FWIW, if you read the article, they are not making energy from a random distribution of heat energy, but rather by a using a temperature differential. Since the generation of the electricity (across the differential) results in a corresponding change in thermal entropy, energy is conserved and no laws are violated.
The fact of life in the video game industry is that once you been in the industry for 10 years and/or over 30 years old, you're no good to the cheap bean counters who run a lot of these game companies.
I work on Mortal Kombat at Midway and we look for guys with deep experience. Our team has youngsters but we also have a lot of guys in their late 30's and early 40's with 15-20 years of game development under their belt each. Guys who have made it from hand-assembled machine code into hex and actual hardware (chip / board) design, to macro-assembler, to C, to C++ with modern OS programming on games. These are great guys with lots of experience, the ability to learn new systems (which they've proven several times), who have the knowledge and memory of getting things done within the constraints of limited resources (and trust me, no matter how fast/big these consoled get, the resources are still VERY limited), and who have learned to manage heavy crunch-time without burning out!
The key thing outside of school is always keep learning new stuff, have an exit strategy to get into the next job, manage your career that benefits your situation the best and stay healthy.
This is true at any job, but it's still nice to know there are some places in the industry where it's OK to have experience. If you do things right at a place that appreciates you, you shouldn't NEED an exit-strategy.
From what I remember the main skills you need are huge amounts of 3D vector geometry to a level that is mainly taught only in the Physics field.
Actually, you'll get a lot more of the vector geometry from taking the math course "Linear Algebra" (which is basically a whole semester of matrix and vector math). As far as the stuff you need to do game physics, I'd also advise taking calculus (differential and integrations) and differential equations. You'd also be well served by a logic (mathematical logic) course. As far as physics classes, you'd probably be fine with one or two semesters of basic newtonian / classical Physics (i.e. Physics 101 / 201). You probably don't need to take quantum physics or relativity to write most games.
I've been programming games for around 20 years now.
Those Electrical & Computer Engineering Courses have been surprisingly useful to me over the years with understanding PS2, XBOX, GCN at the hardware level. DMA, signalling latencies, assymetric multiprocessor (IOP vs EE), stream-processors, etc. They also continue to be very useful for understanding the structure and interaction of the hardware on XBOX360, PS3 and Wii. Heck even my signals and systems engineering classes have been great for understanding everything from audio processing to graphics shaders.
Then again, I'm a low-level systems programmer on games and I do a lot of optimization and hardware bashing. The game-play programmers have less need for the engineering classes but every game team working on a console game REQUIRES at least one low-level guy who understands the hardware at a computer engineering level.
I also work in the games industry and here are the degrees which are probably most useful to you if you want a job programming games:
Bachelor of Science: Computer Science
Bachelor of Science: Mathematics
Bachelor of Engineering: Electrical Engineering (computer or digital design emphasis)
They're certainly not the only degrees to get but they do stand out on a resume as someone likely to be able to handle game programming. Those "BA:Film Appreciation" resumes with "I wanna kode a gr8 game idea I had" go straight into the trash.
Also very important is experience -- any experience. For guys who have never worked on a commercial game, being able to show demos of personal or even class projects covering aspects of game programming on graphics, sound programming, networking, etc will vastly improve your hireability as a beginning game programmer (not to mention probably get you a better starting salary). Being able to describe in depth some of the techniques will get you pretty far on an interview.
Now what's interesting is that while the Game Programming degree will get you some of the experience and prossibly a cool demo, there is still a stigma that the Game Programming degree covers mainly some practical applications and doesn't cover enough theory to allow you to delve into solving new and more complex issues outside of the learned practical applications. Therefore, your best bet is to take one of the tradition degrees and if possible AS ELECTIVES take classes from the Game Programming track.
Google suggested a new standard for ATX power supplies that is supposed to have again, substantial power savings.
Typical PSU's waste between 30-45% of the power consumed in a PC. (55-70% efficiency). This causes extra heat as well which ends up costing more for ventilation and air conditioning in heavily packed server rooms.
You can purchase "80-plus" PSU's which get 80% or better efficiency from 20% to 100% load but they will cost you a little more money up front. You can also purchase HE (High Efficiency) PSU's form most manufacturers that cost a little more but do not quite meet the full specs for "90-plus". Most PSU's have more difficulty with efficiency when producing lower loads (compared to their full output potential) so it's important to have a PSU that's properly sized for your PC.
For example, the Antec Neo HE 500 (High Efficiency) gets up to 85% efficiency at loads over 150W but loads under 100W can be as low as 70% efficiency.
Scouts will still be allowed to sing songs around the campfire during camping trips assuming that all use of music is properly accounted for with performance fees and royalties paid to the RIAA.
The beverage is basically flavored iced green tea which has been shown to promote weight loss. All teas (oolong, black, green) will promote weight loss to some extent but you still need to eat sensibly and excercise to maintain optimum health.
BTW, I drink a cup of green tea a day either hot or iced and this is where I buy mine from:
Nothing will be an iPod killer unless they grab the mindshare. Kids aren't asking their parents for MP3 Players for XMas. They're asking for iPods - specifically and by name. An iPod killer can't just be as good as an iPod. It has to be way better and have people know what it is for it to be a success.
If you upgrade when performance triples, the Core 2 Duos with 4MB L2 are at least 3-4X faster for any tasks that will multithread in raw CPU than your current CPU. They're almost 2X faster in single threaded tasks as well.
Actually it's more like like claiming someone who managed to cover the distance from Point A to Point B, 200 miles apart, in four hours is therefore guilty of speeding since the drivers at Microsoft could never drive that distance in that amount of time while going at or under the speed limit because they have to stop for bathroom breaks, a drive-thru at McD's, and to refill the gas tank along the way. They claim it's simply not possible that he was more efficient *AND* legal in his effort.
FWIW, I bought an XPS600 from Dell with the "NForce for Intel" MB and I was getting write errors on my hard-drives until I disabled both hardware TCP/IP acceleration and SATA command queuing on the NVidia chipset. The computer doesn't run any noticeably slower (real world maybe 1-2%) but it's 100 times more stable. To me, stability is more important than minor speed bumps that corrupt my data anyday.
FWIW, If you look at the Mac Pro (dual Xeons == four cores) and price a similarly configured four core workstation from Dell, the Mac ends up being about a thousand dollars cheaper.
Embedded systems could really use a processor like this. Like my toaster
Sony plans on developing a toaster with a CELL Processor. It will have DRM and only toast Sony brand bread though which will come in an odd size and cost twice as much as any other bread.
But how many people out there were claiming we wouldn't be having any new low-level programmers because kids these days grow up with Windows and Macs rather than Apple IIe and C64's?
Sony's not the only one with bad customer support. My sister got a Dell computer and her media reader drive broke. It took over three hours on the phone with someone in India who was barely comprehensible and *DEFINITELY* reading off a script to get them to replace a $8 part. My time is worth way more than that. I've bought 3 Dells in the past but this experience with their "customer service" pretty much means I'll never buy another Dell again.
FWIW, I program low-level hardware on video game consoles so I do lots and lots of MP programming (on custom chips). I've written code on multi-CPU designs from the Sega Saturn (2 CPU), the Atari Jaguar (3 CPU's - 1 CISC / 2 RISC), PS2 (2 RISC CPU's EE/IOP + 2 DSP's VU0/VU1) not to mention tons of weird custom chips and hardware in various systems. I am also developing on XBOX360 (SMP) and PS3 (which is primarily Assymetric if you don't count the "hyperthreading" on the PPU).
The Symmetric in SMP refers to the fact that each process can run the same tasks.In an assymmetric setup, there may be a processor dedicated to the kernel or other tasks.
I was referring to the symmetry on a hardware level, not software. In an SMP system, you have processors that can run identical binaries (the same tasks). Here you are in agreement with me, just stating it differently. It's possible to take any SMP hardware and write an OS or software for it that explicitly schedules specific tasks to certain CPU's which uses the hardware in a non-Symmetric manner but when we refer to Assymetric MP on a hardware level, there are usually physical differences between the CPUs (again, my example of binary compatibility - which is common in game consoles) or the CPU's are nearly identical but connected to resources in a non-Symmetric manner (i.e. the VU0 and VU1 on the PS2) so they require differenct tasks to be sent to each one.
To me SMP and ASMP is more important to examine from a hardware level than from the software design. This any system that can run SMP from hardware can be made into an ASMP system using software, while a system that is truly ASMP from a hardware design may be anywhere from just very difficult to physically impossible to run as a SMP system in software.
It most definitely has nothing to do with speed.
This point is also 100% in agreement with what I said. My whole post was that speed or lock-state execution wasn't necessary for SMP, just the ability to run similar tasks (i.e. identical binaries).
SMP means the processors are similar (i.e. can run identical binary code). They do not need to run in lock-step synchonization to be SMP. Indeed, it is currently possible to halt a single processor in a dual processor system (two sockets) so a similar case already exists in current SMP systems.
25% per core at 100% capacity. a core at 50% capacity would be half 25% or 12.5%... 37.5% total.
Mathematically that's what you'd expect but 50% CPU load doesn't mean 50% the power - it completely depends on what state a 50% load puts the CPU in (lowered clock or same clock / sleep-state during idle / etc). Plus you have to remember that a halted-state CPU still consumes more power than a CPU that is completely powered down. This is why it takes 45% rather than 37.5%. You can't apply such simple principals to power calculations.
MS licensing is for currently for physical CPU's not for cores. Right now a dual xenon (two CPU's) counts as two CPU's in MS licensing terms but a dual-core (two CPU's within a single die or processor socket) is one CPU under MS licensing terms.
The article is very light on details but the one picture implies power control at the core level. For example if core-1 is running a 100% workload and core-2 has a 50% workload, core-3 and core-4 can be halted resulting in a power load of only 45% the total 4-core max load.
Slashdot Mirror
I love the math in the article - the following is a totally true statement:
"I've seen people at 0.04 seconds. That is less than half a second," he said.
FWIW, if you read the article, they are not making energy from a random distribution of heat energy, but rather by a using a temperature differential. Since the generation of the electricity (across the differential) results in a corresponding change in thermal entropy, energy is conserved and no laws are violated.
Replace "Christian God" in your arugmant with "Invisible Pink Unicorn" and your argument still makes exactly the same amount of logical sense.
The fact of life in the video game industry is that once you been in the industry for 10 years and/or over 30 years old, you're no good to the cheap bean counters who run a lot of these game companies.
I work on Mortal Kombat at Midway and we look for guys with deep experience. Our team has youngsters but we also have a lot of guys in their late 30's and early 40's with 15-20 years of game development under their belt each. Guys who have made it from hand-assembled machine code into hex and actual hardware (chip / board) design, to macro-assembler, to C, to C++ with modern OS programming on games. These are great guys with lots of experience, the ability to learn new systems (which they've proven several times), who have the knowledge and memory of getting things done within the constraints of limited resources (and trust me, no matter how fast/big these consoled get, the resources are still VERY limited), and who have learned to manage heavy crunch-time without burning out!
The key thing outside of school is always keep learning new stuff, have an exit strategy to get into the next job, manage your career that benefits your situation the best and stay healthy.
This is true at any job, but it's still nice to know there are some places in the industry where it's OK to have experience. If you do things right at a place that appreciates you, you shouldn't NEED an exit-strategy.
From what I remember the main skills you need are huge amounts of 3D vector geometry to a level that is mainly taught only in the Physics field.
Actually, you'll get a lot more of the vector geometry from taking the math course "Linear Algebra" (which is basically a whole semester of matrix and vector math). As far as the stuff you need to do game physics, I'd also advise taking calculus (differential and integrations) and differential equations. You'd also be well served by a logic (mathematical logic) course. As far as physics classes, you'd probably be fine with one or two semesters of basic newtonian / classical Physics (i.e. Physics 101 / 201). You probably don't need to take quantum physics or relativity to write most games.
I've been programming games for around 20 years now.
Those Electrical & Computer Engineering Courses have been surprisingly useful to me over the years with understanding PS2, XBOX, GCN at the hardware level. DMA, signalling latencies, assymetric multiprocessor (IOP vs EE), stream-processors, etc. They also continue to be very useful for understanding the structure and interaction of the hardware on XBOX360, PS3 and Wii. Heck even my signals and systems engineering classes have been great for understanding everything from audio processing to graphics shaders.
Then again, I'm a low-level systems programmer on games and I do a lot of optimization and hardware bashing. The game-play programmers have less need for the engineering classes but every game team working on a console game REQUIRES at least one low-level guy who understands the hardware at a computer engineering level.
I also work in the games industry and here are the degrees which are probably most useful to you if you want a job programming games:
Bachelor of Science: Computer Science
Bachelor of Science: Mathematics
Bachelor of Engineering: Electrical Engineering (computer or digital design emphasis)
They're certainly not the only degrees to get but they do stand out on a resume as someone likely to be able to handle game programming. Those "BA:Film Appreciation" resumes with "I wanna kode a gr8 game idea I had" go straight into the trash.
Also very important is experience -- any experience. For guys who have never worked on a commercial game, being able to show demos of personal or even class projects covering aspects of game programming on graphics, sound programming, networking, etc will vastly improve your hireability as a beginning game programmer (not to mention probably get you a better starting salary). Being able to describe in depth some of the techniques will get you pretty far on an interview.
Now what's interesting is that while the Game Programming degree will get you some of the experience and prossibly a cool demo, there is still a stigma that the Game Programming degree covers mainly some practical applications and doesn't cover enough theory to allow you to delve into solving new and more complex issues outside of the learned practical applications. Therefore, your best bet is to take one of the tradition degrees and if possible AS ELECTIVES take classes from the Game Programming track.
Google suggested a new standard for ATX power supplies that is supposed to have again, substantial power savings.
Typical PSU's waste between 30-45% of the power consumed in a PC. (55-70% efficiency). This causes extra heat as well which ends up costing more for ventilation and air conditioning in heavily packed server rooms.
You can purchase "80-plus" PSU's which get 80% or better efficiency from 20% to 100% load but they will cost you a little more money up front. You can also purchase HE (High Efficiency) PSU's form most manufacturers that cost a little more but do not quite meet the full specs for "90-plus". Most PSU's have more difficulty with efficiency when producing lower loads (compared to their full output potential) so it's important to have a PSU that's properly sized for your PC.
For example, the Antec Neo HE 500 (High Efficiency) gets up to 85% efficiency at loads over 150W but loads under 100W can be as low as 70% efficiency.
Scouts will still be allowed to sing songs around the campfire during camping trips assuming that all use of music is properly accounted for with performance fees and royalties paid to the RIAA.
The beverage is basically flavored iced green tea which has been shown to promote weight loss. All teas (oolong, black, green) will promote weight loss to some extent but you still need to eat sensibly and excercise to maintain optimum health.
c Path=139_143
BTW, I drink a cup of green tea a day either hot or iced and this is where I buy mine from:
http://www.denstea.com/index.php?main_page=index&
Dude, you've had a cellphone last more than two years? Hahaha!
Or until they can be on the front page of every newssite when Oprah and Bono go out shopping for one together.
c le/2006/10/13/AR2006101300161.html
http://www.washingtonpost.com/wp-dyn/content/arti
Nothing will be an iPod killer unless they grab the mindshare. Kids aren't asking their parents for MP3 Players for XMas. They're asking for iPods - specifically and by name. An iPod killer can't just be as good as an iPod. It has to be way better and have people know what it is for it to be a success.
Current system is a P4 2.7 GHz
If you upgrade when performance triples, the Core 2 Duos with 4MB L2 are at least 3-4X faster for any tasks that will multithread in raw CPU than your current CPU. They're almost 2X faster in single threaded tasks as well.
Actually it's more like like claiming someone who managed to cover the distance from Point A to Point B, 200 miles apart, in four hours is therefore guilty of speeding since the drivers at Microsoft could never drive that distance in that amount of time while going at or under the speed limit because they have to stop for bathroom breaks, a drive-thru at McD's, and to refill the gas tank along the way. They claim it's simply not possible that he was more efficient *AND* legal in his effort.
FWIW, I bought an XPS600 from Dell with the "NForce for Intel" MB and I was getting write errors on my hard-drives until I disabled both hardware TCP/IP acceleration and SATA command queuing on the NVidia chipset. The computer doesn't run any noticeably slower (real world maybe 1-2%) but it's 100 times more stable. To me, stability is more important than minor speed bumps that corrupt my data anyday.
FWIW, If you look at the Mac Pro (dual Xeons == four cores) and price a similarly configured four core workstation from Dell, the Mac ends up being about a thousand dollars cheaper.
Embedded systems could really use a processor like this. Like my toaster
Sony plans on developing a toaster with a CELL Processor. It will have DRM and only toast Sony brand bread though which will come in an odd size and cost twice as much as any other bread.
Yes I know this is a troll...
But how many people out there were claiming we wouldn't be having any new low-level programmers because kids these days grow up with Windows and Macs rather than Apple IIe and C64's?
Sony's not the only one with bad customer support. My sister got a Dell computer and her media reader drive broke. It took over three hours on the phone with someone in India who was barely comprehensible and *DEFINITELY* reading off a script to get them to replace a $8 part. My time is worth way more than that. I've bought 3 Dells in the past but this experience with their "customer service" pretty much means I'll never buy another Dell again.
FWIW, I program low-level hardware on video game consoles so I do lots and lots of MP programming (on custom chips). I've written code on multi-CPU designs from the Sega Saturn (2 CPU), the Atari Jaguar (3 CPU's - 1 CISC / 2 RISC), PS2 (2 RISC CPU's EE/IOP + 2 DSP's VU0/VU1) not to mention tons of weird custom chips and hardware in various systems. I am also developing on XBOX360 (SMP) and PS3 (which is primarily Assymetric if you don't count the "hyperthreading" on the PPU).
The Symmetric in SMP refers to the fact that each process can run the same tasks.In an assymmetric setup, there may be a processor dedicated to the kernel or other tasks.
I was referring to the symmetry on a hardware level, not software. In an SMP system, you have processors that can run identical binaries (the same tasks). Here you are in agreement with me, just stating it differently. It's possible to take any SMP hardware and write an OS or software for it that explicitly schedules specific tasks to certain CPU's which uses the hardware in a non-Symmetric manner but when we refer to Assymetric MP on a hardware level, there are usually physical differences between the CPUs (again, my example of binary compatibility - which is common in game consoles) or the CPU's are nearly identical but connected to resources in a non-Symmetric manner (i.e. the VU0 and VU1 on the PS2) so they require differenct tasks to be sent to each one.
To me SMP and ASMP is more important to examine from a hardware level than from the software design. This any system that can run SMP from hardware can be made into an ASMP system using software, while a system that is truly ASMP from a hardware design may be anywhere from just very difficult to physically impossible to run as a SMP system in software.
It most definitely has nothing to do with speed.
This point is also 100% in agreement with what I said. My whole post was that speed or lock-state execution wasn't necessary for SMP, just the ability to run similar tasks (i.e. identical binaries).
SMP means the processors are similar (i.e. can run identical binary code). They do not need to run in lock-step synchonization to be SMP. Indeed, it is currently possible to halt a single processor in a dual processor system (two sockets) so a similar case already exists in current SMP systems.
True and that's why I said "currently". At the time of Win2000 there were no multi-core single-socket CPU's that ran the OS.
25% per core at 100% capacity. a core at 50% capacity would be half 25% or 12.5%... 37.5% total.
Mathematically that's what you'd expect but 50% CPU load doesn't mean 50% the power - it completely depends on what state a 50% load puts the CPU in (lowered clock or same clock / sleep-state during idle / etc). Plus you have to remember that a halted-state CPU still consumes more power than a CPU that is completely powered down. This is why it takes 45% rather than 37.5%. You can't apply such simple principals to power calculations.
MS licensing is for currently for physical CPU's not for cores. Right now a dual xenon (two CPU's) counts as two CPU's in MS licensing terms but a dual-core (two CPU's within a single die or processor socket) is one CPU under MS licensing terms.
In other words, MS counts sockets, not cores.
The article is very light on details but the one picture implies power control at the core level. For example if core-1 is running a 100% workload and core-2 has a 50% workload, core-3 and core-4 can be halted resulting in a power load of only 45% the total 4-core max load.