Paper on Performance Optimization

[Neo]

Cool rigs.  :U

An update on this paper: It's now been sitting with an editor at Journal of Parallel and Distributed Computing for 6 weeks still not being reviewed.  I think (though I'm just speculating) he doesn't want it published but doesn't have a legitimate argument against it, so he's just making us wait until we get fed up and retract it.  :(  At least it sounds like the journal manager is none too pleased about the situation.

[jj2007]

An update on this paper: It's now been sitting with an editor at Journal of Parallel and Distributed Computing for 6 weeks still not being reviewed.

Neo, don't despair. One of my papers has been lying around for three years, and is now in print. Another one was received about one year ago, and I am pretty sure it will be accepted. Six weeks is a *very* short delay :P

[Neo]

An update on this paper: It's now been sitting with an editor at Journal of Parallel and Distributed Computing for 6 weeks still not being reviewed.

Neo, don't despair. One of my papers has been lying around for three years, and is now in print. Another one was received about one year ago, and I am pretty sure it will be accepted. Six weeks is a *very* short delay :P

Six weeks is a short delay for reviewers, but all the editor has to do is pick a few names and send the paper to those people.  We even sent a list of possible reviewers after 4 weeks, in case he's legitimately looking for reviewers.  Barring any catastrophes, this editor has simply been refusing to send the paper out for 6 weeks.  A second, much less significant, paper we sent to the same journal 2 weeks ago was sent out for review after about 4 days.  :(

To further put it in perspective, another paper we had on a big improvement to our preprocessing was fully reviewed 2 days after submitting it.  Another one took a few months, which is more the norm, but it was still sent out for review after only a couple days.

[Rockoon]

Speaking of cool rigs, AMD has recently released an affordable line of 6 core processors. Currently just two models, the $200 1055T at 2.8ghz and the $295 1090T at 3.2ghz. Both are socket AM3 processors.

The 1090T has an unlocked multiplier while the 1055T does not, but both are overclockable. Many reviewers have pushed the 1055T to a stable 4ghz using the stock heatsink and fan that comes with the processor. The advantage of the unlocked multiplier of the 1090T is that you can overclock the processor without also overclocking the system bus.

That $200 price point of the 1055T is very attractive IMHO.. can put together a very nice system for under $500 (~$100 motherboard, ~$100 memory, ~$100 case and power supply)

[Neo]

Hmm... it'll still be tough for them to compete with the 8-logical-core Intel Core i7's, which you can get for $260.  The Core i7's also made a big leap in per-core performance, so hopefully AMD has something to counter.

[Rockoon]

Actually the $280 i7 860's and 920's are ranking on par with the $200 1055T in benchmarks

The i7 930 does rank a bit higher, but thats $290.

The 1090T beats all of these for $295

http://www.cpubenchmark.net/high_end_cpus.html

[Neo]

REJECTED!... again, after another 5 months of waiting  ::)

Looks like academia is fine with sticking its head in the sand.  One reviewer said the optimizations were obvious, but if that were true, countless authors of published papers and books would be guilty of academic fraud for blatantly ignoring them.

[hutch--]

 :tdown

I am sorry to hear that but I have a cynical view that academia is a plot by the obscure to try and look profound in the absence of talent. Mediocrity abounds in a world where beeing seen to be politically correct is more important than delivering in terms of performance or output.

Find another context to publish your designs and ideas and don't allow the idiot fringe access to it.