by Anna Filatova
08/23/2005 | 11:46 PM
Sean M. Maloney, Executive Vice President and General Manager of the mobility group at Intel started the Mobility keynote this afternoon with the postulation of the new slogan: broadband at low cost to everyone.
In fact, this is not just a slogan, this is also a major goal. During the previous keynotes, Intel speakers have already mentioned that the mobile market is growing rapidly, and more and more devices go mobile. It is true not only for the notebook field, but also for the portables and handhelds market, which is evolving at enormous pace.
According to some studies, over 1.5 billion people today use mobile phones. And a pretty big number of these people use smart phones or tried to connect to the internet from their mobile phone at least once.
However, if we take a look at the chart above, we will see that practically no one has broadband. Surprising, isn’t it? And this is a global picture. You may think that it represents the emerging markets, which we have already discussed in the previous article today, but in fact it is not. The next slide clearly shows that the broadband is penetrating the US home environment very rapidly, so that it got 5x more popular today than it used to be in 2002.
Right now we can see that broadband is getting more and more important, although ensuring its availability to all the users that need it is a serious challenge. You see now that it not for nothing the slogan quoted above actually exists.
Computing is going mobile, and it is happening very quickly. The level of intelligence is growing, but it has to be ubiquitous and it has to be broadband.
There are two things that need to happen to keep mobile computing going. First of all there should be an open standard wireless technology, and secondly, Intel needs t be truly dedicated and even fanatical about the performance of their solutions and their power consumption level. These are the key features most Intel customers care about in the first place.
Right now Intel and partners are shipping solutions based on the existing Sonoma platform. The next mobile computing platform that is coming out real soon is going to be Napa platform – the first dual-core mobile supporting Intel Centrino technology.
This platform will be based on Intel’s dual-core mobile Yonah CPU. Since computers fundamentally imply the multi-tasking possibility, having an extra core will be extremely useful. The Intel mobile 945 Express chipset with integrated graphics core will offer better graphics performance. The Napa based notebooks will be much smaller in size and will boast better and more intelligent power management system. They will support such technologies as Enhanced Intel Deeper Sleep, Intel Dynamic Power Coordination and Intel Advanced Thermal Manager.
They even demonstrated two laptops based on the Napa platform that boasted all the above listed features. One of them was designed by Panasonic, and another one, the thin and wide one, by Samsung. The remarkable thing about the Panasonic solution is that it offers new battery design that is smaller, lighter and more efficient in terms of its life time. In particular, the new Panasonic solution offers about 30% improvement in the battery life.
In fact, Napa is moving into the market even faster than the predecessors. According to Intel, their partners have already developed over 220 different designs for different market segments that are already on the way and should get into mass production shortly.
Enterprise mobility is considered to be very important. Just like the home environment, the office also goes mobile. The enterprise has specific concerns that they like to address. To take care of these concerns and to make sure they can satisfy the needs of the mobile enterprise customers, Intel and Cisco established a mobility alliance.
The major questions this mobility alliance will be and already is addressing are network security and ease of use in the enterprise environment. Within this alliance they will be delivering enhanced VOIP (Voice Over IP) quality technology and optimal AP selection. According to the Cisco representatives, these will both be included into the business class wireless suite 2006 developed by this new alliance.
Mobile technologies have also moved quickly into the consumer market segment. To illustrate this statement Intel demonstrated two systems running office applications side by side. One of them was based on the dual-core mobile platform and we could clearly see the benefit of the dual-core. The benefits I am talking about implied not just a simple processing performance, but the ability of the system to handle the network.
One more example of the dual core advantages in the mobile platform was a picture rendering test in Cinebench 2005:
Image rendered on a system with dual-core processor (left) and single-core processor (right)
Speaking about mobile phones, we should point out first of all that intelligence is beginning to matter more inside phones. Of course, performance also matters. So, taking these two characteristics into account we can say that the so-called smart phones are becoming a very large category. And this is exactly what Intel claims:
Smart phone market is exactly the market segment Intel has been offering new microprocessors for. One of the latest innovations here was a new cell phone they demonstrated today - the world’s thinnest phone with a full keyboard:
Besides the ready-to-go cell phone model with a new Intel micro-processor inside, they also showed the next generation of processors for mobile phones, smart phones and PDAs. So far this processor is only available as a prototype:
This baby features H264 decode, supports full digital video and ensures great performance rate. So far it runs at 1248MHz, however, Intel is not quite sure yet about the real configurations or working frequencies the solution will support in the actual products. All in all, however, the latest technologies in the cell phone field indicate that the performance grew up by at least 4x, while the power consumption of these devices dropped down significantly.
Obviously, there is strong interest in video for portables and mobile devices, too. But despite the full digital video support most of these devices offer, you don’t want to decode and download videos, as it is not optimal and not really efficient from the performance and resources consumption point of view. The major aim is to enable broadcast video. And it looks like Intel has already found a solution. We could see the world’s first DVBH demo.
It could do 16 video channels straight offering clear image at 24fps only. This is going to be a global standard technology, and not something you have just in the US. Intel is already working with partners - CrownCastle and DiBcom - on solutions like that.
In conclusion to the Mobility keynote presentation Intel officials were talking about the need for an open wireless standard and the benefits of WiMAX in different application fields.
Since the first days these wireless technologies came into life, the cost per bit has dropped down significantly and keeps reducing. Even though Wi-Fi came into this world later than W-CDMA, it grew up much faster, and now occupies the dominating position.
A great example of how rapidly wireless is taking over is the fact that Intel established partnership with UPS and SBC in the USA in terms of installing hot spots and providing wireless access for UPS and SBC.
A pretty nig challenge here is lack of the signal reach, however, Intel is working hard on this one, too.
During the presentation, Intel tried to demonstrate not just the technology itself, but how people actually use it. They tried contacting people in four different remote parts of the world (China, Argentina, Canada and the Netherlands) using Intel wireless enabled solutions.
I personally liked the navy captain from Rotterdam, who as woken up in the middle of the night and looked real funny, especially when he was trying to talk about WiMAX advantages while it seemed he barely knew what WiMAX really is :)
Well, I believe you would like to hear some real processor news now and to see some real processors. Let’s do it! At the session devoted to the dual-core architectures, Intel said and showed quite a few exciting things. But let’s start from the very beginning: I am sure revising a few key things will only do us good.
It is two independent execution cores in one single processor. There are two different design configurations driven by different efficiencies and optimizations. The first one is monolithic implementation, i.e. when we have one physical piece of silicon. The second one is multi-chip, that is when there are two pieces of silicon. As you can see from a picture below, different production methods are used for each of them:
Of course, there are advantages in each production method. Monolithic chips are easier to cut from the wafer and put into the package as a single piece of silicon. With multi-chip design it is more challenging to put two chips together into a single die, but o the other hand it gives us more flexibility in combining chips. That is from the production or technology standpoint.
For the end-user there is really no difference. There is no way to tell by looking at the package what is inside.
Right now, as you know, there are two dual-core processors available. These are Pentium D and Pentium Extreme Edition. Here is a brief summary of their features, just to refresh your memory :)
This is exactly what the roadmap shows:
Blue – single core, Green – multi-core (>=2), Brown – multi-core (>=4)
However, this roadmap also shows a lot of other new products, which we are going to talk more about right now.
Next generation of the Pentium D processors is going to be based on 65nm core aka Presler. The core will consist of two separate pieces of silicon and will feature twice as large L2 cache. All the other features will remain the same as the previous generation Pentium D processor manufactured with 90nm technology.
The 65nm Pentium D will be available in Q1 2006.
After that there is going to be Conroe processor, which we have already talked about in the first IDF article earlier this afternoon.
In the single-core desktop segment we should soon see a new 65nm Pentium 4 processor aka Cedar Mill. Look how it compares with the today’s 90nm Pentium 4:
Mobile dual-core processor manufactured with 65nm technology aka Yonah, which I have already talked about during the Napa platform discussion, will feature Intel Smart cache. Smart cache means shared cache between the two cores. Since we have two cores and we have a single bus, there will also be shared single bus interface. This way both cores can share the same copy of data from the L2 cache. Besides that, the L2 and data cache unit feature improved pre-fetches, i.e. we can do pre-fetches on the per-thread basis, thus ensuring better bus utilization. It also has bandwidth adaptation buffer, i.e. each core takes 4 cycles to adapt.
The difference between independent and shared caches is the following. In case of independent caches the data is transferred from one core to another via the FSB. In case of shared caches the data is transferred directly between the caches, you avoid the bus traffic and synchronization time to get on the bus. This is important for multi-processing systems, because this way you reduce the number of bus cycles involved.
As you can see from the roadmap this solution is also coming in Q1 2006.
For the DP (dual-processor) server market Intel will have DP Dempsey based on 65nm technology. It will have a 2MB L2 cache, ad can run 4 threads simultaneously, as supports Hyper-Threading. The anticipated schedule for this processor is also Q1 2006.
DP and MP Paxville will be made with 90nm process. They will have single bus interface running at 667MHz although there will be independent caches, not shared ones. They will also support Intel Virtualization technology.
These CPUs are coming out in H2 2006.
DP Sossaman will be based on the Yonah core and will be targeted at the server applications, where higher density smaller form-factors are of more importance. This CPU will feature shared L2 cache just like Yonah. According to the roadmap, we can expect them in H2 2006.
Multi-processor Tulsa manufactured with 65nm process will be equipped with 16MB large L3 cache, and will be used in those platforms where large caches ensure significant performance improvement.
For the Itanium 2 family there is Montecito, a CPU composed of 1.7 billion transistors. Each core has its own 1MB L2 and 12MB L3 cache, that makes the total amount of cache memory equal to 26MB.
This solution should be available this year already.
I will continue reporting the latest news from IDF tomorrow, so stay tuned for more news!