San Francisco (CA) – Day 1 of the Fall Intel Developer Forum 2007 has come to an end and we are digesting the news and information we have received so far. TG Daily's Rick Hodgin sums up his impressions about the key products at the show, Penryn and Nehalem.

Pat Gelsinger took the stage early Tuesday morning to introduce the day’s kick-off keynote speaker, Intel’s CEO Paul Otellini. But before he handed off to Otellini he had something rather surprising to say: He referred to IDF 2007 as the “Industry Developer Forum” and not the “Intel Developer Forum“. This proclamation echoed a consistent theme repeated throughout the day: Intel is very interested in supporting their partnerships and alliances.

In truth, Intel came across today as a new company. They appeared to be something less than the cold, hard stand-alone company they have often been perceived. That fact was also echoed internally by more than one senior level executive at a late evening event held for the press to mingle with Intel executives.

In fact, visions of a company that’s not only reaching out externally, but also reaching out internally were conveyed. It seems the internal mindset of even a few years ago has been stripped away. Replaced with a more flexible, dynamic Intel, which at least one senior executive indicated is the reason why Intel has delivered on many products since those years with Netburst, the now famous micro-architecture of the Pentium 4 which nearly brought the company down to its knees.


Extreme to mainstream

The central theme for Otellini’s 1.5 hour message was wrapped up in this phrase: The extreme to the mainstream. The idea is taking what’s extreme today and making it mainstream tomorrow. He talked about this being true in all areas, not just mobile space, or not just server space. But rather it is the goal everywhere. He also cited history as an indicator of where we were, what we have now and what‘s likely to come as things progress. Imagine a typical Starbuck’s today. Countless notebooks and WiFi connectivity. Five years ago that would have been rare. And ten years ago, impossible. All, according to Otellini, thanks to Intel’s contribution in connectivity, something he called “Internet everywhere.”


Penryn

While there were many key ingredients given for Intel’s success, the first processor discussed was Intel’s upcoming mobile chip, Penryn, which we will see on November 12. The updated Core architecture is key, as is the underlying 45nm process technology.

 

Otellini provided an overview of the history of the insulating layer which, in modern CPUs, is only five molecular layers of silicon dioxide (SiO2) thick. He explained that as far back as 15 years ago, Intel’s engineers saw this layer as problematic. The continued scaling of the insulating layer could not continue forever. And, we found out later in the day with Dr. Gordon Moore’s keynote, that five molecular layers is about the lowest you can go in practice. It’s a form of wall, and Intel was right up against it.

Penryn will come with a hafnium-based hi-k dielectric and a metal gate solution. This technology helps to reduce gate leakage by 10x over existing models, according to Intel. This 10x reduction is not in an overall leakage reduction, but only gate leakage. We were told future solutions will address aspects of the remaining leakage, though no details or timeframes were disclosed.

It’s interesting to note that Gordon Moore talked about hard limits being reached in 10-15 years. Up until now, he said, the walls which have been out there in front of the designers have always managed to fall down somehow. But now, they are real walls, ones which might truly stop further development until they are solved.

Penryn also comes with some fairly significant architectural changes, which vary somewhat from what we have understood Intel’s tick-tock strategy to be. "Ticks" are described to be process revisions while "tocks" are supposed to be new processor architectures. Intel showed several slides indicating the same thing – but Penryn will be introducing fairly significant architectural changes that are much more than a refresh.

These include Intel’s Wide Dynamic Execution engine with a new, untested Radix-16 divider (the first significant divider change in many years). Also, there are new components for faster OS responses and virtualization. There is the advanced smart cache of 12 MB, which is also 24-way set associative. Add to that Intel’s Smart Memory Access, new SSE4 instructions and a super-shuffle engine. And finally, there is a new deep power down mode (C6), which and allows logic units to off-load their state into other logic units for a truer power-down.

Clock per clock, Penryns are faster, Intel said. It is unclear how much of this comes from the 50% larger L2 cache and how much comes from new architectures. Otellini, however, did indicate Penryns deliver 20% greater performance due to the 45nm hi-k/metal gate solution.


Nehalem

The big surprise of the day? Two live demonstrations of Nehalem in a physical machine. In the morning, we had a simple Nehalem system setup running Windows XP and a few applications. Otellini told us that Nehalems had been taped out for about a month, and what we saw was three week old "A0" silicon. He then held up a wafer showing wide, native quad-core Nehalems that were in use. Intel representatives made sure that they weren't using the AMD term “native quad-core” when describing Nehalem. Instead, they referred to it as "quad-core on a single die".

 

Another demonstration focused on the hyper-threading capability of Nehalems: Each core can support two threads. Intel showed off aa 16-thread machine, running two threads on eight cores in a dual-socket, quad-core configuration. The Nehalems were running 3D demos showing their I/O prowess with the GPUs.

Nehalem will adopt nearly everything AMD has been touting with K8 and Barcelona. The architecture will see on-die memory controllers, "QuickPath" communication between sockets (formerly called "CSI" and comparable to HyperTransport), including a multi-ported crossbar for communicating in 4-way configurations. Nehalem will also come with a significantly larger L2 cache. It will be a full 12 MB, arranged in dual-banks of 6 MB per dual-core. While Nehalem is a single-die quad-core design, its cache is still broken out into two half-size segments.

Nehalem is a behemoth at 820 million transistors, yet its die is only 107 mm2 in size. That’s 36 mm2 smaller than Penryn, which is expected to carry also 820 million transistors in its quad-core version (and 410 million in dual-core models.)

 

Other notes

During Otellini's keynote, the audience was given a glimpse into what qualities he believes "Intel is exceptional at".  These included three main points.  First was silicon process technology.  Otellini talked about their hafnium solution, which was discovered in 2003 and later made ready for production as Gen1 in Penryn.  He said no other competitor has developed a hafnium solution.  He also said that Intel has found a way to break down every wall they've encountered today.  Second was a related two-fold impact: Intel's core microarchitecture and to an ever increasing degree, chipsets.  This core microarchitecture claim does not refer specifically to the Core or Core 2 brand name, but rather the continued push of making better products.  And lastly is Intel's ability to scale and diffuse products into the marketplace.  Otellini said Intel's ability to create a product and then push it into the global markets with such volume is one key factor in their success.

Currently mobile CPUs account for nearly 40% of all desktop + mobile processors shipped by volume, according to Otellini.  That number is expected to reach the crossover point in 2009 where at least 50% of those CPUs shipped will be mobile.  That's up from 20% shipped in prior to Centrino's release, according to Intel.  This also includes some ultra-mobile PC initiatives, of which Intel will release details on day 2.

Otellini said in addition to being 100% lead-free and 100% halogen-free by the end of 2008, Intel will also begin reducing the size of its form factor by up to 60% in future products.  These reductions are not just for mobile products.  Otellini said Intel is looking to address future markets which will begin using x86 based processors, as well as smaller form factor servers in the future.  Some early Penryn and Nehalem blade designs were shown in one session which were about half the size of traditional blades.


Conclusion

The first day of IDF was an absolute bustle of activity. Of course, Intel is using IDF as an event to rally developers and convince the press that it has great technology in the works. And, just like in the previous 18 months, the company created an impressive display and from what we have seen on day one, Intel has a product line-up that AMD should be concerned of. Tick-tock appears to have been working flawlessly so far.

 

Read all stories from TG Daily in our IDF Fall 2007 Wrap-up. 

 

 

Will someone please explain to ignorant people...

Sep 19, 2007 02:27     

...like me, the efficiency of building rectangular shaped processors on circular wafers?!?!
strobe report abuse vote down vote up Votes: +0

Go read and stop being ignorant

Sep 19, 2007 03:04     

Wafers need to be circular because it allows for an even distribution of the processing chemicals. Imagine putting a fluid on something round and spinning it. The fluid will be distributed evenly. If you try the same thing on a square material you get build ups at the corners and an uneven processes. It does seem wasteful because you lose chips near the edges but thats how it has to be done.
Andrew report abuse vote down vote up Votes: +0

circular wafers

Sep 19, 2007 03:20     

The reason they are circular is because they use a lens to project the "light" onto chemicals to layer circuits onto the wafers layer by layer. If you will notice that your flashlight is round to project the light most efficiently, and I think the same applies to wafers. The light that they project onto the "mask" that produces the wafers would be most efficient if it is rounded in shape (straight line from the bulb, which covers the largest area in a circle).

na report abuse vote down vote up Votes: +0

wafer manufacturing

Sep 19, 2007 03:51     

is why its round. They grow the silicon wafers, then make them into chips. I suppose they could cut it before processing but why?
doc_simple report abuse vote down vote up Votes: +0

i think it's just for the presentation....

Sep 19, 2007 04:00     

other wise... it looks stupied.... anyone?!?!?
Dor report abuse vote down vote up Votes: +0

Re-Will someone please explain to ignorant people.

Sep 19, 2007 04:04     

Hi strobe, yeah... its the vice versa strategy of putting round pizza into square boxes :) I assume you know that silicon is extracted from the sand (which is Silicon Dioxide, SiO2)we find on the beach. What they do is, bring sand and put into a big oven and heat upto till it melt, at this stage... sand is very reddish and looks similar to lava. At this stage, a needle with a perfect Silicon edge is lowered to the oven and takes it back up slowly. What is happening is, crystallisation of Silicon. As in, hot silicon in the melting pot follows the structure in that needle and keeps on growing around it, then cool down. This is a bit slow process. It will go for sometime. Eventually, ones all the silicon adopts the structure of the needle (silicon), it looks like a silver cylinder. Later... this cylinder will be sliced for rest of the fabrication process. Ones they slice it, each slice may look similar to what you see in the pics of this article (of course, after photo-lithography of the micro-circuit), which is called Wafer Disc. What they mean by 300mm (12 inches) is...the diameter of this Wafer Disc. Hope this explnation helps. For more, very accurate details.. please google "Wafer Fabrication".

Alien report abuse vote down vote up Votes: +0

Circles and Rectangles

Sep 19, 2007 04:08     

Yes, please someone explain. I also don't get it.
Rob report abuse vote down vote up Votes: +0

Sounds like some exciting stuff.

Sep 19, 2007 04:09     

That's a good point, strobe. lol. I'm kinda wondering that myself. Hopefully someone can enlighten us on that. Intel has really been on the ball since the release of the Core2. All I gotta say is AMD needs to get their act together quick if they don't want to lose a big chunk of market share to Intel. I'm especially excited to see the performance of their Penryn Quad Core in notebook computers.
winkgood report abuse vote down vote up Votes: +0

yea i always wondered that too!

Sep 19, 2007 04:16     

i always wondered that too. I mean, look how many of the cores are circumcised (ok wrong word but still)
Wongsta report abuse vote down vote up Votes: +0

Re: Will someone please explain to ignorant people

Sep 19, 2007 04:23     

well, for one it is very shiny, and O_O yeah shiny, round is the only thing that makes shiny cooler so yeah XD
SirCabbage report abuse vote down vote up Votes: +0

square die on circular wafer......

Sep 19, 2007 04:56     

Even though the it is not the most optimal design for fitting a square on circular die; there are 2 main reasons: a) design simplification. for EDA tools which place the transistors, a well defined square boundary works best to use xy coordinate system b) wafer dicing, ie cutting the die from wafer. a square die means 2 dicing passes to get all the die's sorted. Any other configuration is sub-optimal. c) lastly and most importantly, why a square and not a rectangle. I am not entirely sure, but my gues would be SYMMETRY.

notsoignorant report abuse vote down vote up Votes: +0

circular wafer

Sep 19, 2007 05:03     

Interesting question... I looked into it on Wikipedia and found THIS: http://en.wikipedia.org/wiki/Czochralski_process The wafers are slices of a long cylinder of semiconductor material. And these cylinders are created via the Czochralski process, which requires constant rotation of the seed crystal. Hence, to make this procedure go smoothly, the shape needs to be circular.
knowitall report abuse vote down vote up Votes: +0

round hole...square peg

Sep 19, 2007 05:13     

The silicon wafer is cut from a continuously-formed crystal pulled from a pool of molten material... http://www.patentstorm.us/patents/6045610.html http://hardware.gotfrag.com/portal/story/31641/ ...so the source is a cylinder. In the photolithography process used to create the dies, the step-and-repeat process is essentially square. Little is given up when creating die images in the areas to be discarded, and the steppers are probably more accurate and have better yield away from their travel limits. Other die shapes are proposed, however in the following: http://www.freepatentsonline.com/5744856.html

livingston report abuse vote down vote up Votes: +0

round wafers

Sep 19, 2007 06:01     

that's the way they are grown in the lab I believe
SonOfAdam report abuse vote down vote up Votes: +0

Ok

Sep 19, 2007 06:02     

Easy, it's about the way the waffers are made, how Si is grown. There are several techniques, Czochralski's, Bridgman's and zone fusion technique. Basically, they all imply than the grown Si lingot has a round shape, much like a cilinder. After their got the cilinder, they simply slice it in waffers. It would be more difficult and consequently more expensive to change this processes to make them rectangular, at least that is my opinion.

siro report abuse vote down vote up Votes: +0

@Strobe

Sep 19, 2007 06:03     

thats the only way the silicon can be easily produced, check out this article: http://www.sudhian.com/index.p...s/show/619
Jay report abuse vote down vote up Votes: +0

It's simple really...

Sep 19, 2007 06:06     

A circular wafer has faster dies on it. This very scientific, you see.
Ryan report abuse vote down vote up Votes: +0

Good question - why round dies for square chips?

Sep 19, 2007 06:24     

I look forward to the answer too
Pat report abuse vote down vote up Votes: +0

ofcourse its round!

Sep 19, 2007 06:48     

its the donut, the most efficient and common shape in the universe. mmm.. donuts..
Homer report abuse vote down vote up Votes: +0

Nilla Waffers MMMMM!

Sep 19, 2007 07:24     

Damn, all this wafer talk has got me hungry! Damn I love me some Nilla Waffers with a cold glass of milk. Ya heard me? MMMMM, I'm starting to get stiffy thinkin' 'bout dem mo fuggers. Hold on, I be back in a minute...gotta go put the cat in the litter box - he trying to doonk in one of my damn shoe boxes again! He don't be trying to on purpiss, ya see, he got one eye and it's crosseyed. It be looking up all the time. He's really a sad kitty. I got him at the local pet store, he was half price 'cause of his fugged up eyes - I mean his eye. $2.00 dollars for a cat...shhhiiit, you can't beat that. Oh, he can't meow either...he just squeaks, but you know he's trying to. He loves doritos. I let him lick my fingers to get all that dorito cheese off of them. Hey, that gives me idea. I be back in a minute...I'm going to go try, I mean test something out real quick. Come 'ere kitty. I'm back, and hell yes, me and kitty gonna get along just fine together! Word up! Yeah, that Nehalem is a behemoth at 820 million transistors, yet its die is only 107 mm2 in size. DAAAAAAAAAAAAMN! Intel 'bout to fugg up AMD, fo sho! Well, I appreshyate everbodees time for reeding my postestes. Can't wait fo the new chips...but I guess I hafs to wait 'cause da shet ain't out yet! Oh dag! Well, piss out to everbody and I'd like to thank everyone for coming out tonight. Gonna go look at me sum more porn. Hell yeah...YA HEARD ME! Word! Peece Damn, this some good weed!

Testicle Taffy report abuse vote down vote up Votes: +0

Hi, just on a completely different topic than the c...

Sep 19, 2007 07:39     

Hi, just on a completely different topic than the circular wafers... what is the actual advantage in having all four cores on one die as opposed two dual-core dies stuck together? I'm trying to decide between buying (waiting for) an AMD phenom quadcore and one of these penryns. And would the advantage of being "native" quad-core be offset by the 45nm-ness of the Penryn? Any information would be great! Thanks!

Zero report abuse vote down vote up Votes: +0

Not only are they round....

Sep 19, 2007 07:57     

Not only are they round and many processors are clipped off the edges, but when a new architecture is involved usuially only a few 'squares' in the circle actuially work. The company my dad works for, which deal in high speed, low voltage, and extremely small amplifiers, only gets about 4-5 working processors per wafer on their first set of wafers. Granted these are smaller wafers, but by the end of a product line they reach almost 100%, which is still a few hundred processors per wafer compared to the origional 5. I would imagine the same is true for the even more complicated processors Intel is working on.

Caeden report abuse vote down vote up Votes: +0

Eng. Tech.

Sep 19, 2007 13:49     

Those aluminum (or gold) wires between dice are fragile, remember PlayStation 1 reliability (bad)? It was the wiring; both on one die is Better!
Ernest Blake report abuse vote down vote up Votes: +0

Its more cost effective to have a round waffer...

Sep 19, 2007 17:17     

The reason is the cost efficiency. It is more efficient to use the wafers as is(grown round) then to cut them into rectangles thus creating another step in production and more of a chance of bad dies. For every step there is, there is more of a chance for more of the dies to be bad. Also it creates more costs and make the price go up for us. They would also have less dies per waffer thus increasing the cost again. This method has been around since the firs processor and will stay for a long time since it offers the most processors per waffer and simplifies the process. And with 45nm there will be more processors per waffer and then more with 32nm. I for one cannot wait for Nahlem. Its going to be very promising.

Jimmy report abuse vote down vote up Votes: +0

I forgot....

Sep 19, 2007 17:23     

Nehalem looks amazing. I keep hearing about AMDs 4x4 solution and I know Intel would plan to go to it as well. I'm more excited then when I built my first PC. 08 seems like it will be very interesting...
Jimmy report abuse vote down vote up Votes: +0

Thanks, Alien...

Sep 19, 2007 22:51     

That was an excellent explanation! And thank you as well to the rest of the ummm..."well intentioned" explanations. They were great entertainment. Peace!
strobe report abuse vote down vote up Votes: +0

They make other people want to buy Intel CPUs, and ...

Sep 20, 2007 03:16     

They make other people want to buy Intel CPUs, and buy Intel shares, and sell AMD shares, and let's hope it is Samsung who gets AMD and not the Chinese rumor as I'd like this race for performance to go on, even pushed by mainstream AMD (can you call them high end anymore?). My next system upgrade will have a CPU with native quad core and hyper threading, and base speed of at least 3 GHz. Whoever gets there first, will get my money. Please don't flame me.

Virtualban report abuse vote down vote up Votes: +0

I thought X38 was to have 2 x 16x PCIE slots

Sep 21, 2007 00:51     

After all the Hype why would Intel produce a similar chipset to the 975 chipset. Is'nt it time to move to True dual PCIE 16x setup?? or has the presenter of the IDF got it wrong when he claimed that the X38 has only one 16x PCIE slot which can be piped
cybertraxx report abuse vote down vote up Votes: +0

Why circular wafers

Sep 21, 2007 07:27     

Its because when, molten silicon pulled out to make ingot....(1) cylinderical geometry is more symmetric than if the cuboid is pulled out. (2) The consequences of first is to provide omni-present desired crystal growth across the entire surface. (3) While, if you pull a cuboid, the zone near to edges might be different than the central zone in terms of crystal orientation [110].
rawat report abuse vote down vote up Votes: +0

Shame....

Sep 21, 2007 14:52     

Bye, bye AMD... I REALLY can't see them surviving as the editors of TGDaily digusted a week or so ago with Intel doing things like this and AMD's huge debt they will not be able to repay... Absolutely brilliant! Gotta love Intel! PS: I'm REALLY shocked that they already have a running Nehalem... it is amazing (Unlike AMD that always promise but cannot back their claims with tangible epuipment), but, oh yeah, Intel "stole 60% of the market with anti-competitive behaviour" (And not with plainy solid technology for the last few years (As according to AMD...)

Nixo report abuse vote down vote up Votes: +0

I make the chips....

Sep 21, 2007 18:52     

The wafer growing is accurate. But also, the comment up top by Andrew is accurate, the chemicals, acid etch, etc. are spun onto the wafers. This provides an even flow. And the non-used edges of the wafers are still used, in other ways....the are disected to look at how the layes are going together, so that any possible improvments can be made. The cool shiny is just a nice side benefit.
Chipmaker report abuse vote down vote up Votes: +0

A thought

Sep 22, 2007 04:16     

Although I do know why wafers are round, and of the chemical buildup on the edges of the wafer, I always wondered why don't they use it for some low volume smaller mm^2 products; for example special high end chipsets (Seeing that chipsets for example is always a manufacturing process behind, making low volume high end chipsets). With it being smaller mm^2, it might still be far enough from the edge of the wafer not to be affected by the chemical buildup on the edges. They can then ask premium $'s due to low volumes, high-end, next generation products which would sell: For example, to gamers, $'s is secondary, but performance isn't.

Nixo report abuse vote down vote up Votes: +0

Think some of your numbers are wrong...

Sep 24, 2007 00:46     

I do not believe Nehalem is scheduled to be 12 megs of cache, indeed Nehalem will have fewer transitors than Penryn due to reduced L2 cache. If you look at a Penryn die /- 60 of the die is cache, where as on Nehalem in it /- 35-40% of the die. The use of Quicklink reduces the need for such a large L2 cache.
The_Vorlon report abuse vote down vote up Votes: +0