Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
slivym
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
45 ms
·
151.
▲
by
slivym
8y ago
I work in the FPGA industry and I've supervised university projects, seen lots of research, hired PhD grads and stuff. I've got to agree that one of the most frustrating parts of FPGA research is that it's almost uniformally
152.
▲
by
slivym
8y ago
TVs already have this, it's referred to as motion smoothing and people hate it. TVs with high refresh rates create interpolated frames to smooth the video when the source material is at a lower frame rate. It works for some situations
153.
▲
by
slivym
8y ago
The tennis racket jello shot is 2500 FPS and then played at 30 FPS and interpolated to 120 FPS. The 300k number is just pure screw up in the article. The actual paper states it is 300k frames total generated not frames per second.
154.
▲
by
slivym
8y ago
I don't know if you've worked in a large corporation, but during my time in one we would constantly get senior management talking about how we were doing really well and going to do great things and we were making history.... whil
155.
▲
by
slivym
8y ago
Frankly, this all seems like quite a lock of fuss over nothing. Firstly, pad locks are generally very easy to just break open with a set of bolt cutters. Particularly, if you look at more secure traditional pad locks practically all of them
156.
▲
by
slivym
8y ago
I think the flip side to this is that the communities are becoming so large now that they aren't really a community. I sometimes go on the /r/UK and the level of discussion gets lower and lower as the community dissolves. Rat
157.
▲
by
slivym
8y ago
Google demonetized videos that their sponsors wanted demonetized. Now, I don't know, but I find it rather far fetched that Google's large sponsors are exclusively part of the 'radical political left', but even if they ar
158.
▲
by
slivym
8y ago
On point 1 take for example Ubiq, it'll cost you say 10x the 1h rate or $4.7k, and the 24 hour volume is 130k, that seems like a pretty good ratio to me. Or Bytecoin - let's say 100x since it's harder to get the capacity - co
159.
▲
by
slivym
8y ago
Running industrial equipment off your local residential power grid is both against the terms and conditions of every power company under the sun and also incredibly stupid to do.
160.
▲
by
slivym
8y ago
I think that is a really bad line of logic. There are lots of things that have been around for a long time and make some people plenty of money that are bad things. Pyramid schemes fit your definition - should I start buying in? The opposit
161.
▲
by
slivym
8y ago
There are plenty of good use cases for these, and I'm sure there are legitimate ways people can decide they're happy with the trade-off. However, for me this screams of a Facebook style scandal coming sooner or later. People buyin
162.
▲
by
slivym
8y ago
Good idea in theory, in practice it's a shit show. The current level of language support is "As long as you know how to write what you want in HDL, then you can write some C that uses custom pragmas and design patterns and attribu
163.
▲
by
slivym
8y ago
I'm not so sure - it's a fantastic fit for GPUs, FPGAs make sense for stream processing for video encoding - but that's more of a discrete device play where you can plug the video feeds directly into the FPGA daughter board.
164.
▲
by
slivym
8y ago
What I'm really interested to know is how this looks from a power-performance perspective. An Arria 10 can draw 20-40 watts (I'm not sure that 90w is really realistic), and the TDP for the Xeon is 165W. When I saw the discussion a
165.
▲
by
slivym
8y ago
It takes in the order of seconds to reconfigure. JIT is not really viable.
166.
▲
by
slivym
8y ago
I would give a very strong health warning about taking those numbers seriously. What that 2nd reference doesn't make clear is that it is NOT a standard Matrix Multiply Operation, it's an 11 row by 16 column Matrix Multiplication.
167.
▲
by
slivym
8y ago
Using OpenCL to program an FPGA is a significantly more difficult task than programming in HDL. Atleast for what would be relevant to a Xeon co-processor. The OpenCL flow is just terrible at getting to the performance levels you need to rea
168.
▲
by
slivym
8y ago
It's not really all straight forward though. Whilst it's true that it's cheaper to get a MoviePass for a month than to buy 2 tickets in that month I can't remember the last time I paid full price for a movie ticket anyw
169.
▲
by
slivym
8y ago
MoviePass and Amazon Prime aren't the same, because Amazon Prime produces a lot of its own content, and is the consumption point for the other content it serves. Amazon buys the rights to shows, and produces shows - so it has a fixed c
170.
▲
by
slivym
8y ago
On the flip side I've managed teams in organisations that see rockstar developers as indispensable. The result is you get the smartest people doing the dumbest stuff for the company- working almost completely against the core goals of
171.
▲
by
slivym
8y ago
Whilst it's true that modern chips have problems with routing clocks, that's not really a limiting factor in chip size. You split the design into clock regions and have clock crossing logic. There's obvious ways this happens
172.
▲
by
slivym
8y ago
That's really hilarious, because Intel's entire automotive division is run by the guy who came from the Mobile Eye acquisition. For those that don't remember - Mobile Eye was one of Tesla's tech suppliers until they spli
173.
▲
by
slivym
9y ago
The problem isn't how risky your career at Google is, it's about the correlation of that risk with the risk of your investments. The most likely risk of working at Google is if you lose your job. Well losing your job is highly cor
174.
▲
by
slivym
9y ago
It seems strange for Musk to make a stupid for musk to claim that profitability in Q3/4 is important when they're talking about his debt. The maturing debt will not be covered by the small profits. If there are profits, which I su
175.
▲
by
slivym
9y ago
I think as mentioned in the article - the family started making public statements and Tesla figured "Oh, you don't want PRIVACY (read: to STFU) then we're going to be TRANSPARENT (read: slander your dead relative to boost our
176.
▲
by
slivym
9y ago
This would be trivial to solve with a vertical mounted LIDAR. You vertically mount it and if the horzontal distance measured level with the bumper of the car are significantly closer than the horizontal distance at a lower angle you can cla
177.
▲
by
slivym
9y ago
I think when you look at it from that perspective Tesla are in real trouble. Essentially what they're publicly stating is "Our car detected it was in a dangerous situation, gave warnings to the driver but took no action". Tha
178.
▲
by
slivym
9y ago
What do you mean by buyback in terms of revenue? * For every $X we will spend $0.0001 buying back tokens? * For every $X we will buy back Y coins? * For every $X we will buy back Y coins at a set value? I guess it's got to be the first
179.
▲
by
slivym
9y ago
Why does advanced mean theoretically better? From what I've seen from industry discussions, other companies seem to believe that Musk is automating processes at such a high cost that it would be more cost effective to actually employ h
180.
▲
by
slivym
9y ago
Lot's of companies compete in lots of sectors. If you buy a Ford van you're not going to go to chevy. But that doesn't change the fact that Ford's primary business is normal cars and chevy competes in that industry. If I
More ›