Mixed Signals

Our first Google Wave invites have arrived, and we've spent the first hours of the day getting our feet wet in the surf (and you could join us: read on). "IM on speed" was the reaction from the newsdesk, as we watched mulitple people around the world typing responses to a discussion thread in real time time. There are robots. There is a video from 'Dr Wave', who wears a white coat and waves a lot on screen. There is a very heavy scent of alpha in the air.

We'll get back with more of our experiences as we get to grips with the salty software - so far, we know that Chrome, Firefox and IE 7/8 seem to work OK, but Opera doesn't.

We also have two Google Wave invite nominations to give away. These aren't invites - you don't get those until Google processes the nominations: "we've got a lot of stamps to lick" as Google says when excusing the delay - but they're the next best thing.

What we don't have is a good single sentence answer to the reasonable question "But what IS Google Wave?". This is where you can help - and win.

Leave a talkback to this blog post with your best one sentence description, and we'll despatch our nominations to the best two received by 5:30pm, UK time, Monday. To boost your chances, come up with the best use you can think of for Wave and include it in your post - but we'll be looking for that elusive capsule definition above everything else.

Usual rules: editor's decision is final, we'll be in contact via your community ID on Tuesday, and feel free to make multiple entries - although anything spammy will result in instant disqualification.

Remember our Floridian pals xG, with their ground-breaking voice over IP wireless system? Remember that they said recently that anyone who was a journalist or analyst could pop down to see them and check out their test installation?

RIch Tehrani, "VoIP industry expert, visionary, author and columnist", and owner of TMCnet took xG up on its offer. He was driven around Fort Lauderdale between a set of six base stations, talking into a phone hooked up to a laptop, and reports that the voice quality was fine. After a while, the phone overheated and shut down.

Which as first demonstrations go, is fair enough. For fans of xG, though, that's all it showed. It doesn't qualify the headline on the piece that "xG Technology xMax Works As Advertised" (remember the thousand square kilometers from a single base station? the "fastest roll-out in history" promised in 2006? Between its launch in 1991, and 1993, GSM put on a million customers) nor to justify any comparisons with LTE and WiMAX. Nor, to be honest, with bog standard 1991 vintage GSM.

All it demonstrates that you can have a conversation within a few square miles on one handset, if you surround it with six base stations. There was a hand-over demonstrated, but as you can see from the various videos connected with the article, not much more. It would be brutal of me to say I could demonstrate much the same with a couple of 5 watt VHF/UHF walkie-talkies: brutal, but quite accurate.

The article, unfortunately, doesn't go into technicalities. Tehrani didn't ask, and xG didn't offer to say, whether the system used xG's proprietary xG Flash Signal single-cycle modulation.

There were some interesting snippets, such as each base station being able to handle 252 handsets, 14 in each of 18 channels. (GSM has around 500 channels, depending on bands available, each supporting up to 16 calls).

The handsets are able to change channel, base station or whatever, thirty three times a second - which is a very high rate, much higher than other systems, and one that may be dictated by the need to keep dodging interference in the 900 MHz unlicensed band that xG is designed to use - and commercialise on, because operators who use xG's equipment don't pay any auction or licence fees. (There is no such band in Europe, by the way, nor any equivalent that could support xG).

900MHz is full of interference - baby alarms, cordless phones, video links and so on. How call density will vary with interference hasn't been stated, as being able to hop to another channel within 30 milliseconds isn't much good if all the other channels are also being interfered with, or have conversations on. Wiith just one handset on the system, then as long as one of eighteen frequencies was clear the demonstration would have worked. And whether interference in one channel can knock out all 14 calls there - again, not specified.

There was no explanation given for the handset getting so hot it shut down - or rather, "it was charging" and "there's no power management" didn't really add up. Early stuff does get hot; this handset, though is their second and has been in development for years.

One of the few technical details to have escaped from xG to date is that there's a very high peak power transmitted from the handsets, which would cause a lot of heat and not a lot of battery life. Without any more information, it's not clear - or indeed even guessable - how battery life will pan out, or how the system would work under a more moderate power regime.

The published specifications for the base station suggest that it transmits one watt per channel for a total of eighteen watts, but consumes 1350 watts in doing so. That's interesting efficiency.

The company has added a flip antenna to their TX-60 handset, which makes sense - 900 MHz has a reasonably long wavelength of around 35 cm, and many early GSM handsets which worked on the same frequencies had external antennas. Although GSM outside America still uses 900MHz, modern handsets have dispensed with external antennas due to improvements in receiver circuitry and overall system design.

Other things not mentioned? Security, authentication, billing systems, bandwidth for data, international use or maximum cell coverage area (one cell was quoted as "6 to 8 miles, bigger than their GSM equipment", although GSM is specified to work at up to 22 miles).

If we get answers to that lot, then we may be able to say what xG is really up to and whether it has a commercial chance. The company may have to move fast, though; its latest half-yearly financial report (PDF) does not bode well for the long term.

So, there I was in San Francisco, listening to Intel tell me all about the bright, spangley future of mobile. The keynotes were full of standard issue happy videos of happy young people doing happy young things with mobile devices, socially networking with video and the great taste of Internet.

Wouldn't disbelieve a word of it, guv, except for one thing. I was sitting at the very epicentre of networked wireless technology, and I couldn't even send an instant message from my phone.

Technically, there was no problem. My phone, the constantly pleasing Android-based G1, has no difficulty in speaking American airwaves, and it is stuffed to the gills with software that tweets, chatters, displays moving pictures, geolocates and so on and so forth. The problem was me: for some reason, I dislike being held to ransom and having my money extorted. And with T-Mobile wanting to charge me £7.50 a megabyte for data, there is no other way to describe data roaming charges.

Seven pounds fifty. A megabyte. My iGoogle home page? 300Kbytes? That'll be two pounds fifty. Just to see my own home page. A fiver to check ZDNet UK's home page (naughty of us to have pictures, I know).

It gets worse, of course, when you consider what the extra costs to T-Mobile are for data roaming. Using the Internet, of course, the local operator just has to hook me up to its local connection point, which it can economically do for its own customers for the usual amounts. In my case, that local operator was T-Mobile – who will also sell me hotspot access via Wi-Fi in San Francisco for eight dollars a day, no particular data limit mentioned.

Extortion isn't too strong a word, especially when you consider how competition has been so carefully excluded from the equation. I tried to do the smart thing and buy an American pay-as-you-go data-only SIM, but my handset is locked to TMob and the local TMob shop assured me that such a thing was "impossible".

I know that many, if not most, companies now forbid data roaming for their employees, because the charges can easily run into thousands of pounds for even moderate use. I know that none of the UK people I talked to at IDF were using data roaming – although some had unlocked handsets that allowed the "impossible" task of using local SIMs for data. And these are the people who are directly involved in creating the exciting new world of mobile which, we are assured, will push technology for the people into ever greater heights of electrowonder.

It doesn't matter how good and cheap the technology is, if you can't afford to use it because of extortion. And it is doubly painful when you're abroad, which is exactly when you most need all those wonderful online information tools.

I have been at parties with mobile phone executives who openly laugh about the rip-off of data roaming charges. That left a very bad taste. What baffles me is whether they've ever made the calculation of how much money they don't make – or are there so many thousand-pound bills floating around that they don't want to bother to collect a hundred ten-pound bills instead?

But the current state of affairs is wrong – criminally, caustically, catastrophically wrong. It is poisoning the hopes of the mobile Internet, it is showing up the lack of international regulation (three cheers here for the EU, which is slowly enforcing sanity on the robber cartels), it is massively stupid in a way only telcos can be massively stupid. It is an enormous insult to customers, developers and manufacturers. It spits in the face of the future. If anyone can suggest a way to break this conspiracy down, I'm more than ready to hear it.

And if anyone from T-Mobile would like to explain to me why it is right to charge me five pounds to look at a single web page, I would love to hear that too. Bring it on.

One of the major themes at this year's IDF in San Francisco is the use and abuse of power. As Intel delicately drip-feeds information about its chips into the world, the picture builds up: these new devices are far cleverer about managing themselves, watts and all, than any before them, and that cleverness extends into and out of the chips themselves.

Take Clarksfield, the Core i7 Mobile processor. Four cores running at 2GHz, taking about ten watts apiece, and if you're running software that uses all four cores equally then that's how it's going to be. Use fewer cores, and giant transistors inside the chip forcibly disconnect the idlers from the power bus. Run software that wants a single thread but uses it a lot, and the connected core is given more volts and a faster clock frequency until it's using a great deal more than the ten watt standard allocation.

That sounds simple enough, but there are lots of subtleties. The operating system may want the hardware to be running in a general low power mode to prolong battery life, so it may not be appropriate for the chip to aggressively maximise performance like that. Or a certain pattern of core usage may violate temperature rules: running too hot reduces chip life by years. So Clarksfield has a complex rules-based power control system that takes all of the factors into account, works out the best way to configure the chip for each combination, and constantly monitors internal and external information to keep things running as best it can. Incidentally, this means that if your computer has poor or faulty cooling, your chip will throttle back to save its life at the expense of performance – which is why Intel is also testing fluid dynamic models of case cooling.

Look inside Mooresfield, the next generation handheld chip, and you'll find the same ideas but cast in a very different way. A lot of the power in a smartphone or mobile internet device is used outside the chip itself, in running the camera or card interfaces, in processing audio or outputting video, and there are lots of variations depending on how the device is being used at any one time. Conversely, you get a lot of benefit if you can do things very fast for a decent slurp of power, then shut down to the lowest level possible until the next time. In such cases, if you run the processor hard for a bit then idle it, the memory and other buses connected to it also stop using power when the main chip is idle – more benefits. But again, you can't get too hot during your peaks of use.

By having a smart power controller – in this case, outside the chip in the support circuitry -- that understands the various usage modes of the phone and the conditions in which it finds itself, Intel configures and controls what it calls power islands, combinations of state that are best suited for particular times and tasks. That controller takes information not only from the operating system but from the various components within the phone, so it knows how to divide up tasks that make best use of what's available.

These ideas have much in common. Hardware, software and user work together to decide on the best way to use power, and apply rules and policies that reflect the complexities of being an Internet-connected, multimedia-capable device which has thermal, battery and other practical limits on what it can do, when. The user's role in this is minimal – deciding at the highest level how to trade off performance for power consumption – and then only for battery powered devices. The smarts live in the operating system and in the controlling systems.

A particularly pleasant side-effect of this work is that because Intel sees Linux as a major part of its various strategies, it has to provide open information and open code that demonstrate exactly what it's up to. This information is hard-won, and by sharing it the company is helping the whole industry to make progress into smarter, more efficient power management and higher performance at a lower cost. Such openness, even if encouraged by circumstance rather than philosophy, is actually doing what intellectual property laws claim to do but so often fail at: making new information patently obvious to everyone, for the advancement of all.

A true and proper use of power.

I said it was going to be one of those IDFs. It was, more than I could have guessed.

To begin at the beginning... but when was that? The 9am first keynote of the day, a little Otellini in the morning? But there'd been a 7:30 breakfast briefing before then. The breakfast briefing, then – but there'd been a 6:30am phone call set up. And that was just four hours after I'd sent the last copy back to the UK from Day Zero. Which was itself out of time, as my planned evening writing was disrupted by the discovery of high level Intellites at the hotel bar, an unusual event which deserved investigation.

In the name of journalism, you understand. That and jetlag.

The 2:30am copy, you'll have read. The 6:30am phone call: timezones, bosses, meetings. You know.

Which leaves us with the 7:30am breakfast briefing. This was called by HP, who promised to (here I quote exactly) "unveil a new product that taps the power of Intel multi-core processors to enable a new generation of communication and collaboration, to achieve unheard of levels of ROI and environmental impact".

Apart from "unheard of levels of environmental impact" being the sort of thing we're trying to avoid, that seems a pretty straightforward promise. What we got was HP's version of Skype (you can tell this because it's called Skyroom and does HD video conferencing from PCs). Which had previously been "unveiled" in March, where HP gave exactly the same demo they did this time. Admittedly, that demo wasn't in the basement of a hotel overpacked with grumpy hacks trying to digest an American breakfast which clearly didn't want to be there either.

Tscha, HP, you've always had a reputation for overselling events, but this is beginning to assume the settled shape of accepted fact.

Then keynotes. The newly promoted Sean Maloney gave a rapid canter intro to the show, and then it was Otellini Time.

But before I mention Paul Otellini's keynote — Moblin, Atom, embedded chips, graphs go up and down (rarely with numbers), smiling people juggle exciting technology — I think we should take a moment to examine that new force in semiconductor physics, Maloney's accent.

A Sarf Londoner by birth and mildly Irish by nature, Maloney's adventures in the ultraworld of Intel management have had their effect on his voice. In one sentence — indeed, in one word — it can fly from Brockley to the Bay Area via stopovers in Boston and Brooklyn, only to make the return trip in the final syllable.

This is an interesting poser. If Maloney's accent can move 10,000 kilometers in the space of a syllable, around 300ms, then his words must be able to travel at around a tenth of the speed of light.

That's fast enough to cause detectable relativistic effects on their weight – and, if my schoolboy physics is correct (It's San Francisco, you can get schoolboys on room service), then a word travelling at 0.1 times c has around 0.5 percent more mass than at rest. Half M vee squared over cee squared, and all that.

That may not seem much. But if every word Maloney utters is even slightly weightier than those said by others, he's got a built-in advantage that will add up over time to an unstoppable force. And indeed, this is what we observe.

Back with Otellini, whose accent is immutable, his opening keynote was masterfully presented and with a few interesting nuggets to sieve out of the stream of marketing. For example: Intel is copying ARM's business model almost exactly, by selling its Atom chip as a virtual component for inclusion in other people's chips and encouraging a huge community of developers. It's also copying Apple — no crime in that — by going down the app store route. In five years, Otellini said, he could see Intel selling more processors like this than in the mainstream of Nehalem, Xeons et al.

There's a lot to chew over there: all I'll say for now is that anyone following this battle should be very, very careful to check everyone's claims against reality. Powerpoint predictions have a habit of turning into 'real' data, because it's so tricky to measure what's being claimed — let alone in comparable form.

The Q&A at the end — well, Europe came up and the insistence that Intel had indeed been naughty in shutting out the opposition by exerting commercial pressure on its customers. Had that happened, asked one hack. "We do not do that", answered Otellini – a curious use of the present tense to answer a question about the past, as another hack noted later.

Intel continues to present itself as being unfairly treated by an evil European judiciary, unable to answer its accusers while they're free to misrepresent whatever evidence they like. Well, perhaps.

About the European accusations, I know nothing more than you do. But in the days when I was involved in a small chip-buying start-up, it was standard practice for large chip companies to use every weapon they could to shut out the competition. (And one that backfired: we needed one part, was told we could only have it if we bought another part with it instead of from the competition, so we went out and found a much better alternative for both. That's harder if you're facing a monopoly and a single-sourced part).

Then it was Maloney's keynote, which veered between rich farce — him trying to order a drink from a giant iPhone-like gaming console that towered over him — and thundering technology, like the experimental PCI Express SSDs that were delivering over a million random IO operations a second, or around 4Gbps throughput. That's the next game-changer, right there: memory mapped mass storage. Bearing in mind that some Intel products can now map a terabyte of memory, there's a revolution (if solid state storage can be said to revolve) waiting to happen.

I'm not sure whether Larrabee, Intel's mysterious manycore graphics chip, is rich farce or gamechanger. Maloney's demonstration of it doing so-so graphics, and his refusal point blank to give any technical information whatsoever about the chip itself, is rather suggestive of the former. I'm not sure when Intel last had a chip out with software developers and absolutely no technical info available to anyone else, but it ain't normal, Martha.

I wish I had time to cover the rest of the day's events. The whiteboards which dot the show floor, on which people are encouraged to write down their ideas — and which have more on them than perhaps Intel realises. The "BUMMM! Bum bum bum BUMMMM!" T-shirts which are supposed to echo the Intel Inside sting, but instead make Brit journos giggle like schoolgirls (This being San Francisco, you can order... oh, but never mind). We even hatched a plot to buy a bunch of these — the T-shirts, not the schoolgirls — and give them to the tramps and panhandlers which surround the conference centre, but that could be construed as bad taste. And rightly.

There was an Intel-sponsored Maroon 5 concert. Nothing more need or should be said.

But now, I must go off for more keynotes. There are many untold stories from yesterday, and I'll get around to them later. Any day when you get high praise from one Intel exec and a severe dressing-down from another needs to compost a bit, I think.

At least until one's safely home.