KK7NIL

20 ms·

271.
by KK7NIL 2y ago
1) No other US company is capable of using the parts for leading edge semiconductors. Case in point: GlobalFoundry had EUV machines but returned them to ASML as they weren't able to compete technically and capital wise. They would be
272.
by KK7NIL 2y ago
The "companies all of their employees moved to" don't make leading edge semiconductors with US based R&D, so I don't know why you're acting like they'd be able to do what Intel does. This isn't about I
273.
by KK7NIL 2y ago
At least some of the equipment (EUV machines being the big example but even things like high bandwidth oscilloscopes and many other tools we never hear about) are already export restricted (export is still possible but requires approval). W
274.
by KK7NIL 2y ago
Just to add to that: AMD did used to make their own chips, but spun it off as GlobalFoundry many years ago to avoid bankruptcy when times got tough. GF was on the leading edge for a long time and even had EUV machines but actually decided
275.
by KK7NIL 2y ago
> Atkinson says that while the CHIPS Act is a start, the United States is likely to need additional funding, support, or incentives if the goal is to see competitive leading-edge semiconductor fabs owned by U.S. corporations. “We’ve hit
276.
by KK7NIL 2y ago
I agree with your goal of getting more high quality publications but it's not at all clear to me that bonding would effectively lead to that. I think it's more likely that it would reward incremental studies that don't questi
277.
by KK7NIL 2y ago
Yes, we really need to make scientific publishing even more expensive and bureaucratic./s I appreciate your creativity but the real solution is to fund replication studies and improve statistics education in the sciences.
278.
by KK7NIL 2y ago
That's right but "the Bode plot of an integrator is a line with a -20 dB/decade slope" doesn't really roll off the tongue ;)
279.
by KK7NIL 2y ago
You are correct, I initially did not explicitly separate the noise according to its frequency (my mistake). Integration only helps with high frequency error and can actually worsen low frequency error, more details in my second reply to ken
280.
by KK7NIL 2y ago
> One can't cancel out random errors by integrating. An ideal integrator has a response of 1/s. That's just a 1st order low-pass filter with the pole at 0. Therefore, it will filter out high frequency noise. > Take a st
281.
by KK7NIL 2y ago
Would you like to expand on what I missed and how my summary is still wrong? Genuinely asking.
282.
by KK7NIL 2y ago
You are correct, I forgot to separate between long-term and short-term random effects. Short-term random effects (as in, the part of the gyro's random walk error significantly higher in frequency than the inverse of the integration per
283.
by KK7NIL 2y ago
We can assume the error will have a random (whether it's actually truly random or merely pseudo-random doesn't matter here, just assume it's indistinguishable from truly random for this discussion) and a non-random component.
284.
by KK7NIL 2y ago
> surely all the little errors compound? Random errors (i.e. noise) cancel out in the long run thanks to integration. You're then only left with systematic offset errors which can presumably be calibrated out to a large extent.
285.
by KK7NIL 2y ago
Indeed, analog circuits are much more process sensitive (while digital circuits are defect sensitive), which makes their design slower and much more closely coupled to the specific process node that is being targeted. This is why now a days
286.
by KK7NIL 2y ago
I'm about 1800-2000 on lichess and have almost complete aphantasia and I do feel like it really hurts me, especially in very sharp middlegame positions. When calculating I have to explicitly list out what squares are attacked by what p
287.
by KK7NIL 2y ago
The Bible is a collection of dozens of books written by many people across about a millennium. Anyone who's read it will know that it contains many different views and philosophies, especially when comparing the first books of the old
288.
by KK7NIL 2y ago
I think there are simpler explanations: 1) historically, wealth has often been obtained by violence, corruption and other less than ethical means; 2) even if it was obtained legitimately, most societies expect the wealthy to donate a signif
289.
by KK7NIL 2y ago
That's just not how taxes work: https://www.irs.gov/filing/federal-income-tax-rates-and-brac...
290.
by KK7NIL 2y ago
> We've got 8 high HBM, but will we have 100 in even 8 years or 1000 in another 8 more? No, probably not, but that's not the point. The point is that now memory and logic (in case of HBM) can be further decoupled, allowing us t
291.
by KK7NIL 2y ago
> Instead of one path forward there are now many. This is a feature, not a bug. Leading edge nodes are becoming harder and harder to develop on, particularly outside of memory and logic. This is because of limitations in terms of Ft, max
292.
by KK7NIL 2y ago
The only scopes using frequency mixing are the LeCroy scopes over 33 GHz and these have been commercial failures. All the other scopes on the market use time interleaving (time multiplexing) which has much better characteristics but require
293.
by KK7NIL 2y ago
I think the discussion has been all about high speed serial, so it's correct to say that the signal bandwidth starts at DC (or very close to it, there's often AC coupling) due to NRZ or PAM encoding.
294.
by KK7NIL 2y ago
There is no Nyquist frequency in analog. The Nyquist frequency only appears when we're going between a continuous (analog) system and a discrete (digital) system. For a real receiver circuit you'd want to sample at least once a sy
295.
by KK7NIL 2y ago
For amplitude (the vertical part of the eye), yes. For jitter (the horizontal part), which is usually the most critical, especially for NRZ protocols like USB 3.1, no, you need clock recovery.
296.
by KK7NIL 2y ago
No because modern high speed serial uses clock recovery for the timing (implemented as a first or second order PLL in the receiver), so you need to implement this in your scope too. In a real-time scope this can be done in software but in a
297.
by KK7NIL 2y ago
Not that I'm aware of. You could generate an eye diagram with an old sampling scope or BERT for probably less than $2k, but those likely wouldn't have the proper clock recovery, which means your horizontal would be sort of pointle
298.
by KK7NIL 2y ago
The old rule of thumb that you want the bandwidth to be 3 to 5x your baud rate is no longer true. It has been found that BW as low as 0.7x has very little effect in TDECQ and so the newest optical ethernet standards have moved to that for s