2021 Mclaren F1 Team

[godlameroso]

On this second point in bold, notice that only a few races ago, Daniel was told thw opposite. He was told that he was braking too late! (I cannot remember if Portugal or Imola). So really it's not that the steering/braking are an unchangeable characteristics; they are changing track to track and Lando is adjusting to them by instinct. Whether it is through experience with the chassis or the way the car works the tires or better relationship with his engineer, Lando is adjusting better with the car.

I agree with you on this. Lando is instinctively, subconsciously and naturally able to make all those adjustments needed in his driving technique around various types of corners. As you say he is not going by a preset template. He is able to apply different methods/styles/techniques across the lap around different types of corners seamlessly.

Ricciardo will get there. As you know braking is highly complicated in these cars where the brakes are not just used to stop the car but to harvest energy as well with the aid of MGU-K. Then there is the engine braking. He is one driver who needs to completely understand every single trait of the braking system to get everything out of the car. I feel once he figures out the brakes he will be up there with Lando. Another point i feel is that he still does not know where the limit lies in slow corners and hence very tentative and ginger with turn in and throttle application. I feel he just needs time to put all the pieces of the puzzle together.

The more you do something emotionally involving the greater the fidelity of the map in your brain.

It's like building a tire model on the computer but it's really your brain. You give a grip value(s) for each bit of slip angle, and you build a statistical model in your mind for tire behavior. More data means more refined model, this model exists subconsciously. Thus through experience you can build a model of the tires in your brain and exploit more grip than someone with a less detailed model.

[PlatinumZealot]

These new simulators will capitalize on advances in machine learning. There a next generation of specialized machine learning processors to come out so they are likley monitoring those before they commit to the next few years.

[McFAN]

"MARK HUGHES: WHY F1 MAY HAVE ANOTHER SHOCK CONTENDER IN BAKU"

Mark Hughes is very bullish about McLaren's prospects for this weekend
Let's hope he's on to somthing

https://the-race.com/formula-1/mark-hug ... er-in-baku/

[Herr_Koos]

Carlando were (jokingly) already bullish after Monaco. The "rocketship" comment was only partly tongue-in-cheek, I feel.

Fingers crossed. Another podium would be brilliant, and hopefully Dan will take to this track better than Monaco.

[mclaren111]

RaceFans:

Baku is the biggest test of outright power unit performance so far this year. It also comes slightly later in the season than usual – most drivers will be on their sixth race with the same unit – so we could see some making early switches to fresh hardware to ensure optimal performance at this power-sensitive track.

Following their switch to the class-leading Mercedes power units this year, McLaren will be one to watch. After Sunday’s Monaco Grand Prix, Verstappen and Carlos Sainz Jnr talked up the performance of Lando Norris’ McLaren on the straights, calling it a “rocket”.

[BMMR61]

Mark Hughes is one of the better pundits, as much detail as Windsor but less fanciful. Most of his predictions tend to be close to the mark. Hoping for another podium, if I were a betting man I’d take a win and a place for Baku. Still will need things to run their way though. Both cars finishing in the top 5 or 6 would be nice. With the circuit being somewhat different from anything else it makes for an intriguing lead up to the weekend. The Barcelona aero updates should be well understood now so getting a good compromise for the tight sector will bring McLaren into play I hope.

[101FlyingDutchman]

Let’s not get ahead of ourselves. Happy with an improved showing from RIC and be in contention of best of the rest or be best of the rest

[nzjrs]

These new simulators will capitalize on advances in machine learning. There a next generation of specialized machine learning processors to come out so they are likley monitoring those before they commit to the next few years.

This is advertising copy. New ML accelerators and specialised processors come out continually - their performance usually tracks GPU performance (certainly withing much less than an order of magnitude) but the current/forever chip battleground is on power efficiency. We can kind of extrapolate the performance curve a couple of years without much imprecision.

The discussion about how ML / AI will impact this type of simulation and my opinion is not fully formed yet but my intuition and from my professional experience in a related ML field is "not as mich as most people think"

[PlatinumZealot]

Mark Hughes is one of the better pundits, as much detail as Windsor but less fanciful. Most of his predictions tend to be close to the mark. Hoping for another podium, if I were a betting man I’d take a win and a place for Baku. Still will need things to run their way though. Both cars finishing in the top 5 or 6 would be nice. With the circuit being somewhat different from anything else it makes for an intriguing lead up to the weekend. The Barcelona aero updates should be well understood now so getting a good compromise for the tight sector will bring McLaren into play I hope.

Mark Huges is hit and miss. Don't read his articles for technical information. He writes well so he lulls the reader into his hypotheses quite well, but he has been wrong many many times on the techincal front. I mostly read his articles because of his paddock stories.

[Chicane]

Mark Hughes is one of the better pundits, as much detail as Windsor but less fanciful. Most of his predictions tend to be close to the mark. Hoping for another podium, if I were a betting man I’d take a win and a place for Baku. Still will need things to run their way though. Both cars finishing in the top 5 or 6 would be nice. With the circuit being somewhat different from anything else it makes for an intriguing lead up to the weekend. The Barcelona aero updates should be well understood now so getting a good compromise for the tight sector will bring McLaren into play I hope.

Mark Huges is hit and miss. Don't read his articles for technical information. He writes well so he lulls the reader into his hypotheses quite well, but he has been wrong many many times on the techincal front.

Spot on mate.

Too many sophisticated words with little technical substance.

[godlameroso]

These new simulators will capitalize on advances in machine learning. There a next generation of specialized machine learning processors to come out so they are likley monitoring those before they commit to the next few years.

We are at a point where biometric information can be used as data. If you measure brainwaves, galvanic skin responses, heart beat, it can be converted into an algorithm that translates into how much the driver is focusing, you can see which parallels exist in terms of steering or pedal responses in both the simulator and track to make more accurate tire models.

[godlameroso]

These new simulators will capitalize on advances in machine learning. There a next generation of specialized machine learning processors to come out so they are likley monitoring those before they commit to the next few years.

This is advertising copy. New ML accelerators and specialised processors come out continually - their performance usually tracks GPU performance (certainly withing much less than an order of magnitude) but the current/forever chip battleground is on power efficiency. We can kind of extrapolate the performance curve a couple of years without much imprecision.

The discussion about how ML / AI will impact this type of simulation and my opinion is not fully formed yet but my intuition and from my professional experience in a related ML field is "not as mich as most people think"

It's starting to cost some serious hydrocarbons to run all the data we take for granted.

[Big Tea]

These new simulators will capitalize on advances in machine learning. There a next generation of specialized machine learning processors to come out so they are likley monitoring those before they commit to the next few years.

This is advertising copy. New ML accelerators and specialised processors come out continually - their performance usually tracks GPU performance (certainly withing much less than an order of magnitude) but the current/forever chip battleground is on power efficiency. We can kind of extrapolate the performance curve a couple of years without much imprecision.

The discussion about how ML / AI will impact this type of simulation and my opinion is not fully formed yet but my intuition and from my professional experience in a related ML field is "not as mich as most people think"

It's starting to cost some serious hydrocarbons to run all the data we take for granted.

And to cool the machines. Data processing is a huge heat source

[djos]

Modern data Centers are very efficient and use technologies like free-air cooling for much of the year instead of traditional refrigeration technologies. Those are kept in reserve for days that are too hot to use outside air.

Gone are the days of freezing your nuts off at 21c while working in a DC. Even the older DC’s more mostly run at 23c to reduce energy consumption.

[godlameroso]

This is advertising copy. New ML accelerators and specialised processors come out continually - their performance usually tracks GPU performance (certainly withing much less than an order of magnitude) but the current/forever chip battleground is on power efficiency. We can kind of extrapolate the performance curve a couple of years without much imprecision.

The discussion about how ML / AI will impact this type of simulation and my opinion is not fully formed yet but my intuition and from my professional experience in a related ML field is "not as mich as most people think"

It's starting to cost some serious hydrocarbons to run all the data we take for granted.

And to cool the machines. Data processing is a huge heat source

Big tech should look at itself, the chip manufacturers are bending over backwards to make more energy efficient chips, why isn't big tech taking a page from F1 and using energy recovery from their heat output?