trinity's a complete bomb now, right?

 

[fazers_on_stun]

http://images.anandtech.com/graphs/graph6347/50410.png


It looks like even a Antec 380 watt for 35-45$ can handle a Trinity overclocked just fine. and just like Noob said Most PC's are on idle(and this is a fact) and when it is it uses LESS power then the Intel solution.

Except that Chris found a different result in today's overclocked/undervolted article:

So, it's probable that we're seeing a difference in configuration. It looks like Anand is using the Gigabyte A85X board and perhaps an older driver version. I'm on the MSI board and Cat 12.8, with a different Intel setup as well. On the Windows desktop, after 10 minutes on each config, I get 59 W for Intel and 67 W for AMD at idle.

Again a difference of 8 watts at idle, except this time in favor of the Intel rig..

 

[-Fran-]

If you ask me, I'd wait for a proper Asus MoBo to do the power testings, haha.

Also, I'd say it's a BIOS/UEFI problem. They usually keep high voltages for components and don't implement power savings until people have massive access to the boards I think.

Cheers! 😛

 

[esrever]

bases on the different reviews I seen, most had trinity below the i3 in idle power consumption.

 

[jimbaladin]

http://www.kitguru.net/components/graphic-cards/zardon/gigabyte-f2a85x-up4-amd-a10-5800k-review-w-discrete/19/

Kitguru reckons the A10-5800K beats the i3-2105 at 1080, every game they tested with a 7970 GHz edition.

 

[esrever]

http://www.kitguru.net/components/graphic-cards/zardon/gigabyte-f2a85x-up4-amd-a10-5800k-review-w-discrete/19/

Kitguru reckons the A10-5800K beats the i3-2105 at 1080, every game they tested with a 7970 GHz edition.

nice find

 

[-Fran-]

From the link: "This particular engine is specifically dependent on the graphics card."

That's like... "This Ferrari Enzo is just as fast as this Toyota Corolla at 65 MPH!", lol.

We need more games to conclude something useful on Trinity + discrete. Let's be more patient until more reviews show up with discrete comparisons. You know, including Hybrid XFire and Discrete only.

Cheers!

EDIT: Here's the Enzo with no speed limit, lol.
http://www.xbitlabs.com/articles/cpu/display/amd-a10-5800k_7.html#sect0

 

[cgner]

From the link: "This particular engine is specifically dependent on the graphics card."

That's like... "This Ferrari Enzo is just as fast as this Toyota Corolla at 65 MPH!", lol.

lmao

 

[esrever]

From the link: "This particular engine is specifically dependent on the graphics card."

That's like... "This Ferrari Enzo is just as fast as this Toyota Corolla at 65 MPH!", lol.

One of the many reasons I find owning a ferrari is a waste.

 

[cgner]

One of the many reasons I find owning a ferrari is a waste.

u mean intels "extreme" processors?

 

[esrever]

u mean intels "extreme" processors?

that too but at least you can use the cpu to the max of its abilities without breaking the law.

 

[-Fran-]

Well... Its not ilegal to get as fast as you can to 65mph, as long as you dont burn rubber on a cops face, lol.

But yeah, just like Intels Xes, the Enzo needs a circuit to strech its legs fully.Just like almost any 400hp + car Id say 😛

I expect PD to be a Mustang... With Vishera being the V6, this should be a stock V8 at least, lol.

Cheers!

 

[cgner]

To me, its like bulldozer was 1.8L civic, and now they give it intake, headers, exhaust. They improve it and fix bottlenecks.

 

[jdwii]

To me, its like bulldozer was V8 mustang, and now they give it intake, headers, exhaust. They improve it and fix bottlenecks.

how can anyone compare crap to a V8 from ford.

 

[jdwii]

Oh i see what you did lol 😉

 

[esrever]

bulldozer is like an 18 wheeler trying to go up a mountain. Most people want sport cars.

 

[de5_Roy]

we NEED competition in cpu arena, same as in gfx card arena...
http://techreport.com/review/23663/as-amd-struggles-intel-chip-prices-stagnate

 

[Regor245]

Vtec vs V8

I know Americans hate honda 😀

 

[noob2222]

http://semiaccurate.com/2012/10/01/amds-trinity-faces-off-with-intels-ivy-bridge/

i7 3770k @ 3.5GHz (3.9GHz Turbo) vs A10-5800k @ 3.8GHz (4.2GHz Turbo) in CPU only workloads:

http://semiaccurate.com/assets/uploads/2012/09/Ivy-Bridge-versus-Trinity.png

http://semiaccurate.com/assets/uploads/2012/09/Trinity-Performance-Deficit.png

Hence why I'm not expecting much out of PD.

so 1/3 of amd's top cpu doesn't impress you vs intels top cpu. this is 1/2 of the core, with no l3 cache. Guess what benefits from l3 cache? Cinebench and games for certain.

Also looking at the power consumption figures in Chris' test here and Anand's on AT, what happened with the RCM (resonant clock mesh) technology?? The low idle power figures seem more like AMD having finer-grained power gating capability, and shutting off more of the CPU when not in use.

the a10 5800 is a 100W part, the A10 5700 is a 65W part. they are within 200 mhz of each other and priced the same. I stated earlier that most likely RCM is touchy at high clock speeds for power consumption. we need some tests with the a10 5700.