It's worth bearing in mind that as you increase the amount of RAM, particularly in high performance systems like those with i9 processors, that the system has to reduce the memory access speeds accordingly.
AIUI
A processor has a limited number of memory channels. On Intel desktop chips (not sure about the AMD side) each channel can drive up to two modules with up to two "rank"s each. Each "rank" normally consists of 8 chips and the more ranks are present on a channel the higher the loading on the bus and the slower the timings needed to make the channel work reliably.
Server chips often use "registered" memory which allows a larger number of ranks (both more modules and more ranks per module) but adds an extra clock cycle of delay to memory reads.
So higher capacity chips like the ones Samsung are promising mean more memory at the same speed or the same amount of memory at a higher speed.
Officially the first i7 parts to support 128GB were the broadwell-e parts released in may 2016 which is over two years ago. There are also reports that while not officially supported 128GB did in fact work on the older haswell-e parts.
So the GPs claim is perfectly plausible.
The question with these new modules is how much will they cost. Will the cost difference between 4x32GB and 8x16GB be more or less than the cost difference between a maintstream desktop CPU/MB and a high end desktop CPU/MB.
Anything involving analysis of 3D structures (the stuff I remember using was an electromagnetic simulator called HFSS). Doubling your mesh resolution leads to increasing memory usage by a factor of 8 so you can very quickly eat up as much memory as you can throw at it in search of more accurate and/or higher frequency results.
Still the US is a comparably sized (whether you measure by land area, population or economy size) area to choose a place to live and work within to the EU as a whole.
Afaict Australia isn't really much different from the US or Canada in terms of advanced authorisation requirements for people arriving by air (which we all know is how most people travel internationally). They just chose to consider their online travel authorisations Visas rather than denying they are Visas like the US and Canada do.
As it happens, other government agencies are pretty picky about the safety of toys
Which is why so many products say "this is not a toy" on the box.
There are plenty of drones in between "kids toys" and "large commercial vehicles". Devices that are big enough to be potentially hazardous, big enough to accommodate automated flight stabilisation, automated waypoint flight, automated return home, long distance remote control and long distance remote vision yet small and cheap enough that any non-dirt poor adult can by go out and buy one.
Fortunately they have found companies prepared to donate access to Aerial imagary (afaict actual sattelite imagary is usually too low resoloution to be much use), the two big ones seem to have been Yahoo and Microsoft.
such a project is already years behind what Google, Apple, and others are already doing.
Depends on the area.
In the first world openstreetmap has as you say had to start from a position of trying to catch up. In some areas they have caught up and even overtaken the propietary mappers, in others they are still behind.
OTOH my understanding is that there are places in the world where openstreetmap contributors were the first to produce a map of the streets in the area.
Normally laws apply to those within the jurisdiction of a government.
However the EU has decided that these laws apply to anyone interacting with an EU client, regardless of whether or not the site operator has a presense in the EU. In response a bunch of smaller american news sites seem to have decided not to take the risk and just block anyone who appears to be from Europe.
in addition to just about every other problem you can think of.
LOL
Public transport is great when you have lots of people trying to go to or from the same places and those people are travelling light.
It sucks pretty badly when you are trying to travel between two points in suburbia. It sucks terribly when you want to go out into the countryside or when you want/need to take more than a handful of stuff with you. I haven't got around to learning to drive yet (it's expensive) but i'm painfully aware that I will have a major problem when my dad is no longer able to drive.
Remember averages include a wide range of cases. At one extreme you have someone who is young and fit, has an occupation that doesn't involve much physical stress and where they succeed in doing the procedure keyhole. At the other extreme you have someone who is old, works a manual job and has some complications during surgery that require them to fully open the abdomen.
defence? why would the publicly available front end of a website need to be sent securely?
Because otherwise some shithead can alter it in transit. The attacker has no need to compromise the credit card details entry form if they can compromise the flow of pages leading up to the credit card entry form.
Afaict it depends what you mean by "effectively it's own Internet".
On the one hand china blocks a large number of big name western search/social/entertainment sites that dominate the Internet experiance in the west. On the other hand they certainly have not cut off communication completely.
The EU seems to be doing everything they can to follow their lead.
The EU seems to be taking rather a different approach, rather than blocking foreign corps it threatens them with legal sanctions (which it may or may not be able to enforce). This has resulted in a few sites (mostly smaller american news sites afaict) refusing to serve Europeans but all the big american players are still active here.
Why do delivery drivers leave packages on people's porches?
Because it's cheaper.
If you require a person to person contact (which is basically what signed-for packages seem to imply, there doesn't seem to be any requirement for the person signing to be the recipiant) to hand off the package then your delivery provider may need to make multiple delivery attempts. If you are prepared to have your package left without a person to person contact then the delivery provider only has to make one attempt. That translates into signature required services being more expensive than non signature required services.
The flip side of that of course is that packages left without making contact, especially ones left outside have a much higher chance of loss or damage.
So ultimately it comes down to money, does it cost the merchant more to pay for "signature required" delivery service or to replace the packages that go missing due to use of non signature-requried services?
One difference between the US and the UK is that in the UK we have letterboxes in our doors that are owned by the householder and can be used by any delivery service. The americans have mailboxes at the end of their garden which are owned by and can only be used by the postal service. I suspect this results in a greater proportion of packages being left outdoors.
They don't do that (as standard) in the U.K.
Many suppliers in the UK do send packages with no signature required.
Depends which corner of academia you are in. Some fields almost everyone used LaTeX, some fields almost everyone uses Word (with third party extensions), some fields there is a mixture.
TQFPs can be hand soldered with normal soldering equipment. The trick is not to even try to solder the individual pins one at a time. That way lies madness.
Put down the chip on the footprint, line it up and put a blob of solder on one corner, then check it is still lined up, if not re-heat the blob and carefully push the chip into position. Now solder the rest of the chip, don't worry about shorts at this stage, just make sure there is solder joining the pins to the pads. Often it's easiest to drag a blob down the side of the chip (some people can get a blob to neatly solder all the pins as it passes by, but that is a lot harder than the youtube videos make it look)
Finally come in with the solder wick and clean up the shorts.
Some people like to use magnification, personally I have found it more of a hinderance than a help when hand soldering.
What *does* matter a lot is the PCB finish, if the PCB has a crappy finish that doesn't solder well then you will have problems with open circuits. It's also helpful to use a footprint with slightly longer pads when desiging a board for hand soldering.
The crazy thing is that the "master" and "slave" terminology for IDE was misleading from the start. Neither drive is involved in communication to the other.
(note: this is a UK perspective, other places may vary)
Universities have to work with the students they can get.
I think you and your co-students were lucky to catch the height of the 80s microcomputer boom, the time when computers booted into BASIC, when using a computer pretty much meant programming it.
Then the windows PCs with their pre-canned applications and no obvious programming language swept in. Leaning to program now meant not just finding a suitible book, it often meant buying the programming language and computer. School computers were locked down by IT departments worried about security, home computers were often regarded as important tools by parents rather than something kids could safely experiments with. The proportion of computer science applicants with prior programming skill plummeted.
Things are improving again now. There have been high-profile drives to teach kids to program, good free tools are becoming more widespread. Maybe in a few years time it will once again be reasonable for Computing departments at Universities to assume that students will come in with at least some programming skills.
Even if that happens though, not all university students learning to program are computer science students. Scientists and engineers increasingly need programming skills as well.
IMO where both google and openstreetmap fall down as a street map is that you have to zoom in way too far to see all the street names. I believe this is a result of the tiles being auto-generated, rather than someone manually packing the information in.
It's tolerable when using the map interactively because you can zoom in to check the road names in your immediate vicinity and out to get the bigger picture, but it's a big problem if you are trying to print maps.
But my main thing is, aside from substantial investment in good L glass....is that IMHO, for the most part, you're not going to change the laws of physics any time soon...and not reduce the size significantly for the lens types out there, so, why come up with a new mount?
Putting a DSLR lens mount on a mirrorless camera would be giving up the main advantage of a mirrorless camera. The lens flange can be closer to the sensor because there is no need to leave room for a mirror assembly.
Having the lens flange closer to the sensor allows for the rear element of the lens to be closer to the sensor.. This makes the overall system smaller and lighter and particually benefits wide angle lenses which can often use a simple single-group design on a mirrorless where they would have to use a retrofocus design on a SLR.
Physically flying a plane from A to B is the first step. People have been doing that for years.
Then there is the question of margins can it fly from A to B with enough redundancy and energy reserves to satisfy the safety authorities for commerical flights?
Then there is the question of payload, can it fly from A to B while carrying the required payload?
Then there is the question of speed, can it meet the time requirements?
Then there is the question of weather, can it fly from A to B in most weather conditions or only on a still dry clear day?
Then there is the question of if it can do all of the above at the same time?
Only when it can actually do the job does it make sense to consider whether it is cost-effective or not.
They bought the division from Nokia, it feels like no one here who isn't Finnish doesn't realise Nokia makes and continues to make things
It is difficult to keep track of the intricacies of corporate takeovers. Especially when the companies obfuscate the issue by licensing their brands and the journalists lie in their headlines.
and continues to make things that are more than just mobile phones.
But most of those things aren't products that are visible to your average consumer or even your average techie. They are products and services sold to Niche markets.
I guess the Finnish media reports more on the ongoing happenings related to Nokia, so people realise that the "Nokia sold to Microsoft" headlines were a lie, while those same happenings don't get reported so much in the global tech-consumer press because they don't relate to any of the hot-button tech consumer issues of the day.
and they've had mobile devices on the market for years.
My understanding is that with the exception of the N1 tablet (which was not sold in the west and seems to have been discontinued) all of the "Nokia" mobile devices on the market since 2014 are/were either from Microsoft or from "HMD global" (who bought part of Nokia's former handset division from MS).
Polyethelene is less dense than water so if it's clean and empty then it will float. However an item in the bag or even a bit of sand washed inside could easilly push it over the edge into sinking.
There are a lot of jerks with high reputation on SO
Indeed the stack exchange reputation system is basically useless for measuring how much of a jerk someone is.
An upvote costs the voter nothing and gains the votee 5/10 (depending on whether question or answer) points. A downvote costs the voter 1 and the votee 2. Plus the association bonus gives you enough rep to upvote but not to downvote. So users with a moderately high rep on one stack exchange site can upvote across the entire network but can only downvote on sites where they have earned rep locally.
So a lot more upvotes are handed out than downvotes and each upvote gains a user more than each downvote costs them.
Furthermore comments don't factor into the scoring system, so people can be as much of a jerk as they like in the comments without risking any reputation.
Slashdot Mirror
It would seem more sensible to me to simply not draw the unwanted energy from the panels in the first place.
It's worth bearing in mind that as you increase the amount of RAM, particularly in high performance systems like those with i9 processors, that the system has to reduce the memory access speeds accordingly.
AIUI
A processor has a limited number of memory channels. On Intel desktop chips (not sure about the AMD side) each channel can drive up to two modules with up to two "rank"s each. Each "rank" normally consists of 8 chips and the more ranks are present on a channel the higher the loading on the bus and the slower the timings needed to make the channel work reliably.
Server chips often use "registered" memory which allows a larger number of ranks (both more modules and more ranks per module) but adds an extra clock cycle of delay to memory reads.
So higher capacity chips like the ones Samsung are promising mean more memory at the same speed or the same amount of memory at a higher speed.
Officially the first i7 parts to support 128GB were the broadwell-e parts released in may 2016 which is over two years ago. There are also reports that while not officially supported 128GB did in fact work on the older haswell-e parts.
So the GPs claim is perfectly plausible.
The question with these new modules is how much will they cost. Will the cost difference between 4x32GB and 8x16GB be more or less than the cost difference between a maintstream desktop CPU/MB and a high end desktop CPU/MB.
Anything involving analysis of 3D structures (the stuff I remember using was an electromagnetic simulator called HFSS). Doubling your mesh resolution leads to increasing memory usage by a factor of 8 so you can very quickly eat up as much memory as you can throw at it in search of more accurate and/or higher frequency results.
Youâ(TM)re not paying for redundancy at the POP level.
Right but 1.5 million customers affected seems a bit higher up the network than "POP level"
Still the US is a comparably sized (whether you measure by land area, population or economy size) area to choose a place to live and work within to the EU as a whole.
Afaict Australia isn't really much different from the US or Canada in terms of advanced authorisation requirements for people arriving by air (which we all know is how most people travel internationally). They just chose to consider their online travel authorisations Visas rather than denying they are Visas like the US and Canada do.
As it happens, other government agencies are pretty picky about the safety of toys
Which is why so many products say "this is not a toy" on the box.
There are plenty of drones in between "kids toys" and "large commercial vehicles". Devices that are big enough to be potentially hazardous, big enough to accommodate automated flight stabilisation, automated waypoint flight, automated return home, long distance remote control and long distance remote vision yet small and cheap enough that any non-dirt poor adult can by go out and buy one.
You need satellite map
Fortunately they have found companies prepared to donate access to Aerial imagary (afaict actual sattelite imagary is usually too low resoloution to be much use), the two big ones seem to have been Yahoo and Microsoft.
such a project is already years behind what Google, Apple, and others are already doing.
Depends on the area.
In the first world openstreetmap has as you say had to start from a position of trying to catch up. In some areas they have caught up and even overtaken the propietary mappers, in others they are still behind.
OTOH my understanding is that there are places in the world where openstreetmap contributors were the first to produce a map of the streets in the area.
Normally laws apply to those within the jurisdiction of a government.
However the EU has decided that these laws apply to anyone interacting with an EU client, regardless of whether or not the site operator has a presense in the EU. In response a bunch of smaller american news sites seem to have decided not to take the risk and just block anyone who appears to be from Europe.
in addition to just about every other problem you can think of.
LOL
Public transport is great when you have lots of people trying to go to or from the same places and those people are travelling light.
It sucks pretty badly when you are trying to travel between two points in suburbia. It sucks terribly when you want to go out into the countryside or when you want/need to take more than a handful of stuff with you. I haven't got around to learning to drive yet (it's expensive) but i'm painfully aware that I will have a major problem when my dad is no longer able to drive.
(and no, I'm not in the USA)
Remember averages include a wide range of cases. At one extreme you have someone who is young and fit, has an occupation that doesn't involve much physical stress and where they succeed in doing the procedure keyhole. At the other extreme you have someone who is old, works a manual job and has some complications during surgery that require them to fully open the abdomen.
defence? why would the publicly available front end of a website need to be sent securely?
Because otherwise some shithead can alter it in transit. The attacker has no need to compromise the credit card details entry form if they can compromise the flow of pages leading up to the credit card entry form.
China is effectively already its own internet
Afaict it depends what you mean by "effectively it's own Internet".
On the one hand china blocks a large number of big name western search/social/entertainment sites that dominate the Internet experiance in the west. On the other hand they certainly have not cut off communication completely.
The EU seems to be doing everything they can to follow their lead.
The EU seems to be taking rather a different approach, rather than blocking foreign corps it threatens them with legal sanctions (which it may or may not be able to enforce). This has resulted in a few sites (mostly smaller american news sites afaict) refusing to serve Europeans but all the big american players are still active here.
Why do delivery drivers leave packages on people's porches?
Because it's cheaper.
If you require a person to person contact (which is basically what signed-for packages seem to imply, there doesn't seem to be any requirement for the person signing to be the recipiant) to hand off the package then your delivery provider may need to make multiple delivery attempts. If you are prepared to have your package left without a person to person contact then the delivery provider only has to make one attempt. That translates into signature required services being more expensive than non signature required services.
The flip side of that of course is that packages left without making contact, especially ones left outside have a much higher chance of loss or damage.
So ultimately it comes down to money, does it cost the merchant more to pay for "signature required" delivery service or to replace the packages that go missing due to use of non signature-requried services?
One difference between the US and the UK is that in the UK we have letterboxes in our doors that are owned by the householder and can be used by any delivery service. The americans have mailboxes at the end of their garden which are owned by and can only be used by the postal service. I suspect this results in a greater proportion of packages being left outdoors.
They don't do that (as standard) in the U.K.
Many suppliers in the UK do send packages with no signature required.
Depends which corner of academia you are in. Some fields almost everyone used LaTeX, some fields almost everyone uses Word (with third party extensions), some fields there is a mixture.
TQFPs can be hand soldered with normal soldering equipment. The trick is not to even try to solder the individual pins one at a time. That way lies madness.
Put down the chip on the footprint, line it up and put a blob of solder on one corner, then check it is still lined up, if not re-heat the blob and carefully push the chip into position. Now solder the rest of the chip, don't worry about shorts at this stage, just make sure there is solder joining the pins to the pads. Often it's easiest to drag a blob down the side of the chip (some people can get a blob to neatly solder all the pins as it passes by, but that is a lot harder than the youtube videos make it look)
Finally come in with the solder wick and clean up the shorts.
Some people like to use magnification, personally I have found it more of a hinderance than a help when hand soldering.
What *does* matter a lot is the PCB finish, if the PCB has a crappy finish that doesn't solder well then you will have problems with open circuits. It's also helpful to use a footprint with slightly longer pads when desiging a board for hand soldering.
The crazy thing is that the "master" and "slave" terminology for IDE was misleading from the start. Neither drive is involved in communication to the other.
(note: this is a UK perspective, other places may vary)
Universities have to work with the students they can get.
I think you and your co-students were lucky to catch the height of the 80s microcomputer boom, the time when computers booted into BASIC, when using a computer pretty much meant programming it.
Then the windows PCs with their pre-canned applications and no obvious programming language swept in. Leaning to program now meant not just finding a suitible book, it often meant buying the programming language and computer. School computers were locked down by IT departments worried about security, home computers were often regarded as important tools by parents rather than something kids could safely experiments with. The proportion of computer science applicants with prior programming skill plummeted.
Things are improving again now. There have been high-profile drives to teach kids to program, good free tools are becoming more widespread. Maybe in a few years time it will once again be reasonable for Computing departments at Universities to assume that students will come in with at least some programming skills.
Even if that happens though, not all university students learning to program are computer science students. Scientists and engineers increasingly need programming skills as well.
IMO where both google and openstreetmap fall down as a street map is that you have to zoom in way too far to see all the street names. I believe this is a result of the tiles being auto-generated, rather than someone manually packing the information in.
It's tolerable when using the map interactively because you can zoom in to check the road names in your immediate vicinity and out to get the bigger picture, but it's a big problem if you are trying to print maps.
But my main thing is, aside from substantial investment in good L glass....is that IMHO, for the most part, you're not going to change the laws of physics any time soon...and not reduce the size significantly for the lens types out there, so, why come up with a new mount?
Putting a DSLR lens mount on a mirrorless camera would be giving up the main advantage of a mirrorless camera. The lens flange can be closer to the sensor because there is no need to leave room for a mirror assembly.
Having the lens flange closer to the sensor allows for the rear element of the lens to be closer to the sensor.. This makes the overall system smaller and lighter and particually benefits wide angle lenses which can often use a simple single-group design on a mirrorless where they would have to use a retrofocus design on a SLR.
Physically flying a plane from A to B is the first step. People have been doing that for years.
Then there is the question of margins can it fly from A to B with enough redundancy and energy reserves to satisfy the safety authorities for commerical flights?
Then there is the question of payload, can it fly from A to B while carrying the required payload?
Then there is the question of speed, can it meet the time requirements?
Then there is the question of weather, can it fly from A to B in most weather conditions or only on a still dry clear day?
Then there is the question of if it can do all of the above at the same time?
Only when it can actually do the job does it make sense to consider whether it is cost-effective or not.
They bought the division from Nokia, it feels like no one here who isn't Finnish doesn't realise Nokia makes and continues to make things
It is difficult to keep track of the intricacies of corporate takeovers. Especially when the companies obfuscate the issue by licensing their brands and the journalists lie in their headlines.
and continues to make things that are more than just mobile phones.
But most of those things aren't products that are visible to your average consumer or even your average techie. They are products and services sold to Niche markets.
I guess the Finnish media reports more on the ongoing happenings related to Nokia, so people realise that the "Nokia sold to Microsoft" headlines were a lie, while those same happenings don't get reported so much in the global tech-consumer press because they don't relate to any of the hot-button tech consumer issues of the day.
and they've had mobile devices on the market for years.
My understanding is that with the exception of the N1 tablet (which was not sold in the west and seems to have been discontinued) all of the "Nokia" mobile devices on the market since 2014 are/were either from Microsoft or from "HMD global" (who bought part of Nokia's former handset division from MS).
Polyethelene is less dense than water so if it's clean and empty then it will float. However an item in the bag or even a bit of sand washed inside could easilly push it over the edge into sinking.
There are a lot of jerks with high reputation on SO
Indeed the stack exchange reputation system is basically useless for measuring how much of a jerk someone is.
An upvote costs the voter nothing and gains the votee 5/10 (depending on whether question or answer) points. A downvote costs the voter 1 and the votee 2. Plus the association bonus gives you enough rep to upvote but not to downvote. So users with a moderately high rep on one stack exchange site can upvote across the entire network but can only downvote on sites where they have earned rep locally.
So a lot more upvotes are handed out than downvotes and each upvote gains a user more than each downvote costs them.
Furthermore comments don't factor into the scoring system, so people can be as much of a jerk as they like in the comments without risking any reputation.