This week is the annual International Solid-State Circuits Conference (ISSCC) where chip companies from all walks of life present papers documenting everything from shipping architectures to future research projects. Intel has always had a large presence at the conference and this year is no different. I'm still trying to get my hands on some of the actual papers being presented but Intel invited some press to a pre-brief on the high level announcements from the conference. 

One such announcement is a test SoC called Rosepoint. It's a 32nm dual-core Atom SoC with an integrated WiFi transceiver. Despite the high levels of integration we see in smartphone SoCs, WiFi is typically serviced by an external combo chip that integrates WiFi and Bluetooth among other radio technologies. Rosepoint brings the WiFi functionality on-die. The name of the game in the mobile SoC space is integration, making Rosepoint a research project with significant real world implications.

Integration is nothing new of course. AMD, Intel, NVIDIA, Qualcomm, TI and all of the others playing in the SoC space have been slowly integrating more functionality on-die over the past decade. Intel claims the difficulty in bringing WiFi on-die is mitigating interference between the RF transceiver and the rest of the SoC. The details of Rosepoint's architecture and how Intel was able to reliably integrate the two are likely described in the ISSCC paper. If I can get my hands on it I'll see about updating this post.

Rosepoint is important because of Intel's dillema as it enters the smartphone SoC space. Most high-end smartphone SoCs sell in the $14 - $25 range, a significant reduction compared to the $50 - $1000 Intel is used to getting for CPUs. Even if you look at smartphone-sized x86 CPUs, Intel can typically get somewhere between $50 - $100 for the CPU. Then add another $20 - $30 for the chipset and margins start looking very nice. Intel can't guarantee > 60% margins selling ~$20 smartphone SoCs. At the same time, Intel based smartphones wouldn't sell very well if they were significantly more expensive than the competition. This puts Intel in a difficult position: settle for lower margins (and upset wallstreet) or figure out a way to offer more value by integrating other parts of the bill of materials. 

Offering and integrating radios where possible is clearly one step, although we'll likely see integrated cellular baseband before we see on-die WiFi. Intel's recent restructuring left the new mobile & comms group with a mandate to deliver an ultra low-power WiFi solution that could work in a smartphone. The first Intel based WiFi in smartphones will begin as a discrete chip but it's clear that integration is on Intel's mind.  The other options for Intel to bring some of that precious BOM in house is to offer reference platforms and/or use software as a differentiator.

In the early days Intel would just sell a CPU and rely on third parties for the rest of the chips on the motherboard. Then came Centrino and the new platform-centric Intel. Expect to see a similar effort in smartphones.

Comments Locked


View All Comments

  • Southernsharky - Monday, February 20, 2012 - link

    It would be nice if they turned the Atom into something other than a steaming pile of [rap before they added more features to it.
  • thrawn3 - Monday, February 20, 2012 - link

    In a smartphone with the much simpler OS the current Atom design should be quite nice. I still use a single core Atom netbook. Sure it is slow but speed isn't everything and ARM proves that.
  • retrospooty - Monday, February 20, 2012 - link

    "It would be nice if they turned the Atom into something other than a steaming pile of [rap before they added more features to it."

    I dunno man, they are about 1 process shrink (to 22nm) and a few power improvements from a dominating chip. Don't count Intel out. They have already mentioned they are working to integrate cellular radio, and now wifi into the SOC. That lowers total power significantly if its all in one. Throw that on a tablet and add Windows 8 with x86 enterprise apps and its a total dominator in the business sector. I would buy one as a consumer so I can run my real apps on it. I am sure most people would as well.
  • extide - Friday, March 2, 2012 - link

    Plus the 22-nm atom should bring us a next gen, out of order, atom core. I have no doubt Intel is going to put a LOT of resources behind getting atom into a competitive position. It's certainly going to be fun to watch!
  • name99 - Monday, February 20, 2012 - link

    Anand, are you in a position to ask Intel what the capabilities of this WiFi part are?

    If it is just another low-power WiFi part, like every other one we have seen (1x1:1, 20MHz, no short guard interval) then Intel is trying to compete purely on price, and while that may get them into the Noname Brand Crapulon 350X, it won't be enough to get them into the upper tier smartphones and tablets.

    Much more interesting is if this part (and related paraphernalia like the driver) can do more --- ideally a lot more --- than current low-power WiFi parts. There are many possibilities here.

    At the PHY level
    - they could utilize short guard intervals (something to boast about, but maybe not worth the power consumption)
    - they could use a more aggressive soft-value, iterative FEC decoding scheme (gets more range, allows for higher speeds at mid-ranges, but again burns power and maybe not worth it)
    - could provide for a 2x2:1 architecture --- now we're talking. This would incorporate two antennas even though only one spatial stream is supported; the system can use switch and stay (very low power) or maximum ratio combining (still low power) for diversity. Hook that up to supporting STC and you have a system that should have much greater range than the current of smartphones and tablets.

    At the MAC level
    - they could use intelligence and statistics (what is the cost of the next data frame, and the probability that there will be a hidden node collision) to decide whether to use CTS/RTS rather than a dumb strategy like a blind "always use if the data size is greater than x bytes".
    - they could use aggressive aggregation (at both MAC and PHY levels) along with intelligent fragmentation (to keep the fragments corrupted from noise at a pre-determined level) and boost the goodput from the usual 50% or so of PHY rate to more like 80%+ of PHY rate.

    These MAC improvements have the great advantage that they don't actually burn any more power. They do require more working RAM for the WiFi system, but it's 2012, not 1987, and we have multiple MB of ram on chips in the form of extremely fast cache --- we should be able to allow the IO system a decent amount of working memory which doesn't even need to be especially high speed.
  • deeceefar2 - Monday, February 20, 2012 - link

    One thing that you left out is software. Intel is also a pretty good software developer when it comes to drivers and their related attachments.

    I suspect this will also be something that Intel uses to differentiate itself from the other WiFi vendors. Since driver related development is difficult, expensive, and goes largely unseen unless it's breaking, I suspect this is very attractive to device makers. They can focus on their development efforts on differentiation and not invest it into driver maintenance which is a throw away if they choose another vendor in the future.
  • damianrobertjones - Monday, February 20, 2012 - link

    ...Need this now. Both are stupidly slow with that darn single core
  • Visual - Tuesday, February 21, 2012 - link

    Nah, just a CPU bump will still not save them. They need a GPU to be worth it to me. I need something to replace a tm2.
  • Hector2 - Tuesday, February 21, 2012 - link

    You two didn't read Anand's Medfied review last month ?

    "Intel is able to deliver performance better than a dual-core Cortex A9 from a single HT enabled core"

Log in

Don't have an account? Sign up now