Ontario 9w and 18w versions + pics

[James]

Originally Posted by muziqaz

to be honest, that game on core i5 was running with a bit higher details I think, because skyscrapers had lights on. Though sometimes textures were disappearing from i5 screen
And I think SB integrated GPU might be able to accelerate those html5 stuff in the same manner as Zacate.

That was the difference between daytime and night time. I don't think the lights were actually showing any light sources or anything lol.

A good indication of a gpu's prowess is when it has to render flames. If you check out 33s onward you'll see the Zacate system hardly falters. That's good.

In this particular example, the Zacate APU is probably running 5x-8x faster than the i5. No doubt AMD are putting their best case forward but based on this we're looking at Zacate being 1.5x-2x faster than SB's best (yes I mean desktop i7 with 12 EU's and turbo).

[muziqaz]

Originally Posted by aaronspink

Its the future like 3dnow was the future!

I don't really know you, but by the looks of it, you are not that silly to make such comments people who are not ignorant enough know that 3dnow failed because Intel had more meat than AMD, and it is extremely hard to introduce something new, when your bigger competitor wants to do everything his way.

[Sheldon]

Originally Posted by Palliot

Have a look how erratic 2D acceleration benchmarks (or driver performance) can be:

http://www.tomshardware.com/reviews/...i,2539-10.html

You really cant detemine GPU (or CPU) performance from that. Its actually possible an nvidia 7050 IGP would beat the socks of both i5 and bobcat in that same test. Now do you want one? I can sell you one if you want

Also remember that the 5000 series has 2D acceleration troubles with that version of Catalyss 9.12. 10.6 went a way into rectifying that.

http://www.tomshardware.co.uk/ati-2d...w-31910-3.html

But I accept your premise on that it doesn't always indicate the performance correctly. What I was trying to say with the video is that the i5 was on equal terms with the benchmark, in that it too was being accelerated through it's onboard GPU.

[muziqaz]

Palliot, i think it was a driver glitch. just dropping frames or something

[James]

Originally Posted by Palliot

Dont know whats up with that, but looks like a driver issue to me.

So what you're saying is, Recommended ATI or nVidia gpu?

Hard to believe a Core i5 cant play that properly if there isnt a driver glitch. The fact bobcat can play it is hardly impressive given the above requirements.

So Bobcat plays it flawlessly and is "hardly impressive". What does that make the i5, which can't even run it at 10fps?

Sooner or later you'll realise that graphics drivers don't make themselves, and intel don't really bother with them either. Unless it's say, 3dmark.

[dragonfli]

Originally Posted by muziqaz

And I think SB integrated GPU might be able to accelerate those html5 stuff in the same manner as Zacate.

you can do it for sure with discrete if money or watts wasn't a rpoblem, but what can you do with low voltage celeron or atom at the same price??

zacate doesn't has to be the fastest, just very good value

[muziqaz]

Originally Posted by dragonfli

you can do it for sure with discrete if money or watts wasn't a rpoblem, but what can you do with low voltage celeron or atom at the same price??

zacate doesn't has to be the fastest, just very good value

yes, it will be very good value for people who are facebooking, twittering, and youtubing.

[muziqaz]

Charlie has a photo of the chip upside down
http://www.semiaccurate.com/2010/09/...obcat-systems/
anyone likes to count?

[redpriest]

Originally Posted by aaronspink

Its the future like 3dnow was the future!

Or the future like x86-64 ^_^

[Sparky]

So what if it was a driver glitch. Point is, that the Intel had issues with framerate, along with the driver issues that will inherently follow on the Intel platform.

You're not going to see those sort of issues on the AMD product - sure, driver work there is not all that spectacular on the crossfire setups (GSOD, crossfire profiles), but due to the shear size of the driver team, it's always going to be better than the Intel solution in these simpler parts.

Of course, the Intel guys will just spin it like they always have, it's either going to be delayed, or SB can easily be modified to *devastate* Bobcat or the like... Hopefully Intel has the same attitude as you guys, because they're going to be in for a little surprise, methinks.