_ihaque

The problem is not a brief surge of startup energy; it is the total sustained power you need to heat the water for the shower.

Let's say you want to run a single shower at 2.5gpm, the water comes into the home at 55F (12.8C, pretty temperate!) and you want your shower at 100F (37.8C), so a 25C rise. 2.5gpm = 9.463 L/min = 158mL/sec. Water's heat capacity is 4.184* J/mL/degC, so the sustained thermal power you need to deliver is:

4.184 x 158 x 25 = 16.5kJ/s = 16.5kW

At 240V, 16.5kW would be almost 70A.

What makes a heat pump heater more efficient is that it multiplies the electrical power input into a larger thermal power output, by "borrowing" thermal energy from the surroundings. So you could get (for example) 15kW thermal power from 5kW electrical power. In practice the heater will be tanked, rather than tankless, for two reasons:

1. The effective power will probably be less than that.

2. That was an estimate to run just one shower and you might want to run more than that!

So the tank lets you buffer the power draw over time; there will be some thermal loss over time, but insulation is cheap, and it's hard to fight a 3-5x amplification from the heat pump.

* Yes, yes, per gram/K but 1mL of water is close enough to 1g for this.

This is an uncharitable take on the original comment which literally said:

> the fact this is how hot it will be from now on

That sounds to me like an acknowledgement that, in fact, the climate has changed.

> Isotopes of a given element have the same number of protons (92 for uranium), but different numbers of neutrons (143 for U-235 and 146 for U-238), meaning they behave the same chemically but differ in mass or radiation emissions.

A fun but off-topic note: "behave the same chemically" is only approximately true. For heavy atoms like the ones discussed in the article, it's basically true. But for hydrogen, adding one neutron doubles its mass and you can get real effects on chemical reaction rates.

https://en.wikipedia.org/wiki/Kinetic_isotope_effect

And of course, an obligatory "In The Pipeline" link on how it's used in drug discovery: https://www.science.org/content/blog-post/isotopes-get-your-...

Based on the description of the wiring to the motor (24V, GND, POT1, POT2, NC), it doesn't sound like the original setup would have been drawing much power through the pot either -- there's probably something else on the other end of that wire that is doing modulation based on the sense resistance, and the motor is itself drawing power from the 24V line. So while it's true that there should be a check for the allowable limits on the digipot, I don't think it's actually being used to sink much power.

The linked article doesn't say that much about what the bootstrap load is or how it worked, but luckily for us @kens wrote a whole article about it: https://www.righto.com/2020/10/how-bootstrap-load-made-histo...

If this (or, more accurately, a drug developed downstream of this technology) worked, they absolutely would.

A great example of this in relatively recent history is the treatment of hepatitis C. The treatments pre-circa-2011 were pretty crappy: interferon/ribavirin had poor cure rates and bad side effects. But it was still better than hepatitis destroying your liver, so people dealt with it.

Then, in 2011, as the culmination of years of trials, telaprevir (from Vertex/J&J) [1] and boceprevir (Schering-Plough/Merck) [2] were approved and were DRAMATICALLY BETTER than interferon/ribavirin.

...and then just two years later both of these drugs got nuked by the approval of sofosbuvir (aka Sovaldi, from Pharmasset/Gilead) [3], which has _cure_ rates in excess of 90%. Telaprevir and boceprevir were pulled from the market because there was simply no more market for them once sofosbuvir hit. Scientific competition at its finest.

There is absolutely a dark side to pharmaceutical pricing and licensing, but please don't let the existence of that dark side cloud your vision of the transformative impact that those of us in biotech R&D want to have (and in many cases, have had). I was at Vertex when the early trial data on telaprevir started coming out and it became clear that we might be able to offer patients real hope who did not have it before.

[1] https://en.wikipedia.org/wiki/Telaprevir [2] https://en.wikipedia.org/wiki/Boceprevir [3] https://en.wikipedia.org/wiki/Sofosbuvir

I think you mean SMT (simultaneous multithreading), as opposed to SMP (symmetric multiprocessing)?

Those wondering about the mouthful of a name that is "orforglipron" may be interested in the INN (international nonproprietary name) standard for generic drugs. That name is a "generic" or "nonproprietary" name.

Those typically have a "stem" which indicate the mechanism by which the drug works. In this case, it's "-glipron", which you can parse as "glipr" -- GLP-1 receptor -- and "on" -- it's an "agonist", or something that turns "on" the receptor. So "-glipron"s are drugs which agonize, or activate, the GLP-1 receptor. There are other gliprons in trials, including danuglipron (https://en.wikipedia.org/wiki/Danuglipron) from Pfizer, and many others that don't have assigned INNs.

Semaglutide and liraglutide also agonize GLP-1, so why aren't they gliprons? Because chemically they are analogs of the naturally occurring GLP-1 peptide, so they get the -glutide (GLUcagon-like pepTIDE) stem.

I don't know why tirzepatide got that name. It's a peptide, so "-tide" makes sense, but it's actually listed as a "various" exception in the master document of INNs: https://iris.who.int/bitstream/handle/10665/379226/978924009....

No, it's not a reformulation of tirzepatide. Orforglipron is a small molecule with a molecular mass of 883 g/mol and tirzepatide is a large molecule peptide six tims larger, with a molecular mass of 4813 g/mol.

The Wikipedia pages are decent -- the picture makes it fairly obvious how much smaller orforglipron is than tirzepatide even if you don't know how to read chemical structures; just count atoms!

https://en.wikipedia.org/wiki/Orforglipron

https://en.wikipedia.org/wiki/Tirzepatide

We definitely do (as in, I myself have) hire folks from "cold" inbound applications.

Not that I know of! Note that the positions we list are typically in different teams so it's worth reading the descriptions to make sure you're picking whichever one is most appropriate to your experience+interests (and, as a corollary, if you weren't a fit for a previous position, you may be one for one that comes up in the future).

I should point out that we are hiring! Software engineering, data science/ML, and IT positions are available in Salt Lake City, Toronto, New York, and London:

https://job-boards.greenhouse.io/recursionpharmaceuticals

Yes, we (and plenty of others) have prior data showing that perturbations (things that you can do to cells, like knocking out genes or putting drug-candidates on them) have different effects at different times, so we are excited to see the potential for time course imaging.

Note that "video" is a little different here than the way we all usually think of it - think "minutes or hours between frames" not "frames per second".

My team did part of the work described in TFA, AMA.

These are tests of circulating cell-free DNA (cfDNA), not circulating cells. The fraction of circulating _cancer cells_ is too low (outside of blood cancers or metastatic cancers) to be technically feasible to detect; ditto in pregnancy. By contrast, while there isn't much cfDNA (DNA fragments floating around in blood plasma outside of cells) -- on the order of 1-10 ng/mL -- there's enough to extract reliably, and critically, the fraction of that that is tumor-derived (or fetally-derived, for prenatal testing) can be high. It's already high single to low double digit percent at 10-12 weeks of gestation, and can go much higher in some cancers.

But the biggest challenge for these tests is that this "tumor fraction" can be very very low in early stage cancers, which is why stage I sensitivity tends to be quite poor.

Biological microscopes illuminate the sample from below, as the samples are typically transparent. Metallurgical microscopes illuminate reflective samples from above.

*"Below" meaning "on the opposite side from the objective" - you illuminate _through_ the sample.

Hi Udayan - nice to see another shot at electrowetting microfluidics in this space.

1. How does your technology or product offering distinguish itself from prior attempts like NeoPrep (née Advanced Liquid Logic) from Illumina or Voltrax from ONT?

2. How "hackable" is the offering for power users to implement custom protocols on the instrument and consumables?

Neon is produced from (cryogenic) separation of air, so, yes?

For anyone interested in why the system looks the way it does, I recommend patio11's insightful recent blog post, which is nominally about payroll systems but ends up discussing the history of taxation in the US because that's the origination of many of these systems: https://www.bitsaboutmoney.com/archive/payroll-providers-pow...

A relevant quotation for much of the discussion here:

> And, relevant to the question of whether Intuit controls U.S. tax policy: it can’t, because that would imply they have wrested control from Norquist. Norquist considers a public filing option a tax increase by stealth and opposes it automatically.

Level 2? Certainly not.

A typical L2 charger will provide in the range of 6-10kW (AC, marginally less ends up going to the battery after conversion losses). 10kW is only 13hp; forget acceleration - the car draws more than that at highway speeds.

(You can come at this result another way too - an L2 charge may take 6-10hr to refill the battery from empty. But the car would not be able to drive 6-10hr at highway speed starting at 100% charge! So the L2 must be delivering less power than the car consumes at cruising speed.)

I believe the 387 was an actual coprocessor; the _487_ may be what you're thinking of -- the 487SX contained a whole 486DX inside and just disabled the host processor [1].

[1] https://en.wikipedia.org/wiki/X87#80487