Procedurally generated spaceships: https://github.com/a1studmuffin/SpaceshipGenerator
Pretty cool looking.
1 Like
i am here to give a PSA that if you havenāt kept up with xkcd you really should.
itās a wonderful comic.
3 Likes
2 Likes
ā¦ Forty-EIGHT cores with 7nm architecture? Way to go AMD! 
2 Likes
That canāt be right !! My intuition tells me tunneling is going to be an issue thereā¦Iām no CPU architect but how do they deal with that? Is it software to correct errors?
For example, even with a .01%(warning not real math) chance thatās still 10,000 times a second on averageā¦ 7nm is only a 70 angstroms and at that point youāre exponentially approaching significant probabilities as you move those traces closer.
Itās been awhile since Iāve done transmission problems though so I could be way off.
1 Like
What application needs lots of processing power, but doesnāt particularly care about absolute accuracy in every single calculation? Robotics. Machine learning. Iām sure there are others. It may even be useful in large scale simulations where they can work in some sanity check code or they have that trait of highly-repetitive processings. One error is simply factored out of the overall trend.
1 Like
I almost tagged you because I figured you would have an answer.
That makes sense though, a lot of noise can be filtered out these days just using data analysis. For example getting subpixel resolution for a point on an image. I suppose you could push the hardware way past its limits and just correct for it after the fact in software.
1 Like
It sounds like youāre still thinking in terms of getting a 100% ironclad correct final answer. My apologies if I misunderstand.
The point here is that many applications donāt need that ironclad guarantee. It is in their nature to be tolerant of inaccuracies.
If Iām running a simulation where Iām only sure of my input data +/- 1%, then does it matter if the computer is 99.99% accurate? My input data is the weak link, so getting 100% accuracy from inaccurate data isnāt particularly important.
In robotics, if Iām trying to crank out solutions to kinematics equations as fast as possible, then I can either have 100% accurate solutions (well, within the limits of my math) every N milliseconds or I can have 99.99% accurate solutions every N microseconds. Unless Iām doing eye surgery with a robot arm the length of a football field, I can probably live with that modest precision loss and will be thrilled to have a better reaction time. Or be able to handle more complex kinematics problems.
Where the ironclad guarantee is needed, a traditional architecture would be used.
I wonder if these new architectures would allow for every calculations to be run three, five or seven times, with the machine using the most common answer. Or the average. Whatever. Selectable precision. It would be a bit like ray tracing. Keep casting rays until you are happy with the image quality.
Now Iām reminded of unumsā¦
2 Likes
Not at all. I totally understand that garbage in = garbage out and that any ācorrectionā is still has some sort of error. When I spoke of subpixel resolution as an example that is what I had in mind, where we still had ~few % error from the correct answer.
What I was thinking is that while electrons may be lost due to tunneling, statistically if we know we will have the ārightā answer 98% of the time or whatever it is, we can call that the right answer because it happened 98% of the time. What I was trying to get at is that even though we may lose some data/calculations to tunneling we can still correct for it after the fact. It wonāt be a linear improvement due to the time spent averaging and filtering out bad data, but still faster than larger cpu architectures. I know Iām probably not 100% correct(who is?), but as long as Iām on the right line of thinking Iām okay with it.
1 Like
The Tesla Model S can swim:
3 Likes
I love the tweet Elon posted after. Man I wish I could buy a Tesla just for Elon being awesome
4 Likes
Hmmm, seems AMD/ATI are catching up fast.
AMD RX 480 Gaming Benchmarks Leaked, Rivals Fury & 980 ā Full Specs & Hands-On Preview Surface
Read more: http://wccftech.com/amd-rx-480-gaming-benchmarks-leaked/#ixzz4CQdk52U4
2 Likes
Well, that was nostalgic. I wonder if someone else remembers the Brackenwood animations of a decade past?
1 Like
Respect to the British that the exit from the EU.
I think Iāve shared these guys someplace else on here, but have an inspirational promo 
3 Likes
The Carl Saganās Cosmos documentaries. Still incredible, even if they are old.
3 Likes
Please no.
3 Likes
Explain, I do not know what people say in your the country.