Yeah, going to optical when copper will do doesn't make economic sense. There's no need for the bigger bandwidth of optical as long as the transceivers are fixed and non-upgradeable, as they would be in consumer equipment. Even in the high-end space, there seems to be lots of 10GigE over copper these days.
I'll add that "science has generated net positive fusion" is not true. He's probably thinking of JT-60, which achieved conditions where, if deuterium/tritium (DT) had been used as the fuel, it would have meant breakeven. However JT-60 doesn't have tritium handling facilities, so they can only use deuterium and extrapolate to DT's higher cross section. Still, it's a remarkable achievement.
In my mind, the biggest obstacle to fusion power is finding/designing materials that will last several years in a commercial reactor. And we have hardly anything at the moment.
That's correct. I run in limited mode, and have a RunAs command set up for administrative tasks. It's not possible to copy/paste or drag between explorer windows started with different privileges.
I don't see this happening anytime soon. FPGA performance is nowhere near the $/performance ratio that would make sense for the consumer sector. Right now, FPGAs are doing what they've been doing for quite some time: custom logic or glue logic. I see the Atom+FPGA products as aimed toward the embedded space, where applications can benefit from hardware-deterministic timing and customizable interface logic; definitely not for number-crunching purposes.
And in general, FPGAs are only good at very repetitive tasks that can be effectively pipelined. You might have 600MHz performance on paper, but once your design gets complex, the on-chip routing starts to get taxed and becomes the bottleneck. Every successive generation of FPGAs improves on both logic and interconnect, but price/performance ratio and the limited number of applications preclude widespread adoption. And in the meantime, conventional CPUs and GPUs will continue to improve.
I hate LabVIEW myself, but it's pretty entrenched in the non-EE engineering disciplines, and also to some extent in physics. NI is very good at selling things.
The collider and hot fusion people keep saying 'fund us more and we'll get huge energy breakthroughs, but the reality always seems to fall a long way short.
I don't think the collider/HEP folks have ever promised energy breakthroughs; that's not what they're interested in. Collider science is pure science.
The DOE fusion budget, on the other hand, has basically been falling monotonically since the 1970s.
The way the story title is written is very bad. Strictly speaking, MiniBooNE has shown an excess of events which do not match Standard Model predictions, which may or may not be explained by the presence of an additional neutrino.
From the abstract, there is "a probability for consistency with the background-only hypothesis of 0.5%", which puts it into the "very suggestive" category rather than the "done deal" category.
For a high-speed serial connection like SATA, it's more useful to consider an eye diagram, which tells you how much margin you have in voltage and time. The clock and data go along the same line; the receiver does clock recovery and maybe error correction. The SATA spec has a target BER (bit error rate) of something like 10^-15, so at gigabit rates you would expect an error every 10^6 seconds (~12 days).
That said, it's ridiculous that SATA cables could affect digital audio, because obviously the CPU and memory that the audio data pass through leave a much more significant audio signature. Getting lower CAS RAM makes the soundstage deeper and broader, and my Phenom X6 enhanced the highs without muddying the... seriously, how deep does this rabbit hole go?
This reminds me of a guy in a thread who was convinced that files ripped from newfangled "Blu-Spec" CDs were better than files ripped from regular CDs. He posted a few tracks for comparison, and of course the CRCs and MD5s were the same. When he said he didn't trust the CRC, a bit compare was done. Still insisted there was a difference, gotta keep an open mind, etc. I'm convinced that these people are on the wrong side of the "sufficiently advanced technology is indistinguishable from magic" divide.
Brain fart: two orders of magnitude greater temperature to reach one order of magnitude lesser reaction rate, or 100 times the temperature for 1/10 the reaction rate.
Take a 25W resistor and pass a 10 amp current through it and it gets VERY hot; pass a 1 amp current at 10 times the voltage, and it doesn't.
V = IR. You can't do what you have proposed, it's impossible. If you reduce the current by a factor of ten, the voltage must decrease by a factor of ten as well, the resistance being constant.
My friend once snaggged some nVIDIA chips at a MacWorld Expo. This would've been around 2000, 2001? I took a hammer to one later and it was definitely real.
Slashdot Mirror
Yeah, going to optical when copper will do doesn't make economic sense. There's no need for the bigger bandwidth of optical as long as the transceivers are fixed and non-upgradeable, as they would be in consumer equipment. Even in the high-end space, there seems to be lots of 10GigE over copper these days.
I'll add that "science has generated net positive fusion" is not true. He's probably thinking of JT-60, which achieved conditions where, if deuterium/tritium (DT) had been used as the fuel, it would have meant breakeven. However JT-60 doesn't have tritium handling facilities, so they can only use deuterium and extrapolate to DT's higher cross section. Still, it's a remarkable achievement.
In my mind, the biggest obstacle to fusion power is finding/designing materials that will last several years in a commercial reactor. And we have hardly anything at the moment.
That's correct. I run in limited mode, and have a RunAs command set up for administrative tasks. It's not possible to copy/paste or drag between explorer windows started with different privileges.
I don't see this happening anytime soon. FPGA performance is nowhere near the $/performance ratio that would make sense for the consumer sector. Right now, FPGAs are doing what they've been doing for quite some time: custom logic or glue logic. I see the Atom+FPGA products as aimed toward the embedded space, where applications can benefit from hardware-deterministic timing and customizable interface logic; definitely not for number-crunching purposes.
And in general, FPGAs are only good at very repetitive tasks that can be effectively pipelined. You might have 600MHz performance on paper, but once your design gets complex, the on-chip routing starts to get taxed and becomes the bottleneck. Every successive generation of FPGAs improves on both logic and interconnect, but price/performance ratio and the limited number of applications preclude widespread adoption. And in the meantime, conventional CPUs and GPUs will continue to improve.
Sounds like they're pricing it close to actual cost.
You do realize that 75k puts one in nearly the top 10% of personal income, right? What are we supposed to do after we "wake up," go into finance?
I hate LabVIEW myself, but it's pretty entrenched in the non-EE engineering disciplines, and also to some extent in physics. NI is very good at selling things.
The collider and hot fusion people keep saying 'fund us more and we'll get huge energy breakthroughs, but the reality always seems to fall a long way short.
I don't think the collider/HEP folks have ever promised energy breakthroughs; that's not what they're interested in. Collider science is pure science.
The DOE fusion budget, on the other hand, has basically been falling monotonically since the 1970s.
Why'd you go with XMOS over, say, additional FPGA fabric?
And it's only 60K logic elements, making it clear that this device is not intended for number crunching.
The way the story title is written is very bad. Strictly speaking, MiniBooNE has shown an excess of events which do not match Standard Model predictions, which may or may not be explained by the presence of an additional neutrino.
From the abstract, there is "a probability for consistency with the background-only hypothesis of 0.5%", which puts it into the "very suggestive" category rather than the "done deal" category.
Background info:
http://en.wikipedia.org/wiki/Liquid_Scintillator_Neutrino_Detector
http://en.wikipedia.org/wiki/MiniBooNE
For a high-speed serial connection like SATA, it's more useful to consider an eye diagram, which tells you how much margin you have in voltage and time. The clock and data go along the same line; the receiver does clock recovery and maybe error correction. The SATA spec has a target BER (bit error rate) of something like 10^-15, so at gigabit rates you would expect an error every 10^6 seconds (~12 days).
That said, it's ridiculous that SATA cables could affect digital audio, because obviously the CPU and memory that the audio data pass through leave a much more significant audio signature. Getting lower CAS RAM makes the soundstage deeper and broader, and my Phenom X6 enhanced the highs without muddying the... seriously, how deep does this rabbit hole go?
This reminds me of a guy in a thread who was convinced that files ripped from newfangled "Blu-Spec" CDs were better than files ripped from regular CDs. He posted a few tracks for comparison, and of course the CRCs and MD5s were the same. When he said he didn't trust the CRC, a bit compare was done. Still insisted there was a difference, gotta keep an open mind, etc. I'm convinced that these people are on the wrong side of the "sufficiently advanced technology is indistinguishable from magic" divide.
He also seems to have books on "The Physics of Miracles," "The Physics of UFOs," and "The Physics of Spirits."
Brain fart: two orders of magnitude greater temperature to reach one order of magnitude lesser reaction rate, or 100 times the temperature for 1/10 the reaction rate.
Helium-3 fusion would be nice, but it's even harder to do than DT fusion.
http://www.kayelaby.npl.co.uk/images/p548.jpg
Look at the curve that says DT, and compare it to the one that says 3He-3He. Note how you need twice the temperature to reach half the reaction rate.
In my view, we've got to build DT fusion reactors before we can even hope to make helium-3 ones.
You can get the best of both worlds: buy WoG direct from 2dboy, then put your serial into Steam.
I didn't do it intentionally but it worked for me.
I read the HAPL page you linked and it looks like a direct-drive program. You probably got superseded by HiPER, the European project.
Focus Fusion is definitely fringe-y.
Back when I was in school, the top sharer (one standout guy) had a measly terabyte. Technology marches on.
They do measure sea levels with satellites, just not with lasers.
http://news.bbc.co.uk/2/hi/science/nature/6922312.stm
Take a 25W resistor and pass a 10 amp current through it and it gets VERY hot; pass a 1 amp current at 10 times the voltage, and it doesn't.
V = IR. You can't do what you have proposed, it's impossible. If you reduce the current by a factor of ten, the voltage must decrease by a factor of ten as well, the resistance being constant.
Yeah, this jibes pretty nicely with Virtex-6 availability. Now if Xilinx can just get the Spartan-6 out the door...
My friend once snaggged some nVIDIA chips at a MacWorld Expo. This would've been around 2000, 2001? I took a hammer to one later and it was definitely real.
NVIDIA's already got an ARM platform: Tegra.