Ivybridge should match LLano in graphics

[ydnas7]

Speak of the devil.

I should hope this is lower spec Llano, the white band approximates HD3000 performance.

 

[Lepton87]

I should hope this is lower spec Llano, the white band approximates HD3000 performance.

Where do you see anything that would indicate that those charts have anything to do with Liano performance? Those charts show the performance of radeon 6450 which is a 160sp barts derived discrete card with GDDDR5 memory. Liano is supposed to have up to 480sp.

 

[piesquared]

And, the HD3000 IQ looks like pre Voodoo 5 era so not exactly anything for intel to be proud of.

 

[Lonbjerg]

And, the HD3000 IQ looks like pre Voodoo 5 era so not exactly anything for intel to be proud of.

Really?
http://www.youtube.com/watch?v=4crvOyq4NMM
Why do people say so stupid things?

 

[Arzachel]

Really?
http://www.youtube.com/watch?v=4crvOyq4NMM
Why do people say so stupid things?

It was a hyperbole.


I'm definitely looking forward to a 14" laptop with hybrid crossfire graphics and a SSD. Gaming is the most CPU intensive task I do, so a souped-up Athlon II X4 is plenty of fine for me. Although, depending on how Bulldozer does, I might wait for Trinity and IB.

 

[Topweasel]

Where do you see anything that would indicate that those charts have anything to do with Liano performance? Those charts show the performance of radeon 6450 which is a 160sp barts derived discrete card with GDDDR5 memory. Liano is supposed to have up to 480sp.

Yeah I was going to say isn't that basically the chip in the Zacate. Llano at worse is using like 6530.

Edit: never mind the 2 bottom A models and an E model Fusion based on husky (10.5) cores is using the 6450. But primarily the other 8 Chips are all using 320SP (double) 6500 series IGP's or higher.

 

[chewietobbacca]

The bottom Llano chips use what's basically the 6450, but the higher end use 65xx to 66xx level chips (minus the GDDR5 and all that jazz of course)

Anywho, last I saw, Llano was shipping so has there been an ETA on this? I'd love to finish my HTPC in the next month or two, and Llano would be perfect... more CPU performance than Zacate but no need for overwhelming CPU performance, and the extra GPU performance means light gaming (such as indie titles) will be smooth

 

[IntelUser2000]

The one about the image quality, there's another slide that shows the left pic is running on lower settings which is why it doesn't show details and such.

 

[ElFenix]

The one about the image quality, there's another slide that shows the left pic is running on lower settings which is why it doesn't show details and such.

if it's comparing playable settings then it's a valid comparison, imho.

 

[IntelUser2000]

if it's comparing playable settings then it's a valid comparison, imho.

You are right, but its not a image quality problem, it just performs less.

 

[ydnas7]

http://newsroom.intel.com/docs/DOC-2032

it looks like ivy is alive
for same performance as Sandy, IB will use about half the power
peak performance will only increase by about 20%.

so apart from increased EUs, IB might give much more battery life and/or performance.

what was the turbo frequency of of SB becomes the base frequency of IB. (with the turbo being just an additional 20%)

so a SB laptop may be 12 EU at 650 mhz base and 1350 mhz turbo may become IB
16 EU at 1350 mhz base and 1650 mhz turbo

 

[tincart]

You are right, but its not a image quality problem, it just performs less.

Depends who you're talking to. If I get a person who is not computer-literate to play a game on an HD3000 and then a 6450 and have the game automatically detect graphics settings (presumably choosing higher settings for a discrete card), then they will say the difference was "image quality" and not performance.

Comparisons are going to be made by people who buy OEM products and don't bother installing new components such as discrete cards. For many of them, the question may be "how pretty is the game when I launch it?"

We have a different way of comparing image quality and performance that is more accurate, but may not be applies by many other consumers.

 

[Nemesis 1]

Ivy bridge 6 months after Llano? I really don't think so lol.

Why? Theres probably alot of people who have ES IB as we speak . around a 100 or so . Intel is releasing first QT 2012. Why you doubt this goes beyond all reasoning. Intel been pretty up front and honest siince 2006. Weres as AMD NV . not so much so they lie very well.

 

[Idontcare]

Why? Theres probably alot of people who have ES IB as we speak . around a 100 or so . Intel is releasing first QT 2012. Why you doubt this goes beyond all reasoning. Intel been pretty up front and honest siince 2006. Weres as AMD NV . not so much so they lie very well.

I have to agree here, IB is likely to be available in Jan 2011, 12 months after SB was released.

That will be really close to 6 months after Llano is available for purchase on Newegg. (which I am waiting, not so patiently, for as I really really want to replace my aging laptop that is now holding together with double-sided sticky tape and external fans for cooling)

 

[Markfw]

You mean Jan 2012 ?

 

[Nemesis 1]

http://newsroom.intel.com/docs/DOC-2032

it looks like ivy is alive
for same performance as Sandy, IB will use about half the power
peak performance will only increase by about 20%.

so apart from increased EUs, IB might give much more battery life and/or performance.

what was the turbo frequency of of SB becomes the base frequency of IB. (with the turbo being just an additional 20%)

so a SB laptop may be 12 EU at 650 mhz base and 1350 mhz turbo may become IB
16 EU at 1350 mhz base and 1650 mhz turbo

Ya this is hard to nail down . You have to remember the IGP on IB is like a core. Intel has not defined publicly yet all there is to know .
What we do know is that a core that runs less voltage gains greater performance up to 37% intel could keep the cpu cores at present performance levels and cut power usage by 50%+ to those cores and really push the IGP harder to gain that 37% increase in performance . even tho the igp is ondie it has its own power plane .
If I were intel thats the way I would play it . As cpu performance only has to be better than a stars cpu performance which is nothing for IB to accomplish. So intel does seem to have a leg up here. Further more until I see llano performance I lol at 3x faster graphics. Zacata had alot of hype but it isn't all that at all.

Intel had me a little confussed over the atom chip . But after yesterday all the dots lined up.

This graphics stuff is really trolling territory. Me and another member here went at it pretty hard over tegra1 and 2 performance . I was backing Imagination tech processor. When Tegra 2 was released AT was hyping the hell out of it. Than Apple comes with there chip. And AT didn't hype it nearly as much even tho it walked all over tegra2. I see the players and how they play .

 

[Idontcare]

You mean Jan 2012 ?

lol, yeah Time keeps flying along!

 

[jimbo75]

Yeah....if intel had actually shown a dx11 benchmark I might even have believed anything possible. They didn't though, suggesting that they have had to hack so many bits off Ivy just to get a demo out asap.

Like I said, it was already 3 months late.

45nm Jan 26 2007 - http://www.extremetech.com/article2/0,2845,2087985,00.asp

32nm Feb 10 2009 - http://www.youtube.com/watch?v=vTJLQuHUYFw&feature=player_embedded

22nm May 4th 2011 - http://www.youtube.com/watch?v=SB706hhCDZc&feature=player_embedded

Now what about Otellini? What's he saying about it?

Well at IDF last September he was saying http://www.nordichardware.com/news/69-cpu-chipset/41165-intel-already-making-22nm-processors.html

It's on track for delivery in the second half of next year

But at the last conference call he was saying http://www.xbitlabs.com/news/other/...cant_Technology_Announcement_of_the_Year.html

We remain on track to begin production on our 22nm silicon process technology by the end of this year.

Delivery second half and begin production by the years end? Nah.

IB is 3 months late, right now. They might be able to pull it in but if the chip isn't ready then you aren't going to start baking it. Plus they have extra SB to sell, and they will release server first anyway making the whole point moot.

 

[Nemesis 1]

so a SB laptop may be 12 EU at 650 mhz base and 1350 mhz turbo may become IB
16 EU at 1350 mhz base and 1650 mhz turbo

Looks good . I see it like this. Intel goes after the full 37% performance increase for the IGP .

So I would use these numbers based off of yesterday news and nothing more .

IB IGP base clock 1100 top turbo 1850. the sb has already been overclocked to 1.9 with a .2 voltage increase.

 

[Nemesis 1]

Yeah....if intel had actually shown a dx11 benchmark I might even have believed anything possible. They didn't though, suggesting that they have had to hack so many bits off Ivy just to get a demo out asap.

Like I said, it was already 3 months late.

45nm Jan 26 2007 - http://www.extremetech.com/article2/0,2845,2087985,00.asp

32nm Feb 10 2009 - http://www.youtube.com/watch?v=vTJLQuHUYFw&feature=player_embedded

22nm May 4th 2011 - http://www.youtube.com/watch?v=SB706hhCDZc&feature=player_embedded

Now what about Otellini? What's he saying about it?

Well at IDF last September he was saying http://www.nordichardware.com/news/69-cpu-chipset/41165-intel-already-making-22nm-processors.html

But at the last conference call he was saying http://www.xbitlabs.com/news/other/...cant_Technology_Announcement_of_the_Year.html

Delivery second half and begin production by the years end? Nah.

IB is 3 months late, right now. They might be able to pull it in but if the chip isn't ready then you aren't going to start baking it. Plus they have extra SB to sell, and they will release server first anyway making the whole point moot.

INtel never GAVE a DATE . There is no differance between 2h of 2011 and the end of 2011 None zero . We been threw this with John once already on BDs release. If you would have tried this in the AMD thread with John you would have got the ban hammer.

So your making all this up out of the fact intel didn't demo DX11. LOL at your silliness.

Now this is trolling Quote IB is 3 months late, right now. They might be able to pull it in but if the chip isn't ready then you aren't going to start baking it. Plus they have extra SB to sell, and they will release server first anyway making the whole point moot.End quote You best have a link showing were intel ever said the public would recieve IB in 2011. Me and IDC covered this along time ago about intel saying we would see a 22nm chip 2011. Intel never named that chip . Never happened . IDC was thinking IB . I was thinking server SB (IB) or Knights corner.

Not only that I seen 22nm IB already in the demo so we did see it in 2011. But thats just a play on words as your doing. The way intel presented made it sound like a 2011 release. So me and IDC debated on this point long ago.

 

[jimbo75]

INtel never GAVE a DATE . There is no differance between 2h of 2011 and the end of 2011 None zero . We been threw this with John once already on BDs release. If you would have tried this in the AMD thread with John you would have got the ban hammer

The hell are you talking about? Are you Otellini? Who in the AMD thread got banned?

There IS a BIG difference between shipping 2nd half and beginning production by the end of the year.

End of year = last 3 months at BEST. You cannot ship in 2011 if you only start production at the end of 2011 so cry more. If you can't take the truth stay the hell out of the forums.

PS maybe you should look at the thread title, it's not actually an intel only circle-jerk thread, you have one of those already so feel free to vacate this one.

 

[jimbo75]

Looks good . I see it like this. Intel goes after the full 37% performance increase for the IGP .

So I would use these numbers based off of yesterday news and nothing more .

IB IGP base clock 1100 top turbo 1850. the sb has already been overclocked to 1.9 with a .2 voltage increase.

You do realise that 37% increase was at very low voltages? Forget about seeing anything near 37% increases unless you are buying a smartphone.

 

[jimbo75]

Me and IDC covered this along time ago about intel saying we would see a 22nm chip 2011. Intel never named that chip . Never happened . IDC was thinking IB . I was thinking server SB (IB) or Knights corner.

Who's talking about 2011? You won't see IB in January either. Or Feb or March. At best they will do a laughable paper launch just so they can say they are still "on track", but there will be no volume IB until 2nd half 2012.

You wanna know how fast that chip was in the demo? 2 GHz. That's an awful lot of catching up to do if it's gonna be 4GHz in the oven for the 2nd half of this year don't you think?

 

[Nemesis 1]

Its hard for me to understand some in this topic. Intel released SB in jan of 2011 . If they release IB in qt1 2012. IB is in cadance with SB . If intel slipped its tick tock timeline it had nothing to do with SB or IB. That a freaken fact.

 

[jimbo75]

Its hard for me to understand some in this topic. Intel released SB in jan of 2011 . If they release IB in qt1 2012. IB is in cadance with SB . If intel slipped its tick tock timeline it had nothing to do with SB or IB. That a freaken fact.

I don't understand which part of it you aren't getting. Intel could say "IB wil be released tomorrow", would you believe that?

So why believe it will be ready for q1 next year?