Some quick rules of thumb (not intended to be accurate to the day's range but get you in the right range)
Euro=about a buck (it's at $1.25ish now, but was as low as $0.80 last year.
Yen=about a penny (ranged from about 1 USD=JPY60 ($0.02) to 1 USD=JPY200 ($0.005)
Pound about $2.00 (it's been closer to $1.50 in the past (and it might have enven aproached parity pretty recently).
Go check something accurate before you make a transaction, but those generally work for reading a news story and trying to place currencies in rough USD terms.
Those little 4 cyl engines only pull 200 HP at well over 6000 rpm (hp=torque*rpm/5625). Those little four bangers have pretty flat torque curves so HP ramps nicely once you pass about 6000 rpm. The trick is never race anything with a V8 from a standstill , as you have to drop the clutch in the powerband to get a decent start. Also, the designer must use a shifty, tight ratio transmission, Bently drivers would hate it. The old Bentleys were designed to not require shifting very often, becuase the first customers didn't know how, in the traditional spirit new bentleys have 400-600 Ft Lbs of torque at most RPMs to provide the same feel. Some drivers don't like driving rev happy cars around town as you have to shift pretty regularly. I personally love my close ratio Honda (Integra GSR) but it really wouldn't be fun to drive if you didn't shift at redline on a regular basis.
Absolutly, but I have some hope. First is that there aren't that many people who can do engineering. Coming from the US DoE college studies, no more than 7% of college grads have the brains to do engineering (that was the percent of undergrads who declared engineering/phy sci majors in 1999/2000 probably about the peak) from the US DoE undergrad study it's a pdf sorry. I think CS has dropped significantly since then. And my own experience shows that a fair number of undergrads don't finish the program. Let's say 5% of graduates make it to graduation in a hard science program (no MIS majors or social sciences). From another study on the site it looked like the national HS graduation rate is about about 4 mil/yr with about 2 million of those entering college each year (45%-50%) which means that about 2% of a population can do engineering. Assuming that within our lifetimes everyone (at least those smart enough to do hard sciences) in India and China gets the opportunity to attend college and they have a similar birth rate. We should have about 3 million (100k annual gradsX30yr career) practicing hard science types in the country (probably about the same in western Europe), China and India each add about 10 million. Yeah that's a lot but most of theirs will be moving towards projects that are required for internal use.
Also, I would think that good engineers would have some benefits from a bigger population of practicing engineers Unlike other types of jobs (manufacturing really doesn't benefit from this)(more competition but also more opportunites to improve on the ideas generated elsewhere them). Oh and engineers aren't quite that cheap, they go for more like 10k-20k, still low but rising quickly. A ton of lower skill (phone support type stuff) throws off the averages. IBM lost some internal docs relating to the exact costs of skilled workers and managment types (it was about 1/4 of IBM's US salary for the same position).
I still don't know why those get no respect, they are really nice cars, I bought an Integra and love it, but every time I see a del Sol I think about how much it would rock if I dropped my engine in one. Happily the S2000 seems to be much better liked, after I move somewhere that doesn't have a high of 3F right now, I'll be getting one.
What's happening appears to me to be a combination of three things:
First, output is increasing due to productivity, managers are using employee fear to not hire yet, this will probably hold for up March-June if demand for goods stays healthy. This happens every upturn, and will continue until execs get worried about losing market share to others who hire (or spend on capex).
Second, there is a shift in production to India and it will last a long time, but right now it is in a bubble (figure that it's 1998-1999 on that one) it will turn around after a few more notable bad experiences with highly skilled trades (there are only so many good engineers/coders in any place it's likely that just like in the.com boom, lots of less effective people are hopping on the band wagon with the really good ones. There will be a ton of jobs that move over there after the bubble bursts, but look at Japan as an example, it took a lot of production through the 1970s and 1980s and now it buys a lot of other things, we may not export many cars to Japan, but luxury goods, CPUs, and Software sell pretty well over there.
Finally, your last point is very true, and there will be better markets for some things over here. Techs that understand how business works (go take an accounting class or two) and can improve or improvise solutions and explain them to managers in a language that they understand will do very well regardless of markets. There were a ton of stupid projects that were implemented in the late 90s, having someone who understands both that this product works, and will help the business in the following ways will be worth their weight in gold to their employeer (even if it ends up being them).
If you are very sharp, and willing to eat dirt for about a decade, analog engineers will be in heavy demand for the forseeable term, there are a ton of old guard retiring, and that is something that you pretty much have to have experience doing to do it well.
One of the main causes for the Renaissance was increased trade brought Europeans in contact with the scholars of the Arabic world who had preserved and improved upon the ideas of the Greeks and Romans (as well as a few of their own).
I would assume this is because unlike those under say 45 (too pick a number out of the air, the average people over that age do not have a ton of computer experience, and those that do probably started on a Unix like system. So you probably have two groups retireees who have used Unix their whole computing lives, and would give us whippersnappers a lesson in vi (or Emacs) or retireees who didn't want to expend any more of their pension on a computer to read email surf the net, and who ever runs their system can fix things remotely and lock them down so much better than an older hardware windows box. For a new, budget constrained user who doesn't need Wal-Mart (or your local hypermart) software compatability a Linux machine is a pretty attractive deal, I'm wishing I'd done it for my grandmother.
If he's gaining in the polls, on a longer trend, it's because a lot of conservatives (think Reagan Democrat types, not Kudlow types) are mad at Bush over his illegal alien amnesty plan and the budget deficits. That and the front runner's implosion last week have caused a lot of voters to at a minimum think about the other candidates running.
It doesn't make a whole lot of difference for big blockbusters with ad budgets in the tens of millions (everyone already knows about them), but for a low budget independant film the few mentions on an awards show might well get them 10 or 20 times the exposure that their ad budget could have ever delivered. Last year's big winners on the the money side were The Pianist and Y tu Mama, Tambien. Both grossed significantly more after the awards wer handed out(the Pianist was only in release a few weeks before the oscars). I would assume that some of the more independant film houses will probably have oscar night where they (re)show smaller award winners for those who missed them.
Exchange rates move a little each time that $8k salary is paid, as well as everytime you buy a Nokia phone, or other foreign made product. In the US we have (for the past 30 or 40 years) bought significantly more than we exported, but that is only half the story, a big part of our imports are oil related. What has kept things in balance (and even driven the dollar up) is that foreigners have generally wanted to invest in our financial products (equities, Treasury and Corporate debt, physical plants) this is what economists call the capital account, and has run a surplus (roughly balancing with the current accout deficit) until very recently after the bubble popped that cut some investment, but over the past year or so, money has been moving to europe because interest rates were significantly higher. The exception to this is Asian exporting countries that wish to prolong their export growth, who have been buying financial products here (mostly Treasury debt) if they stop, the dollar will fall significantly (bringing those $25k salaries a whole lot closer to the $8k salary (in dollar terms). Business follows things in trends, and just like railroads, the internet, Japan, and many other huge changes, there will be significant production developed there, but right now it is a bubble and it will pop eventually. The good news is that this will significantly reduce the excitemet assocated with India, the bad news is that something like 90% of the investmet in a bubble change occurs after it pops (just at a much slower rate).
Even more important than the upgrades are the fact that as people develop software, they will develop it for the platform they are using. This was the trick that made MS rich, and why they killed Netscape, and are so worried about Linux (if you doubt they are worried, go to a campus and see how many CS students are getting free MS software (windows and VB)). If they don't get the lock in of developers, the applications (which is what users ulitmately care about) will be developed for something other than WIndows, then users choose the other platform to use the applications. This is why MS is pretty happy to keep letting very poor users pirate Windows.
Most big software companies, and a few semi companies (Linear is one) maintain those typse of margins, too. Although part of the benefit is that a ton of tech companies don't have to count options as expenses under the assumptions used by accountants. If they did MS's beautiful margins look a whole lot worse. Still quite high, but not 50% operating. The trick to keeping those types of margins is keeping competition to a minimum.
If you read a bit further down they break out each segment (client=windows, information worker=office and some other things like project, server tools=VB+SQL+Windows server+exchange+CALs for each, MSN=subs+ads on MSN.com, CE/Mobility=microsoft's share of cellular and PDAs, business sol.=Navision and Great Plains, home and ent=Xbox+game softare+the hardware they make) client and info worker carry operating margins (revenue less cost of rev, less R&D (developer costs) less selling and marketing (ads and salespeople) less G&A (admin finance and accounting) of well over 90% on average. Since the majority of MS costs are below the cost of revenue line, there is no reason to assume that client and IW are too different from average. From my own research XBox adds about $250-$500 million of that cogs amount depending on the quarter (christmas is a big expense). If you want it in a spreadsheet email me and I'll draw you some pretty charts and stuff.
Microsoft lumps everything together for cost of goods sold or, basically, wholesale value. However, Oracle has gross margins (what the company keeps after paying the value of what they sold, in a simple example a retail store's wholesale cost would be what they pay their supplier) of 95% or more and MS claims twice the operating margin (profit after basically all costs except taxes are out) of 85%-90% for their windows and office divisions, which is about twice Oracle's overall operating profit for databases and applications. It's not exactly apples to apples but it's pretty close. Figure a billion of windows and office cost much less than $50 million (I'd guess that its below $10 million).
Why is it that we are willing to make very big decisions on so little information, a lot of people are heavily influenced by the horsepower number on their car (or even just the displacement volume of the engine) which is as bad or worse than just looking at the clock of the processor. It doesn't take a whole lot of research to get a torque curve for most performance oriented cars, why are we willing to make a decision on so little info?
Developing a modern Computer CPU (no licesning or embedded stuff) requires about $1-2 billion. You have two choices with your investment, make an elegant solution that has 0 inital software compatability, or work with an archaic slipshod architecture, that will likely be compatable with virtually all consumer software written in the last decade. Which one are you going to bet your billion on? Intel's trying the first choice, with Itanium, and even in enterprise software, which has a much better record for diversity of hardware supported, and so far their only sales have come from their co-developer, and a little house that does something with graphics and was tottering near bankruptcy three years ago.
The quick answer is that a longer pipeline lets a processor boost it's clock speed significantly. However, if you mis predict, you have to go back and search for the right data (if it's not in cache you have to go to ram, not in ram HD, then it takes roughly one cycle for each stage in the pipeline and you're back in action. I've heard branch prediction has an upper 90% hit rate, but that still means a few percent of the time you have to wait, in this case 30 cycles to accomplish the next step.
In general Intel's design philosiphy has been to trade everything for a faster clock speed, even if it doesn't actually increase the overall system performance. I think it was on average about 40% that you lost going from P3 to PIV. Consider the Pentium M (it's an PIII done on a modern process with some optimizations for power). Intel doesn't like to mention it but I think they run at or under 2 Ghz and are quite competitive with the best PIVs out there.
Not to mention that you generally start with a "home field advantage" with the local press, you usually have to hurt them in some way to get them to be against you. Look at the P-I and MS or the Mercury News coverage of MS.
I would guess that it was supposed to be a Democrat only server, in which it makes sense that a ranking democrat would have hired that admin. The article isn't very clear about what the server was supposed to be used for.
The Juicy bits are in the form of the level of influence certain organizations have over which nominees come to the floor for a vote, and the exact wording of why certain nominies should not. Leaking that to the media, gives those organizations a bloody lib, if you will.
Slashdot Mirror
Some quick rules of thumb (not intended to be accurate to the day's range but get you in the right range)
Euro=about a buck (it's at $1.25ish now, but was as low as $0.80 last year. Yen=about a penny (ranged from about 1 USD=JPY60 ($0.02) to 1 USD=JPY200 ($0.005) Pound about $2.00 (it's been closer to $1.50 in the past (and it might have enven aproached parity pretty recently). Go check something accurate before you make a transaction, but those generally work for reading a news story and trying to place currencies in rough USD terms.
Those little 4 cyl engines only pull 200 HP at well over 6000 rpm (hp=torque*rpm/5625). Those little four bangers have pretty flat torque curves so HP ramps nicely once you pass about 6000 rpm. The trick is never race anything with a V8 from a standstill , as you have to drop the clutch in the powerband to get a decent start. Also, the designer must use a shifty, tight ratio transmission, Bently drivers would hate it. The old Bentleys were designed to not require shifting very often, becuase the first customers didn't know how, in the traditional spirit new bentleys have 400-600 Ft Lbs of torque at most RPMs to provide the same feel. Some drivers don't like driving rev happy cars around town as you have to shift pretty regularly. I personally love my close ratio Honda (Integra GSR) but it really wouldn't be fun to drive if you didn't shift at redline on a regular basis.
Absolutly, but I have some hope. First is that there aren't that many people who can do engineering. Coming from the US DoE college studies, no more than 7% of college grads have the brains to do engineering (that was the percent of undergrads who declared engineering/phy sci majors in 1999/2000 probably about the peak) from the US DoE undergrad study it's a pdf sorry. I think CS has dropped significantly since then. And my own experience shows that a fair number of undergrads don't finish the program. Let's say 5% of graduates make it to graduation in a hard science program (no MIS majors or social sciences). From another study on the site it looked like the national HS graduation rate is about about 4 mil/yr with about 2 million of those entering college each year (45%-50%) which means that about 2% of a population can do engineering. Assuming that within our lifetimes everyone (at least those smart enough to do hard sciences) in India and China gets the opportunity to attend college and they have a similar birth rate. We should have about 3 million (100k annual gradsX30yr career) practicing hard science types in the country (probably about the same in western Europe), China and India each add about 10 million. Yeah that's a lot but most of theirs will be moving towards projects that are required for internal use.
Also, I would think that good engineers would have some benefits from a bigger population of practicing engineers Unlike other types of jobs (manufacturing really doesn't benefit from this)(more competition but also more opportunites to improve on the ideas generated elsewhere them). Oh and engineers aren't quite that cheap, they go for more like 10k-20k, still low but rising quickly. A ton of lower skill (phone support type stuff) throws off the averages. IBM lost some internal docs relating to the exact costs of skilled workers and managment types (it was about 1/4 of IBM's US salary for the same position).
Thanks for the advice, I'll probably try GS or SoM for something different.
I guess it does, hopefully the relavent fact here is that arbitrage opportunites are used up as the prices begin to equalize.
I still don't know why those get no respect, they are really nice cars, I bought an Integra and love it, but every time I see a del Sol I think about how much it would rock if I dropped my engine in one. Happily the S2000 seems to be much better liked, after I move somewhere that doesn't have a high of 3F right now, I'll be getting one.
What's happening appears to me to be a combination of three things: .com boom, lots of less effective people are hopping on the band wagon with the really good ones. There will be a ton of jobs that move over there after the bubble bursts, but look at Japan as an example, it took a lot of production through the 1970s and 1980s and now it buys a lot of other things, we may not export many cars to Japan, but luxury goods, CPUs, and Software sell pretty well over there.
First, output is increasing due to productivity, managers are using employee fear to not hire yet, this will probably hold for up March-June if demand for goods stays healthy. This happens every upturn, and will continue until execs get worried about losing market share to others who hire (or spend on capex).
Second, there is a shift in production to India and it will last a long time, but right now it is in a bubble (figure that it's 1998-1999 on that one) it will turn around after a few more notable bad experiences with highly skilled trades (there are only so many good engineers/coders in any place it's likely that just like in the
Finally, your last point is very true, and there will be better markets for some things over here. Techs that understand how business works (go take an accounting class or two) and can improve or improvise solutions and explain them to managers in a language that they understand will do very well regardless of markets. There were a ton of stupid projects that were implemented in the late 90s, having someone who understands both that this product works, and will help the business in the following ways will be worth their weight in gold to their employeer (even if it ends up being them).
If you are very sharp, and willing to eat dirt for about a decade, analog engineers will be in heavy demand for the forseeable term, there are a ton of old guard retiring, and that is something that you pretty much have to have experience doing to do it well.
Should I get Fire Emblem or Golden Sun II after I beat Tactics Advance. Thank you.
One of the main causes for the Renaissance was increased trade brought Europeans in contact with the scholars of the Arabic world who had preserved and improved upon the ideas of the Greeks and Romans (as well as a few of their own).
I would assume this is because unlike those under say 45 (too pick a number out of the air, the average people over that age do not have a ton of computer experience, and those that do probably started on a Unix like system. So you probably have two groups retireees who have used Unix their whole computing lives, and would give us whippersnappers a lesson in vi (or Emacs) or retireees who didn't want to expend any more of their pension on a computer to read email surf the net, and who ever runs their system can fix things remotely and lock them down so much better than an older hardware windows box. For a new, budget constrained user who doesn't need Wal-Mart (or your local hypermart) software compatability a Linux machine is a pretty attractive deal, I'm wishing I'd done it for my grandmother.
If he's gaining in the polls, on a longer trend, it's because a lot of conservatives (think Reagan Democrat types, not Kudlow types) are mad at Bush over his illegal alien amnesty plan and the budget deficits. That and the front runner's implosion last week have caused a lot of voters to at a minimum think about the other candidates running.
It doesn't make a whole lot of difference for big blockbusters with ad budgets in the tens of millions (everyone already knows about them), but for a low budget independant film the few mentions on an awards show might well get them 10 or 20 times the exposure that their ad budget could have ever delivered. Last year's big winners on the the money side were The Pianist and Y tu Mama, Tambien. Both grossed significantly more after the awards wer handed out(the Pianist was only in release a few weeks before the oscars). I would assume that some of the more independant film houses will probably have oscar night where they (re)show smaller award winners for those who missed them.
Exchange rates move a little each time that $8k salary is paid, as well as everytime you buy a Nokia phone, or other foreign made product. In the US we have (for the past 30 or 40 years) bought significantly more than we exported, but that is only half the story, a big part of our imports are oil related. What has kept things in balance (and even driven the dollar up) is that foreigners have generally wanted to invest in our financial products (equities, Treasury and Corporate debt, physical plants) this is what economists call the capital account, and has run a surplus (roughly balancing with the current accout deficit) until very recently after the bubble popped that cut some investment, but over the past year or so, money has been moving to europe because interest rates were significantly higher. The exception to this is Asian exporting countries that wish to prolong their export growth, who have been buying financial products here (mostly Treasury debt) if they stop, the dollar will fall significantly (bringing those $25k salaries a whole lot closer to the $8k salary (in dollar terms).
Business follows things in trends, and just like railroads, the internet, Japan, and many other huge changes, there will be significant production developed there, but right now it is a bubble and it will pop eventually. The good news is that this will significantly reduce the excitemet assocated with India, the bad news is that something like 90% of the investmet in a bubble change occurs after it pops (just at a much slower rate).
Even more important than the upgrades are the fact that as people develop software, they will develop it for the platform they are using. This was the trick that made MS rich, and why they killed Netscape, and are so worried about Linux (if you doubt they are worried, go to a campus and see how many CS students are getting free MS software (windows and VB)). If they don't get the lock in of developers, the applications (which is what users ulitmately care about) will be developed for something other than WIndows, then users choose the other platform to use the applications. This is why MS is pretty happy to keep letting very poor users pirate Windows.
Most big software companies, and a few semi companies (Linear is one) maintain those typse of margins, too. Although part of the benefit is that a ton of tech companies don't have to count options as expenses under the assumptions used by accountants. If they did MS's beautiful margins look a whole lot worse. Still quite high, but not 50% operating. The trick to keeping those types of margins is keeping competition to a minimum.
If you read a bit further down they break out each segment (client=windows, information worker=office and some other things like project, server tools=VB+SQL+Windows server+exchange+CALs for each, MSN=subs+ads on MSN.com, CE/Mobility=microsoft's share of cellular and PDAs, business sol.=Navision and Great Plains, home and ent=Xbox+game softare+the hardware they make) client and info worker carry operating margins (revenue less cost of rev, less R&D (developer costs) less selling and marketing (ads and salespeople) less G&A (admin finance and accounting) of well over 90% on average. Since the majority of MS costs are below the cost of revenue line, there is no reason to assume that client and IW are too different from average. From my own research XBox adds about $250-$500 million of that cogs amount depending on the quarter (christmas is a big expense). If you want it in a spreadsheet email me and I'll draw you some pretty charts and stuff.
Microsoft lumps everything together for cost of goods sold or, basically, wholesale value. However, Oracle has gross margins (what the company keeps after paying the value of what they sold, in a simple example a retail store's wholesale cost would be what they pay their supplier) of 95% or more and MS claims twice the operating margin (profit after basically all costs except taxes are out) of 85%-90% for their windows and office divisions, which is about twice Oracle's overall operating profit for databases and applications. It's not exactly apples to apples but it's pretty close. Figure a billion of windows and office cost much less than $50 million (I'd guess that its below $10 million).
That's because most people are sleeping most soundly causing the least impact on surrounding residents and reducing the likelihood of resistance.
Why is it that we are willing to make very big decisions on so little information, a lot of people are heavily influenced by the horsepower number on their car (or even just the displacement volume of the engine) which is as bad or worse than just looking at the clock of the processor. It doesn't take a whole lot of research to get a torque curve for most performance oriented cars, why are we willing to make a decision on so little info?
Developing a modern Computer CPU (no licesning or embedded stuff) requires about $1-2 billion. You have two choices with your investment, make an elegant solution that has 0 inital software compatability, or work with an archaic slipshod architecture, that will likely be compatable with virtually all consumer software written in the last decade. Which one are you going to bet your billion on? Intel's trying the first choice, with Itanium, and even in enterprise software, which has a much better record for diversity of hardware supported, and so far their only sales have come from their co-developer, and a little house that does something with graphics and was tottering near bankruptcy three years ago.
The quick answer is that a longer pipeline lets a processor boost it's clock speed significantly. However, if you mis predict, you have to go back and search for the right data (if it's not in cache you have to go to ram, not in ram HD, then it takes roughly one cycle for each stage in the pipeline and you're back in action. I've heard branch prediction has an upper 90% hit rate, but that still means a few percent of the time you have to wait, in this case 30 cycles to accomplish the next step.
In general Intel's design philosiphy has been to trade everything for a faster clock speed, even if it doesn't actually increase the overall system performance. I think it was on average about 40% that you lost going from P3 to PIV. Consider the Pentium M (it's an PIII done on a modern process with some optimizations for power). Intel doesn't like to mention it but I think they run at or under 2 Ghz and are quite competitive with the best PIVs out there.
Not to mention that you generally start with a "home field advantage" with the local press, you usually have to hurt them in some way to get them to be against you. Look at the P-I and MS or the Mercury News coverage of MS.
I would guess that it was supposed to be a Democrat only server, in which it makes sense that a ranking democrat would have hired that admin. The article isn't very clear about what the server was supposed to be used for.
The Juicy bits are in the form of the level of influence certain organizations have over which nominees come to the floor for a vote, and the exact wording of why certain nominies should not. Leaking that to the media, gives those organizations a bloody lib, if you will.