Sign in with
Sign up | Sign in

Intel Developer Forum, Day One

Intel Developer Forum, Day One
By

The first day of Intel's Developer Forum saw discussion of the company's new CPU microarchitecture, code-named Sandy Bridge and manufactured at 32 nm. But the most important changes might be how the company develops future products.

It’s September in San Francisco, so it must be time for another Intel Developer Forum.

As is usual in these Intel technology love fests, we’re seeing new CPU rollouts, updates on memory roadmaps, new usage models, and the Showcase, the mini-tradeshow where you can find products and technologies ranging from cloud-based, distributed raytracing to tiny handheld computers.

The big chip news at this year’s  IDF is the rollout of Intel’s new Sandy Bridge architecture. In case you haven’t been following the rumors, Sandy Bridge is the latest “tick” in Intel’s tick-tock model of product development. The “tick” is the creation of a new microarchitecture, always built on an existing manufacturing process. The “tock” is pushing out enhanced versions of existing products, but rolling them out on newer (and ever-smaller) manufacturing processes.

But before we dive into the details of Sandy Bridge, it’s worth a brief visit to IDF, Day Zero.

Experience-Based Product Design? From Intel?

Dr. Genevieve Bell is one of the few Intel Fellows who is a woman. She’s also the only Intel Fellow with a doctorate in anthropology rather than an engineering degree.

Anthropologists study culture--how people interact, live in their environments, and socialize. Dr. Bell is focused on understanding why people love their gadgets and what it is about technology experiences that they crave on a daily basis. That, in turn, allows her and her team to help shape what new user experiences and usage models may become popular a few years from now.

Understanding what people really want to do, as opposed to what they have to endure to do what they want to do, is critical for today’s tech companies. As technology products become more like commodities ($300 laptops, $500 big screen HDTVs), making the right guesses about what products to build becomes as much a part of a company’s core competency as the ability to clone a $2 billion manufacturing plant. Guess wrong, and those big process plants go idle. Guess right, and you rake in the big bucks.

Intel thinks the idea of understanding future user experiences is important enough that it has funded an entire arm of its research organization to this, known as “Interactions and Experiences Research.” Split into design and technology elements, and headed by Dr. Bell, the idea is to understand how users worldwide experience their technology, what they love about it, and what frustrates them.

The key image Bell used in her presentation was an ad from the 1959 release of the Bell Princess Phone.

Nothing about this  ad even mentions using the Princess Phone to actually make phone calls.Nothing about this ad even mentions using the Princess Phone to actually make phone calls.

"It’s little, it’s lovely, it lights.”

But does it make phone calls?

Bell used this as an example of a product that was highly technological, but the marketing was decidedly non-technical. But the Princess phone brought the telephone out of the front hall and into bedrooms, made buyers want to have more than one phone line in the house, and, in an eerily prescient version of today’s smartphones, was sold as a service. You could only rent a Princess phone and had to install and pay a monthly charge for a second phone line. But it revolutionized communication in its day.

Intel’s Experiences and Interaction group is looking for usage models that might represent the next, similar leap. Whether they’ll succeed inside Intel’s engineering results-oriented culture is an open question, but it’s an interesting perspective on how one engineering company is trying to reshape how it develops products for the future.

Display 27 Comments.
This thread is closed for comments
  • 5 Hide
    dragon5677 , September 14, 2010 9:12 AM
    hope the day comes soon when sandy bridge is released with a surprisingly low price.
  • 4 Hide
    letsgetsteve , September 14, 2010 9:15 AM
    really not a fan of the pin change in mobo's but I guess I'm interested
  • -6 Hide
    kashifme21 , September 14, 2010 10:44 AM
    No point of having such powerful hardware, when 95% of current games can be run on a dual core cpu and an 8800gtx gpu @ max settings.

    Intel, Nvidia and AMD should rather pursue game developers to make use of current hardware instead of throwing newer hardware which most likely will be unused.

    Sandy Bridge aint needed for Web surfing or using Word pad any 8yr old CPU is enough for that. Bring on the games or i am not interested in any more upgrades.
  • 1 Hide
    Tamz_msc , September 14, 2010 11:42 AM
    Quote:
    Overall, Sandy Bridge looks to be a solid mainstream offering. Performance enthusiasts should note that LGA 1366 is not going away, and Westmere-based hexa-core CPUs will continue to be offered. Intel even suggested that future LGA 1366 offerings may become available, but wasn’t prepared to make any definitive announcements.

    Unless they are offered at reasonable prices, I see no point in people investing in LGA 1366.
  • 3 Hide
    jfby , September 14, 2010 11:58 AM
    I don't think it's so much that people will want to invest in LGA 1366 but rather Intel will not support anything else.

    The argument to have only 16 lanes available doesn't seem legit; 'mainstream' gamers aren't going to buy a 1366 system at the moment.

    At least people who have bought a 1366 have a potential for an upgrade 2-3 years from now, though I'm sure the better choice will just be a brand new system, again.
  • 1 Hide
    pjmelect , September 14, 2010 12:42 PM
    Intel graphics even at twice the speed, no thanks. AMD could take a big lead over Intel if they integrate ATI graphics in their CPU's.
  • 6 Hide
    ScrewySqrl , September 14, 2010 1:25 PM
    dragon5677hope the day comes soon when sandy bridge is released with a surprisingly low price.


    Low priced? from Intel? You must be joking
  • 5 Hide
    atdhe , September 14, 2010 1:46 PM
    kashifme21No point of having such powerful hardware, when 95% of current games can be run on a dual core cpu and an 8800gtx gpu @ max settings.Intel, Nvidia and AMD should rather pursue game developers to make use of current hardware instead of throwing newer hardware which most likely will be unused.Sandy Bridge aint needed for Web surfing or using Word pad any 8yr old CPU is enough for that. Bring on the games or i am not interested in any more upgrades.


    You know, a computer can be used for more than just gaming, surfing or Word ;) 
  • 0 Hide
    Trueno07 , September 14, 2010 2:33 PM
    more info here:

    http://www.anandtech.com/show/3922/intels-sandy-bridge-architecture-exposed
  • 2 Hide
    wolfram23 , September 14, 2010 2:35 PM
    Glad to be missing this generation, as I just got an i5 750 earlier this year. Won't need to upgrade until at the earliest Ivy Bridge.

    I don't really like how so much die space gets wasted on their P67 platform. Hopefully they make special CPUS that are a little beefier and no on-die graphics for that chipset... tho it's unlikely.
  • -1 Hide
    jdamon113 , September 14, 2010 3:17 PM
    This direction is pointed at the business model, main stream deaktops.
    Yes they sound quite powerfull and will do well, Not sure why intel is leaving the high end market up in arms, I think until intel buys nvidia they will continue to be like this. Until then, we wait for X68 to come through or stick with aging technolgy in the X58 and the chips is supports.
    Today, intel all your managed to do was make me look forward to AMD's Next offerings. I still have run a 775 with an extrime chip. Its still fast and can stand up to a 1366 / 920 . From what I read here and other articals, my next platform will be AMD bulldozer.
  • -6 Hide
    f-14 , September 14, 2010 3:31 PM
    here's an idea to throw at intel, IBM had a 1GHZ cpu before 2000, nothing could cool it adequately at the time so they burned out right away P4 3.8 GHZ was out long before 2010, tell them to design for 4GHZ on their new high end products and their low end never dropping below 3GHZ. if they priced the 4GHZ around $300-$400 and the 3GHZ $100-200 they would sell about 5 times more product. i guess that just makes too much sense and they could gut their marketing department if they did that.
  • 0 Hide
    wa1 , September 14, 2010 5:07 PM
    A new tech = always for gaming and 3D works.
    You still can run ms.word (doc) on Pent.4 PC...
    Sigh...
  • 7 Hide
    nexus9113 , September 14, 2010 5:24 PM
    f-14here's an idea to throw at intel, IBM had a 1GHZ cpu before 2000, nothing could cool it adequately at the time so they burned out right away P4 3.8 GHZ was out long before 2010, tell them to design for 4GHZ on their new high end products and their low end never dropping below 3GHZ. if they priced the 4GHZ around $300-$400 and the 3GHZ $100-200 they would sell about 5 times more product. i guess that just makes too much sense and they could gut their marketing department if they did that.


    You clearly don't understand CPU design and engineering and haven't heard about the debunked "clockspeed myth".

    Clock speeds don't really mean much anymore, it's primarily about the architecture and how the CPU handles data. Hence why a 2.2GHz iX Core can outperform a Quadcore Core2 that might be clocked higher.
  • -5 Hide
    truerock , September 14, 2010 6:42 PM
    I think what I am hearing from Intel is:
    1. AMD isn't really pushing Intel competively - so Intel can be lazy about increasing the speed of its CPUs
    2. Because Intel doesn't need to develop faster CPUs they can try to rip off their customers with obsolete CPUs that incorporate non-CPU technology, i.e. video/graphics

    Intel has always wanted to avoid creating faster CPUs and put other technologies on their CPU chips. Granted, it was AMD that started the "a slower CPU can be better" myth. I understand that multi-CPU and enhanced memory architectures provide better throughput - but, nothing beats a faster CPU clock at most tasks. We have been stuck at less that 4 GHz for 7 years! That is insane. Intel has no viable competetor and is using its monopoly to screw everyone.
  • -8 Hide
    truerock , September 14, 2010 7:01 PM
    I bought my 3.4 GHz Pentium 4 CPU on eBay for $70. This year you can pay $1,200 for a 3.4 GHz Sandy Bridge CPU. I am not impressed. There is something very, very wrong with Intel.
  • 2 Hide
    onecallednick , September 14, 2010 9:39 PM
    What about overclocking? I hear that sandy bridge is going to be tough to OC what with the "system agent" controlling everything all at once.
  • 2 Hide
    warmon6 , September 15, 2010 12:38 AM
    truerockI think what I am hearing from Intel is:1. AMD isn't really pushing Intel competively - so Intel can be lazy about increasing the speed of its CPUs2. Because Intel doesn't need to develop faster CPUs they can try to rip off their customers with obsolete CPUs that incorporate non-CPU technology, i.e. video/graphicsIntel has always wanted to avoid creating faster CPUs and put other technologies on their CPU chips. Granted, it was AMD that started the "a slower CPU can be better" myth. I understand that multi-CPU and enhanced memory architectures provide better throughput - but, nothing beats a faster CPU clock at most tasks. We have been stuck at less that 4 GHz for 7 years! That is insane. Intel has no viable competetor and is using its monopoly to screw everyone.


    There is also a thing called IPC (instruction per clock) ...... The primary reason why a Athlon single core Athlon 64 at 2.2 GHz matched or beat a 3.2GHz HT p4 on most tasks back in the day......

    AMD never started a Myth about MHz speed. Intel started that the MYTH about MHz/GHz with there P4 when they notice that AMD had something that can compete and exceed there own product. Thus hoping that the public thats not educated in computers would see High GHz and would by there product. AMD countered that MYTH by stating that higher MHz does not equal higher performance.

    After a year or more of trying to push this myth, intel droped the rest of the Netburst line (after prescott, there was suppose to be Tejas and Nehalem) and with the core line. Which is much faster than the P4 even with only running 1 core on the core 2 line.

    Now for the Performance between Core i5 Sandy Bridge vs Core i5 Nehalem, I guess you haven't seen them as other sites like Anandtech has bench marks that are single cores and show this.

    http://www.anandtech.com/show/3871/the-sandy-bridge-preview-three-wins-in-a-row/9

    Core i5 2400 @ 3.1 without turbo boost on, beatting core i7 880 AND core i7 980x is impressive to me as it sounds those had turbo on.



    GHz Mean nothing these days. P4's (even though im typing on one as we speak) proved that. 3.8GHz (which was a p4) is the highest shipped cpu and that that was due to heat and amount of power needed to keep thats High GHz. Same issue even today. Even with IBM cpus for severs, they dont reach 3 ghz. (except for an experimental overclocked 500GHz chip thats not made on silicon but thats was 4 years ago.) IPC, better cache system, wider ram bandwidth, will be better for performance on single threaded stuff than higher ghz.
  • 1 Hide
    theoutbound , September 15, 2010 1:10 AM
    I think this is going to be the platform of choice for laptops going forward. Decent integrated graphics and low power draws have me excited for SB based laptops.
  • 1 Hide
    Proximon , September 15, 2010 1:38 AM
    I just remember how those dial telephones made my finger hurt. :) 
Display more comments