© The Financial Times Ltd 2014 FT and 'Financial Times' are trademarks of The Financial Times Ltd.
November 30, 2012 7:34 pm
Science and industry ministers have charted a way forward for Europe’s space activities. Their biennial meeting last week agreed to spend €10bn over the next five years on a wide-ranging programme that includes developing a new version of the Ariane heavy-lift rocket, contributing to Nasa’s next manned spacecraft and working with Russia on robotic missions to Mars.
The gathering in Naples of the 20 member states of the European Space Agency (ESA), an international organisation independent of the European Union, was much more amicable than the EU budget meeting in Brussels. And in contrast to Britain’s desperate bid to keep a lid on its payment to the EU, the UK delegation went to Italy with an offer – gratefully accepted by its ESA partners – to increase its contribution by 20 per cent, to £1.2bn, over five years.
For the first time since the Thatcher government decided in the 1980s to have nothing to do with manned spacecraft, Britain is contributing to a crew vehicle. The amount is relatively small, £16m, but added to other nations’ contributions it will enable ESA to build a service module for Orion, the spacecraft that Nasa is developing to take astronauts not only to the International Space Station but eventually on deep space missions to asteroids and even Mars.
To put matters into perspective, however, the total UK contribution to ESA will still be only about half that of Germany or France and just ahead of Italy’s. Britain remains out of the Ariane rocket programme, on which the big decisions were to work on an upgrade of the current Ariane 5 to produce a 5ME (for “midlife evolution”) by 2017 – and design a longer-term successor, Ariane 6, for use from 2021 or so. Ariane needs upgrading in the face of growing competition in the satellite-launching business, not only from Russia and Asia but also from private US companies.
Mosquitoes can fly in rain but not in thick fog, according to researchers at Georgia Tech, because fog droplets interfere with the insects’ flight control system.
Meanwhile Nasa is cutting back, under budgetary pressure from the Obama administration. One sacrifice is a planned collaboration with ESA on a pair of missions to Mars – which would put an orbiter around the planet in 2016 and land a rover in 2018 – to carry on the good work being done by Nasa’s Curiosity. Fortunately, a new agreement to work with Russia instead has saved this ExoMars mission.
The Naples meeting also agreed to continue or extend many other ESA programmes. They include planetary exploration, space telescopes, Earth observation, a weather satellite and a new generation of telecommunications and broadcasting satellites. Although Nasa is still by far the world’s largest space agency – and China is expanding its space activities very quickly – ESA is enjoying a good run in second place.
Of mice, men and medicine
No mice on Earth have more human DNA than the animals kept by Kymab, an antibody engineering company in Cambridge. Genetic manipulation has given the mice a substantial part of the human immune system – all in the cause of making better medicines.
Allan Bradley, Kymab’s founder and chief scientist, says the mice have about 2.7 million base-pairs (chemical units) of human DNA – roughly 0.1 per cent of the whole genome. As a result, researchers can make a wide repertoire of antibodies by immunising them with the appropriate antigen.
Antibodies are protein molecules made by the immune system to bind to harmful substances – antigens – which may be viruses, bacteria, toxins, abnormal proteins or cancer cells. Sales of antibody-based drugs have boomed over the past 15 years, and are now worth more than $50bn a year globally.
Inoculating live mice is traditionally the best way of making antibodies. In the past these murine antibodies had to be “humanised” in the laboratory, but genetic engineering has made it possible to avoid this step by giving mice a human immune system.
A couple of months after immunisation, when tests show that a mouse is making the desired antibody, its spleen is removed and the antibody-producing B-cells extracted. These are then “immortalised” by fusion with cancer cells and expanded many times in culture.
Kymab is the first spin-out company from the Wellcome Trust’s Sanger Institute, the leading UK centre for genomic expertise, where Bradley used to be director. Wellcome, Britain’s biggest medical research charity, is also the lead investor in Kymab. Although the company will licence its Kymouse technology to drug companies, real success would come from developing its own antibody treatments.
Titan achieves record number of flops
The world of supercomputing has a new speed champion called Titan, writes Ling Ge. The machine at Oak Ridge National Laboratory in Tennessee tops the 40th edition of the biannual TOP500 list of the world’s fastest computers. Its superprocessing prowess will help accelerate the pace of discovery and innovation across a range of scientific fields, from developing more efficient engines to studying climate change and finding cures for disease. Titan, a Cray system, has taken the title from Sequoia, an IBM machine at the Lawrence Livermore National Laboratory.
But how is supercomputing speed assessed? The rankings are based on the Linpack benchmark, which measures the number of “floating-point operations per second”, or flops, that a machine can perform. Titan operates at 17.59 petaflops – the “peta” suffix meaning a quadrillion, or a million billion. Those 17.59 quadrillion calculations per second are equivalent to each of the world’s seven billion people simultaneously carrying out 2.5 million calculations per second.
The architecture of Titan is a combination of gaming and traditional computing technologies. It employs a family of processors called graphics processing units (GPUs) first created for video gaming.
Titan highlights the growing importance of GPU-based parallel computing, in which larger problems are divided into many smaller calculations that are then carried out concurrently. Nvidia, Californian manufacturer of the 18,688 GPUs in Titan, says they account for 90 per cent of the system’s peak performance.
GPUs, designed to drive down energy costs and increase performance, are much less power-hungry than general-purpose central processing units (CPUs). Titan has also just established its green credentials by coming near the top of the Green500 rankings of the world’s most power-efficient supercomputers.
“You simply can’t get these levels of performance, power- and cost-efficiency with conventional CPU-based architectures. Accelerated computing is the best and most realistic approach to enable ‘exascale’ performance levels within the next decade,” says Steve Scott, Nvidia technologist. An exascale computer would run at 1,000 petaflops – about 50 times faster than Titan.
Titan is available to researchers from academia, government laboratories and a broad range of industries, who will use it to model physical and biological phenomena and seek breakthroughs faster than is possible by experimentation alone. A total of 62 out of the top 500 supercomputers use GPU-accelerated CPUs, up from 58 machines in the June list.
The TOP500 list, started in 1993, is a German-US collaboration between the University of Mannheim, Lawrence Berkeley National Laboratory in California and the University of Tennessee, Knoxville. Past champions include Fujitsu’s K computer in Japan, which claimed top spot in 2011 and is now third, and China’s Tianhe-1A computer, which ranked number one in 2010 and now comes eighth.
Copyright The Financial Times Limited 2014. You may share using our article tools.
Please don't cut articles from FT.com and redistribute by email or post to the web.