Beta

×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Comments

top

NVIDIA Launches Tegra K1-Based SHIELD Tablet, Wireless Controller

ponos Use case? (42 comments)

I don't really see the need for gaming on the go, and if such a need exists, isn't it sufficiently covered by existing gadgets (smartphone or non-gaming tablet)? Furthermore, are the current tablet games worth buying hardware specifically for the job? What would be the point of Angry Birds at 120fps?

Anyway, the hardware looks cool, but the fact that no other manufacturer bothered to use the nvidia hardware is a bit disconcerting. If it were the best thing since sliced bread, many designs would have flooded the market.

about two weeks ago
top

New Treatment Stops Type II Diabetes

ponos The quality of the treatment matters (253 comments)

Although a definite cure would be nice, simply improving the quality of glycemic control would be a revolution. Optimal glycemic control can be very complicated, especially for patients who do not have the courage to follow strict dietary and healthstyle recommendations. Being able to treat a complicated disease with a single daily injection sounds nice. I know it would simplify my job as an MD...

about two weeks ago
top

New Treatment Stops Type II Diabetes

ponos Re:There's another treatment that stops most T2 (253 comments)

I am not a great cook but I do enjoy cooking and I can't think of a texture that you can't achieve with good materials and decent cooking technique. Healthy food really doesn't have to have a bad texture I think. Maybe an investment in a few cooking books and some classes would help you?

about two weeks ago
top

Mozilla Doubles Down on JPEG Encoding with mozjpeg 2.0

ponos HSA decoder (129 comments)

AMD recently presented HSA-enabled jpeg decoding. That would also be an interesting addition. Make these shaders work a little...

http://developer.amd.com/resou...

about two weeks ago
top

When will large-scale IPv6 deployment happen?

ponos Re:We already built a new Internet (305 comments)

See, this is the thing. We let you on the last Internet we built, the one we started with 110 baud modems, and you messed it up.

So we already built a new Internet 2. It's here, it runs on 100 GB/s ports with 40 GB/s campuswide at all the top research universities.

And we're not letting you on it.

Who is "we"? Judging from your attitude you probably weren't even born at the time. But anyway, I wanted to point out that several "private" networks exist for the military, big companies etc. When we say the Internet, we implicitly assume that it can be used to to send email to anyone, browse anything and use all kinds of services. Any big private network cannot be a substitute. So, if you need to come out of your walled garden to post messages in Slashdot, order stuff from Amazon and read your gmail account, maybe your new internet is not much more than a data exchange mechanism. Which is great, really, but you can't call it "Internet 2" because it seems to me that it is missing some key functionality.

about a month and a half ago
top

When will large-scale IPv6 deployment happen?

ponos Re:IPv6 already massively deployed (305 comments)

IPv6 is already massively deployed in the United States.

Example: all major hosting providers support IPv6.

Thank you for this informative post. I was under the impression that at least the Linux network stack and, most probably, the Win 7-8 network stack are fully IPv6 compatible. I know my router is IPv6 compatible and I had used IPv6 briefly in the past (via tunneling) out of curiosity. So, why is everyone so pessimistic? How hard can it be to just enable IPv6 if everything is ready at the user side?

about a month and a half ago
top

$3000 GeForce GTX TITAN Z Tested, Less Performance Than $1500 R9 295X2

ponos Re:AMD Consistent framerate, since when? (151 comments)

the R9 295X2 offered higher and more consistent frame rates

http://cdn.pcper.com/files/ima...

But not "stable", "consistent" or "smooth". This is still a major issue with the core of all AMD cards which hasnt been fixed.
You get what you pay for. Nvidia might be the "expensive" of the bunch, just wish i forked out a little more instead of getting my HD7770.

Do you realize that in the graph you linked no card dips below 50fps at any time? In fact, if you count the occasional peaks crossing the (ridiculously low) 15ms/66fps threshold, the Titan Z shows 6 frames slower than 15ms and the 295X2 shows 4 frames at more than 15ms (if I count correctly). You really can't argue that the Titan Z is smoother. All cards are extremely smooth.

about 1 month ago
top

$3000 GeForce GTX TITAN Z Tested, Less Performance Than $1500 R9 295X2

ponos AMD fp64 rate (151 comments)

I would just like to point out that the 295X2 has superior absolute gaming performance and superior fp32 performance but, just like most gaming NVidia products, the fp64 is crippled at 1/8 fp32 rate at configuration in order to create a profit margin for the costlier "pro" products. The hardware itself is capable of 1/2 fp64 rate and should be superior to the Titan Z if AMD decides to offer "pro-level support".

As proof, consider the fp64 rate of the single-chip AMD W9100, sold at ~$4000, which is 2.6 TFlops (http://www.amd.com/Documents/FirePro_W9100_Data_Sheet.pdf), versus the 2.7 TFlops of the Titan Z (1/3 fp32 rate, see http://en.wikipedia.org/wiki/G...). AMD could unlock the 295X2 at its full potential 5.2 double precision TFlops and release it any day if they want, crushing the Titan Z.

Honestly, instead of the Titan Z, I'd rather buy the AMD W9100 for $4000 and get equivalent double precision compute rate, better perf/W and, most importanty, certification for pro applications and ECC memory. That is certainly worth the extra $1000 in this product segment.

about 1 month ago
top

iOS 8 Strikes an Unexpected Blow Against Location Tracking

ponos Users should start asking for privacy (323 comments)

The adoption of measures protecting privacy depends on user demand. Online commerce has been considered safe enough for years yet exchanging an email or having any online activity is completely unprotected. I was always surprised by lack of interest from users. Maybe the younger users, if they are not yet addicted to making all their life public on facebook et al would put some pressure for simple technical solutions that guarantee a basic level of privacy. Obviously, I don't expect complete protection against three-letter agencies; that's not the point. In that sense, this looks like a step in a desireable direction, even if it is done for the wrong reasons. As a potential customer, I appreciate this effort.

about 2 months ago
top

EU Launches World's Largest Civilian Robotics Program; 240,000 New Jobs Expected

ponos Re:A bad idea for reasons of basic economics (171 comments)

Because the funding source doesn't have a clue nor would it have any interest in spending the money efficiently or effectively. Just because there are more zeroes on the check doesn't mean that more science is being done.

I understand that the private sector can be more efficient for certain things, but these are not gifts nor scholarships. I have applied for EU finding and I can tell you that there is a lot of work that goes into proving that they got their money's worth. You have progress reports to do, intermediate results to publish and paperwork to fill in order to keep the funding. Getting the money without doing any research (=stealing) is not that easy and, in my limited experience, does not occur that often.

How about useful research? Government funding isn't so extremely important, when you want research that actually pays for itself within a few centuries.

That's a philosophical viewpoint that, in my opinion, produces short-sighted research of the kind that will give you incremental iPhone updates but no major breakthrough. I cannot convince you of the validity of this claim, but true science is a high-risk and long-term endeavor of the kind that does not appear favorably in quarterly financial statements. The private sector revolves around the next yearly bonus, not about a project that can pay off 10 or 20 years later. Also, don't forget that private "research" is locked under patents and any useful results do not necessarily benefit the society as a whole (at least for 10+ years). So, even if you assume that research by private organisations compares favorably, it is not truly equivalent.

I don't see why it's so hard to see that. "Hard and competitive" doesn't mean anything of value happens. All those people striving for easy money when they could be doing something productive for society?

How about you try to get that "easy" money. Have a look at the requirements for application in the Horizon 2020 EU research program. You need several AAA laboratories (ideally, with multiple Nature/Cell/Science publications) in order to stand a chance. Now, if you feel that basic research is not "productive", I'm probably wasting my time. At least consider the possibility that big research projects produce side effects that are beneficial but difficult to measure (say, WWW was invented initially for use in CERN).

about 2 months ago
top

EU Launches World's Largest Civilian Robotics Program; 240,000 New Jobs Expected

ponos Re:Is this Slashdot? (171 comments)

How many of these "robot maintainers" are white/Asian, and how many are African-American/Latino?

Yeah, that's what I thought, you right-wing racist prick. All you want to do is put people who don't look like you out of a job. Fuck you.

Hm, I do agree that some people may lose their jobs, but I don't see how you inferred that they would be African-American or Latinos. I did not mention race in my post and, in my value system which is not particularly right-wing, all lost jobs are equal independent of the race of people who lost them. It just so happens that the loss of some jobs in a certain sector may be the result of great progress in other aspects of life. So, I believe the overall balance is or could be positive. At least if you are optimistic about that sort of thing.

about 2 months ago
top

EU Launches World's Largest Civilian Robotics Program; 240,000 New Jobs Expected

ponos Re:A bad idea for reasons of basic economics (171 comments)

This is another example of corporate welfare masquerading as a jobs plan, combined with protectionist sentiment. The central planners will take money out of the productive economy and spend it on a corporate giveaway to favoured interests. Jobs that otherwise would have been created in the productive sector will be lost, while only the 240,000 pork barrel jobs will be noticed by the superficial.

Is there an alternative way of stimulating research in a specific field for the public good? And why wouldn't the proposed approach work? I mean, NASA went to the moon in the 60s and here we are today waiting for some billionaires who hope to one day send some rich kids at a hundred km from the earth's surface. As if that would be a great achievement. And don't even tell me who in the private sector would ever fund obscenely expensive shit like CERN or the ITER fusion reactor. The fact is, if you want basic research, government funding is extremely important. So, while the productive sector is busy developing the iPhone 6 or some other must-have "gadget", someone will have to pay for basic research if you want to get that flying car one day.

And, for what it's worth, getting EU research funding is often so hard and competitive that if you manage to obtain it, it becomes a key item in your resume. Sort of like a prize. So, I fail to see how a highly specialized research program with high barriers to entry will result in pork barrel jobs.

about 2 months ago
top

Intel Announces Devil's Canyon Core I7-4790K: 4GHz Base Clock, 4.4GHz Turbo

ponos Market segmentation (157 comments)

Sounds like a really cool product. It's a pity intel has over-segmented their product line and I can't get this chip with ECC. That would be cool. In fact, ECC should some day become a standard feature. But that's just wishful thinking.

about 2 months ago
top

EU Launches World's Largest Civilian Robotics Program; 240,000 New Jobs Expected

ponos Is this Slashdot? (171 comments)

Everyone speaks about a possible losss of jobs or trademark issues. Am I the only one thinking that robot technology is cool? This is the kind of shit that could allow exploration of the oceans and eventually space, prosthetic help for sick people, cheaper and more efficient mass production etc. Plus, it would probably generate some interesting by-products, like advanced algorithms, maybe a new programming language or new processor types. And it gives jobs to young people with PhDs.

PS Jobs are being lost and created all the time. Think robot maintainer, robot programmer, robot police (?) (the "Turing"?), robot designer. And, anyway, if a job can be taken by a robot it probably isn't very interesting or creative to begin with. If I had a choice, I'd rather be doing the creative stuff.

about 2 months ago
top

The Flaw Lurking In Every Deep Neural Net

ponos Re:Errors (230 comments)

In this case however, it should be noted that the humans are ALSO in error. They see both images as the same, when the images are in fact not the same.

Actually, the human is the benchmark here. Being able to recognize two different photos as coming from the same person is a feature, not a bug. That's the whole point of running a neural net classifier. Otherwise you just "diff" the photos and only accept byte-identical ones as similar. Mathematically correct, but not very useful in real life.

about 2 months ago
top

Why You Shouldn't Use Spreadsheets For Important Work

ponos Spreadsheets can become complex (422 comments)

The point is that spreadsheets can become very complex and they don't have any serious provisions for code review and debugging. Spreadsheets are not meant to be numerically stable and bug-free. They are meant for presenting and manipulating simple data. People abuse spreadsheets for database work and data transformation that are best done in a combination of SQL + something like R. I have been confronted with big spreadsheets in my research activities and the first thing I do is just convert the raw data in CSV, load it in SQL or R and take it from there, writing proper code in the process.l

about 2 months ago
top

Why You Shouldn't Use Spreadsheets For Important Work

ponos Re:OOXML or Excelception (422 comments)

How can you possibly expect a CSV to contain a formula? It's a text file.

about 2 months ago
top

Wikipedia Medical Articles Found To Have High Error Rate

ponos Caveat emptor (200 comments)

In fact, almost any source of information in an ever-changing science and practice will contain statements that can be contested. Even major medical textbooks disagree in details or between editions. So, I don't thing anyone expects a wikipedia article to be absolutely accurate because such an article rarely exists even in the "peer-reviewed" domain. In practice, I used wikipedia as a decent source of information several times. I suppose I would have noticed glaring omissions or errors and I'm not even looking for critical pieces of information (say, chemotherapy protocols). For professionals, wikipedia is a tool like any other. For patients, things can be a little more complicated. In the end, I think wikipedia is much better than wacky sites offering "natural cancer treatments" or other scams. So, overall, nothing to see here.

about 2 months ago
top

The Flaw Lurking In Every Deep Neural Net

ponos Re:Well what do you know (230 comments)

The main advantage of learning algorithms like neural nets is that they can automagically generalise and produce classifiers that are relatively robust. I wouldn't be surprised at all if a neural net misclassified an extreme artifical case that could fool humans (say, some sort of geometric pattern generated by a complicated function or similar artificial constructs). Here, however, it appears that the input is really, really similar and simple to recognize for humans. Obviously the researchers have recreated a "boundary" condition, but the fact that this becomes manifest in real-life examples is a bit worrying for the validity of the algorithm in general situations and especially its scalability in much bigger projects were similar cases may arise more frequently.

about 2 months ago
top

The Flaw Lurking In Every Deep Neural Net

ponos Re:Errors (230 comments)

I don't think a computer ai will be perfect, either, because "thinking" fuzzily enough to develop intuition means it's going to be wrong sometimes. The interesting thing is how quickly we get pissed off at a computer for guessing wrong compared to a human.

But we do expect some levels of performance, even from humans. You have to pass certain tests before you are allowed to drive a car or do neurosurgery. So, we do need some, relatively tight, margins of error before a machine can be acceptable for certain tasks, like driving a car. An algorithm that has provable bias and repeatable failures is much less likely to be acceptable.

The original article also mentions the great similarity between inputs. We expect a human to misinterpret voice in a noisy environment or misjudge distance and shapes in a stormy night. However, we would be really surprised if "child A" is classified as a child, while similar looking "child B" is mislcassified as a washing machine. Under normal conditions, humans don't do these kind of errors.

Finally, even an "incomplete" system (in a goedelian sense) can be useful it it is stable for 99.999999% of inputs. So, fuzzy and occasionally wrong is OK in real life. However, this will have to be proven and carefully examined empirically. We can't just shrug this kind of result away. Humans are known to function a certain way for thousands of years. A machine will have to be exhaustively documented before such misclassifications are deemed functionally insignificant.

about 2 months ago

Submissions

ponos hasn't submitted any stories.

Journals

ponos has no journal entries.

Slashdot Login

Need an Account?

Forgot your password?
or Connect with...

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>