Bulldozers first screens

[Postmodum]

Btw... doesn't the PC Mark Vantage results include some GPU scores as well?

If so ... that sucks because, the slight "edge" we see between 2600k and the Bulldozer, can be from the gains that the 6670 has vs integrated gpu in the 2600k :x

[Mumak]

Those red part of bars are called MARKETING POWER

Btw... don't the PC Mark Vantage results include some GPU scores as well?

If so ... that sucks because, the slight "edge" we see between 2600k and the Bulldozer, can be from the gains that the 6670 has vs integrated gpu in the 2600k :x

[EvilOne]

Do you guys waiting on BD honeslty think it has a chance against sandy bridge e with quad channel memory? I hope Im wrong but Im sure some of you all think that BD will be a huge let down.

If the IPC gain is as high as the PCmark leak indicates do you think Sandy Bridge E - 6 core has a chance against a Interlagos Bulldozer - 16 core? I hope your wrong but im sure some of you all think that BD will be a huge let down.

[hurleybird]

Do you guys waiting on BD honeslty think it has a chance against sandy bridge e with quad channel memory? I hope Im wrong but Im sure some of you all think that BD will be a huge let down.

Do you not listen at all? Interlagos is AMD's analogue to E.

Zambezi vs. Sandy = eight threads vs. eight threads

Interlagos vs. E = sixteen threads vs. sixteen threads

If the IPC gain is as high as the PCmark leak indicates do you think Sandy Bridge E - 6 core has a chance against a Interlagos Bulldozer - 16 core? I hope your wrong but im sure some of you all think that BD will be a huge let down.

Neutered 6 Core E's will come first, but you better believe that if AMD releases a higher performing 16-core desktop Interlagos, Intel will respond with an eight core E. Remember, at the chip level E is an octo-core!

[Skratch]

If the IPC gain is as high as the PCmark leak indicates do you think Sandy Bridge E - 6 core has a chance against a Interlagos Bulldozer - 16 core? I hope your wrong but im sure some of you all think that BD will be a huge let down.

You have no case since for the last 7 years intels chips have out performed AMDs,its always a waiting game from the amd side.

Intel will be in 22nn before amd even gets BD out the door.

I dont have to wait 5 years to find out if sandy E is going to be better than BD

[Dresdenboy]

Do you not listen at all? Interlagos is AMD's analogue to E.

Zacate vs. Sandy = eight threads vs. eight threads

Interlagos vs. E = sixteen threads vs. sixteen threads

You likely meant "Zambezi" (Bulldozer), not "Zacate" (Bobcat).

[hurleybird]

I do indeed. Just coming off a night shift, does it show?

[Postmodum]

Those red part of bars are called MARKETING POWER

Yes, but the red part is 3DMark Vantage...
What i don't know is if the PCMark Vantage, doesn't test the GPU too?

Because, if it does... it sucks big time, since the BD is using a far better graphic solution ... enough to make a difference in the final score, that way, scoring higher than the 2600k, even if it's by a "nail" ...

Meh... hate these benches from companies... i'm still to see a constructive benchmark from Intel or AMD where they take the same spec'ed system and testing it head to head ...

[Skratch]

I hope BD comes out swinging I honeslty do.

[mak1skav]

I am a huge fan of Intel processors but i don't hide that i have great expectations and hopes for this BD thing. I have stopped buying computer related products based on a company fanboism 15 years ago so now i just want to buy the best performance my few money can provide me.

Most of the times the competition improves the price and the performance of a product and so the buyer benefits from it. That's why i am hoping for great results from the "famous" Bulldozer CPU architecture.

I really don't understand why so many people try to predict the performance and the worst thing is that they try to convince all other members that they are true. Every time i login in the forums i see new posts and i am loading the thread anxious to read any real news but in vain. Sadly though every time i discover that there are just more rumors, speculations and battles about who is the best performance predictor.

Even if you are big fans of Intel company like me, don't you want to buy more products from your be-loving company cheaper?

Don't forget competition makes things better.

[Skratch]

what does quad-channel memory have to do with that? that doesnt improve performance.

16 threads with 8 cores needs as much bandwith as you can give them.

How is memory bandwith not going to inprove performace?

[flyck]

Do you guys waiting on BD honeslty think it has a chance against sandy bridge e with quad channel memory? I hope Im wrong but Im sure some of you all think that BD will be a huge let down.

yes because SB-E will be so great...? it will have 6cores clocked @ 3.33GHz... If BD is faster than 990x the difference between SB-E and BD will be minimal. (since SB-E will be <10% faster than the 990x)

[XRL8]

Wow, what a trollfest.
a Quick reminder for those detached from reality.

AMD`s attempt at "cores" and threading got MASSIVELY different from intel in bulldozer.Only rational way to compare them is the die area/power consumption and price.
OR if you want to cling on to that.Look at it this way.:
AMD`s 8Threads vs Intels 8 Threads
8core zambezi vs 4core/8thread sandybridge (built on the same process node, so its kind of comparable at least).

Second, in this benchmark, first part of it is the part that is interesting to us, its non igp/gfx card related.
And in that part 2600K is around 8core Zambezi.So we know that these 2 cpus get similar score in a benchmark.Lets dissect a little.
What this benchmark "scores", well, its a platform benchmark, it does everything from single threading to weak multithreading ,hard disk benchmarks and bandwith ones.
It surely IS NOT a benchmark aimed at bulldozer`s strong points.
So im asking how on earth can anybody claim:
a) Its a "FAIL" on amds front ? ,is sandy bridge a "FAIL" ?
b) How can anyone get any gaming or other purely computational performance estimate form it ?

And as for SandyE, its differrent league.For a single SandyE system you will be able to build two Sandy or Bulldozer ones...AND its half a year later.

[Skratch]

yes because SB-E will be so great...? it will have 6cores clocked @ 3.33GHz... If BD is faster than 990x the difference between SB-E and BD will be minimal. (since SB-E will be <10% faster than the 990x)

We will have to wait it out,I just hope BD will clock high since most sandys are doing 4.7-5.0 ghz on air

[Manicdan]

^^ @XLR8, i agree

theres 3 real tests
1-2 thread (mostly single threaded stuff, superpi or something better)
3-5 threads (mostly 4 threads running, games)
unlimited threads (no logical cap, like CB 11.5 and other rendering software)

with how different the approach is between brands, the battle is at all 3 tests rather than just one or two. and were at the point where both CAN be equally competitive at all 3 test segments. if it WILL is another question yet to be answered.

[undone]

Man, suddenly remind me of this OLD rumor, dig out then find:

http://www.xtremesystems.org/forums/...d.php?t=265710

PC Mark:
x6 1100T 6430b
i7 980x 7150b
i7 2600k 7700b
AMD Zambezi 7950b

Cinebench R11.5
x6 1100T 5.9b
i7 980x 8.95b
i7 2600k 6.7b
AMD Zambezi 11b !!!

3D06 CPU
x6 1100t 5990b
i7 980x 7388b
i7 2600k 6600b
AMD Zambezi 8800b!

So the slide are not unreliable according to this old leak. Zambezi may be only slightly higher than 2600k in PCMark, but who knows the Cinebench & 3dmark06 score comparison is reliable or not. If they're all true then Zambezi may have a great succeed after all.

[Postmodum]

Sorry but... doesn't the PC Mark Vantage has graphical test too? So why do you say it's not related?
And if you have a stronger graphic solution, your score will be higher...

If in the graph... there was a note saying that there's not graphic tests ... i would be ok with it ...
But like everyone else, i'm speculating about this ... and if the PC Mark Vantage tests your entire system, graphic solutions will have a good impact on the final score ... helping that difference between BD and SB ...

That's just my way of analyzing the graph ... i think it's correct no?

( No, i'm not a fanboy ... and i really want to go from my i7 to the BD ... before you start trolling )

[flyck]

We will have to wait it out,I just hope BD will clock high since most sandys are doing 4.7-5.0 ghz on air

If only intel could do something with that. Gulftown is overclocking to those speeds also... yet both SB and gulftown are running against their TDP much sooner. (SB has some room now... but only if you don't use AVX supported code)

So the slide are not unreliable according to this old leak. Zambezi may be only slightly higher than 2600k in PCMark, but who knows the Cinebench & 3dmark06 score comparison is reliable or not. If they're all true then Zambezi may have a great succeed after all.

or this slide is made out of the former slide? :-)

[ACK]

...and according to this thread, I should buy BD?

[Postmodum]

Undone, good catch ... it really looks to be a good association

[Olivon]

Sorry but... doesn't the PC Mark Vantage has graphical test too? So why do you say it's not related?
And if you have a stronger graphic solution, your score will be higher...

If in the graph... there was a note saying that there's not graphic tests ... i would be ok with it ...
But like everyone else, i'm speculating about this ... and if the PC Mark Vantage tests your entire system, graphic solutions will have a good impact on the final score ... helping that difference between BD and SB ...

That's just my way of analyzing the graph ... i think it's correct no?

( No, i'm not a fanboy ... and i really want to go from my i7 to the BD ... before you start trolling )

We don't know wich test was used

If it's the global score, GPU plays a role for sure (according to PC Mark Vantage white paper from Dresdenboy) but we know nothing, so ...

[Postmodum]

Thanks Olivon ... ehehehe
So what we can see there is BD with 6670 having higher pcmark score than a 2600k with igp ... and that difference can be from the graphic solution used Meh ...

Let's see

Thank you again for answering my questions ... eheheh i did the question twice :P

[-Sweeper_]

from the PCMark's white paper

PCMark Vantage Main Features
PCMark Suite
o A collection of various single- and multi-threaded CPU, Graphics and HDD test sets with the focus on Windows Vista application tests

as the slide talks about balanced computing, it could mean CPU+GPU performance

[Postmodum]

Yep... so ... i think everyone should hold your horses, because that difference in pcmark may be from the 6670 solution used and it's not about raw cpu/memory/mb/hdd performance ... and it may end up having lower scores than 2600k if no graphical tests were used ...

That's why these benchmarks suck ... using different systems to show an advantage ... meh! :P

[FlanK3r]

Btw... doesn't the PC Mark Vantage results include some GPU scores as well?

If so ... that sucks because, the slight "edge" we see between 2600k and the Bulldozer, can be from the gains that the 6670 has vs integrated gpu in the 2600k :x

they compare not GPU 6670 with IGP in SB, but with x6 1100T!

A8,A6 (APU Llanos chip) are compare with IGP 2600K!