Which was put there before they started collecting taxes. I'm betting if they have a do over on positioning their warehouses they would put them in low population states (aka just move it across the border to OK or NM).
Of course, collecting sales tax probably didn't hurt their business, but I personally know that I frequently check for alternatives on large purchases because saving $20-50 is worth the 30 seconds it takes to find another online retailer that doesn't charge taxes in TX.
They are in a tax-free zone in TX, which is 99% of the reason they planted here. No taxes on inventory.
And if you look at something like iMazing, you'll see how to do device management right. Even Xcode's device management interface blows away the crapfest that is iTunes (not comparable features or anything, it accesses certain developer features, but those are also mostly available via iMazing)
"Your region" is an arbitrary area designation and would be something closer to you than me.;) I'm positive at least 1 of those is closer to you than me and therefore falls into the "you should take care of this" rather than someone further away. "You" also being generic.
Removed party affiliation from the ballot, as well as randomizing the ballot order for a position are all that are needed to largely kill the 2 party system. Both can be argued as being necessary to free and open elections, as both lead to bias in voting. Obviously removing party affiliation also removes party line voting, probably the single largest contributor to the 2 party system.
Various portions of Africa, Europe, Asia, South America, and most of the Middle East (deserves its own listing) are all in turmoil. A full listing is too long to go through at this point, which I think says all that is necessary.
Had the FTC included 1 founding rule in their charter, the 10% rule, these types of situations could be avoided. The 10% rule is simple: if your company via a merger or acquisition will exceed 10% of any market, you may not merge/acquire.
It admittedly destroys the bigger plays on Wall Street, but I'm pretty sure no one cares except those 0.001%. It certainly plays to bigger competition and will require proper regulation of imports to prevent unfair trade practices.
As for the military, has it ever occurred to you that most of us actually don't want there to be a giant hegemonic power (you) throwing around your military weight in the world?
Has it occurred to you that perhaps we (giant hegemonic power) really would rather not be throwing around our military weight? Step up and do your part and and ensure peace in your regions, we'd all rather be home. Well, most of us as there's always a few in every country that would love to dominate their neighbors.
Banding occurs even on BD content, in very specific scenes. It seems to happen generally in dark scenes, or scenes that are computer generated. Real world scenes that are sufficiently lighted have enough color variance that there usually is no banding opportunity as there are no smooth gradation of color across a large enough area to notice. House of Cards, for instance, should have very little computer generated content and I imagine is reasonably well lit comparatively, at least from what I've seen of it.
Regarding the bandwidth, I don't think it is a bandwidth issue but the fact that I was watching some really low hit movies, and I'm guessing between Netflix's low buffering and higher latency regarding this content combined worked against me. I really wish Netflix would keep at least 2 minutes buffered, plus the entire stream that you've already watched, for a variety of reasons. Conversely, I've never had a youtube problem....:)
I was testing hulu again a few months ago, and while the PQ was acceptable on some B&W shows, it still stuttered occasionally, on a 50Mbps connection. I also played an HD episode I'd recently missed, and that PQ was noticeably softer than even my provider's feed (this particular show was not OTA, which has a noticeably better PQ). Note, I'm watching this on a calibrated Panny ZT60, and yes, it is noticeable, even to my crappy degrading eye sight. It's one of those things that once you know what to look for, you can't unsee it. Banding (lack of color depth) is the biggest issue with 1080P that I see currently, and 4K doesn't address that either, although HDR does. That might be the only driver for 4K, honestly, because the regular 4K screens once properly calibrated are no better than 1080P plasma, and in many ways worse. (Again, it's that once you know what to look for, you can't unsee blocking, ghosting, or banding and it seriously annoys me, anyways) I was looking at the OLED 4K HDR sets recently and my first thought was "How do these look once they're properly calibrated?"
What is the performance per watt of the new 18 core Xeon chip compared to the best Xeon that existed in 2010?
I don't know, but the benches for the new 40 core Broadwell EP chip indicate that with almost 7 time the cores, the speed up is only about 5 times faster at a rough guess. The per core performance isn't much better, but what they did manage was to reduce the power requirements significantly, I think. So yes, in 6 years, they did manage to double total performance depending upon workload. Obviously single threaded performance didn't scale up, but multi-threaded tasks will benefit significantly at a reduced power load. The new CPUs are estimated to cost more than double ($2K+) compared to the older CPUs (about $1K at the time, between $90 and $300 today).
At this point you'd have to start adding in consideration of budget. Dual socket hexcore systems in 2010 would have been at a comparable price, although using more than double the power, and have half the performance, for example (if you were using Intel). For certain task types, you might have been better off buying AMD multi-core systems with 4+ sockets, although I have no performance numbers for those in that date range.
Why would I compare anything but the top-end of 2010 with current systems? I'm not interested in some cost-conscious CPU from 2010. I'm interested in whether the CPU improvements are continuing. It is obvious from the facts that CPU improvements, while still occurring, have slowed in their effects. Even a 4 year difference should have marked improvements. 50% is not a marked improvement. I can cut that 50% by merely OC'ing to the same peak clock speed and the improvements drop off significantly (the new CPUs turbo mode hit between 4 and 4.2 GHz, clocking the 980X to 4.2GHz removes a large chunk of the performance gain).
Now, this is not to say that improvements are not occurring to lower cost CPUs, as the high end features start filtering down to the i5s and i3s at a lower cost. But at the upper end, performance has started to plateau for all practical reasons. The slope of the performance curve over time has flattened significantly, although it is still increasing, it is increasingly markedly slower, perhaps doubling every 8 years now (I'm assuming a hexcore CPU by 2018 will double the performance). It is no longer the case where 18-24 months separate a doubling or more in per core performance increases. (Note, we're still seeing this in the low-power ARM arena, but soon they will plateau as well)
Compare the i7-980x to Haswell-E and you'll see the gains.
No, you won't. The TDP for the top end Haswell-E is the same, and the performance is roughly 40% better. That's for a 6 year old processor. The doubling of transistors every 24 (or even 48 months) months with major improvements in performance is obviously gone. What we are seeing is CPU performance increasing down the line for lower end items, like NUCs, laptops, tablets, and phones. That increase is also slowing and I suspect we'll see them plateau within a couple of years barring some major breakthrough in energy efficiency or a new manufacturing technique.
It's kind of like cars or planes. There were lots and lots of designs in the first 20 years or so, then they rapidly migrated to a few known working designs which were then iterated upon and perfected until something new spurred the next iteration of innovation. We are now nearing that iteration and perfection loop for CPUs while we wait for the next discovery.
The cheapest upgrade for any 1366 system is one of the Xeon hexcores. The x58 boards will definitely support them, and several will even allow you to run ECC RAM and/or 8GB sticks. I was looking at grabbing a dual socket MB a few years ago but decided the cost and effort were more than I was willing to deal with. Besides, I like quiet, I think my next box is going to be a NUC.:)
First, you're talking about an i7 920 (4 core, 2.7GHz). The 980X is an entirely different class of processor with 6 cores and 3.45GHz stock, plus it's unlocked.
I'm running an 980x on a dual SATA III Samsung 850s RAID0 system drive (yes, I know, I have my reasons). Trust me, it's not slow at all, in any way shape or form. It's mildly OC'd, running just about 3.8GHz and 1700MHz or so on the RAM, IIRC. I'm 99% certain it will easily meet your 6700K's performance, and maybe not even breaking a sweat. It should run stably over 4GHz and near 2200 MHz if I remove half the RAM, which I'm not willing to do.
I had toyed with grabbing a Xeon 5650 (cheap today) and dropping in 48GB of ECC RAM as it supports 8GB sticks and can be reliably OC'd to near 4GHz while being cooler than the 980X (maybe). But 24GB of RAM has been fine, and it's an aging motherboard that seems silly to throw even a few hundred extra on at this point, not to mention I'd be running pretty far out of documented specs on a number of components, despite assurances that they can handle it just fine. The caps are solid, but they will go at some point and then I will likely part out the remainder of the system to reduce the new system costs.
You're also comparing two chips that don't compare well. Your 980X was $1K back in the day, compare it to i7-5960x which is $1K today. About the same TDP but double the performance.
If you want to use the i7-6700k, compare it to the chip back then that was a similar price, the i7-920. It will be double the performance.
1) from Futuremark, the 5960X was only slightly faster than the 6700K, which is why I chose the much cheaper and lower powered 6700K. The 3Dmark11 benchmarks are 11610 (38%) and 10790 (28%) higher, more or less. These are CPU dominated benchmarks.
2) the claim was saving 50% of the power. It's why I was looking for a lower power chip or something that would double performance. It's just not there. That's not to say the chips aren't better today than they were 6 years ago, as they obviously are, but they're not good enough yet to really be THE reason to upgrade.
I was going to say if the motherboard dies or something, I'd just replace it. Having just reviewed the asking prices of x58 motherboards, I'm tempted to sell my current rig and buy a cheaper, quieter replacement.
As an example, I'm running a 6 year old 980X. It's 130W TDP, and 3dMark11 at 8440, provided it's not OC'd (it is). Looking very very carefully through Intel's latest offerings, the 6700K comes in about 25% more performance (stock) at 95W TDP, and it's the only one that saves on power in the top 10. I'm also dropping 2 cores and 4 threads which for many of my workloads could limit the performance gain even more due to x86s high context switching costs. Note that this is 6 years of CPU improvements, and I can't even get a 50% performance gain. I can get a 40% gain at the same power draw for a mere $1500 CPU cost. I can only guess the more numerous core parts will draw even more power, but at least I might get to a doubling of performance. Yep, not much reason to upgrade the CPU.
There's another factor there, I know of at least 3 people in a major metroplex that have the maximum service they can get and won't be able to consistently stream even at 7Mbps.;) While 7Mbps might be serviceable for some HD, it would still leave people wanting when it comes to detail. Just because they're used to Comcast/TWC etc compression doesn't mean that the picture is good. Quality is not a subjective thing, acceptable levels of quality are. There's also the question of audio quality. But you'd probably say I don't count there either.
Slashdot Mirror
Which was put there before they started collecting taxes. I'm betting if they have a do over on positioning their warehouses they would put them in low population states (aka just move it across the border to OK or NM).
Of course, collecting sales tax probably didn't hurt their business, but I personally know that I frequently check for alternatives on large purchases because saving $20-50 is worth the 30 seconds it takes to find another online retailer that doesn't charge taxes in TX.
They are in a tax-free zone in TX, which is 99% of the reason they planted here. No taxes on inventory.
I went to college directly after high school and lost nearly everything I'd learned in HS as a result.
So you learned something?
And if you look at something like iMazing, you'll see how to do device management right. Even Xcode's device management interface blows away the crapfest that is iTunes (not comparable features or anything, it accesses certain developer features, but those are also mostly available via iMazing)
Nah, I like Apple, I don't like iTunes.
"Your region" is an arbitrary area designation and would be something closer to you than me. ;) I'm positive at least 1 of those is closer to you than me and therefore falls into the "you should take care of this" rather than someone further away. "You" also being generic.
Seriously doubtful, given the amount of money flowing into party campaigns.
Removed party affiliation from the ballot, as well as randomizing the ballot order for a position are all that are needed to largely kill the 2 party system. Both can be argued as being necessary to free and open elections, as both lead to bias in voting. Obviously removing party affiliation also removes party line voting, probably the single largest contributor to the 2 party system.
Yes, please. Anonymous or otherwise, FB needs to be removed as a main gatekeeper for the masses.
Various portions of Africa, Europe, Asia, South America, and most of the Middle East (deserves its own listing) are all in turmoil. A full listing is too long to go through at this point, which I think says all that is necessary.
Had the FTC included 1 founding rule in their charter, the 10% rule, these types of situations could be avoided. The 10% rule is simple: if your company via a merger or acquisition will exceed 10% of any market, you may not merge/acquire.
It admittedly destroys the bigger plays on Wall Street, but I'm pretty sure no one cares except those 0.001%. It certainly plays to bigger competition and will require proper regulation of imports to prevent unfair trade practices.
As for the military, has it ever occurred to you that most of us actually don't want there to be a giant hegemonic power (you) throwing around your military weight in the world?
Has it occurred to you that perhaps we (giant hegemonic power) really would rather not be throwing around our military weight? Step up and do your part and and ensure peace in your regions, we'd all rather be home. Well, most of us as there's always a few in every country that would love to dominate their neighbors.
$3500*365*15=$19M, plus plant costs. Doesn't seem so crazy all of a sudden.
Banding occurs even on BD content, in very specific scenes. It seems to happen generally in dark scenes, or scenes that are computer generated. Real world scenes that are sufficiently lighted have enough color variance that there usually is no banding opportunity as there are no smooth gradation of color across a large enough area to notice. House of Cards, for instance, should have very little computer generated content and I imagine is reasonably well lit comparatively, at least from what I've seen of it.
Regarding the bandwidth, I don't think it is a bandwidth issue but the fact that I was watching some really low hit movies, and I'm guessing between Netflix's low buffering and higher latency regarding this content combined worked against me. I really wish Netflix would keep at least 2 minutes buffered, plus the entire stream that you've already watched, for a variety of reasons. Conversely, I've never had a youtube problem.... :)
...Next time you're on Facebook...
Contrary to what the content creator fantasized would happen, the advertisement in front of the advertisement didn't fool me.
You're on Facebook....
Breaking it is easy, edit your hosts file and make www.amazon.com 127.0.0.1. Done.
I was testing hulu again a few months ago, and while the PQ was acceptable on some B&W shows, it still stuttered occasionally, on a 50Mbps connection. I also played an HD episode I'd recently missed, and that PQ was noticeably softer than even my provider's feed (this particular show was not OTA, which has a noticeably better PQ). Note, I'm watching this on a calibrated Panny ZT60, and yes, it is noticeable, even to my crappy degrading eye sight. It's one of those things that once you know what to look for, you can't unsee it. Banding (lack of color depth) is the biggest issue with 1080P that I see currently, and 4K doesn't address that either, although HDR does. That might be the only driver for 4K, honestly, because the regular 4K screens once properly calibrated are no better than 1080P plasma, and in many ways worse. (Again, it's that once you know what to look for, you can't unsee blocking, ghosting, or banding and it seriously annoys me, anyways) I was looking at the OLED 4K HDR sets recently and my first thought was "How do these look once they're properly calibrated?"
What is the performance per watt of the new 18 core Xeon chip compared to the best Xeon that existed in 2010?
I don't know, but the benches for the new 40 core Broadwell EP chip indicate that with almost 7 time the cores, the speed up is only about 5 times faster at a rough guess. The per core performance isn't much better, but what they did manage was to reduce the power requirements significantly, I think. So yes, in 6 years, they did manage to double total performance depending upon workload. Obviously single threaded performance didn't scale up, but multi-threaded tasks will benefit significantly at a reduced power load. The new CPUs are estimated to cost more than double ($2K+) compared to the older CPUs (about $1K at the time, between $90 and $300 today).
At this point you'd have to start adding in consideration of budget. Dual socket hexcore systems in 2010 would have been at a comparable price, although using more than double the power, and have half the performance, for example (if you were using Intel). For certain task types, you might have been better off buying AMD multi-core systems with 4+ sockets, although I have no performance numbers for those in that date range.
Why would I compare anything but the top-end of 2010 with current systems? I'm not interested in some cost-conscious CPU from 2010. I'm interested in whether the CPU improvements are continuing. It is obvious from the facts that CPU improvements, while still occurring, have slowed in their effects. Even a 4 year difference should have marked improvements. 50% is not a marked improvement. I can cut that 50% by merely OC'ing to the same peak clock speed and the improvements drop off significantly (the new CPUs turbo mode hit between 4 and 4.2 GHz, clocking the 980X to 4.2GHz removes a large chunk of the performance gain).
Now, this is not to say that improvements are not occurring to lower cost CPUs, as the high end features start filtering down to the i5s and i3s at a lower cost. But at the upper end, performance has started to plateau for all practical reasons. The slope of the performance curve over time has flattened significantly, although it is still increasing, it is increasingly markedly slower, perhaps doubling every 8 years now (I'm assuming a hexcore CPU by 2018 will double the performance). It is no longer the case where 18-24 months separate a doubling or more in per core performance increases. (Note, we're still seeing this in the low-power ARM arena, but soon they will plateau as well)
Compare the i7-980x to Haswell-E and you'll see the gains.
No, you won't. The TDP for the top end Haswell-E is the same, and the performance is roughly 40% better. That's for a 6 year old processor. The doubling of transistors every 24 (or even 48 months) months with major improvements in performance is obviously gone. What we are seeing is CPU performance increasing down the line for lower end items, like NUCs, laptops, tablets, and phones. That increase is also slowing and I suspect we'll see them plateau within a couple of years barring some major breakthrough in energy efficiency or a new manufacturing technique.
It's kind of like cars or planes. There were lots and lots of designs in the first 20 years or so, then they rapidly migrated to a few known working designs which were then iterated upon and perfected until something new spurred the next iteration of innovation. We are now nearing that iteration and perfection loop for CPUs while we wait for the next discovery.
The cheapest upgrade for any 1366 system is one of the Xeon hexcores. The x58 boards will definitely support them, and several will even allow you to run ECC RAM and/or 8GB sticks. I was looking at grabbing a dual socket MB a few years ago but decided the cost and effort were more than I was willing to deal with. Besides, I like quiet, I think my next box is going to be a NUC. :)
First, you're talking about an i7 920 (4 core, 2.7GHz). The 980X is an entirely different class of processor with 6 cores and 3.45GHz stock, plus it's unlocked.
I'm running an 980x on a dual SATA III Samsung 850s RAID0 system drive (yes, I know, I have my reasons). Trust me, it's not slow at all, in any way shape or form. It's mildly OC'd, running just about 3.8GHz and 1700MHz or so on the RAM, IIRC. I'm 99% certain it will easily meet your 6700K's performance, and maybe not even breaking a sweat. It should run stably over 4GHz and near 2200 MHz if I remove half the RAM, which I'm not willing to do.
I had toyed with grabbing a Xeon 5650 (cheap today) and dropping in 48GB of ECC RAM as it supports 8GB sticks and can be reliably OC'd to near 4GHz while being cooler than the 980X (maybe). But 24GB of RAM has been fine, and it's an aging motherboard that seems silly to throw even a few hundred extra on at this point, not to mention I'd be running pretty far out of documented specs on a number of components, despite assurances that they can handle it just fine. The caps are solid, but they will go at some point and then I will likely part out the remainder of the system to reduce the new system costs.
You're also comparing two chips that don't compare well. Your 980X was $1K back in the day, compare it to i7-5960x which is $1K today. About the same TDP but double the performance.
If you want to use the i7-6700k, compare it to the chip back then that was a similar price, the i7-920. It will be double the performance.
1) from Futuremark, the 5960X was only slightly faster than the 6700K, which is why I chose the much cheaper and lower powered 6700K. The 3Dmark11 benchmarks are 11610 (38%) and 10790 (28%) higher, more or less. These are CPU dominated benchmarks.
2) the claim was saving 50% of the power. It's why I was looking for a lower power chip or something that would double performance. It's just not there. That's not to say the chips aren't better today than they were 6 years ago, as they obviously are, but they're not good enough yet to really be THE reason to upgrade.
I was going to say if the motherboard dies or something, I'd just replace it. Having just reviewed the asking prices of x58 motherboards, I'm tempted to sell my current rig and buy a cheaper, quieter replacement.
As an example, I'm running a 6 year old 980X. It's 130W TDP, and 3dMark11 at 8440, provided it's not OC'd (it is). Looking very very carefully through Intel's latest offerings, the 6700K comes in about 25% more performance (stock) at 95W TDP, and it's the only one that saves on power in the top 10. I'm also dropping 2 cores and 4 threads which for many of my workloads could limit the performance gain even more due to x86s high context switching costs. Note that this is 6 years of CPU improvements, and I can't even get a 50% performance gain. I can get a 40% gain at the same power draw for a mere $1500 CPU cost. I can only guess the more numerous core parts will draw even more power, but at least I might get to a doubling of performance. Yep, not much reason to upgrade the CPU.
There's another factor there, I know of at least 3 people in a major metroplex that have the maximum service they can get and won't be able to consistently stream even at 7Mbps. ;) While 7Mbps might be serviceable for some HD, it would still leave people wanting when it comes to detail. Just because they're used to Comcast/TWC etc compression doesn't mean that the picture is good. Quality is not a subjective thing, acceptable levels of quality are. There's also the question of audio quality. But you'd probably say I don't count there either.
A company called VSI which consists of former DEC VMS engineers is currently porting VMS to x86-64.
I wonder if it'll be the same smashing success as Solaris on x86?