Trinity vs Ivy Bridge

CarstenS said:
Intel actually promises an almost 3x increase for "Vantage Performance" performance compared to a somehow strange Ci7 2600k with HD2000 graphics - at least for desktops:
http://www.xbitlabs.com/news/cpu/di...y_Bridge_Performance_Numbers_to_Partners.html
Click to expand...
No it actually states it's non-k 2600 (which indeed exists, just lacks the unlimited turbo and only has hd2000 - yeah like that would make sense...).
Given HD3000 seems to be slightly more than 50% faster in vantage compared to HD2000 from a quick search this would still mean HD4000 is nearly twice as fast as HD3000.
Don't know exactly if these results are really that meaningful (don't see much about the setup, did they use same memory frequency, same driver?) but that would be a somewhat larger increase than previously expected.
If they actually intend to ship it in chips that matter (the k editions don't count) for the desktop I don't know...
 
from 3DCenter.de:

http://phx.corporate-ir.net/External.File?item=UGFyZW50SUQ9MTI1NDIwfENoaWxkSUQ9LTF8VHlwZT0z&t=1

look at the footnotes for Slide 16 on page 14:

Slide 16:
Projections and testing performed by AMD Performance Labs. The score for the 2012 AMD A4-4355M (ULV-17w) on the "Pumori" reference design for PC Mark Vantage Overall benchmark is projected to score 3525 while the 2012 AMD E2-1800 on the "Torpedo" reference design scores 2757. The 3DMark Vantage Performance score for the same configurations was 1700 for the 2012 AMD A4-4355M (ULV) and the 938 for the 2012 AMD E2-1800. Scores rounded to the nearest whole number.

Projections and testing performed by AMD Performance Labs. The score for the 2012 AMD A6-4455M (ULV-17w) on the "Pumori" reference design for PC Mark Vantage Overall benchmark is projected to score 4200 while the 2011 AMD A6-3400M (35w) on the "Torpedo" reference design scores 4545. The 3DMark Vantage Performance score for the same configurations was 2012 AMD A6-4455M (ULV) 2355 and the 2011 AMD A6-3400M scored 2292. Scores rounded to the nearest whole number.
Click to expand...

:freak: :???: typical AMD marketing.
 
** Shrug ** What's wrong with that? One chart showing more performance at the same TDP and the other showing slightly less performance at half the TDP. They're clearly and legibly marked. Or are you just making a joke about the erroneously numbered footnote?
 
Zaphod said:
** Shrug ** What's wrong with that? One chart showing more performance at the same TDP and the other showing slightly less performance at half the TDP. They're clearly and legibly marked. Or are you just making a joke about the erroneously numbered footnote?
Click to expand...

Yeah, I saw nothing wrong with that either, seems pretty legit to me..
 
mczak said:
No it actually states it's non-k 2600 (which indeed exists, just lacks the unlimited turbo and only has hd2000 - yeah like that would make sense...).
Click to expand...

OEMs use that config though, pairing it with discrete graphics. They seem to prefer it over 2600K with HD Graphics 3000 enabled.
 
Hi;

the slides and footnotes are perfectly fine.

My :???: was directed at AMD's marketing which gave away (AFAIK) the presentation slides i posted before without the footnotes and therefore painted the ULV-trinity in a far worse light than what we see now.
 
Hi. Collecting some news from the web.

2602f956_75a.jpeg

mi1Gr.png
 
Actually, there is one glaring inconsistency I didn't notice before:
The score for the 2012 AMD A4-4355M (ULV-17w) on the "Pumori" reference design for PC Mark Vantage Overall benchmark is projected to score 3525
The score for the 2012 AMD A6-4455M (ULV-17w) on the "Pumori" reference design for PC Mark Vantage Overall benchmark is projected to score 4200
Click to expand...
The former is certainly much less impressive compared to the A6-3400M (35w) on the "Torpedo" reference design scoring 4545 and the E2-1800 (another slight boost to Zacate?) on the "Torpedo" reference design scoring 2757.
 
Last edited by a moderator:
That's PCMark so it relies more on CPU than GPU performance, and the ULV Trinity models are probably all very low clocked with tight turbo clocks, whereas the Llano A6-3400M goes up to 2.3GHz for single-core.

Nonetheless, the "similar performance to Llano at half the TDP" statement stands true. The PCMark score difference between that Llano A6 and the Trinity A6 is around 8%. Furthermore, there may still be higher performance Trinity parts at 17W.
 
Sure. Going by their stated results. It is just a little strange that they listed two different sets of scores for what at first glance at the slides appears to be the same 17w platform. Substituting in the lower set of scores in the performance per watt comparison makes that look significantly worse (but conversely, doing the opposite would make the direct performance comparison at the same TDP look better).

I assume that just means that not all 17w AMD reference platforms are created alike in terms of actual power consumption.
 
Last edited by a moderator:
Zaphod said:
Sure. Going by their stated results. It is just a little strange that they listed two different sets of scores for what at first glance at the slides appears to be the same 17w platform. Substituting in the lower set of scores in the performance per watt comparison makes that look significantly worse (but conversely, doing the opposite would make the direct performance comparison at the same TDP look better).

I assume that just means that not all 17w AMD reference platforms are created alike in terms of actual power consumption.
Click to expand...

Not sure what you mean...one is an A4 dual core and the other is an A6 quad core. The A4 will obviously be a cheaper and lower performing part.
 
42ZqK.jpg


Finally, some more info.
Trinity seems to be pretty high clocked.
Unlocked multiplier "K" versions at launch. That shows some confidence, at least.

GPU has 384 VLIW4 shaders, probably with 24 TMUs and 8 ROPs, and pretty high clocked at 800MHz too.
 
That's a 33% clock boost on the GPU, although I guess it might be a Turbo clock. Either that or it's the PowerTune effect.
 
ToTTenTranz said:
42ZqK.jpg


Finally, some more info.
Trinity seems to be pretty high clocked.
Unlocked multiplier "K" versions at launch. That shows some confidence, at least.

GPU has 384 VLIW4 shaders, probably with 24 TMUs and 8 ROPs, and pretty high clocked at 800MHz too.
Click to expand...
Looks pretty good imho, it's encouraging the clocks got quite high even with the 65W version. Should be enough to beat old Llano's at least (though I guess part of the high gpu clock also has to do with the better cpu/gpu turbo). Of course, the cpu still won't beat anything above maybe a pentium dual core but more couldn't be expected.
 
jimbo75 said:
Not sure what you mean...one is an A4 dual core and the other is an A6 quad core. The A4 will obviously be a cheaper and lower performing part.
Click to expand...
I simply wondered why use two different SKUs if they're really both the same in terms of power? That will simply make their own marketing slides look worse than they have to.
 
mczak said:
Of course, the cpu still won't beat anything above maybe a pentium dual core but more couldn't be expected.
Click to expand...

Are you saying two Piledriver modules @ 3.8->4.2GHz aren't enough to beat a dual-core Sandybridge i3 @ 3.1GHz?!
I find that way too pessimistic.


Zaphod said:
I simply wondered why use two different SKUs if they're really both the same in terms of power? That will simply make their own marketing slides look worse than they have to.
Click to expand...
They're probably going into different price points.. But yeah, they should have at least mentioned that in the slide (they probably did, during the presentation).
 
You must log in or register to reply here.

Similar threads

D
  • Locked
A quick XBOX 360 vs. PS3 Comparison
2
Replies
24
Views
8K
MrSingh
M
A
Full transcript of John Carmack's QuakeCon 2005 Keynote
3 4 5
Replies
98
Views
32K
nemisis
N
Back
Top