vardump

31 ms·

571.
by vardump 3y ago
Higher end home users are doing 25, 50 and 100 Gbps. Hardware is actually rather affordable now. Although there's just not enough PCIe bandwidth in current consumer hardware for 100 Gbps...
572.
by vardump 3y ago
Assume monthly profit per subscriber is $30. Not sure whether that's accurate. (2.6e6 * $30 * 12 * 5) / 100 = $4.68B. So far there has been 154 launches to put 5600 satellites in low earth orbit. So if each satellite costs $300k,
573.
by vardump 3y ago
When MP2 came out, my computer was barely able to play a song in 44.1 kHz 16 bit mono. I think bitrate was 192 kbps, but not sure. (Later on I was so surprised MP4 didn't replace MP3!)
574.
by vardump 3y ago
Musk denies this in his usual "colorful" style: https://twitter.com/elonmusk/status/1776272471324606778 Elon Musk @elonmusk · 2h Reuters is lying (again) So this is non-news.
575.
by vardump 3y ago
Good. IMHO, PCIe bandwidth is problematic on consumer devices. Can’t have multiple high bandwidth devices.
576.
by vardump 3y ago
Any word about performance?
577.
by vardump 3y ago
Any sources for those claims? Shouldn't it reduce labor costs when you have just one part instead of hundreds?
578.
by vardump 3y ago
Why would there be any difference? Most modern cars are write-offs once there's any structural body damage. Fender benders don't damage the relevant parts.
579.
by vardump 3y ago
Probably way faster than that. 3000 fps on a standard 2024 CPU without a GPU wouldn't shock me. (If optimized for AVX2/AVX-512.) But this is running on a low end FPGA. No ultra-wide vector FP units, no massive SRAM caches, etc.
580.
by vardump 3y ago
Being constrained to a low-end FPGAs is quite a handicap. Even high-end ($10k+) FPGAs would be a severe disadvantage compared to where ASICs were 10+ years ago. So no point to be snarky.
581.
by vardump 3y ago
Do need to have Parallels and Windows installation on macOS. But other than that, can run a native version of Visual Studio on ARM64.
582.
by vardump 3y ago
While this might help in the short term, in the long term this is going to be detrimental to the US car industry by slowing their path to competitive EVs. The Chinese are hungry and their EV industry is developing fast. Soon those cars are
583.
by vardump 3y ago
Compared to now, back then it was a lot more you had to figure out on your own. I remember coding a game on C64 in the eighties. Just to figure out how to print the players score so that it is sufficiently fast was a challenge. Dividing b
584.
by vardump 3y ago
In the era before internet, you either figured things out on your own or you couldn't get anything done. If you were very lucky, some magazine might have mentioned it. Another way out was to just use disassembler if some other software
585.
by vardump 3y ago
Sure, but the point was that nothing of the original logic will be recoverable, once the FPGA loses power. Unless you make an ASIC that's effectively an FPGA.
586.
by vardump 3y ago
For military use FPGAs with encrypted bitstreams is a pretty interesting application. The arming process provides the key and once the single-use weapon hits the target, there's no way for the enemy to get to the original bitstream. So
587.
by vardump 3y ago
No one said it's fine.
588.
by vardump 3y ago
Not saying you're wrong, but what else is there that just works and is usable instantly when you open the lid? While my work Windows laptop might be faster, it's certainly not the one I'm going to pick in a pinch or when I wa
589.
by vardump 3y ago
That's actually a pretty normal way to do things. Optimization for JITting, let the CPU hardware to do the heavy lifting instead of putting conditional jumps everywhere. It's useful for other things as well. I've used SIGSEGV
590.
by vardump 3y ago
Nowadays it's often hard to even notice a 4 MB file, especially when you're on a gigabit connection or better. Of course unfortunately not everyone has a connection like that. But can understand how someone might have missed it.
591.
by vardump 3y ago
LZW would do very badly on a list of binary chess moves. Huffman (and even more so arithmetic encoding or ANS (asymmetric numeral systems [0]) would be significantly better, if you're careful how you encode data. [0]: https:/&#x2
592.
by vardump 3y ago
Kinda, but with a different dynamically determined dictionary for each successive position.
593.
by vardump 3y ago
They should first ban growth hormones and overuse of antibiotics.
594.
by vardump 3y ago
That's... crazy. When I've been in safety critical industry, you had records for everything .
595.
by vardump 3y ago
Also Censor Design's Wonderland XIII was pretty amazing at the time: https://youtu.be/oEiQxFL2Tzk?t=431 Then again the following is even more mind blowing coming from an unexpanded C64 (yes, from a real SID!): https:&
596.
by vardump 3y ago
Ah you're right. Anyways I wonder why it's still so slow. 60-120 cycles sure looks like a CORDIC implementation, but perhaps not.
597.
by vardump 3y ago
Pretty weird Intel's sine computation latency hasn't changed all that much over the years. Latencies have been pretty similar for 20 years. EDIT: That's a paper for a software library, not the CPU's internal implementati
598.
by vardump 3y ago
There's no russophobia except in Kremlin's propaganda. European nations just can't take any chances with aggression and untrustworthiness of the current Russian leadership. That's just common sense. This is nothing again
599.
by vardump 3y ago
Good catch. Somehow missed it when quickly skimming the article.
600.
by vardump 3y ago
2nm, sure, but what about SRAM register files and caches? SRAM has pretty much stopped scaling already. SRAM scaling is where I'd love to see some advances.

More ›