Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×

Comment Re:Software? Really? (Score 1) 89

I find this decision difficult to understand. I was forced to get a BlackBerry for work, ended up with a Z30. The hardware seems decent, but the absolute worst thing has always been the software.

I've heard that the corporate side for controlling, monitoring, and updating remote handsets is very good - and this is who they mostly sell it to.

Comment Translation (Score 1) 97

This certainly has consequences for B2G OS. For the community to continue working on B2G OS they will have to maintain a code base that includes a full version of Gecko, so will need to fork Gecko and proceed with development on their own, separate branch.

Translation: We found it is useless so we've thrown it over the fence.

Comment Re:In other news... (Score 1) 304

IFPI, RIAA, and BPI have sued users who have the audacity to recall verses and lyrics from songs they heard on Youtube and the radio.

RIAA spokesperson commented, "Each time one of these social deviants hums, sings, or otherwise repeats our intellectual property to their peers represents an enormous loss of revenue to our members. We intend to aggressively pursue legal action whenever possible to recoup all potential losses. It's clear the recent declining revenue in our industry is directly caused by these unauthorized reproductions of our copyrighted material"

They backed down this time but I'm sure they'll keep trying

Submission + - Working Round The 'Big Data Bottleneck' In Modern CPUs (thestack.com)

An anonymous reader writes: Modern CPU architecture is built to retrieve large chunks of data, to limit the number of time-consuming journeys between the central controller and the location of the data in memory banks. When you're fetching the first data block of a picture for Photoshop, bringing along the adjacent block makes sense, because you're probably going to need it. But when you're making ten calls to a 'sparse' dataset, where each of the items you want is resident in different memory allocation, and none of them have any relevant adjacent data, the architecture is fighting the intention.

Researchers from MIT have addressed the problem by creating a C++ extension that gathers these requests into one queue for each core, and then forces the cores to swap and negotiate which requests they can most efficiently handle for the minimum number of journeys to memory. In earluy tests, access to sparse datasets has been increased by up to four times using this method, and promises even greater increases with a dedicated architecture. Contributing researcher Vladimir Kiriansky explains why the teams called the extension 'Milk', and why the name also explains the challenge: "It’s as if, every time you want a spoonful of cereal, you open the fridge, open the milk carton, pour a spoonful of milk, close the carton, and put it back in the fridge."

Slashdot Top Deals

It seems that more and more mathematicians are using a new, high level language named "research student".

Working...