Will Nvidia skip a new GPU release?

[Shadow Reaper]

You should be nice to me because I am good enough and smart enough, and dog gone it, people like me.

To be fair, the part you are talking about was done online with a free RF Link Calculator, back 6 or 7 years ago. That is the easy, fun part. Now there are so many link calculators online it’s hard to decide on just one. Certainly, 1 watt is enough. The trick is in the gain at the space end. With calculations like this, it is standard practice to put in 7 db “extra” so there are never any problems. IIRC I put in 11 db. It helps that back then, “Software Defined Radio” was running something like -32 db sensitivity. By now it’s probably much higher. It uses a pencil beam which is pretty wide by the time it reaches Earth, and includes a starfinder that just needs to find our planet—something many orders magnitude within its capability. The squiggle motors drive the pointing with sub-micron accuracy, which is again, easily within their capability.

What Musk uses are lasers, which is far better but unnecessary for this application. To talk to Mars you absolutely need lasers and a very advanced starfinder.

View: https://youtu.be/xNx_gU57gQ4?si=F0Ao6Emd7kWkBbkL

 

[Radegast74]

You should be nice to me because I am good enough and smart enough, and dog gone it, people like me.

To be fair, the part you are talking about was done online with a free RF Link Calculator, back 6 or 7 years ago. That is the easy, fun part. Now there are so many link calculators online it’s hard to decide on just one. Certainly, 1 watt is enough. The trick is in the gain at the space end. With calculations like this, it is standard practice to put in 7 db “extra” so there are never any problems. IIRC I put in 11 db. It helps that back then, “Software Defined Radio” was running something like -32 db sensitivity. By now it’s probably much higher. It uses a pencil beam which is pretty wide by the time it reaches Earth, and includes a starfinder that just needs to find our planet—something many orders magnitude within its capability. The squiggle motors drive the pointing with sub-micron accuracy, which is again, easily within their capability.

What Musk uses are lasers, which is far better but unnecessary for this application. To talk to Mars you absolutely need lasers and a very advanced starfinder.

View: https://youtu.be/xNx_gU57gQ4?si=F0Ao6Emd7kWkBbkL

Stuart Smalley FTW!

 

[Shadow Reaper]

- YouTube

Auf YouTube findest du die angesagtesten Videos und Tracks. Außerdem kannst du eigene Inhalte hochladen und mit Freunden oder gleich der ganzen Welt teilen.

youtube.com

 

[Yex]

Oh absolutely. GPU and ram have skyrocketed. On the verge of TMI, basically I moved in with the gf and rented my properties out for the past 3 years. One of the sacrifices I made was to box the computer up and only console game as there simply isn't enough room for me to have a full setup in her place. Unfortunately we haven't worked out and I'll be moving back into my own place. On the bright side, I've been collecting rent for 3 years as a passive income that I haven't touched and as we're all probably aware, rent prices have gone through the roof( no pun intended) which worked out nicely for me. Now with star citizen making leaps and bounds in those past few years I've missed and going forward, I thought it's time to splash the cash and get that monster setup. Probably part of a midlife crisis but I'm not quite old enough to call it that haha.

What I read was:
> I am single
> I own property
> I have disposable income
> early 30's

My 2 pence probably isn't worth anything, but I like my mac pro I basically never play on big desktop because I work at my desk and sitting at my desk feels like work. I think most people feel the same, work life balance is real.

Sitting in the garden with my other half & cat while playing baldurs gate is like.. Peak
It has all day battery even playing games like stellaris, and well.. I take it around charity demo stuff at uni's, and play a lil slay the spire while I wait in lobby.

I regret building my desktop in 2016 for SC release 'soon'
:|
ARM isn't the future, it's the present
Risc5 is the future
Amd64 is just, the worse option. And that's coming from someone who did work bare metal datacenters

 

[Yex]

AI finally works and wonder oh wonder solves all of mankinds problems in the blink of an eye

Mankind's problems is mankind
I agree with @CitizenDad it didn't enhance productive, we just do the same thing but lazier.

Internally I get cc'd on an email thread the ceo wrote with ai, a director responds with ai, the head of sales chips in with ai and myself & my engineers go.. Guys that's not how our product works.. Then they get mad because ai said it did. Then I have to sit in a board room tip toeing around ego's like broken egg shells explaining why the copy paste code it spat out looks real but doesn't actually do anything useful.

The topic I'm thinking of exactly is they thought some math function could eliminate hullucinations in production. But it doesn't work like that.. You cannot just apply levenstein to individual calls all expect it to fix everything :|
Rinse repeat that 3 years, if you say no, you get laid off.
I don't do the meetings anymore my team just spend 3wks implementing it instead of fixing real issues :))))

Send help :)))))

 

[Yex]

4 PhDs in the room all said it should work

:| you don't agree with everything I say but I am a PhD in the room :|
Published, award winning ;_; why don't you love me shadow

 

[Thalstan]

Mankind's problems is mankind
I agree with @CitizenDad it didn't enhance productive, we just do the same thing but lazier.

Internally I get cc'd on an email thread the ceo wrote with ai, a director responds with ai, the head of sales chips in with ai and myself & my engineers go.. Guys that's not how our product works.. Then they get mad because ai said it did. Then I have to sit in a board room tip toeing around ego's like broken egg shells explaining why the copy paste code it spat out looks real but doesn't actually do anything useful.

The topic I'm thinking of exactly is they thought some math function could eliminate hullucinations in production. But it doesn't work like that.. You cannot just apply levenstein to individual calls all expect it to fix everything :|
Rinse repeat that 3 years, if you say no, you get laid off.
I don't do the meetings anymore my team just spend 3wks implementing it instead of fixing real issues :))))

Send help :)))))

three years? Eesh, your executives have a significant retention rate for what you tell them. It seems we have to remind them every 4-6 months... Every time a new freaking shiney thing comes along and says it will save time and money (ie, they can lay people off for the cost of a netflix subscription...even if it doesn't work right)

 

[Yex]

three years? Eesh, your executives have a significant retention rate for what you tell them. It seems we have to remind them every 4-6 months... Every time a new freaking shiney thing comes along and says it will save time and money (ie, they can lay people off for the cost of a netflix subscription...even if it doesn't work right)

You know what they say, if somewheres always hiring then they're always firing.

I have a nice smile or so I'm told. So they keep me

 

[Thalstan]

Been at my company 20 years. First 6 months I was there, I told them they should really make an effort to consolidate and mesh their databases so they work together. They should also make an effort to design holistic systems so that new systems will work with legacy systems and legacy systems can be upgraded/updated without causing issues.

NO…we don’t need to spend that effort. We are a small company and our data needs are not that complex. If we have duplicate data sets or incompatible databases, that’s okay. We need to ensure that everyone has the data they need for their program. Besides, we need to keep our costs down and those would be big projects. (Maybe 5 big programs and 3-4 smaller ones that would need integration into a unified system). It would also take 2 years…

Been pitching unified structures every 5 years or so as we keep growing…from 40 folks to SIGNIFICANTLY more.

Now suddenly they want to use AI to start mining and analyzing data…and are finding out that with all the separate programs and lack of a unified system design standard over the years, they can’t do it.

Example, we have one program that looks at certain types of outages,and another platform that looks at a different type of outage. Being able to look and cross reference outages from both systems is an obvious way to see how things might be related, but because program 1 was run by “Sally”, and program 2 was run by “Susan”, they chose different structures, different vendors, and they are completely incompatible with each other. Yet, neither wants to migrate to the other’s platform because they will lose status. This means an entirely new program that will incorporate both…. Most of the datasets they will use are the exact same data sets (duplicate work) with only about 10 percent of it being different. It would not have been hard to add the functionality of second system to the first when it first was developed, ut they refused to do it as it might give the other group access to “their” data.

They will need to spend money to combine 15 major programs and about 30 midrange ones (and today’s minor programs were the major programs 20 year ago). Now they are looking at significantly higher costs and a timeframe of a decade to get it all changed over.

Funny thing, I work for a company of engineers, run by engineers. This should have been an easy sell, but it involved fiefdoms, and it involved electrical engineers vs computer systems engineers and apparently, they could not get along.

 

[Shadow Reaper]

:| you don't agree with everything I say but I am a PhD in the room :|
Published, award winning ;_; why don't you love me shadow

It’s worse than you know. I’m a celtophile. I love the Irish. I’ve spent months in pubs from Limerick, to Cork to Dublin, and of all the Irish I know and love, it is only you I don’t like. I think it’s because you don’t shower enough.

But, I can like you tomorrow.

 

[Yex]

It’s worse than you know. I’m a celtophile. I love the Irish. I’ve spent months in pubs from Limerick, to Cork to Dublin, and of all the Irish I know and love, it is only you I don’t like. I think it’s because you don’t shower enough.

But, I can like you tomorrow.

I knew it you love me
I'll stand in the rain so I will wash all the muck aff

 

[Yex]

Been at my company 20 years. First 6 months I was there, I told them they should really make an effort to consolidate and mesh their databases so they work together. They should also make an effort to design holistic systems so that new systems will work with legacy systems and legacy systems can be upgraded/updated without causing issues.

NO…we don’t need to spend that effort. We are a small company and our data needs are not that complex. If we have duplicate data sets or incompatible databases, that’s okay. We need to ensure that everyone has the data they need for their program. Besides, we need to keep our costs down and those would be big projects. (Maybe 5 big programs and 3-4 smaller ones that would need integration into a unified system). It would also take 2 years…

Been pitching unified structures every 5 years or so as we keep growing…from 40 folks to SIGNIFICANTLY more.

Now suddenly they want to use AI to start mining and analyzing data…and are finding out that with all the separate programs and lack of a unified system design standard over the years, they can’t do it.

Example, we have one program that looks at certain types of outages,and another platform that looks at a different type of outage. Being able to look and cross reference outages from both systems is an obvious way to see how things might be related, but because program 1 was run by “Sally”, and program 2 was run by “Susan”, they chose different structures, different vendors, and they are completely incompatible with each other. Yet, neither wants to migrate to the other’s platform because they will lose status. This means an entirely new program that will incorporate both…. Most of the datasets they will use are the exact same data sets (duplicate work) with only about 10 percent of it being different. It would not have been hard to add the functionality of second system to the first when it first was developed, ut they refused to do it as it might give the other group access to “their” data.

They will need to spend money to combine 15 major programs and about 30 midrange ones (and today’s minor programs were the major programs 20 year ago). Now they are looking at significantly higher costs and a timeframe of a decade to get it all changed over.

Funny thing, I work for a company of engineers, run by engineers. This should have been an easy sell, but it involved fiefdoms, and it involved electrical engineers vs computer systems engineers and apparently, they could not get along.

Sounds like a bit of a mess, a potential clinical case of "this is how it's always been done".

MBA led companies can be chaos of NEWNEWNEW while engineering led can be a case of "if it ain't broke don't fix it".

Imho I'd go for young engineers who tend to be a good blend of both running a company. Only problem with that is they're too honest half the time, business needs a level of bullshit to get by - when you show up to a conference giving a talk and it's on a different topic they changed but didn't tell you.

Reach out to a data lake provider and talk them through, validate it with a poc in your off time and have their sales team convince management.
Datalakes are fucking magic, sales folks are good at selling. Building in house is expensive, so if a third party offering is fast and "free" (never free).. Win win

 

[Mudhawk]

Send help :)))))

Choose: Rum or Vodka in the emergency supplies?

Every time a new freaking shiney thing comes along and says it will save time and money

In my experience upper managements are like mangy cats.
They get easily distracted by all that glitters.
Still there's nothing cute about them.
And you do not want them sitting in your lap...

I have a nice smile or so I'm told.

They said the same about the Cheshire cat...

This should have been an easy sell, but it involved fiefdoms, and it involved electrical engineers vs computer systems engineers and apparently, they could not get along.

I feel you. We have turf wars between the guys that bolt security systems to the walls vs. the guys that programm the keys.
And they are in the same department.

and of all the Irish I know and love, it is only you I don’t like.

You know, from a logical standpoint you should review that sentence.
Unless you are able to not like someone you love.
Then again, you do not "just" like someone if you love them.
Soooo... maybe it works out..?

 

[Yex]

They said the same about the Cheshire cat...

Teeth, all the way to the back

Choose: Rum or Vodka in the emergency supplies?

Rum if they're out of whiskey.
House master used to give us all a 'wee dram' on a Saturday before sending us off to the socials. Probably not the way the do it anymore, but haven't stopped me liking a good sherry cask since.

 

[BUTUZ]

Bringing this thred back on topic -