Someone correct me if I'm wrong, but I believe all AMD CPUs include virtualization extensions. I know my cheepie $400 AMD HP machine does and I do quite a bit of work in a virtualized XP instance without issue. The machine is certainly a commodity laptop, but it serves me very well for developing on the go (not that I'd ever actually recommend one to someone)
The stock 2GB of RAM can be a pain at times (but doable) so I'd suggest at least 4GB if you want a comfortable virtualized development environment.
But you're ignoring the costs of gaining approval to actually sell a drug in a given country. In the US, gaining FDA approval on a new drug easily costs an order of magnitude more than research and marketing combined. The level of regulation and oversight of every tiny detail of the whole process is incredible and this adds significant cost. Your safety is the main reason why drugs are so expensive, not research or marketing.
Not just that, having to carry the thing would be a pain. To put it into perspective, the average (unloaded) M16A4 weapon system weighs somewhere around 7-8 lbs and a loaded 30 round aluminum magazine is about a pound. An M4 Carbine weighs even less. Basically, 10 lbs is your rifle and ammo. I can't imagine too many Marines would be thrilled to have to hump an extra 10 lbs of potentially useless gadgetry all day, every day. 10 lbs is a lot heavier than it sounds.
0.8% unemployment? 2%? Really? I know that sounds great, but that's no more healthy than 15% unemployment. In fact, if you really had that low of an unemployment rate, the rate of inflation would be through the roof as employers paid increasingly ridiculous salaries to try to fill positions. Such a low unemployment rate simply means you're lacking an employable workforce. You want there to constantly be people in transition otherwise the economy has no where to grow. That's just ECON 101.
When was the last time you woke up thinking "I hope I don't get eaten today" or "I hope I can kill myself some food today so I don't starve"? Seeing as you have the time to comment on/. I'd assume never. Why? Because society provides you everything you need. What other species on this planet has such a complex society that ensures the survival of its members? None
Look at it this way - even with such a low rate of reproduction (unlike insects or bacteria as others suggest), humans have been able to populate the the entire planet with an ever increasing rate of population growth. It is because of our intelligence and technological innovation that we are able to maintain such a large population. If it weren't for things like modern farming techniques and refrigeration, the human population would have hit a ceiling and started dying off a while ago.
I guess we're so far removed from being concerned about survival that we forget that all this gadgetry we have evolved from the need to survive.
Have you tried using 64-bit Vista on the desktop though? It's really not that straight forward. Not everyone bothers with WHQL certification, and not every application plays nice on a 64-bit platform. I put it on my laptop for a period of time but had issues with a number of 3rd party apps and drivers behaving incorrectly. I was able to manage most of the time (the majority of my problems were with games), but the show stopper for me was the OpenVPN drivers - I just couldn't get the OpenVPN drivers to work at all. All-in-all, it just wasn't worth the headache (didn't really need 64-bit other than for the shits and giggles)and back went 32-bit XP. I'm not sure who is more to blame here - MS or the 3rd parties - but in any case it wasn't a pleasant experience.
OTOH, I've been running 64-bit Ubuntu for years now without an issue (except for Flash, but that's been resolved by Adobe). To the user, there is no difference between the 32-bit or 64-bit install and everything 'just works'. I realize this has a lot to do with the nature of Linux, the use of package management and my specific hardware, but I have to agree 64-bit Windows does feel like a bastard child in comparison.
Its actually very common to use coaxial like structures on PCB boards by placing sensitive signal traces between ground layers in a grounded copper pour. This works very well for shielding analog signals, but it doesn't work for high frequency digital signals. The grounded copper surrounding the trace creates a significant amount of capacitance that needs to be overcome every time there is a change in state. Once you get into the hundreds or thousands of megahertz, you start to consume more and more power to keep the rise times to acceptable levels. Eventually, you get to a point of diminishing returns where you're consuming huge amounts of power to drive signals and spending extremely large amounts of design time trying to minimize inductance and capacitance amongst traces - this is basically where we are now.
Also notice how we've all but abandoned parallel buses in PCs. This is because it is almost impossible with copper to ensure the multiple signals of a parallel bus will arrive at their destination at the same time - creating what is called a race condition. At higher frequencies, you'll start seeing data bits from the same message arriving before and after the clock signal which results in an unreliable interconnect. We have reached this point years ago, and that's why everything is using serial today - SATA, PCIe, USB, etc - serial buses eliminate the possibility for a race condition at the cost of requiring much higher clock speeds for the same throughput. Optical interconnects on the other hand could be used to much, much higher frequencies before we start seeing race conditions thus allowing for a very high speed parallel bus.
I don't think adding fiber to a PCB would really increase the cost of a PCB all that much, at least not when it becomes a common practice. I don't think the fabrication process would be all that much different from copper in the end; I don't see it being much more than routing a transparent plastic instead of copper. Granted, you'd be more limited in your routing paths, but that's nothing an extra layer wouldn't fix. So yeah, I guess it would increase costs a bit, at least at first, but it would probably reach the commodity level pricing we are at now pretty quickly. (The real cost increase will probably be the connectors and not the PCB itself.)
The point is we have just about maxed out copper and there is no other long term option but to move to optical interfaces.
While TFS fails to mention it, a major part of TFA is that they are partnering with the likes of Freescale, Qualcomm, and Texas Instruments. These 3 companies are all semiconductor manufacturers, like AMD and Intel, and produce some of the most exciting and leading edge ARM based mobile multimedia solutions out there. You can't deny that it's odd the x86 architecture goes completely unrepresented.
IMO, ChromeOS is probably going to be geared more for cheap 'internet appliances' where ARM is much better suited. If you don't need to worry about platform compatibility, why would you even consider a legacy ISA like x86 that consumes more power and runs hotter? I'm sure they're going to support x86 simply because it currently dominates the available hardware, but I'd bet COTS ChromeOS devices will mostly come packing ARMs.
1. He claims that tempratures have been trending downwards for the past 11yrs - this can be debunked by a simple google search [google.com.au] and is laughable to anyone who has looked at the temprate records.
Did that, am I missing something? It looks like he is right - NASA's data shows a pretty clear leveling off after 2000. Now I do see headlines trumpeting '2008 was the 7-10th warmest year on record', but seeing as how that totally disregards any rate of change or relative comparison to the previous 5 years, it is an intellectually dishonest claim at best. But I'm no climate scientist either, so maybe I'm looking at it wrong.
My biggest issue is, and maybe you can clear this up for me, what is the level of certainty that higher CO2 levels cause higher global temperatures? Now I know CO2 is a greenhouse gas in laboratory experiments, atmospheric CO2 levels have been increasing exponentially, and global temperatures have been increasing over the past 100+ years, but I don't understand how one can claim, with a relatively high level of confidence that the three are related. I can definitely see the good hypothesis there, but I just don't see anything truly supporting the hypothesis, at least not to the 'sure thing' degree many seem to claim. Furthermore, say if the Earth is on a cooling trend over the past decade even though CO2 levels are at ever increasing record highs, wouldn't that bring the hypothesis into question? Not saying it turns it on its head by any means, but I'd think it would, as part of good process, beg for additional consideration. But again, I'm no climate scientist so maybe I'm missing something.
I guess I just don't buy the 'it's a shut case' when we're talking about such an incredibly complex system - such claims raise my bullshit meter and make me question what is being discussed. Given that thought, it seems a bit dubious that the US House would, uncharacteristically, pull legislation out of committee to rush it through a vote - especially when the bill could actually pass.
the incentive is to extract oil as quickly as possible
That's actually not true at all. The idea is to extract oil at a controlled rate given the estimated rate of consumption and the desired market price. This is because oil producers know they have a limited supply of oil, and extracting it as quickly as possible will only flood the market causing prices to drop, and thus severely hurting their profits in the long term.
Furthermore, you fail to take into account new oil extraction technologies. Oil companies spend large sums on researching methods to better extract oil. This equates to new methods that allow more oil to be extracted from exiting fields and previously unknown or unreachable fields to be tapped.
The argument that peak oil will cause a sudden, catastrophic jump in energy cost is seriously flawed. In reality, it will (is) gradually wean the world off oil. Energy prices have been on a steady increase for sometime now and has thus spurred individuals and corporations alike to seek more efficient ways of doing things.
I wish I had mod points because I agree with you completely - I am forced to announce my burning hatred of Windows every time I have to blindly enter my 40+ character WPA password into an XP machine... twice.
My biggest gripe is, why do I need to hide my wifi password anyway? If someone can get close enough to me to be able to read the key on my screen, why would I care if they access my wireless? I just can't think of a single instance of not wanting my WPA key to be shown.
I have one with Verizon and I'm fairly sure you can restrict access to it by handset if you like. I've never bothered because I really couldn't care less if some stranger uses a bit of my bandwidth. Furthermore, it only allows a limited number of calls at one time - 5 IIRC - so it's not like the entire neighbourhood will be using your internet connection. Lastly, the range isn't that great, (remember, this thing is using licensed spectrum) so if you live somewhere where you need this (ie in a non-densely populated area) I highly doubt your neighbours can connect to it anyway. Heck, I can't connect to it from more than a few feet outside my house.
I do agree the cost is kind of high (I feel less as bad for myself seeing as those in the UK really get screwed with a monthly fee) but if you need service, you need service. I'd rather cough up the extra cash to have good coverage and live where I want to live (aka bumfuck). The only other option would be a cellular repeater with an antenna mounted on your roof and that'd be significantly more expensive and dependant on the reception already available.
What about the TI OMAP3 platform that uses a GPU from Imagination Technologies/PowerVR? Or what about Freescale's i.MX515 platform? Not to mention both of these devices currently use the ARM Cortex A-8, while Tegra is still using the ARM11. Oh, and the best part is these devices are on the market now.
But yeah, there is a lot of competition in the high performance, low power market and NVidia is just the newest entrant. Frankly, as an embedded developer who is currently evaluating such a solution, there is nothing I've heard about Tegra that particularly peaks my interest. Sure NVidia is a big name in the consumer computer market, but that doesn't mean that translates to the embedded world. I'm much more interested in what is coming from TI and Freescale, and they both have excellent OSS support.
Sigh, I hate this common misconception and complete lack of understanding of how drug and medical research work in the US.
First, the commercials for drugs you see are for only the top selling, highly competitive drugs that depend on patient awareness of the products in the market. These are things like allergy medicine, sleeping aids, anti-depression drugs, herpes remedies, cures for erectile dysfunction, etc. These drugs are all the top earners for their respective companies as they are the most used drug categories out there. In order for these companies to remain profitable and continue to produce less popular drugs (for things like rare cancers and the such), they have to actively compete with the other products on the market.
Next, in the US, medical research is the cheap part. In fact, compared to the entire price of bringing the entire product to market, it's dirt, dirt cheap. The cost of sales and advertisement is also dirt cheap in the big picture. Yes, there is a lot of public funding into research of compounds and medical procedures and the like that go to universities and other such research groups. So these groups will produce hundreds of compounds that show promise in the lab, but now what? This is where the biotech and pharma companies come in. Usually, they'll sign some kind of agreement, offering to pay X amount of cash and provide Y amount of royalties to the research group and will pick up where they left off to bring it to market. Now there is no guarantee the compound will make it to market - it may not work, it may kill people, who knows.
So now that the companies have licensed rights to these compounds, they continue the work on bringing it to market per the FDA requirements. This usually includes 2 additional phases of testing, each phase containing multiple trials. These trials often contain 1,000's and even 10,000's of patients, each costing the company money. There is also the need for whole departments of data managers to keep all of the data pouring in in order. In the end, it winds up costing 100's of millions of dollars to bring a single compound to market - regardless if it is the next top selling allergy medicine or the cure for some rare cancer that only a handful of people in the world will need. Now this doesn't even include the compounds that fail somewhere in the pipeline - basically equating to 10's or 100's of millions of dollars being pissed away for nothing. After all of these expenses, they usually only have a few years left on the patent to make a return before the drug goes generic
My point is, you're not paying for the research, you're paying for the FDA certification that it is safe. The requirements in most other countries are much more lenient, and this is why you'll see most US drug trials being conducted in Europe, Australia, and New Zealand.
Oh, and the government has much, much less money than all the private investors willing to put money into these things. Once you make research public, and part of the public domain, you lose this private funding, and thus you lose the majority of your funding.
I'm sorry, this has nothing to do with ending our dependency on oil or curbing carbon emissions - it's about creating another tax that will be invisible to the consumer.
First of all, it will be nearly impossible to get that kind of fuel mileage for a fleet average in such a short period of time without outright dictating what your customers buy. And yes it is a very short period of time - the next few model years are already slated for production, parts are ordered, and contracts are signed - putting together a production car takes a good amount of time. It will require designing new, lighter frames and more efficient power-trains. They could probably bang out a couple models, but they'd most likely be half-assed and avoided by customers (this is what happened with the last time this came around - anyone remember cars in the late 70's and 80's?).
The main issue, however, is consumers LOVE huge inefficient cars, and auto makers make HUGE profits off of them. It will undoubted be cheaper and easier for them to simply take their time getting to the new restrictions and continue selling their current line-up. The government will slap them with a fine which they'll gladly pay and pass along to the consumer.
In the end, the government has more of your money and your support, and the world isn't any greener.
I actually visit The Pirate Bay on my phone somewhat regularly - when I'm on the go and think of something I want, I search TPB from my phone, then SSH into my home machine to download it with rtorrent. So no, I don't actually download torrents over 3G, but I do search for torrents over 3G.
The Euro has always been worth more than the Dollar. Since the financial crisis though, the difference has actually lessened significantly. (~$1.50 per Euro then vs ~$1.30 per Euro now).
Now $100 Canadian is like $0.06 US
(PS, it's fun to pick on other nation's currencies, even if it's completely baseless - especially Canada)
I have seen memtest give a 'false' positive - I had been running a pair of DIMMs and decided to upgrade and add another pair. I wasn't able to find a new pair that matched the first, so I was forced to use a different manufacturer. With the new pair installed, the system became very unstable and memtest reported a huge number of errors. At first I thought the new memory was bad, but after further testing I found each pair would test out fine by itself, and would only fail when both were installed at the same time.
The point is, failing a memtest doesn't always mean you have bad memory, but could also indicate a memory controller incompatibility, incorrect BIOS setting, or even a motherboard/memory controller failure.
If the hardware requirements are simple enough, I would seriously consider doing the design in-house.
I have my own start-up company and faced the same situation when I first started. I am a computer engineer though, and can do the PCB design myself, but would rather outsource as much as possible. Long story short, after getting a number of quotes, a production ready PCB design (sans software) was around $50k for a fairly straight forward design. I don't know about you, but I didn't have that kind of cash to dump into a single design (I'm working with 4 discrete boards now).
Anyway, I started with PCB123 as it is free to get started and very easy to use. The downside is it does lock you into their PCB manufacturer (Sunstone Circuits) but it is a good place to start for prototype and initial production runs.
Furthermore, for prototypes and small scale production board assembly, I recommend hand assembly. Get yourself a pair of tweezers, a tub of solder paste (I prefer Kester Easy Profile 256, but I work in a RoHS exempt industry so this may not work for you), an electric skillet and a SMT rework station. For small batches, this is by far cheaper and faster than machine assembly - It's just not worth the time and cost of having an assembly house setup their equipment to only assemble a handful of boards. When you're ready for a final production run with a few thousand units, you'll obviously need the machine assembly - until then stick with hand assembly. I assure you hand SMT assembly is much easier than it may seem; I actually prefer it to hand through-hole assembly and now avoid through-hole like the plague on any designs I'll need to hand assemble.
Once you have a working design, you can dump your design files onto a production house and have the thing produced - many places you just need to send them your gerber files and bill of materials and they'll take care of the rest - PCB123 offers such a service via their partners. I would suggest staying local and avoiding China. I've heard many horror stories and what you save in production costs, you'll lose in quality, shipping costs, and turn-around times. When starting off, being able to call someone for support who speaks the same language as you and works the same hours as you is nearly priceless. Perhaps look to China for things like enclosures - I can't really give you advice there as that is not something I've had to deal with.
If you're capable of writing the embedded software for such a device, I'd be willing to bet you'd be capable of designing and building the device as well. It may take a bit of extra effort and require you to learn some new things, but I really don't think it is as difficult as it may seem. Look for any development kits that can do what you need, and then copy and modify their schematics for your own design - this is why part manufacturers produce these kits. If you don't feel comfortable doing this yourself, reach out to any electrical engineering friends you might have, or hire an EE as a part-time contractor. Your best bet may be to do the bulk of the leg work yourself, and hire an EE to look over your shoulder and give you guidance.
It works VERY well. Decoding a 1080p H.264 video using software on my dual-core 3GHz machine pegs the cores back and forth to around 80-90%, but plays fairly well. Using an mplayer build patched to support VDPAU, my CPU remains idle (clock drops to 1GHz, and 1-3% CPU usage) and plays equally as well, if not better. Furthermore, I was still able to go about my business with no noticeable impact on performance, even when using hardware-dependent Compiz. I have not though, tried this on multiple streams.
Oh, and this is using some cheap NVidia 8600 something-or-other card that I picked up new for ~$50. I, for one, was truly impressed by VDPAU and what it means for low-cost HD content.
How's your German? or Russian? What, still speaking English? Still able to open you big stupid mouth and not get thrown in prison and/or shot? Oh, that's right, you're just another spoiled, naive, and unappreciative jerk off.
I'm sorry, I try really hard not to feed the trolls...
Libertarians are very socially liberal, not conservative. This is a key part of their views - the government shouldn't have a say in what individuals are allowed to do with themselves. As for the hot button issues, Libertarians are generally in favor of legalized abortions, same-sex marriage, decriminalization of vices (drugs use, prostitution, gambling), etc.
I should also note, Bob Barr is not a good example of a mainstream Libertarian. A lot of people were pretty surprised by his nomination to the Libertarian presidential ticket.
I think there is a misunderstanding here either in the definition of what a CPU/GPU is or in the difference between the two.
A 'CPU' in this context is a general purpose processor core like the x86. It's job is to be able to do everything reasonably well - namely fetch and decode instructions, read and write to memory, and do some simple math. A 'GPU' is a special purpose processor core that is designed to only be able to crunch numbers, but do it very, very well. A CPU is already like a GPU in that it can do all of the GPU's functions, just not very well. Adding more such functionality to a CPU is just going to make a crappy CPU in the long run. And adding more general purpose functionality to a GPU is just going to make for a crappy GPU.
The devices you named (Fusion, etc) are not CPUs, or GPUs, but system-on-chips. They contain both a CPU and GPU (along with some other system functionality) on a single die. A 'CPU' is not the thing you plug into your motherboard anymore, but simply a single part of it. This is where things are going, and where they should be going. All of the SSE, MMX, etc crap needs to be offloaded from the CPU and given to one or more GPUs or DSPs on the die. The point is, we are moving to more discrete, specialized processors that can do fewer tasks, but do them very well - not some monolith do everything processor. CPUs and GPUs are very different animals and will remain as such - it just may be less clear to the consumer as they are squeezed into a single piece of silicon.
Take a look at embedded processors as an example - namely the TI OMAP3 series. On a single die, fit inside a ~2cm^2 package, contains an ARM core, a GPU, and a DSP. This architecture of using a small and simple RISC general purpose CPU and off loading the more complex functions onto special purpose processor cores is allowing for mobile devices to decode 720p video and render 3D graphics in real time while consuming ~1 watt. My two year old laptop can barely do this while sucking down 50+ watts and heating the entire room.
You are right the GPU is becoming more integral to the whole system, but the CPU and the GPU are not converging, but rather quite the opposite - this is the beginning of an era where discrete functionality will be broken off into specialized processors cores optimized for single tasks.
Slashdot Mirror
Someone correct me if I'm wrong, but I believe all AMD CPUs include virtualization extensions. I know my cheepie $400 AMD HP machine does and I do quite a bit of work in a virtualized XP instance without issue. The machine is certainly a commodity laptop, but it serves me very well for developing on the go (not that I'd ever actually recommend one to someone)
The stock 2GB of RAM can be a pain at times (but doable) so I'd suggest at least 4GB if you want a comfortable virtualized development environment.
I prefer to play Freecell and manage my servers over SSH on my phone while driving to work in the morning - good thing they didn't ban that yet!
But you're ignoring the costs of gaining approval to actually sell a drug in a given country. In the US, gaining FDA approval on a new drug easily costs an order of magnitude more than research and marketing combined. The level of regulation and oversight of every tiny detail of the whole process is incredible and this adds significant cost. Your safety is the main reason why drugs are so expensive, not research or marketing.
Not just that, having to carry the thing would be a pain. To put it into perspective, the average (unloaded) M16A4 weapon system weighs somewhere around 7-8 lbs and a loaded 30 round aluminum magazine is about a pound. An M4 Carbine weighs even less. Basically, 10 lbs is your rifle and ammo. I can't imagine too many Marines would be thrilled to have to hump an extra 10 lbs of potentially useless gadgetry all day, every day. 10 lbs is a lot heavier than it sounds.
0.8% unemployment? 2%? Really? I know that sounds great, but that's no more healthy than 15% unemployment. In fact, if you really had that low of an unemployment rate, the rate of inflation would be through the roof as employers paid increasingly ridiculous salaries to try to fill positions. Such a low unemployment rate simply means you're lacking an employable workforce. You want there to constantly be people in transition otherwise the economy has no where to grow. That's just ECON 101.
When was the last time you woke up thinking "I hope I don't get eaten today" or "I hope I can kill myself some food today so I don't starve"? Seeing as you have the time to comment on /. I'd assume never. Why? Because society provides you everything you need. What other species on this planet has such a complex society that ensures the survival of its members? None
Look at it this way - even with such a low rate of reproduction (unlike insects or bacteria as others suggest), humans have been able to populate the the entire planet with an ever increasing rate of population growth. It is because of our intelligence and technological innovation that we are able to maintain such a large population. If it weren't for things like modern farming techniques and refrigeration, the human population would have hit a ceiling and started dying off a while ago.
I guess we're so far removed from being concerned about survival that we forget that all this gadgetry we have evolved from the need to survive.
Have you tried using 64-bit Vista on the desktop though? It's really not that straight forward. Not everyone bothers with WHQL certification, and not every application plays nice on a 64-bit platform. I put it on my laptop for a period of time but had issues with a number of 3rd party apps and drivers behaving incorrectly. I was able to manage most of the time (the majority of my problems were with games), but the show stopper for me was the OpenVPN drivers - I just couldn't get the OpenVPN drivers to work at all. All-in-all, it just wasn't worth the headache (didn't really need 64-bit other than for the shits and giggles)and back went 32-bit XP. I'm not sure who is more to blame here - MS or the 3rd parties - but in any case it wasn't a pleasant experience.
OTOH, I've been running 64-bit Ubuntu for years now without an issue (except for Flash, but that's been resolved by Adobe). To the user, there is no difference between the 32-bit or 64-bit install and everything 'just works'. I realize this has a lot to do with the nature of Linux, the use of package management and my specific hardware, but I have to agree 64-bit Windows does feel like a bastard child in comparison.
When my shit turns purple and smells like rainbow sherbet.
Its actually very common to use coaxial like structures on PCB boards by placing sensitive signal traces between ground layers in a grounded copper pour. This works very well for shielding analog signals, but it doesn't work for high frequency digital signals. The grounded copper surrounding the trace creates a significant amount of capacitance that needs to be overcome every time there is a change in state. Once you get into the hundreds or thousands of megahertz, you start to consume more and more power to keep the rise times to acceptable levels. Eventually, you get to a point of diminishing returns where you're consuming huge amounts of power to drive signals and spending extremely large amounts of design time trying to minimize inductance and capacitance amongst traces - this is basically where we are now.
Also notice how we've all but abandoned parallel buses in PCs. This is because it is almost impossible with copper to ensure the multiple signals of a parallel bus will arrive at their destination at the same time - creating what is called a race condition. At higher frequencies, you'll start seeing data bits from the same message arriving before and after the clock signal which results in an unreliable interconnect. We have reached this point years ago, and that's why everything is using serial today - SATA, PCIe, USB, etc - serial buses eliminate the possibility for a race condition at the cost of requiring much higher clock speeds for the same throughput. Optical interconnects on the other hand could be used to much, much higher frequencies before we start seeing race conditions thus allowing for a very high speed parallel bus.
I don't think adding fiber to a PCB would really increase the cost of a PCB all that much, at least not when it becomes a common practice. I don't think the fabrication process would be all that much different from copper in the end; I don't see it being much more than routing a transparent plastic instead of copper. Granted, you'd be more limited in your routing paths, but that's nothing an extra layer wouldn't fix. So yeah, I guess it would increase costs a bit, at least at first, but it would probably reach the commodity level pricing we are at now pretty quickly. (The real cost increase will probably be the connectors and not the PCB itself.)
The point is we have just about maxed out copper and there is no other long term option but to move to optical interfaces.
While TFS fails to mention it, a major part of TFA is that they are partnering with the likes of Freescale, Qualcomm, and Texas Instruments. These 3 companies are all semiconductor manufacturers, like AMD and Intel, and produce some of the most exciting and leading edge ARM based mobile multimedia solutions out there. You can't deny that it's odd the x86 architecture goes completely unrepresented.
IMO, ChromeOS is probably going to be geared more for cheap 'internet appliances' where ARM is much better suited. If you don't need to worry about platform compatibility, why would you even consider a legacy ISA like x86 that consumes more power and runs hotter? I'm sure they're going to support x86 simply because it currently dominates the available hardware, but I'd bet COTS ChromeOS devices will mostly come packing ARMs.
1. He claims that tempratures have been trending downwards for the past 11yrs - this can be debunked by a simple google search [google.com.au] and is laughable to anyone who has looked at the temprate records.
Did that, am I missing something? It looks like he is right - NASA's data shows a pretty clear leveling off after 2000. Now I do see headlines trumpeting '2008 was the 7-10th warmest year on record', but seeing as how that totally disregards any rate of change or relative comparison to the previous 5 years, it is an intellectually dishonest claim at best. But I'm no climate scientist either, so maybe I'm looking at it wrong.
My biggest issue is, and maybe you can clear this up for me, what is the level of certainty that higher CO2 levels cause higher global temperatures? Now I know CO2 is a greenhouse gas in laboratory experiments, atmospheric CO2 levels have been increasing exponentially, and global temperatures have been increasing over the past 100+ years, but I don't understand how one can claim, with a relatively high level of confidence that the three are related. I can definitely see the good hypothesis there, but I just don't see anything truly supporting the hypothesis, at least not to the 'sure thing' degree many seem to claim. Furthermore, say if the Earth is on a cooling trend over the past decade even though CO2 levels are at ever increasing record highs, wouldn't that bring the hypothesis into question? Not saying it turns it on its head by any means, but I'd think it would, as part of good process, beg for additional consideration. But again, I'm no climate scientist so maybe I'm missing something.
I guess I just don't buy the 'it's a shut case' when we're talking about such an incredibly complex system - such claims raise my bullshit meter and make me question what is being discussed. Given that thought, it seems a bit dubious that the US House would, uncharacteristically, pull legislation out of committee to rush it through a vote - especially when the bill could actually pass.
the incentive is to extract oil as quickly as possible
That's actually not true at all. The idea is to extract oil at a controlled rate given the estimated rate of consumption and the desired market price. This is because oil producers know they have a limited supply of oil, and extracting it as quickly as possible will only flood the market causing prices to drop, and thus severely hurting their profits in the long term.
Furthermore, you fail to take into account new oil extraction technologies. Oil companies spend large sums on researching methods to better extract oil. This equates to new methods that allow more oil to be extracted from exiting fields and previously unknown or unreachable fields to be tapped.
The argument that peak oil will cause a sudden, catastrophic jump in energy cost is seriously flawed. In reality, it will (is) gradually wean the world off oil. Energy prices have been on a steady increase for sometime now and has thus spurred individuals and corporations alike to seek more efficient ways of doing things.
I wish I had mod points because I agree with you completely - I am forced to announce my burning hatred of Windows every time I have to blindly enter my 40+ character WPA password into an XP machine... twice.
My biggest gripe is, why do I need to hide my wifi password anyway? If someone can get close enough to me to be able to read the key on my screen, why would I care if they access my wireless? I just can't think of a single instance of not wanting my WPA key to be shown.
I have one with Verizon and I'm fairly sure you can restrict access to it by handset if you like. I've never bothered because I really couldn't care less if some stranger uses a bit of my bandwidth. Furthermore, it only allows a limited number of calls at one time - 5 IIRC - so it's not like the entire neighbourhood will be using your internet connection. Lastly, the range isn't that great, (remember, this thing is using licensed spectrum) so if you live somewhere where you need this (ie in a non-densely populated area) I highly doubt your neighbours can connect to it anyway. Heck, I can't connect to it from more than a few feet outside my house.
I do agree the cost is kind of high (I feel less as bad for myself seeing as those in the UK really get screwed with a monthly fee) but if you need service, you need service. I'd rather cough up the extra cash to have good coverage and live where I want to live (aka bumfuck). The only other option would be a cellular repeater with an antenna mounted on your roof and that'd be significantly more expensive and dependant on the reception already available.
What about the TI OMAP3 platform that uses a GPU from Imagination Technologies/PowerVR? Or what about Freescale's i.MX515 platform? Not to mention both of these devices currently use the ARM Cortex A-8, while Tegra is still using the ARM11. Oh, and the best part is these devices are on the market now.
But yeah, there is a lot of competition in the high performance, low power market and NVidia is just the newest entrant. Frankly, as an embedded developer who is currently evaluating such a solution, there is nothing I've heard about Tegra that particularly peaks my interest. Sure NVidia is a big name in the consumer computer market, but that doesn't mean that translates to the embedded world. I'm much more interested in what is coming from TI and Freescale, and they both have excellent OSS support.
Sigh, I hate this common misconception and complete lack of understanding of how drug and medical research work in the US.
First, the commercials for drugs you see are for only the top selling, highly competitive drugs that depend on patient awareness of the products in the market. These are things like allergy medicine, sleeping aids, anti-depression drugs, herpes remedies, cures for erectile dysfunction, etc. These drugs are all the top earners for their respective companies as they are the most used drug categories out there. In order for these companies to remain profitable and continue to produce less popular drugs (for things like rare cancers and the such), they have to actively compete with the other products on the market.
Next, in the US, medical research is the cheap part. In fact, compared to the entire price of bringing the entire product to market, it's dirt, dirt cheap. The cost of sales and advertisement is also dirt cheap in the big picture. Yes, there is a lot of public funding into research of compounds and medical procedures and the like that go to universities and other such research groups. So these groups will produce hundreds of compounds that show promise in the lab, but now what? This is where the biotech and pharma companies come in. Usually, they'll sign some kind of agreement, offering to pay X amount of cash and provide Y amount of royalties to the research group and will pick up where they left off to bring it to market. Now there is no guarantee the compound will make it to market - it may not work, it may kill people, who knows.
So now that the companies have licensed rights to these compounds, they continue the work on bringing it to market per the FDA requirements. This usually includes 2 additional phases of testing, each phase containing multiple trials. These trials often contain 1,000's and even 10,000's of patients, each costing the company money. There is also the need for whole departments of data managers to keep all of the data pouring in in order. In the end, it winds up costing 100's of millions of dollars to bring a single compound to market - regardless if it is the next top selling allergy medicine or the cure for some rare cancer that only a handful of people in the world will need. Now this doesn't even include the compounds that fail somewhere in the pipeline - basically equating to 10's or 100's of millions of dollars being pissed away for nothing. After all of these expenses, they usually only have a few years left on the patent to make a return before the drug goes generic
My point is, you're not paying for the research, you're paying for the FDA certification that it is safe. The requirements in most other countries are much more lenient, and this is why you'll see most US drug trials being conducted in Europe, Australia, and New Zealand.
Oh, and the government has much, much less money than all the private investors willing to put money into these things. Once you make research public, and part of the public domain, you lose this private funding, and thus you lose the majority of your funding.
I'm sorry, this has nothing to do with ending our dependency on oil or curbing carbon emissions - it's about creating another tax that will be invisible to the consumer.
First of all, it will be nearly impossible to get that kind of fuel mileage for a fleet average in such a short period of time without outright dictating what your customers buy. And yes it is a very short period of time - the next few model years are already slated for production, parts are ordered, and contracts are signed - putting together a production car takes a good amount of time. It will require designing new, lighter frames and more efficient power-trains. They could probably bang out a couple models, but they'd most likely be half-assed and avoided by customers (this is what happened with the last time this came around - anyone remember cars in the late 70's and 80's?).
The main issue, however, is consumers LOVE huge inefficient cars, and auto makers make HUGE profits off of them. It will undoubted be cheaper and easier for them to simply take their time getting to the new restrictions and continue selling their current line-up. The government will slap them with a fine which they'll gladly pay and pass along to the consumer.
In the end, the government has more of your money and your support, and the world isn't any greener.
I actually visit The Pirate Bay on my phone somewhat regularly - when I'm on the go and think of something I want, I search TPB from my phone, then SSH into my home machine to download it with rtorrent. So no, I don't actually download torrents over 3G, but I do search for torrents over 3G.
The Euro has always been worth more than the Dollar. Since the financial crisis though, the difference has actually lessened significantly. (~$1.50 per Euro then vs ~$1.30 per Euro now).
Now $100 Canadian is like $0.06 US
(PS, it's fun to pick on other nation's currencies, even if it's completely baseless - especially Canada)
I have seen memtest give a 'false' positive - I had been running a pair of DIMMs and decided to upgrade and add another pair. I wasn't able to find a new pair that matched the first, so I was forced to use a different manufacturer. With the new pair installed, the system became very unstable and memtest reported a huge number of errors. At first I thought the new memory was bad, but after further testing I found each pair would test out fine by itself, and would only fail when both were installed at the same time.
The point is, failing a memtest doesn't always mean you have bad memory, but could also indicate a memory controller incompatibility, incorrect BIOS setting, or even a motherboard/memory controller failure.
If the hardware requirements are simple enough, I would seriously consider doing the design in-house.
I have my own start-up company and faced the same situation when I first started. I am a computer engineer though, and can do the PCB design myself, but would rather outsource as much as possible. Long story short, after getting a number of quotes, a production ready PCB design (sans software) was around $50k for a fairly straight forward design. I don't know about you, but I didn't have that kind of cash to dump into a single design (I'm working with 4 discrete boards now).
Anyway, I started with PCB123 as it is free to get started and very easy to use. The downside is it does lock you into their PCB manufacturer (Sunstone Circuits) but it is a good place to start for prototype and initial production runs.
Furthermore, for prototypes and small scale production board assembly, I recommend hand assembly. Get yourself a pair of tweezers, a tub of solder paste (I prefer Kester Easy Profile 256, but I work in a RoHS exempt industry so this may not work for you), an electric skillet and a SMT rework station. For small batches, this is by far cheaper and faster than machine assembly - It's just not worth the time and cost of having an assembly house setup their equipment to only assemble a handful of boards. When you're ready for a final production run with a few thousand units, you'll obviously need the machine assembly - until then stick with hand assembly. I assure you hand SMT assembly is much easier than it may seem; I actually prefer it to hand through-hole assembly and now avoid through-hole like the plague on any designs I'll need to hand assemble.
Once you have a working design, you can dump your design files onto a production house and have the thing produced - many places you just need to send them your gerber files and bill of materials and they'll take care of the rest - PCB123 offers such a service via their partners. I would suggest staying local and avoiding China. I've heard many horror stories and what you save in production costs, you'll lose in quality, shipping costs, and turn-around times. When starting off, being able to call someone for support who speaks the same language as you and works the same hours as you is nearly priceless. Perhaps look to China for things like enclosures - I can't really give you advice there as that is not something I've had to deal with.
If you're capable of writing the embedded software for such a device, I'd be willing to bet you'd be capable of designing and building the device as well. It may take a bit of extra effort and require you to learn some new things, but I really don't think it is as difficult as it may seem. Look for any development kits that can do what you need, and then copy and modify their schematics for your own design - this is why part manufacturers produce these kits. If you don't feel comfortable doing this yourself, reach out to any electrical engineering friends you might have, or hire an EE as a part-time contractor. Your best bet may be to do the bulk of the leg work yourself, and hire an EE to look over your shoulder and give you guidance.
Just my $0.02
It works VERY well. Decoding a 1080p H.264 video using software on my dual-core 3GHz machine pegs the cores back and forth to around 80-90%, but plays fairly well. Using an mplayer build patched to support VDPAU, my CPU remains idle (clock drops to 1GHz, and 1-3% CPU usage) and plays equally as well, if not better. Furthermore, I was still able to go about my business with no noticeable impact on performance, even when using hardware-dependent Compiz. I have not though, tried this on multiple streams.
Oh, and this is using some cheap NVidia 8600 something-or-other card that I picked up new for ~$50. I, for one, was truly impressed by VDPAU and what it means for low-cost HD content.
How's your German? or Russian? What, still speaking English? Still able to open you big stupid mouth and not get thrown in prison and/or shot? Oh, that's right, you're just another spoiled, naive, and unappreciative jerk off.
I'm sorry, I try really hard not to feed the trolls...
Libertarians are very socially liberal, not conservative. This is a key part of their views - the government shouldn't have a say in what individuals are allowed to do with themselves. As for the hot button issues, Libertarians are generally in favor of legalized abortions, same-sex marriage, decriminalization of vices (drugs use, prostitution, gambling), etc.
I should also note, Bob Barr is not a good example of a mainstream Libertarian. A lot of people were pretty surprised by his nomination to the Libertarian presidential ticket.
I think there is a misunderstanding here either in the definition of what a CPU/GPU is or in the difference between the two.
A 'CPU' in this context is a general purpose processor core like the x86. It's job is to be able to do everything reasonably well - namely fetch and decode instructions, read and write to memory, and do some simple math. A 'GPU' is a special purpose processor core that is designed to only be able to crunch numbers, but do it very, very well. A CPU is already like a GPU in that it can do all of the GPU's functions, just not very well. Adding more such functionality to a CPU is just going to make a crappy CPU in the long run. And adding more general purpose functionality to a GPU is just going to make for a crappy GPU.
The devices you named (Fusion, etc) are not CPUs, or GPUs, but system-on-chips. They contain both a CPU and GPU (along with some other system functionality) on a single die. A 'CPU' is not the thing you plug into your motherboard anymore, but simply a single part of it. This is where things are going, and where they should be going. All of the SSE, MMX, etc crap needs to be offloaded from the CPU and given to one or more GPUs or DSPs on the die. The point is, we are moving to more discrete, specialized processors that can do fewer tasks, but do them very well - not some monolith do everything processor. CPUs and GPUs are very different animals and will remain as such - it just may be less clear to the consumer as they are squeezed into a single piece of silicon.
Take a look at embedded processors as an example - namely the TI OMAP3 series. On a single die, fit inside a ~2cm^2 package, contains an ARM core, a GPU, and a DSP. This architecture of using a small and simple RISC general purpose CPU and off loading the more complex functions onto special purpose processor cores is allowing for mobile devices to decode 720p video and render 3D graphics in real time while consuming ~1 watt. My two year old laptop can barely do this while sucking down 50+ watts and heating the entire room.
You are right the GPU is becoming more integral to the whole system, but the CPU and the GPU are not converging, but rather quite the opposite - this is the beginning of an era where discrete functionality will be broken off into specialized processors cores optimized for single tasks.