What to Expect From AMD at ISSCC 2011

[god_43]

gahh wheres llano?? screw zacate bring on llano/ontario

?? Ontario and zacate are the same core arch?

[mAJORD]

not much to say about llano, it's K10 + evergreen based GPU fusion..

Bobcat and Bulldozer are their newest architectures since Athlon/K7, so plenty of juicy bits to go into

[JkS]

I don't know about you lot, but I'm very interested in knowing about the power saving features. I really hope that they can come up with something innovative there (my power bills are starting to concern me ).

[Sn0wm@n]

cant wait for it

[Johnny87au]

?? Ontario and zacate are the same core arch?

iirc i thought ontario uses less TDP then zacate, i thought it goes ontario>zacate>llano in terms of power..

[The Coolest]

Ontario are 9W 1GHz dual core or 1.2GHz single core.
They also have a much slower GPU, at 280MHz compared to 500MHz in Zacate.

[marten_larsson]

Ontario and zacate is the same chip, just binned differently. It's like saying a x4 and x3 are different cpus - they are but they're not.

[EvilOne]

Ontario and zacate is the same chip, just binned differently. It's like saying a x4 and x3 are different cpus - they are but they're not.

Yeah,

hope they give some Llano info out instead of those 2.

[FlanK3r]

Surely that's not stable? I need 1.525v for 3.9GHz stable

Looks stable


Max is 4300 MHz aircooled stable (1.48V), problems are not temps, but one core is whorse than others. But some hard benchmarks I can at 4400 MHz. And validation 4894 MHz with normal ambients 20 C.
I had overall 5 Thubans, 2 I solt it for my friends. I have still others Thuban 3 chips. But really, this from PRIME screen (and 4300 MHz max stable) is the best from all! Better than 1100T Thubans. Only MAD222 has one better. Think, I was lucky 1015DPMW is great

[Tomasis]

flankr, wow

[[XC] Oj101]

Ah ok, I can also Prime away at high speeds for an hour, but without 1.525v I get a BSOD within about three days of running World Community Grid.

Please PM me your background

[saaya]

I hope amd has at least as good power gating as Intel...
Without it, even awesome Ipc boosts won't make them king of the hill...

[JkS]

Anyone know if there is a stream that goes along this event?

[rcofell]

Afaik, like most academic conferences there likely won't be, at least the sub-sessions for sure.
The best source is probably going to be whatever tech sites pick up on, as well as whatever AMD publishes on their blog.

[JkS]

Afaik, like most academic conferences there likely won't be, at least the sub-sessions for sure.
The best source is probably going to be whatever tech sites pick up on, as well as whatever AMD publishes on their blog.

Thanks for the info.

[crazydiamond]

Looks stable
[IMG]xhttp://img683.imageshack.us/img683/5721/prime1hstable3954mhz277.png[/IMG]

Max is 4300 MHz aircooled stable (1.48V), problems are not temps, but one core is whorse than others. But some hard benchmarks I can at 4400 MHz. And validation 4894 MHz with normal ambients 20 C.
I had overall 5 Thubans, 2 I solt it for my friends. I have still others Thuban 3 chips. But really, this from PRIME screen (and 4300 MHz max stable) is the best from all! Better than 1100T Thubans. Only MAD222 has one better. Think, I was lucky 1015DPMW is great

<3 week 1015

[Mechromancer]

Any news yet? Did any tech site actually go?!

[Raqia]

Any news yet? Did any tech site actually go?!

Some news re: bulldozer

http://www.eetimes.com/electronics-n...z?pageNumber=2

AMD Bulldozer, Intel server CPUs
Advanced Micro Devices provided more details on its new Bulldozer core first described at Hot Chips in August. AMD senior engineer Hugh MacIntyre said the core enables 3.5 GHz performance is same power and thermal envelope as AMD's prior core design.
The core delivers linear performance across a range of frequencies and 0.8-1.3V voltages it will need to operate. It uses 213 million transistors in a 30.9mm2 block with 11 metal gates in a 32nm SOI process, he said.
A separate paper described Bulldozer's 40-entry instruction out-of-order scheduler and execution unit that can issue up to four instructions per cycle. The unit helps the core meet its target of delivering 90 percent of performance of past AMD cores with a significant reduction in area and power, said Michael Golden, another AMD engineer.

...although there's nothing that wasn't in the abstract save the 90% performance figure which we sort of knew already.

[BeepBeep2]

90% speed of Phenom II per core
Add 2 cores
Better power gating?
Shrink to 32nm?

It seems they will need 5-6 ghz stable on air at the very least to compete with Sandy Bridge single threaded...

[flippin_waffles]

Some news re: bulldozer

http://www.eetimes.com/electronics-n...z?pageNumber=2



...although there's nothing that wasn't in the abstract save the 90% performance figure which we sort of knew already.

I think this part is particularily interesting:

The core delivers linear performance across a range of frequencies and 0.8-1.3V voltages it will need to operate

[nunchucka]

I think this part is particularily interesting:

i find the 3.5ghz to be very interesting as it tends to give this screenshot some credibility. why would they choose 3.5ghz in a screenshot taken before this info was released?

http://www.xtremesystems.org/forums/...&postcount=261

[flippin_waffles]

i find the 3.5ghz to be very interesting as it tends to give this screenshot some credibility. why would they choose 3.5ghz in a screenshot taken before this info was released?

http://www.xtremesystems.org/forums/...&postcount=261

True, and the enthusiasm coming out of AMD seems to be very genuine so they must believe they are sitting on some pretty good products.

[EvilOne]

Some news re: bulldozer

http://www.eetimes.com/electronics-n...z?pageNumber=2



...although there's nothing that wasn't in the abstract save the 90% performance figure which we sort of knew already.

Looks like that guy didnt do his (or any) reaseach about the 90% thing. Its 180% when 2 cores are getting used in same module, and its NOT compared to old/other cores, its compaired to 1 core running in the same module. There is no performance related data (verified) so claiming that its 90% of an old Ph2 core is crazy, JF has already said 100 times that its faster then the old cores anyways lol.

Any solid info came out of ISSCC yet?

[Glow9]

Looks like that guy didnt do his (or any) reaseach about the 90% thing. Its 180% when 2 cores are getting used in same module, and its NOT compared to old/other cores, its compaired to 1 core running in the same module. There is no performance related data (verified) so claiming that its 90% of an old Ph2 core is crazy, JF has already said 100 times that its faster then the old cores anyways lol.

Any solid info came out of ISSCC yet?

By not doing his research you mean not copying down what the engineer said correctly? You in denial or something?

[Sn0wm@n]

90% speed of Phenom II per core
Add 2 cores
Better power gating?
Shrink to 32nm?

It seems they will need 5-6 ghz stable on air at the very least to compete with Sandy Bridge single threaded...

but the thing is its not only the core performance that makes a difference .. its the rest feeding the core ... and i suspect that amd improved alot of things .... and since we didnt have a look at any benchmarks those 90% figures could be the thing we kept hearing about the module penalty .....


so im still not sold on sub par performance until i see it with my own 2 eyes ...


amd spent 10 years on that design .. so lets give them credit for making it better and not worse

By not doing his research you mean not copying down what the engineer said correctly? You in denial or something?

and you think that JF-AMD spewed random stuff out of his mouth on that subject???


he had to be cleared by legal dept way before he could comment on that subject .. and yet his words were clear ... we have an ipc increase compared to older core design ....


so again .. this article isnt good at all ...