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

 



Forgot your password?
typodupeerror

Comment Re:Anyone and everyone mentioning AI (Score 1) 61

This. Hint for them: probabilistic Monte Carlo algorithms are not AI and they have been around forever.

In fact, the Metropolis algorithm was develops by Nicolas Metropolis in the 50s to model neutron flux in thermonuclear explosions. It is now widely used to model photon flux in you favorite movie CGI so in a way nuclear weapons and you beloved Netflix shows have AI inside of them if you go by that lame definition

Comment Re:Your AI designed personal computer! (Score 1) 71

Not good enough. The bunker needs to be buried 2km deep on a remote island in the antarctic and "surveillance" means a battallion of marines, paid 10x their salary in gold backed by an aircraft carrier task force. Maybe, maybe then it will be safe. No guarantees.

Comment Re:Ruby never was that much ... (Score 1) 80

Well, I used to be a big Ruby guy and now I am a big TypeScript guy

That said, Ruby has its uses, I use it as a general scripting language to test things all the time. ruby syntax and metaprogramming is genius grade stuff

My last job was a mid sized company that used Ruby for everything and it was a bad idea. There are lots "Ruby on Rails Boot Camp" types around, they employed them, as opposed to retaining their experts, and now the company has had row after row of lay-offs. The problem is that ruby is good boot camp language for beginners who can't handle anything fancy and they are cheap, so management thinks it is a good idea. Cheap is always good. Lots of cheap glued together with non-coder Physics-grad managers who think SCRUM is a religion is... not.

Comment Re:Miracles (Chips, how do they work?) (Score 1) 126

> Put competent, diverse and industry specific leadership on your board instead of stacking it with multiple CEOs. (Intel's current board is terrible)

Yes and he actually stated that he regret not bringing in more semiconductor experienced people on the board, so it seems that he recognizes the failure on his side

> i. Incentivize rank and file who perform well. "We didn't meet targets as a company so nobody gets anything" is completely unacceptable.

Yup. Nothing kills motivation like the idiot or power-hungry political narcisissist down the hall gets stock options and you get to slave in the cellar.

> ii. Stop the constant layoffs for "underperforming" - nothing kills a company faster.

Dunno, my last company went down because they did NOT let the underperformers loose but played the whole "SCRUM mean everyone is equal and happy and in touch with their feelings" game. As opposed to letting people who are domain experts do their domain expert thing.

> i. Get rid of lines of business that do not contribute to core company. IT consulting services - seriously?

Intel does IT consulting?! Seriously? Why?
The only thing they should consult on is custom chip or interface design for niche markets.

> ii. Build your brand up again. You've lost it to the competition steadily over the past 20 years. Hire an all star marketing team and pay them significantly more than what they are worth.

Please no. Intel is only riding on marketing and they do not innovate. As soon as they have a dominant position they sit on their butts and let the marketeers suck the last drop of blood out of the market. Apples push into ARM silicon that outguns Intel in every measure and AMD breathing into their necks is a very good thing for the world. If you WintelSoft a break they will again dominate he IT world like a cancer for 20 years like they did from the late 90s.The little innovation they tried (Itanium/Itanic) was basically a raging dumpster fire.

Comment Re:Is people really using notebooks for AI? (Score 1) 75

Surely for dev purposes and testing and some document analysis. Apple also makes the Mac Studio a Mini which, if you give them enough RAM, kicks butt with AI. Expensive, but still the cheapest option for the performance

I have a MacBook Pro and a Ryzen with a 12Gb RTX under my desk. The RTX is surely faster with AI, but the Mac can load much bigger models with 48Gb of RAM. The secret sauce is that Apple shares the GPU and system memory, something intel’s old PCI architecture has problems with

The laptop also runs at about 15% of the power of the RTX for 60% of the performance. You do the sums.

If Apple takes this to a low power high performance AI server they will eat NVIDIA s breakfast.

Slashdot Top Deals

Cobol programmers are down in the dumps.

Working...