US Robotics cloud the mind of reporters: US Robotics plans to ship Texas Instruments-based PBCC encoding access points and cards in June. Unfortunately, they clouded the mind of the reporter on this piece who makes it sound as if poor old TI was ostracized by the IEEE.
The article states: For now, TI is the only supporter of the 22Mbps PBCC-22 technology, which it had hoped to get adopted into yet another high-speed WLAN standard under discussion at the IEEE, 802.11g. However, the standards committee voted to use a rival modulation scheme, OFDM (orthogonal frequency division multiplexing) proposed by Intersil Corp. of Irvine, California, according to TI and Intersil. TI's PBCC-22 remains as an optional part of the still-unfinished standard, which means products will not have to support it.
This is incorrect in part. The 802.11g task group approved the 802.11a 54 Mbps form of OFDM, which wasn't designed for the 2.4 GHz space. No one I've spoken to believes that it will deliver that kind of raw bandwidth. If it did, the whole issue of other encodings would be moot. A separate form of OFDM, called OFDM-CCK in some of the standards filings, was actually designed to run at 22 Mbps. That form plus TI's PBCC-CCK were approved as optional encodings. Intersil, the backer of OFDM, will release chipsets later this year that support 802.11g's draft form.
I asked Intersil and Texas Instruments's product folks at the 802.11 Planet conference last November whether we'd see devices that supported both PBCC and OFDM-CCK, and they more or less replied that both sides thought the other's technology was the wrong approach and no sensible manufacturer would put both in one unit because they would see the light on how the other company's technology was weaker.
The characterization that TI lost is just not right. TI won an important concession, as did the industry. Both TI and Intersil can push their technology into the marketplace, which will help decide the winner. You can imagine the series of head-to-head tests between the chipsets in different implementations. TI has always claimed that PBCC does a better job in extending range, and that may be true. Of course, Intersil claims that OFDM does the same thing.
Some nice clarity from News.com: the bulldog Ben Charny, who seems to run these stories down with great tenacity, uncovers the real story. WECA, the group that certifies equipment to carry the Wi-Fi seal, won't specifically test for PBCC interoperability. They only test for mandatory encodings, they say.
Intersil and Silicon Wave announced Bluetooth/Wi-Fi chipset last week at WinHEC: another portent of the future. I'm wondering whether we'll have a single mini-PCI or PC Card that supports GPRS, 802.11(a/b/g), and Bluetooth? It's not outside the realm of physics, especially of all the standards can use the same MAC.
Intel, Loaves, and Fishes (and Blades and Razors): Intel gets all Biblical, like, and talks about their desire to have hot spots everywhere. Intel makes the chips (or will be soon), and wants to encourage public space providers to push it out. This, of course, helps their market in a variety of ways: newer devices will push the sales of newer machines; new motherboards with on-board Wi-Fi or wireless push new sales. Intel gets to diversify, too, and sell more wireless gear, unlatching their gravy train (as they have been increasingly working on the last several years) from the pure processor market. Ultimately, Intel might see a fraction of their revenue from CPUs.