
For my science (ab initio electronic structure) 10+ significant figures are usually required. In some cases it's debatable if 64 bits is enough. FP8 is an irrelevancy. In fact FP32 is mostly an irrelevancy...
180 publicly visible posts • joined 26 Jun 2007
Problem is that he is already a convicted crook:
https://www.foxnews.com/politics/ny-v-trump-jury-verdict-former-presidents-historic-criminal-trial
https://apnews.com/article/trump-trial-deliberations-jury-testimony-verdict-85558c6d08efb434d05b694364470aa0
https://www.bbc.co.uk/news/live/cd9x99p2ek8t
https://www.reuters.com/legal/jurors-begin-second-day-deliberations-trump-hush-money-trial-2024-05-30/
Only a completely brazen liar with no concern for the truth would ... Oh ...
Cloned the repo and after a couple of tries managed to build it. The good news is that it is a **massive** step forward on the earlier flang that I tried a couple of years back, in that it can actually compile modern Fortran and produce executables that work. But it appears to lack a number of basic features without which I won't consider it as a development tool at least - so far flags to enforce array bounds checking and standards conformance have eluded me. But this might be because any decent documentation has also eluded me, flang --help being about the only thing I can find so far.
I remember in a previous life attending a talk on financial modelling, stochastic differential equations and all that stuff. Not my thing at all, but I do remember one interesting observation which was along the lines of financial modelling is unlike climate modelling and weather forecasting because the outputs of the model can directly and measurably affect the thing that is being modelled. Looks like that's no longer true.
This is a very strange article. The offence is abuse of a software licence - in particular not obtaining one to use the software. This is an important issue, and IMO the world in general and academia in particular has got to learn that software doesn't just come out of thin air, if you want high quality and especially if you want maintenance you will have to pay for or contribute to it in some way. But then the article goes off talking about access to academic publications, a totally separate issue. It's of definite importance but I fail to see how retracting a paper due to breaking the law connects with it.
Some years back I was visiting a sub-Saharan country in Africa and got talking with one of the locals over a coffee about how corruption affects their democracy. I can 't remember exactly how he put it, but the essence was that if some local bigwig comes by and says to a villager "I will give you a cow if you promise to vote for me" the villager will take it and vote for the bigwig, because the central government is totally divorced from his existence, while a cow will make a real difference to his family. Then the local finished off by saying something like " but of course that wouldn't happen in a developed world democracy." For some reason I've been thinking about that little encounter more often recently.
A genuine laugh out loud moment that one ...
The doubly worrying thing is that a *good* sum() function is non-trivial, especially if dealing with floating point. So much so that I will roll my own if I really need accuracy, who knows who wrote the library functions for all the compilers I need to support?
Note both Oxford and Cambridge give BAs for all their first degrees - I hold a BA (Oxon) in Chemistry
But a large part of the problem is successive ministers have seen the science and technology budget as there to fund their own pet projects, rather than supporting any coherent national plan. IMO millions have been hived away and wanted that way. Boris Johnson and George Osborne are especially at fault here.
No. Some MPs form the government, but not all. Quoting https://www.parliament.uk/about/how/role/relations-with-other-institutions/parliament-government/
"Parliament and the Government are different. They have different roles and do different things."
"HM Government consists of the Prime Minister, their Cabinet and junior ministers, supported by the teams of non-political civil servants that work in government departments."
And
"Parliament is there to represent our interests and make sure they are taken into account by the Government. The Government cannot make new laws or raise new taxes without Parliament's agreement.
Parliament is made up of people we have elected and people who have been appointed. They sit in two separate Houses:
The House of Commons, where all the people we have elected at the General Election work, as MPs, for the next five years. This includes people in other political parties, as well as those in the winning party who were not chosen to be ministers."
Thus all MPs are members of the House of Commons, but not all MPs are members of the Government
More than that, it's quite possible that the games are analysed afterwards by some software to find indicators as to whether it is a human playing, or if there is computer "assistance". I can't tell you about the Go world, but this certainly happens in Chess - any rated online tournament will likely be sent for analysis, and yes, people have been disqualified and/or banned for use of a computer having been detected by this method. Ken Regan (https://www.buffalo.edu/news/experts/ken-regan-faculty-expert-chess.html) is probably the best known researcher in this area.
Yes, Priti Patel is evil She wants to kill people (https://www.indy100.com/news/priti-patel-resurfaced-clip-death-penalty-ian-hislop-question-time-video-home-secretary-9020006) unlike mini-Trump who merely colludes in trying to get them beaten up (https://www.theguardian.com/politics/2019/jul/14/black-eyes-boris-johnson-plot-attack-reporter-darius-guppy). Both liars though. And these are the people wanting to hand over our personal data ...
I remember that day. I remember as a student sitting in the common room watching the television footage. Crying. Crying for the people of China. Crying for a people being told lies. Crying for all of us whenever we unquestioningly forgo freedom for security. World, don't forget this. This is what the Chinese government is all about. Every time you have dealings with the Chinese government remember Tiananmen Square.
Macho flops/£ and achieved flops/£ are very different things, especially once software development and maintenance costs are factored in, and also electricity bills. But the honest truth is for this project the vast majority of users didn't need or want GPGPUs to do their science, and in many cases GPU versions of the applications just don't exist, so early on a decision was made to be purely CPU based - the machine is there to do science, and if it can't do the science the user base needs it doesn't matter how many macho flops it can do. As a result I can't answer your question as we never did that calculation.
Sigh ... To all those who don't understand the archaic temperature units which are the primary measure in the article can I remind you of
https://www.theregister.co.uk/Design/page/reg-standards-converter.html
so they can be converted into something more comprehensible. For starters 120F=2.9Hn
Just to add some numbers over the period 2014-2016 UK researchers and Innovators received 15.2% of the Horizon 2020 funding, for a total of just under €3.8bn funding. This was the second highest in Europe, just behind Germany (16.7%): https://ec.europa.eu/unitedkingdom/news/horizon-2020-three-years-uk-tops-league-participants_en