Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×

Comment 5045M (Score 1) 30

Given the 5090 has over 21000 cuda cores this is more like a 5045 with half the compute capacity, a quarter of the power limits and around 70% the memory/bus.

I guess on the plus side there's no 12v high fire connector to worry about because these chips are mb integrated and wouldn't draw enough power to worry about anyway.

Comment Rules to live by (Score 3, Informative) 44

Collect a paycheck while you go to interviews. Jump ship when you get a better offer.

A friend of mine put himself on the market every two years whether he wanted a job or not. In retrospect, I feel that was the most insightful and logical decision to be made at the time. As he pointed out, "nothing might come from it, but if they offer me my dream job there's no way I wouldn't take it".

IIRC, he had some rule about not taking a new job unless it paid more than 15% higher than his current job, or some such.

A further consideration is what you do *while* in your job. Always be looking for your next job, but that can also be within the same company. Talk to people at other departments, get a feel for what they do, listen in on a few staff meetings, and maybe volunteer to do some task or other that they are struggling with.

In other words, be curious and don't be afraid to learn new things.

When your department gets cut, lots of times you can switch to a different department. If you have shown that you understand the job requirements and get a recommendation from other people in that department, the company doesn't have to spend all the money needed to interview people.

I started in the OS division, then switched to Networking, then switched to compilers, then became a consultant. Switching allowed me to keep my job for several more years than would have been otherwise possible, and having a background in multiple disciplines was useful in finding jobs at other companies.

You're not actually working for the company, you're working for *yourself*. Lots of people don't understand this, the company can and will "unhire" you for any reason, or no reason. It might not even be your fault.

Always be on the lookout for your next job. You take care of yourself.

(And needless to say you don't go around trumpeting that opinion at the company. You can still have loyalty and still expect to continue working, but just know that your ultimate allegiance is to yourself.)

Comment Re:Epistemology (Score 1) 109

There are really two choices, that General Relativity is horribly wrong by a few orders of magnitude.

False Dichotomy. Newtonian Physics works pretty darn well for a lot of situations, is simple, so much so, we still use it because it is practical. But it is wrong.

For all we know, Einstein's equations may be suffering the same basic problem, we just haven't figured out why it isn't working where we see it not working as expected. But it works for a large number of scenarios that it is still usable.

That is my current working theory ;-)

Comment Re: Royalties (Score 1) 52

You seem to think that only the R are on the take. It is clear, everyone in DC is on the take. The only difference is which team you're on as to whether or not you care (or don't care).

They are all Hypocrites, talking out of whichever side of the mouth gets them elected again, so they can grift off the taxpayers.

"But the other side is worse" you might say. Sure, Fine. The shit sandwich is better over there with you.

Comment Re:So a lot of people realized ... (Score 1) 57

He means takeout as opposed to a poor dine-in experience. It's odd reading this summary being called "curious" when the summary buries the real stat at the end, people are spending less time to pick up their food at these not-restaurants which frees the CS workers to help out with other tasks, that's the curious efficiency boost.

Comment Re: Router (Score 1) 43

Lifespan of the HW is the important detail here. My home is not 99.99999% Uptime required. I don't need "Cisco" grade anything for the "Cisco" grade price. pfSense is good enough for home, it HAS support available (commercial grade) but I do not require it, because I'm not needing 99.99999% Uptime and 1 hour response time.

My current hardware has been running over 5 years. It needs to be replaced. Even Commercial grade needs replacement, even if it is just for obsolescence, eventually. My WiFi Router isn't routing anything, it is just an AP at this point. I use NONE of the features on it, other than getting clients on the network. Neither DHCP nor DNS is served by it. The point being, I do not Trust the OS on it, and cannot upgrade or swap it out due to Proprietary Blobs and OpenWRT is not ever going to run on it, and I don't like the HW that OpenWRT does run on (for other reasons). Therefore, it sits there being an AP serving clients, behind an actual Firewall/Router.

I'm happy you found OpenWRT setup you like. I used to by WiFi Routers that had OpenWRT, and I like the OS. So this is not a bash against that OS. I just have chosen a different route.

Comment Several countries in America are (Score 1) 201

At least, Argentina and Mexico used to observe daylight savings (here: horario de verano, Summer time) until roughly 10 and 5 years ago. They are very different countries, in very far away parts of America. And yes, both switched away from daylight savings.
Of course, many other countries –the most tropical part– has never observed DST to begin with.

Comment Assumptions had changed. (Score 0) 91

For a 2nd point of reference, I note that in the run up to the 2016 election his *reasoning* for his predictions was based on historical trends.

Before the party conventions he predicted Marco Rubio to be the candidate, because historically the position Marco held had been the candidate in many previous elections.

That's a basic fallacy. The "Baltimore Stockbroker" is more than a scam, it's a fallacy that we should watching for in our daily lives. It comes up in finance whenever you hear "this guy predicted the past 5 stock market crashes, and for $10 you can get his book where he predicts the next one". (Compare with the number of people trying to hawk a book on the stock market, the number of published measurements of the stock market (around 200, AFAIK), and the probability that one of these correlates with the stock market crashes of the past.)

Nate Silver is someone with a deep understanding of the mathematics of statistics, but not the fundamental premises on which those statistics are based; IOW, he didn't recognize that the base situation had changed, so that his statistics turned out to be measuring the wrong thing.

In the run up to the 2016 election the online betting pools swing away from Nate's predictions, it's the equivalent of multiple people trying to guess the number of beans in a large jar at a carnival (the average guess is astonishingly accurate). The betting pools were accurate, and all the polling, as it happens, were not.

Furthermore, polling has become tainted by political pressure. If the polls show one candidate comfortably in the lead, no one feels the need to donate to the campaign. If the polling shows the candidates locked in a race too tight to predict, then OMG we better send them some money or else the wrong lizard will get elected. Polls are now a mechanism for encouraging campaign donations.

Nate didn't realize that the assumptions on which his math was based had changed.

I think the same thing happened in the 2024 election.

Slashdot Top Deals

"One Architecture, One OS" also translates as "One Egg, One Basket".

Working...