AMD Piledriver rumours ... and expert conjecture

Page 8 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
Reynod

Reynod

Administrator
Mar 26, 2007
14,921
16
42,215
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 ...
 
Sort by date Sort by votes
On
Well at this point I don't think anybody knows what the whole story is, except perhaps AMD's engineers. However looking at the Anandtech benchmark page, I don't see any multithreaded vs. single threaded tests where BD scales up by 640% (8 x 80%) as alleged in this thread. At most, the Cinebench R10 score goes up by 541% 1 --> 8 cores, while the R11.5 score goes up by 587% 1 --> 8 cores. On a per-core basis, that amounts to 67% to 73% . For SB's HT, which Intel alleges offers up to 30% improvement with 'light' multithreading, R10 shows 48% improvement and R11.5 shows a 56% improvement per thread. Admittedly 2 data points don't make a case, but for these 2 Intel's HT performs higher than the advertised max while AMD's CMT lower than the advertised max.

I could go on and compare by increase in die size - 12% vs. 5%, but that's probably stretching a 2-data-point example too far 😛.
Click to expand...
Might want to go check your figures, http://www.anandtech.com/show/4083/the-sandy-bridge-review-intel-core-i7-2600k-i5-2500k-core-i3-2100-tested/17

Cinebench SB .. with HT@3.4ghz = 22875
without HT@3.3ghz = 20381

without taking into consideration the Ghz difference, thats 12%
normalize to 1ghz, with HT= 6727, without = 6176. Thats a difference of 8% not 56% as you claim

As for bd scaling 640%, cinebench is heavily fpu intensive, bd's weakness, but its still 640%, your forgetting turbo

http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/7

8150 = 1.02, turbo @ 3.93ghz (according to anand's turbo speed tests here) = normalized score of .259*8=2.07, back to 3.6ghz = 7.47 for 100%, actual 5.99 = 80.1% scaling (pretty much what AMD said).
 
Upvote 0 Downvote
On Cinebench, the Core i7-2600K was 14% faster than the AMD FX-8150. This is an unexpected result, since the FX-8150 is an eight-core CPU, while the Core i7-2600K is a quad-core CPU with Hyper-Threading technology. The FX-8150 achieved the same performance level as the Phenom II X6 1100T, and was 10% faster than the Core i5-2500K.

http://www.hardwaresecrets.com/article/AMD-FX-8150-vs-Core-i5-2500K-and-Core-i7-2600K-CPU-Review/1402/11

On Photoshop CS4, Intel CPUs were way faster: the Core i7-2600K (3.4 GHz) beat the AMD FX-8150 (3.6 GHz) by 34%, while the Core i5-2500K (3.3 GHz) beat it by 30 percent. The AMD FX-8150 was 4% faster than the Phenom II X6 1100T (3.3 GHz) on this program.
http://www.hardwaresecrets.com/article/AMD-FX-8150-vs-Core-i5-2500K-and-Core-i7-2600K-CPU-Review/1402/6
 
Upvote 0 Downvote
If any mod knows why the message I restored was deleted, please let me know. From what I can see it was still within lines of this thread, it had a link to info of Trinity, Trinity is a PD part so its on topic, and as well a theory on fazers part that Trinity might get delayed.
Click to expand...

Might have been me ... I was rushing to clean up the posts (as I was hanging for a can of coke) which had nothing to say about the topic and i might have nuked yours ...

If so then just restore it jimmy.

I tend to swing the deletehammer with reckless abandon here.

Note the marks on the wall of the forum where I left dents ...

:)
 
Upvote 0 Downvote
You want to stay resurrected ... or do I check your IP to see you are actually Daddiggle and permaban you again?

Either have something to say about Piledriver or move on.
Click to expand...
I had something to say about Piledriver...but I forgot. I though I should let you know...

Okay I remembered. How exactly is PD's arch going to differ from BD?
 
Upvote 0 Downvote
You want to stay resurrected ... or do I check your IP to see you are actually Daddiggle and permaban you again?

Either have something to say about Piledriver or move on.
Click to expand...
eh who? I'm On a mobile at the moment. Our broadband networks assign the same ip to all its users.
 
Upvote 0 Downvote
Not guilty, so I dunno.
Click to expand...

I doubt it would be you. Wise monkeys never do.

😛

Trinity A0 ES at 3.8Ghz/4.1Ghz TC is said to equal Llano at 2.9Ghz in 3dmark and Cinebench 11.5

http://translate.google.com/translate?sl=zh-CN&tl=en&js=n&prev=_t&hl=cs&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fwww.chiphell.com%2Fthread-307338-1-1.html
Click to expand...

I hope thats not true. Would not be what people are hoping for and might be a glimpse of PD itself.

Might have been me ... I was rushing to clean up the posts (as I was hanging for a can of coke) which had nothing to say about the topic and i might have nuked yours ...

If so then just restore it jimmy.

I tend to swing the deletehammer with reckless abandon here.

Note the marks on the wall of the forum where I left dents ...

:)
Click to expand...

Wasn't mine, was one of Fazers. While the first part was off topic, the rest was on topic about PD, possible Trinity delay and AMD itself.
 
Upvote 0 Downvote
Trinity A0 ES at 3.8Ghz/4.1Ghz TC is said to equal Llano at 2.9Ghz in 3dmark and Cinebench 11.5

http://translate.google.com/translate?sl=zh-CN&tl=en&js=n&prev=_t&hl=cs&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fwww.chiphell.com%2Fthread-307338-1-1.html
Click to expand...

It isn't too hard too see that Trinity's CPU cores are gonna stink big time......even with 5% IPC improvements, it just won't be able to keep up with 4 real Stars cores. Period.
Firstly, they'll need at least 3.6Ghz+ to match the llano cores, but the TDP limits will definitely make that difficult. Secondly, these PD cores aren't even gonna have L3 cache.
According to an IEEE paper i just went through, each BD module is only 30.9mm2, including the 2mb L2 cache, while each integer half core is barely 3mm2.
http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=5746227

What AMD really needs to do is seriously beef up those half integer cores(3 ALU's atleast), as they don't take up too much space......just my thoughts...
 
Upvote 0 Downvote
It isn't too hard too see that Trinity's CPU cores are gonna stink big time......even with 5% IPC improvements, it just won't be able to keep up with 4 real Stars cores. Period.
Firstly, they'll need at least 3.6Ghz+ to match the llano cores, but the TDP limits will definitely make that difficult. Secondly, these PD cores aren't even gonna have L3 cache.
According to an IEEE paper i just went through, each BD module is only 30.9mm2, including the 2mb L2 cache, while each integer half core is barely 3mm2.
http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=5746227

What AMD really needs to do is seriously beef up those half integer cores(3 ALU's atleast), as they don't take up too much space......just my thoughts...
Click to expand...

Your forgetting that trinity will be a 2 core (module) part. Thats two cores vs four cores. Two trinity cores competing against 4 Llano cores means IPC has improved significantly.
 
Upvote 0 Downvote
Your forgetting that trinity will be a 2 core (module) part. Thats two cores vs four cores.
Click to expand...

Trinity is stated to have 4 PD "cores", ie 2 modules. It's AMD's fault if they decided to call it a quad core.

"Two trinity cores competing against 4 Llano cores means IPC has improved significantly."....or so we all hope!!!
 
Upvote 0 Downvote
Jimmy I probably deleted it then came back later and deleted the original (other offending post).

If you see any personal slights just delete the whole post ... irrespective if it has good content in it.
Click to expand...

I understand except when we have the ability to modify the post, we should. Instead of removing the entire thing, remove whats not needed and leave whats useful.

Trinity is stated to have 4 PD "cores", ie 2 modules. It's AMD's fault if they decided to call it a quad core.

"Two trinity cores competing against 4 Llano cores means IPC has improved significantly."....or so we all hope!!!
Click to expand...

Should be interesting to see. Thing is, Trinity will use PD so this should give us a slight idea of how PD will really perform. Take Trinity, turn off th GPU and compare it to a FX-4100. Bam. BD vs PD.
 
Upvote 0 Downvote
Should be interesting to see. Thing is, Trinity will use PD so this should give us a slight idea of how PD will really perform. Take Trinity, turn off th GPU and compare it to a FX-4100. Bam. BD vs PD.
Click to expand...
It will definitely be an interesting comparison, but the future PD for desktops should be a better performer than the PD derived core in Trinity, simply because Trinity will lack the L3 cache that desktop PD will have.

Additionally, Trinity's clockspeed compared to future PD desktop CPU's will almost certainly be limited due to Trinity's TDP having to share a GPU, unless AMD decides that all CPU's from here on will have a GPU integrated.
 
Upvote 0 Downvote
It will definitely be an interesting comparison, but the future PD for desktops should be a better performer than the PD derived core in Trinity, simply because Trinity will lack the L3 cache that desktop PD will have.

Additionally, Trinity's clockspeed compared to future PD desktop CPU's will almost certainly be limited due to Trinity's TDP having to share a GPU, unless AMD decides that all CPU's from here on will have a GPU integrated.
Click to expand...

I see your point, but with some areas L3 does not help at all as we have seen from Llano. But I can see where it would help in other areas vs BD.

But still it would give us a rough estimate of what we should see performance wise.
 
Upvote 0 Downvote
Might want to go check your figures, http://www.anandtech.com/show/4083/the-sandy-bridge-review-intel-core-i7-2600k-i5-2500k-core-i3-2100-tested/17

Cinebench SB .. with HT@3.4ghz = 22875
without HT@3.3ghz = 20381

without taking into consideration the Ghz difference, thats 12%
normalize to 1ghz, with HT= 6727, without = 6176. Thats a difference of 8% not 56% as you claim
Click to expand...

No, in order to put AMD's CMT and Intel's SMT on an equal basis, I looked at how they scale up, ignoring 'real', 'almost real' and 'fake' 😛 core differences by comparing a single thread bench to an 8-thread bench for each CPU. There is no Cinebench review that I'm aware of comparing 4 BD cores to 8, which is why I did it that way - apples to apples.

 
Upvote 0 Downvote
Wow... this thread has to be setting new records for mod derailment 😛

It is nice to hear that AMD is hitting 4.1ghz with first gen samples. Unfortunate that 3.8/4.1 supposedly only matches Llano @ 2.9.

It takes 40% clock advantage to normalize the 20% module scaling inefficiency...

Hopefully AMD can continue to chip away at the IPC gap with better steppings.
 
Upvote 0 Downvote
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom