Thunderbolt goes a bit higher, so a RAID array over thunderbolt might do over 600MB/s...
There's a lot of professional video gear that supports thunderbolt. Probably because a lot of professional video gear targets Mac. That's actually kind of annoying, and the third-party Windows drivers for HFS+ have spotty support. I couldn't get them working with the CF cards recorded on a KiPro Mini, for example. I had to hunt down somebody on-site with a macbook to read the damned things.
The integrated graphics chips in Intel's Ivy Bridge processors is capable of displaying and decoding h.264 video in hardware. They've been on the market since late 2011. I'm not talking about expensive nVidia cards here, I'm saying the integrated graphics Intel has been selling for over a year already supports 4K (although they only actually enabled output/decoding in October 2012, anybody who had Ivy Bridge hardware benefitted).
The downside is that it requires two DisplayPort outputs, and few motherboards ship with more than one. We won't get the ability from an Intel iGPU to output 4K video over a single DisplayPort connection until Haswell, due to ship in June 2013.
High end current nVidia graphics cards such as the GTX 670 support 4K output over a single HDMI or DisplayPort connection, however the official word from nVidia is that they've never actually tried it due to the lack of a display to test with (what limited 4K hardware that's out there doesn't do 4K over a single HDMI or DisplayPort connection).
Really, it's just early days. Generally, midrange quad core chips of today should have no problems decoding 4K video.
They don't look terribly useful for anything like image projection, but they'd be fine for a flashlight.
Of note is that they only hit 200 lumens per watt at 1W. If you crank it up to 15 W, they'll do 1769 lumens, which is only 118 lumens per watt. That's still an improvement over their previous best, the XM-L, which did 1040 lumens at 10 W, but it still means that if you want to maintain that 200 lumens per watt, you're going to need a bunch of LEDs to hit higher brightnesses.
The exhaust velocity on an ion thruster is nowhere near that of a partical accelerator. According to a 1996 NASA document I found, the exhaust velocity of an ion thruster is 31.5 kilometers per second, while particle accelerators are very close to the speed of light. The speed record is 0.999999999976c, which is roughly ten thousand times faster than the ion thruster exhaust velocity. That's not to say that thruster exhaust velocity is the speed limit, since the thruster exhaust velocity is relative to the thruster, not the third party observer measuring the speed of the spacecraft.
At some point, though, you're going to be getting hit pretty hard by interstellar particles. Even in diffuse regions, you get a particle density of 10^-4 particles per cubic centimeter, so if we assume a spacecraft with a surface area of 10,000 square centimeters (one square meter) traveling at 3000 kilometers per second, you're going to be hit by (if my math holds up), 300 million particles per second travelling at a rather high speed, although that's not actually that much energy. I don't know enough to say if that would cause a problem.
Umm, no. They had a hell of a time moving to 22nm and getting volume production up, pretty much your entire post is just paranoid delusion.
Intel just has more money to throw at the problems, and they've managed to get a lead of a few years on the other companies. They only maintain that lead because they keep pushing forward.
The problem is that makes it harder for everybody else to compete, but that's not really Intel's fault.
Consumers need some light multithreading, but single-threaded performance is still king. ECC is not something consumers care about (enterprise does, sure, but not consumers).
Power matters less in desktops, so using more power to do the same thing is not as big a problem there. But in the mobile space, it's a big problem.
The use cases you're talking about aren't really what a typical consumer or even office machine is used for.
Are they delivering in the APU market? The CPU performance is poor in their APU products, and the GPU performance is faster than Intel's, but not fast enough to matter. It's still too slow for gaming, and Intel's iGPUs are fast enough for office use.
If their APUs were matching Intel's power/performance, but still kicking their ass on the GPU side, then it might be something special. As it stands, the tradeoffs are just not worth it.
Amd's APUs are faster than Intel's iGPUs, but much slower than discrete chips. Why shouldn't you have an AMD CPU with a discrete GPU in the notebook market?
Sigh. I realize you're a troll, but I'll respond anyhow: getting half the single-threaded performance limits the usefulness of your CPUs. I haven't seen any reviews for piledriver-based Opteron chips (they're relatively new), but comparing the previous gen Opteron against the Sandy Bridge Xeons, "keep up" still means they're falling behind the top Xeon chips in multi-threaded performance, they're way behind in single-threaded performance, and they use a ton more power.
We already get all of our power from hydro, and there's still lots of untapped capacity. I don't think not being able to buy any is a risk anytime soon.
Benchmarks of the HD4000 would have been useless, as AMD sent a desktop chip, performance would not be representative of the mobile HD4000's performance. AMD could just have easily have sent an AMD CPU for apples-to-apples comparisons.
I think the fact that AMD's server chips need twice as many cores just to keep up with Intel's parts is kind of indicative of the problem, and I really do hope that they have something competitive in the market. Intel's current products are great, but only because AMD kicked them in the ass with the Athlon 64 years ago. If they go too long without a real competitor in the higher end of the market (beyond where ARM can reach), they'll stagnate.
That's not the point, the point is that recommending people use your competitor's product over your own is never something a company should do, unless you're in full mea culpa mode like the Apple Maps fiasco.
In the marketing world, you're always supposed to at least PRETEND that your products are superior.
I'm not trolling, and I've owned a few AMD CPUs in my day (four, I think? Does the Geode count?), I was just flabbergasted that AMD would send out Intel CPUs to review their GPUs with. I mean, eating your own dogfood is kind of a fundamental thing, and when you do something like this, it sends the message that your own products aren't good enough for the purpose. I chose an inflammatory title to highlight how ridiculous this is.
I would actually argue that AMD only holds an advantage at the extremely low-end, below the $40 pricepoint (you can get a dual-core sandy bridge for about that), and even then only if you don't care about power. I feel bad for AMD, because I owned a long string of AMD processors, and they were fantastic products. I started on the K6-2 and ended on the Athlon XP, and they were all great. With the Athlon 64, they really hit it out of the park, and had a fantastic run, and then... nothing. It didn't help that Intel had a fantastic chip with the Conroe, but AMD had flop after flop that wasn't even up to their own previous standards. They managed to get some of the worst problems under control to produce decent chips again, but by then they had fallen far enough behind Intel that they could only compete on price, and that approach was driving them bankrupt (hence the whole selling off Global Foundries).
I don't WANT Intel to dominate the CPU market, but AMD just isn't a credible competitor anymore. They haven't been able to compete on the performance front for years, their server chips have kept some of the hope alive but at this point need twice as many cores just to stay anywhere remotely performance competitive, their power efficiency hasn't been competitive for ages, and while their APU stuff has turned out some interesting products, they have a pretty limited market since they have better GPU performance and worse CPU performance than comparable Intel products, and Intel's GPU performance got "good enough" for the kinds of uses that you'd find those chips in anyhow.
In terms of competing on the ultra-low end prices, now ARM is starting to creep up in that market, and AMD is being sandwiched in-between. I think they're a more credible competitor to Intel at this point, but with Microsoft crippling Windows 8 on ARM by refusing to allow you to run Win32 apps not made by Microsoft, we're never going to see ARM competing for the desktop or laptop market (unless Windows 8 is a big enough flop itself). We might see ARM make inroads into the lower-end of the market, which could be enough to keep Intel on their toes, but that doesn't really leave much room for AMD...
CPUs take years to go from concept to market, and I really do hope that AMD has something fresh in the market that puts them on a competitive playing field with Intel. Being a full process node behind Intel hurts, but the power savings we see from die shrinks on Intel's products isn't enough to make it completely impossible to compete from one process node behind. So I do hope that we can bring some competition to the x86 CPU market in the future, but at the current point in time, they don't really have anything worth buying over its Intel counter-part, and sending out Intel's products to review AMD's GPUs really isn't helping things.
My point is that whenever you see somebody come up with what seems like some new way of generating power and cites some "could be as low as" cost, it invariably ends up costing far more. And since they're not even claiming substantial cost savings over what we're already using for all of our power (in Quebec, anyhow), it's of questionable utility.
The subject might look like I'm trying to troll, but... I'm actually referring to TFA. AMD sent the TechReport reviewer a Gigabyte Z77 motherboard with an Intel i7-3770K processor. So it says on the first page of TFA.
AMD... sent an Intel processor... to review an AMD GPU...
The company proposing this says the cost could potentially eventually be as low as $0.03 per kilowatt hour. Translation: it costs way more than that.
Meanwhile, the next province over from where the company is based in Sarnia, Ontario... HydroQuebec is charging $0.05 per kilowatt hour, today, for real-world use.
The kickstarter funding doesn't necessarily indicate there's a market, but it's also possible that people are kickstarting this thing because they don't know that there are already much better options on the market. The shortcomings I see with the POP (and warning, I'm going to compare it extensively to the NewTrent IMP120D, because I own one and am quite happy with it):
1) It's unnecessarily enormous. Seriously, this thing has 26 amp-hours of capacity, and has a volume of about 127 cubic inches (yes, accounting for the cylindrical shape). A NewTrent IMP120D stores 12 amp-hours, and has a volume of about 15 cubic inches. If we consider roughly comparable storage capacities and say you need two IMP120Ds, the POP is about four times bigger than it should be. The POP is also weighs about as much as three IMP120Ds.
2) It has questionable usability. On top of the fact that the strange placement of the retractable cables makes it look like it would be difficult to access them, the USB ports, battery capacity indicator, and AC plug are all on the BOTTOM of the unit. In order to plug in a USB device, or to check how much power it has left, you have to pick the thing up and turn it over, hoping that the top does't fall off. And since you may have devices plugged into it, you may need to unplug your devices just to turn it over and look at the battery indicators.
3) It's overpriced. The MSRP of the POP portable is $199. Two IMP120Ds would cost $154, and amazon seems to have a special where they throw in a free 5 amp-hour battery pack when you buy an IMP120D.
The concept of the thing is itself not bad. The problem is it has a terrible execution. They could have made the thing less than half the size, put the retractable cables on the side instead of the bottom, and put the battery indicator on top, and right off the bat that would have eliminated many of the problems with their device; part of the reason the thing is so freaking enormous is because they put the thing on raised feet and have a bizarre compartment on the top.
Then why not give it 32KB of RAM instead of 21KB? Or how about using a Z180 (1MB address range) or an eZ80 (16MB address range)? Or how about using a modern processor instead of an ancient Z80? It's not like they haven't got a line of calculators using a completely different architecture (TI-89/TI-92)...
The other consideration that you mention, that sticking with an ancient processor of questionable usefulness causes severe limitations, is part of the whole problem.
Not to mention this thing was already a questionably useful product. I've got a battery with a bit less than half the capacity, it cost less than half as much, and since it's much less wasteful with empty space, easily slips into a bag or pocket. This POP thing is bulky and awkward. I'm just not seeing a market for it.
The TI-82, introduced in 1993, had 28KB of RAM. So, this new calculator is in some respects a downgrade when compared to TI's own product from almost twenty years ago.
WTF, TI. Could you really not afford to slap a 512KB chunk of RAM into that SoC of yours? Really?
And for that matter, the thing still has a crappy old Z80 processor, and not much flash. In fact, this new calculator is worse in every single respect to the 2004 model of TI-89 Ti except for the screen. Less RAM, less performance, less flash...
ARM is targeting the Cortex A7 (their ultra-low-power ARM core) at sub-$100 smartphones. Surely at some point somebody is going to slap one of those smartphones in a calculator body and capture a significant portion of the market...
Well, go back farther then, and pull the data. You'll see from around '98 or '99 until about 2001ish, we were going way faster than Moore's Law, then from 2001 on it was going a little bit slower than Moore's Law but more or less on-par with it, and then from 2009 it's been way below Moore's Law. Is it too early to call an end to the trend? Perhaps, but it depends on technological advancement.
Seagate is saying HAMR will take them to 60TB (where HAMR is expected to reach its limitations) in a 3.5" drive within a decade. This would be 3.9 doublings over a decade, which is roughly 31 months per doubling... It doesn't sound like Seagate thinks they'll keep up with Moore's law either. But even if they did keep up with Moore's law, SSDs are still getting cheaper much faster than Moore's law.
I attribute this to a combination of Moore's Law and increasing demand (economies of scale). If you move to a bigger wafer in your manufacturing process, for example, you can significantly reduce your costs without getting improving density any. Eventually SSD pricing will settle in closer to Moore's law, but it hasn't yet.
But can you trust them? They're actively selling SSDs with 30-40% failure rates (the Petrol and Octane series). They're not new drives, the failure rates are STILL that high, and they keep selling them.
Slashdot Mirror
2560x1440 is cheaply available (I paid $750 for my Dell U2711 over a year ago), and I'll take it over 1920x1200 any day.
Thunderbolt goes a bit higher, so a RAID array over thunderbolt might do over 600MB/s...
There's a lot of professional video gear that supports thunderbolt. Probably because a lot of professional video gear targets Mac. That's actually kind of annoying, and the third-party Windows drivers for HFS+ have spotty support. I couldn't get them working with the CF cards recorded on a KiPro Mini, for example. I had to hunt down somebody on-site with a macbook to read the damned things.
The integrated graphics chips in Intel's Ivy Bridge processors is capable of displaying and decoding h.264 video in hardware. They've been on the market since late 2011. I'm not talking about expensive nVidia cards here, I'm saying the integrated graphics Intel has been selling for over a year already supports 4K (although they only actually enabled output/decoding in October 2012, anybody who had Ivy Bridge hardware benefitted).
The downside is that it requires two DisplayPort outputs, and few motherboards ship with more than one. We won't get the ability from an Intel iGPU to output 4K video over a single DisplayPort connection until Haswell, due to ship in June 2013.
High end current nVidia graphics cards such as the GTX 670 support 4K output over a single HDMI or DisplayPort connection, however the official word from nVidia is that they've never actually tried it due to the lack of a display to test with (what limited 4K hardware that's out there doesn't do 4K over a single HDMI or DisplayPort connection).
Really, it's just early days. Generally, midrange quad core chips of today should have no problems decoding 4K video.
Cree publishes the wavelength distribution for the different colour temps of MK-R:
http://www.cree.com/led-components-and-modules/products/xlamp/arrays-directional/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/Data%20and%20Binning/XLampMKR.pdf
They don't look terribly useful for anything like image projection, but they'd be fine for a flashlight.
Of note is that they only hit 200 lumens per watt at 1W. If you crank it up to 15 W, they'll do 1769 lumens, which is only 118 lumens per watt. That's still an improvement over their previous best, the XM-L, which did 1040 lumens at 10 W, but it still means that if you want to maintain that 200 lumens per watt, you're going to need a bunch of LEDs to hit higher brightnesses.
The exhaust velocity on an ion thruster is nowhere near that of a partical accelerator. According to a 1996 NASA document I found, the exhaust velocity of an ion thruster is 31.5 kilometers per second, while particle accelerators are very close to the speed of light. The speed record is 0.999999999976c, which is roughly ten thousand times faster than the ion thruster exhaust velocity. That's not to say that thruster exhaust velocity is the speed limit, since the thruster exhaust velocity is relative to the thruster, not the third party observer measuring the speed of the spacecraft.
At some point, though, you're going to be getting hit pretty hard by interstellar particles. Even in diffuse regions, you get a particle density of 10^-4 particles per cubic centimeter, so if we assume a spacecraft with a surface area of 10,000 square centimeters (one square meter) traveling at 3000 kilometers per second, you're going to be hit by (if my math holds up), 300 million particles per second travelling at a rather high speed, although that's not actually that much energy. I don't know enough to say if that would cause a problem.
Umm, no. They had a hell of a time moving to 22nm and getting volume production up, pretty much your entire post is just paranoid delusion.
Intel just has more money to throw at the problems, and they've managed to get a lead of a few years on the other companies. They only maintain that lead because they keep pushing forward.
The problem is that makes it harder for everybody else to compete, but that's not really Intel's fault.
Consumers need some light multithreading, but single-threaded performance is still king. ECC is not something consumers care about (enterprise does, sure, but not consumers).
Power matters less in desktops, so using more power to do the same thing is not as big a problem there. But in the mobile space, it's a big problem.
The use cases you're talking about aren't really what a typical consumer or even office machine is used for.
Are they delivering in the APU market? The CPU performance is poor in their APU products, and the GPU performance is faster than Intel's, but not fast enough to matter. It's still too slow for gaming, and Intel's iGPUs are fast enough for office use.
If their APUs were matching Intel's power/performance, but still kicking their ass on the GPU side, then it might be something special. As it stands, the tradeoffs are just not worth it.
Amd's APUs are faster than Intel's iGPUs, but much slower than discrete chips. Why shouldn't you have an AMD CPU with a discrete GPU in the notebook market?
Sigh. I realize you're a troll, but I'll respond anyhow: getting half the single-threaded performance limits the usefulness of your CPUs. I haven't seen any reviews for piledriver-based Opteron chips (they're relatively new), but comparing the previous gen Opteron against the Sandy Bridge Xeons, "keep up" still means they're falling behind the top Xeon chips in multi-threaded performance, they're way behind in single-threaded performance, and they use a ton more power.
We already get all of our power from hydro, and there's still lots of untapped capacity. I don't think not being able to buy any is a risk anytime soon.
Benchmarks of the HD4000 would have been useless, as AMD sent a desktop chip, performance would not be representative of the mobile HD4000's performance. AMD could just have easily have sent an AMD CPU for apples-to-apples comparisons.
I think the fact that AMD's server chips need twice as many cores just to keep up with Intel's parts is kind of indicative of the problem, and I really do hope that they have something competitive in the market. Intel's current products are great, but only because AMD kicked them in the ass with the Athlon 64 years ago. If they go too long without a real competitor in the higher end of the market (beyond where ARM can reach), they'll stagnate.
That's not the point, the point is that recommending people use your competitor's product over your own is never something a company should do, unless you're in full mea culpa mode like the Apple Maps fiasco.
In the marketing world, you're always supposed to at least PRETEND that your products are superior.
I'm not trolling, and I've owned a few AMD CPUs in my day (four, I think? Does the Geode count?), I was just flabbergasted that AMD would send out Intel CPUs to review their GPUs with. I mean, eating your own dogfood is kind of a fundamental thing, and when you do something like this, it sends the message that your own products aren't good enough for the purpose. I chose an inflammatory title to highlight how ridiculous this is.
I would actually argue that AMD only holds an advantage at the extremely low-end, below the $40 pricepoint (you can get a dual-core sandy bridge for about that), and even then only if you don't care about power. I feel bad for AMD, because I owned a long string of AMD processors, and they were fantastic products. I started on the K6-2 and ended on the Athlon XP, and they were all great. With the Athlon 64, they really hit it out of the park, and had a fantastic run, and then... nothing. It didn't help that Intel had a fantastic chip with the Conroe, but AMD had flop after flop that wasn't even up to their own previous standards. They managed to get some of the worst problems under control to produce decent chips again, but by then they had fallen far enough behind Intel that they could only compete on price, and that approach was driving them bankrupt (hence the whole selling off Global Foundries).
I don't WANT Intel to dominate the CPU market, but AMD just isn't a credible competitor anymore. They haven't been able to compete on the performance front for years, their server chips have kept some of the hope alive but at this point need twice as many cores just to stay anywhere remotely performance competitive, their power efficiency hasn't been competitive for ages, and while their APU stuff has turned out some interesting products, they have a pretty limited market since they have better GPU performance and worse CPU performance than comparable Intel products, and Intel's GPU performance got "good enough" for the kinds of uses that you'd find those chips in anyhow.
In terms of competing on the ultra-low end prices, now ARM is starting to creep up in that market, and AMD is being sandwiched in-between. I think they're a more credible competitor to Intel at this point, but with Microsoft crippling Windows 8 on ARM by refusing to allow you to run Win32 apps not made by Microsoft, we're never going to see ARM competing for the desktop or laptop market (unless Windows 8 is a big enough flop itself). We might see ARM make inroads into the lower-end of the market, which could be enough to keep Intel on their toes, but that doesn't really leave much room for AMD...
CPUs take years to go from concept to market, and I really do hope that AMD has something fresh in the market that puts them on a competitive playing field with Intel. Being a full process node behind Intel hurts, but the power savings we see from die shrinks on Intel's products isn't enough to make it completely impossible to compete from one process node behind. So I do hope that we can bring some competition to the x86 CPU market in the future, but at the current point in time, they don't really have anything worth buying over its Intel counter-part, and sending out Intel's products to review AMD's GPUs really isn't helping things.
My point is that whenever you see somebody come up with what seems like some new way of generating power and cites some "could be as low as" cost, it invariably ends up costing far more. And since they're not even claiming substantial cost savings over what we're already using for all of our power (in Quebec, anyhow), it's of questionable utility.
The subject might look like I'm trying to troll, but... I'm actually referring to TFA. AMD sent the TechReport reviewer a Gigabyte Z77 motherboard with an Intel i7-3770K processor. So it says on the first page of TFA.
AMD... sent an Intel processor... to review an AMD GPU...
Talk about lack of faith in your own products.
The company proposing this says the cost could potentially eventually be as low as $0.03 per kilowatt hour. Translation: it costs way more than that.
Meanwhile, the next province over from where the company is based in Sarnia, Ontario... HydroQuebec is charging $0.05 per kilowatt hour, today, for real-world use.
The kickstarter funding doesn't necessarily indicate there's a market, but it's also possible that people are kickstarting this thing because they don't know that there are already much better options on the market. The shortcomings I see with the POP (and warning, I'm going to compare it extensively to the NewTrent IMP120D, because I own one and am quite happy with it):
1) It's unnecessarily enormous. Seriously, this thing has 26 amp-hours of capacity, and has a volume of about 127 cubic inches (yes, accounting for the cylindrical shape). A NewTrent IMP120D stores 12 amp-hours, and has a volume of about 15 cubic inches. If we consider roughly comparable storage capacities and say you need two IMP120Ds, the POP is about four times bigger than it should be. The POP is also weighs about as much as three IMP120Ds.
2) It has questionable usability. On top of the fact that the strange placement of the retractable cables makes it look like it would be difficult to access them, the USB ports, battery capacity indicator, and AC plug are all on the BOTTOM of the unit. In order to plug in a USB device, or to check how much power it has left, you have to pick the thing up and turn it over, hoping that the top does't fall off. And since you may have devices plugged into it, you may need to unplug your devices just to turn it over and look at the battery indicators.
3) It's overpriced. The MSRP of the POP portable is $199. Two IMP120Ds would cost $154, and amazon seems to have a special where they throw in a free 5 amp-hour battery pack when you buy an IMP120D.
The concept of the thing is itself not bad. The problem is it has a terrible execution. They could have made the thing less than half the size, put the retractable cables on the side instead of the bottom, and put the battery indicator on top, and right off the bat that would have eliminated many of the problems with their device; part of the reason the thing is so freaking enormous is because they put the thing on raised feet and have a bizarre compartment on the top.
Then why not give it 32KB of RAM instead of 21KB? Or how about using a Z180 (1MB address range) or an eZ80 (16MB address range)? Or how about using a modern processor instead of an ancient Z80? It's not like they haven't got a line of calculators using a completely different architecture (TI-89/TI-92)...
The other consideration that you mention, that sticking with an ancient processor of questionable usefulness causes severe limitations, is part of the whole problem.
Not to mention this thing was already a questionably useful product. I've got a battery with a bit less than half the capacity, it cost less than half as much, and since it's much less wasteful with empty space, easily slips into a bag or pocket. This POP thing is bulky and awkward. I'm just not seeing a market for it.
The TI-82, introduced in 1993, had 28KB of RAM. So, this new calculator is in some respects a downgrade when compared to TI's own product from almost twenty years ago.
WTF, TI. Could you really not afford to slap a 512KB chunk of RAM into that SoC of yours? Really?
And for that matter, the thing still has a crappy old Z80 processor, and not much flash. In fact, this new calculator is worse in every single respect to the 2004 model of TI-89 Ti except for the screen. Less RAM, less performance, less flash...
ARM is targeting the Cortex A7 (their ultra-low-power ARM core) at sub-$100 smartphones. Surely at some point somebody is going to slap one of those smartphones in a calculator body and capture a significant portion of the market...
Well, go back farther then, and pull the data. You'll see from around '98 or '99 until about 2001ish, we were going way faster than Moore's Law, then from 2001 on it was going a little bit slower than Moore's Law but more or less on-par with it, and then from 2009 it's been way below Moore's Law. Is it too early to call an end to the trend? Perhaps, but it depends on technological advancement.
Seagate is saying HAMR will take them to 60TB (where HAMR is expected to reach its limitations) in a 3.5" drive within a decade. This would be 3.9 doublings over a decade, which is roughly 31 months per doubling... It doesn't sound like Seagate thinks they'll keep up with Moore's law either. But even if they did keep up with Moore's law, SSDs are still getting cheaper much faster than Moore's law.
I attribute this to a combination of Moore's Law and increasing demand (economies of scale). If you move to a bigger wafer in your manufacturing process, for example, you can significantly reduce your costs without getting improving density any. Eventually SSD pricing will settle in closer to Moore's law, but it hasn't yet.
Let me know where you can get any new and current HDD for $20 or less.
But can you trust them? They're actively selling SSDs with 30-40% failure rates (the Petrol and Octane series). They're not new drives, the failure rates are STILL that high, and they keep selling them.
Talk about an unscrupulous company.