Re: Happy times
One of the regular judges, Prof Michael French died recently. Some of the younger people at his memorial event held an egg race in his memory.
28 posts • joined 6 Oct 2012
One of the regular judges, Prof Michael French died recently. Some of the younger people at his memorial event held an egg race in his memory.
I eventually found an open source video editor to suit my needs in an unexpected place. It turns out that you can use the video editing functions in Blender and achieve quite a bit. However the ui is a little challenging!
Whilst in the EU everyone can know with some confidence that the 'bad' things we can do are more limited, particularly with regard to trading with the EU. Anyone choosing where to base their operations then has a choice between a known status inside the EU and one that could take a radical and unexpected stance in an independent uk. Pity the poor middle manager of global co. who suggests they are based in an independent uk ahead of Eire.
I idly wonder if the reason that Sony's phone's keep the camera going is because they (sensibly) have a dedicated camera button, to which users should reasonably expect fast response from the camera app?
If the first surface the laser touches is simply a good quality white TiO2 paint, then IIRC, around 95% of the incoming radiation is reflected, albeit not back at the target. So 30 KW of light delivered would mean that about 1.5 KW needed to be dispersed. So, if you simply filled a suitable ablative matrix with TiO2 powder, then even without a dispersive effect of any smoke you're sorted. e.g. a phenolic resin matrix can do 200KJ/Kg, assuming a rather slow 10 seconds of attack at 1.5KW we have 15KJ to dissipate, so that hot patch needs to weigh 130g. Assuming that's about a 200mm hole, and assuming you can get away with 50:50 resin to filler whilst maintaining reflectivity- I think you might get away with about a 6mm thick panel and you might not even need to repaint the the car's bodywork after the attack (if you leave an air gap) (assuming I haven't messed up the calculation!).
I can't help feeling that researching how to make such lasers is a good idea even without a good end application- but surely this needs to improve by a couple of orders of magnitude to start to be a useful weapon compared to projectiles?
A fellow traveler told me today how he had been taking some Whitby fosils of ancient amonites into the US and the customs inspector attempted to impound them as it's illegal to take Seafood into the US. He got away with it after suggesting the customs office try BBQing and consuming them!
Our experience is that the real villain here is probably the translation company. We've wasted a load of time to try and set up a good way of working with them to get our client and web text translated: it's been a world of pain. Every time we ask any of them "what's the best way of managing the data transfer"- thinking there must be some savvy folk who do this right- we get blank looks. It seems they generally prefer excel, although of course they will change the code table and try and sometimes translate any HTML tags.
Why translation companies should be so bad at this stuff I don't know...
... and here's another wading now!
Having done a bunch of "Cleantech" investments it seemed to me that nearly everything I saw that could make a really big impact was beyond the scope of any but the most long term and substantial VC fund.
The kind of stuff that makes a difference on emissions tends to be competing with an established technology that's had the benefit of many cycles of optimisation. So if my new cement production, metal winning, or energy storage technology has a chance it must reach the point where it is working at levels of optimization that also take several iterations to approach.
Stuff that's going to make a big difference often involves big equipment, that means that those iterations tend to often be much slower than for more traditional venture investments:if the technology involves plant and equipment or significant scale-up to become economic, then each cycle of improvement can be years. In a 10 year standard VC structure you can get 6 years sometimes (assuming you don't invest on day 1 and need some time to realise at the other end). That's a very tight constraint.
Back around 2002/3 lots of large US VC's moved towards cleantech generally- it was interesting to observe that they often moved their focus to software optimisations around the edges of existing processes for exactly these reasons.
Last time I looked at the Akamai data the S.Korean average speeds were still the highest, but still around 49% of connections were sub 10 Mbp. Averages can be deceptive.
Implicit within this story is that it was the responsibility of Nesta to filter from 1000's of applications the ones where the innovation alone has the potential to drive major shifts in business, and then to back those regardless of any of the other circumstances. The telecom industry, and I suspect the inventor, didn't anticipate the iphone phenomenon, yet we expect a government agency to have that kind of foresight?
I was investing myself during the period concerned and saw 1000's of applications, and yet never saw this inventor approach me for seed funding, and from the sound of the article, if efforts were made to raise commercial money they were not successful. Could it be that the inventor was a key limitation to commercialising his own technology? My experience is that no amount of brilliant innovation is sufficient to mean that you can commercialise a technology if the inventor has either attitude, capability or experience limitations. Perhaps we also expect that our government agencies should step in to fix those problems too?
During the years up to its recent change in status I know that Nesta invested in many truly innovative and brave technology ventures, often to the tune of several £100'000s. Every commercial investor in early stage technology knows you must miss some great opportunities in the 1000's you must reject. Putting pressure on such agencies to account for missed opportunities is likely to breed exactly the kind of defensive investment behaviour that will depress their potential for positive impact. What we should be judging Nesta on is those that it did back vigorously and the returns that those investments made.
My experience of Nesta was that it did have its limitations and constraints. However I did work with people who had a good understanding of what they were doing. If we believe that the state should be intervening in our economy through innovative seed stage investments then maybe the more important issue is that instead of improving provision, we've simply given it up. Nesta no longer does seed stage technology investment and as far as I am aware no alternative has been put in place.
8% over base rate is absolutely stupid. Very few small businesses have a cost of capital that low.
On one level I agree absolutely- take away MS operating system cost and an Android convertible is a great alternative- nearly. However, I have tried to use Android as a travelling laptop replacement- it's mouse- compatibility gives it a huge advantage over iOS for desktop-type applications. However, the office-suites proved, for work needs, to be inadequate. All the office alternatives I tried were buggy and all of them had compatibility issues which meant you often couldn't work with other people using Office. Finally, fast task switching is no substitute for windowing when it comes to work tasks.
However, Win8 (Not RT!) on my Atom powered Samsung convertible is a great work combination. I can run all my apps except modern games (the original Call of Duty is about as good as it gets!), I've even done a little light video editing. Moreover in this form factor a proper stylus makes a huge difference- and windows handwriting recognition works very well for me. It's also much much faster than the Android+Tegra 2 I was using before.
The biggest issue with this option seems to be cost. This machine seems to retail at around £650-750 at the moment- that's way too high to gain widespread adoption. I'd love to know how much of that is Intel's fee for the Clovertrail chipset and Microsoft's for Win8?
BT seem to have decided that contention isn't something we should be worrying about anymore: on the business broadband pages they've replaced the term with "throughput" that you measure with their speedtester!
One day we could make it really really efficient and stuff so the power from the hot water could drive the whole datacentre and maybe even make it even more efficient so it could power a town too. Then we could build datacentres just to power our civilisation without any greenhouse gases.
(Sorry couldn't resist!)
"Operations per second per watt" surely that's operations per joule?
Equally flippantly has anyone here ever owned a Fiat car? My recollection is that Italian car+electricity+water=quite a lot of downtime!
Iceotope's approach to mixing liquid cooling with HPC probably makes more sense though.
A typical toggle switch has an activation force of around 1N and a travel (with the old fashioned type) of maybe 10mm, so if we're 100% efficient we'd only get 10mJ of energy. If we assume that piezo mechanism recovers half the energy we get a useable 5 mJ per switching action.
On the other hand
A 3000mAh lithium AA cell has a 15 yr shelf life and provides enough power as 1.6 million activations of that toggle switch. OK they're £2 a go but I can't help thinking that this is solving a non-problem?
FWIW I have one of the samsung convertibles using this processor and can relate how it actually performs.
For normal office-type stuff including multiple chrome tabs, skype, MS Office etc. it runs fine, it also copes OK with basic photo stuff (GIMP) and even seems OK with very basic video editing with MS Live Essentials.
It sucks at any type of 3d game: for instance World of Tanks runs at 3-5 fps even in lowest graphics settings. The only playable 3d game I've found so far is the original counterstrike.
However, battery life is brilliant- 10 hours of real-world usage on mixed office activities.
I remember a former Intel exec bragging about how successful ATOM had been in allowing netbooks to exist without ever being sufficiently successful to impact much on the main processor lines- I think the added RAM, wider bus, hyperthreading means that now this processor is "good-enough" for most everyday PC use.
I remember an Intel guy bragging about how the point of the original ATOM processor was to make sure netbooks were sufficiently poor that they didn't have too much impact on volumes of expensive CPUs. However, with the new Atom Z2760 Clovertrail (when they are available in any quantities) you get ARM-type battery life and decent desktop performance- albeit not for games. That really does make Win RT seem a bit pointless as it runs Win 8 Pro nicely.
Leaving aside the error that those game sales were Billions not Millions of pounds...
Whilst the retailer channel provides solid data on the decline of physical goods. I don't believe many of the major outlets for online game distribution provide this kind of information. For instance Steam, the largest of the distributors for PC games doesn't provide such information.
Is it too cynical for me to speculate on the value of this as a story if the total is heading South rather than staying boringly about the same?
I believe many telephone handset leads still use a special type of wire called "tinsel" where there is a nylon fibre core wrapped in one or more helical windings of copper foil. As you stretch the wire the helix extends to give you a much longer strain before failure- massively reducing cable failures on kinks, bends and stretches.
It seems to have gone out of fashion I suspect because:
It has a higher resistance so isn't ideal for low impedance headphones
It's a complete pain in manufacturing- special insulation displacement connectors are required and it's kind of tricky to solder!
I've seen loads of business plans- and enterprise software ones were the worst- that used this kind of description and yet the founders were confused when you politely said you hadn't any idea what they actually sell. I blame the business academics drumming into folk that you "sell on benefits not features"- that works for some stuff but it kind of presupposes you have some idea what the stuff actually is!
I remember coding late into the night and being a bit disturbed when the machine spoke to me. (No, I was too young to drink then!). I looked out of the window to see a British Gas van in the street and all was clear- the Speccy was demodulating the van's radio set!
Within days of arriving at high school my daughter was given homework that required MS Publisher & MS Access files to be opened- IIRC neither was easy using competitor products.
The new Intel Z2760 Atom CPU is actually a decent compromise- I'm regularly getting 10hrs battery life on my Ativ 500 tablet running Win8 Pro and use it as my main machine (I'm travelling a lot!). Basic office/web browsing is nice and fast and so the bigger laptop stays at home mainly for games! The machine cost me about £550 (albeit in Korea) so, for what I need, it's better than an ultrabook at about half the price. However, I hear this class of machine is hard to get hold of (screens+CPU?) so I think they're too late to make much impact this year- seems to me that ARM's competitive effect has caused Intel to up its game a bit!
There's plenty of room to improve PUE more. Iceotope, a new technology company based in Sheffield, is offering liquid cooled kit which means that the heat is extracted without active cooling and, as you get to drop the fans from each blade, the PUE is (at least in one interpretation of that number) ends up being less than 1.
Patents are a huge help to small innovative businesses- or at least the one I run! They give investors the confidence that the business can drive a good exit for their investment without which really innovative businesses would be impossible to fund properly. Bad patents are bad, and especially in the US there is a legacy of "un-inventive" patents going back years, but it doesn't mean that we'd all be better off without any!
erm I'm pretty sure that Blu-ray is compressed with H264 too but it can be at quite a high bit rate c.f. terrestrial dvb
I seem to remember reading that only 25% of internet-capable TV's are ever actually connected to the internet!