-
-
to think how we have come from then. and if only dad had then known the true value of a Altair 8800 as I have seen them in old 1970s electronics magazines.
Australia was on the cutting-edge of a lot things computing technology. we had a lot of computer scientists here!
way back in the 1960s, then we lost the lot, by the mid 1980s as japan went ahead of us.
we had a lot of military technology in development too. TOW weapon systems.
a wire-guided anti-tank missile, that worked like a fishing reel from the back of the rocket motor.
jet powered military drones , Unmanned Aerial Targets and other rockets too.
we made AWA weather balloon radiosondes in Melbourne that had two vacuum tubes & worked on 73MHz or was it 71 MHz.
we made our own military vacuum tube type back-pack worky torky radios too, 3 to 15 MHz then there was space tracking stations . -
Those predictions were eerily accurate. People take computers for granted, we have desktop computers and we both depend on them way too much and yet also gain freedom from them. Only the terminal was a bit off, although one could argue the cloud model will make that one increasingly true too.
-
One of the funny things I have noticed - in terms of TFLOPS, what you have in your pocket now* was classified as a really good workstation 10 years ago and as a supercomputer 20 years ago.
Yes, I know it's not an accurate comparison, but still, it's interesting.
* - the iPhoneXS claims ~5TFLOPs
-
I ran across an interesting assertion in a book by Richard Hamming that I hadn't seen elsewhere:
That was written circa 1997, as far as I can tell, but it's not clear when the equation was modeled. Obviously it isn't from 1943, even if the epoch begins at that point, since the "fastest computer on the market" wasn't a meaningful thing to talk about then.
The 'limiting asymptote' seems pretty close to the reality we'd be stuck with if SIMD and multicore CPUs hadn't come to the rescue. -
* - the iPhoneXS claims ~5TFLOPs
I find that hard to believe when a RTX 2060 is 5.2 TFLOPS. Of which, the now rather dated 980Ti is 5.6 TFLOPS. Maybe TFLOPS nowadays is like GHz in the early 2000s and only tells part of the story, with more sophisticated processors having a more diverse instruction set capable of doing more real world work per operation? -
The really sad thing is that 20 years ago those 5.5 TFlops were doing things like advanced weather modelling, nuclear physics modelling, major economic models and the like. Now they generate emojis and cleverly make cartoons out of pictures. Can you imagine that futurists of twenty and more years ago were scratching their heads and saying if we can just get enough computer power we can do a lot of really mundane things.
-
I remember when 1 TFLOPS was a rack the size of a refrigerator that costs as much as a really nice car. Fast forward to late 2014 and for just $300, you could upgrade your PC to 3.5 TFLOPS. Ever since then, one big use of that was turning 1080p to 4K with surprisingly good results. Actually, it only used ~600 GFLOPS of that at first, then with software updates and configuration tweaking, that went up to 1 TFLOPS in exchange for better quality.
It's worth noting that putting those TFLOPS to "real use" is as easy as mining Curecoin and Foldingcoin in order to fight diseases like cancer and COVID-19. -
The really sad thing is that 20 years ago those 5.5 TFlops were doing things like advanced weather modelling, nuclear physics modelling, major economic models and the like. Now they generate emojis and cleverly make cartoons out of pictures. Can you imagine that futurists of twenty and more years ago were scratching their heads and saying if we can just get enough computer power we can do a lot of really mundane things.
I really wonder if Gutenberg had Tabloid Media in mind when he brought the Printing Press to the West....
Just kidding of course but I do not think it is really sad but rather uplifting that most everyone can have a computer in its Pocket... it enables some people to do amazing things, and others to gossip.
-
I'm looking forward to the next big paradigm shift in technology.
Where interface standards are king and all devices support them. Where it's trivial to link any technology to any other technology in any way to you want.
If i want to link a button on my mouse to open my garage door, or change my washing machine spin cycle to low speed, this should be a few touchscreen clicks to link an event to an outcome.
Any mouse, washing machine or garage door brand you might purchase would interface just as well as any other brand.
-
The really sad thing is that 20 years ago those 5.5 TFlops were doing things like advanced weather modelling, nuclear physics modelling, major economic models and the like. Now they generate emojis and cleverly make cartoons out of pictures. Can you imagine that futurists of twenty and more years ago were scratching their heads and saying if we can just get enough computer power we can do a lot of really mundane things.
That's a very negative view on things. Sure, we piss away ridiculous amounts of computing power and many washing machines and coffee makers have computers on board they would have killed for 50 years ago. Yet we also do a lot of amazing things with what we have. Right now the biggest supercomputer ever has been constructed out of many thousands of distributed nodes by people volunteering their computing power to fighting Covid. People also do amazing things on a much smaller scale. CNC manufacturing at home has become a possibility and people are doing things that were flat out scifi not too long ago from the comfort of their own home. It's absolutely amazing what one person or a small group can achieve. -
Quote
A TFLOPS is not a TFLOPS is not a TFLOPS/
I know - it's not exactly an accurate metric for comparison
As I have said: Yes, I know it's not an accurate comparison, but still, it's interesting. -
A more consumer friendly version of this is the RTX tensor cores, which does exactly the same, that's how they can claim so much better TFLOPS number while not significantly increasing die size.
This chart shows the Tensor FLOPS separately:
https://en.wikipedia.org/wiki/GeForce_20_series#Chipset_table -
That's silly, they should have just made bigger desks.
-
The really sad thing is that 20 years ago those 5.5 TFlops were doing things like advanced weather modelling, nuclear physics modelling, major economic models and the like. Now they generate emojis and cleverly make cartoons out of pictures. Can you imagine that futurists of twenty and more years ago were scratching their heads and saying if we can just get enough computer power we can do a lot of really mundane things.
That's a very negative view on things. Sure, we piss away ridiculous amounts of computing power and many washing machines and coffee makers have computers on board they would have killed for 50 years ago. Yet we also do a lot of amazing things with what we have. Right now the biggest supercomputer ever has been constructed out of many thousands of distributed nodes by people volunteering their computing power to fighting Covid. People also do amazing things on a much smaller scale. CNC manufacturing at home has become a possibility and people are doing things that were flat out scifi not too long ago from the comfort of their own home. It's absolutely amazing what one person or a small group can achieve.
I actually agree with your comment. But still find it sad that of the billions of processors shipped each year only a fraction of a percent (which is still a huge number) do all of the good things you are talking about. Include Google in the mix and you get over the percentage point level. A somewhat larger number makes the reports we write prettier than the typewriter generated things of 50 years ago, and maybe saves a few man-hours in the process. And that the bulk of them do the silly things I was talking about. -
The really sad thing is that 20 years ago those 5.5 TFlops were doing things like advanced weather modelling, nuclear physics modelling, major economic models and the like. Now they generate emojis and cleverly make cartoons out of pictures. Can you imagine that futurists of twenty and more years ago were scratching their heads and saying if we can just get enough computer power we can do a lot of really mundane things.
That's a very negative view on things. Sure, we piss away ridiculous amounts of computing power and many washing machines and coffee makers have computers on board they would have killed for 50 years ago. Yet we also do a lot of amazing things with what we have. Right now the biggest supercomputer ever has been constructed out of many thousands of distributed nodes by people volunteering their computing power to fighting Covid. People also do amazing things on a much smaller scale. CNC manufacturing at home has become a possibility and people are doing things that were flat out scifi not too long ago from the comfort of their own home. It's absolutely amazing what one person or a small group can achieve.
I actually agree with your comment. But still find it sad that of the billions of processors shipped each year only a fraction of a percent (which is still a huge number) do all of the good things you are talking about. Include Google in the mix and you get over the percentage point level. A somewhat larger number makes the reports we write prettier than the typewriter generated things of 50 years ago, and maybe saves a few man-hours in the process. And that the bulk of them do the silly things I was talking about.
I agree wholeheartedly with CatalinaWOW. Let's be honest. 99% of what people use the "new" technology for is entertainment. Video games. Facebook. Tik Tok. Music videos.
Someone mentioned weather forecasting. I find it beyond comprehension how little weather forecasting has progressed in the last decades. The average person can do as well just looking at a radar map/movie, seeing which way the rain is heading and what the temperatures are, and predict fairly closely what tomorrow's weather will be. The "experts" can't figure out what the weather will be in the next few days with any accuracy. Plus or minus 10 degrees F is ridiculous. And how many times have you received a sudden alert of bad weather coming in the next hour? And even then it fizzles out. They have NO CLUE. But hey, they've got these incredibly powerful supercomputers doing simulations that aren't accurate.
Desktop computers? Sometime take a look at a desktop computer motherboard from the 90's and one from today. They look virtually identical. The core technology hasn't changed. What's changed is tweaking the performance improvement. Faster and more efficient. So people can play video games.
Self driving cars? Are you freakin' serious? Who the hell needs or wants a self driving car? It sounds cool, but that's about it. It's an opportunity for a new industry, so people push it. Like wind and solar. New technology equals new opportunity for new money-making industries. Even if they are incredibly expensive and of limited ACTUAL use.
Nonsense like Alexa? Are you serious? So we can sit on our butts and have a computer do stuff for us? But it's cool and fun so we automatically consider it good. Drones?? 3D printers?? Smartphones?? The list goes on and on.
Where's the REAL and USEFUL advances in technology? Cancer cures? Fixing the real causes of deaths and starvation around the world? Really improving our lives rather than having the entire universe with their heads stuck in their smartphones watching cats playing piano.
At the end of the day, if we're being honest, we'll admit that we LIKE all those fun and entertaining and mindless technologies, so we automatically classify them as GOOD. And we go to great lengths to come up with any possible justification we can think of, even if it's totally irrelevant and ridiculous. And nobody can criticize those technologies cuz we like them. Even if they're freakin' useless.
Hey, I'm an engineer, and I like technology as much as the next guy. But geez, let's get real folks. -
Quote
a wire-guided anti-tank missile, that worked like a fishing reel from the back of the rocket motor.
Can you send me please the White Paper and the Matching Blue Prints for?Quote
Thanks -
Quote
a wire-guided anti-tank missile, that worked like a fishing reel from the back of the rocket motor.
Can you send me please the White Paper and the Matching Blue Prints for?Quote
Thanks
it's not like something new
https://en.wikipedia.org/wiki/BGM-71_TOW -
The 1st expeditions to the Moon had a 'Computer' that was not even as good as a Commodore-64
-
I agree wholeheartedly with CatalinaWOW. Let's be honest. 99% of what people use the "new" technology for is entertainment. Video games. Facebook. Tik Tok. Music videos.
This is pretty much what I was getting at when I replied to CatalinaWOW. If that is how you feel you really need to start looking around and looking a little harder. Weather predictions have become much more accurate even if they're still not perfect. Road deaths have dwindled since the 80s both thanks to on board computers and those used to improve their construction. Cancer has become much, much more survivable in no small part thanks to massive amounts of computing power. The world is a much better place than it was in many regards but it you insist on brushing aside advances while focusing on the inane you'll end up missing the point. Things have massively progressed every decade since the 80s. Who cares people are also using advanced technology for entertainment? Technology becoming a commodity means anything but wasting it. Last but not least, one sure sign of progress is people falling behind and shaking their fists at clouds about how everything now sucks.
Someone mentioned weather forecasting. I find it beyond comprehension how little weather forecasting has progressed in the last decades. The average person can do as well just looking at a radar map/movie, seeing which way the rain is heading and what the temperatures are, and predict fairly closely what tomorrow's weather will be. The "experts" can't figure out what the weather will be in the next few days with any accuracy. Plus or minus 10 degrees F is ridiculous. And how many times have you received a sudden alert of bad weather coming in the next hour? And even then it fizzles out. They have NO CLUE. But hey, they've got these incredibly powerful supercomputers doing simulations that aren't accurate.
Desktop computers? Sometime take a look at a desktop computer motherboard from the 90's and one from today. They look virtually identical. The core technology hasn't changed. What's changed is tweaking the performance improvement. Faster and more efficient. So people can play video games.
Self driving cars? Are you freakin' serious? Who the hell needs or wants a self driving car? It sounds cool, but that's about it. It's an opportunity for a new industry, so people push it. Like wind and solar. New technology equals new opportunity for new money-making industries. Even if they are incredibly expensive and of limited ACTUAL use.
Nonsense like Alexa? Are you serious? So we can sit on our butts and have a computer do stuff for us? But it's cool and fun so we automatically consider it good. Drones?? 3D printers?? Smartphones?? The list goes on and on.
Where's the REAL and USEFUL advances in technology? Cancer cures? Fixing the real causes of deaths and starvation around the world? Really improving our lives rather than having the entire universe with their heads stuck in their smartphones watching cats playing piano.
At the end of the day, if we're being honest, we'll admit that we LIKE all those fun and entertaining and mindless technologies, so we automatically classify them as GOOD. And we go to great lengths to come up with any possible justification we can think of, even if it's totally irrelevant and ridiculous. And nobody can criticize those technologies cuz we like them. Even if they're freakin' useless.
Hey, I'm an engineer, and I like technology as much as the next guy. But geez, let's get real folks. -
Mr. Scram, when the inane is 95% and the useful is 5%, then yeah, it seems reasonable to focus on the inane.
-
A TFLOPS is not a TFLOPS is not a TFLOPS/
If you had not posted it, I would have; there is a considerable difference between vector, scalar, ILP (Instruction level parallelism with CPUs), and MLP (memory level parallelism with GPUs) flops.
Further, a modern low power CPU may have the TFLOPs in execution resources of an older desktop CPU but does it also include enough memory bandwidth to support that level of performance with large datasets? Or is it limited to level 1 cache sized chunks?
-
A TFLOPS is not a TFLOPS is not a TFLOPS/
If you had not posted it, I would have; there is a considerable difference between vector, scalar, ILP (Instruction level parallelism with CPUs), and MLP (memory level parallelism with GPUs) flops.
Further, a modern low power CPU may have the TFLOPs in execution resources of an older desktop CPU but does it also include enough memory bandwidth to support that level of performance with large datasets? Or is it limited to level 1 cache sized chunks?
And all of this is irrelevant if the software hasn't been (or can't be) written to take advantage of the hardware. Which is what all the "ooo, shiny!!!" tech-spurts found last year when the RTX GPU's came out and they learned that little or no software supported it, and in fact it was the rare graphics software problem that COULD take advantage of the specific hardware configurations. I lost interest in the RTX cards earlier this year when it became apparent that only a certain type of raytracing problem really benefited from all that. And my $1,200 RTX card has been sitting in a box ever since. -
Mr. Scram, when the inane is 95% and the useful is 5%, then yeah, it seems reasonable to focus on the inane.
You must feel you have a horrible life.
-
Mr. Scram, when the inane is 95% and the useful is 5%, then yeah, it seems reasonable to focus on the inane.
You must feel you have a horrible life.
Whenever someone runs out of facts, they reach for the personal attacks and discredits. Geez guys, grow up huh?