News Nvidia's Jensen Huang Shares Unusual Management Style: Make No Plans

Status
Not open for further replies.
Always be wary of over-interpreting N=1 case studies. You can't tell if they succeeded because of what's being highlighted, or in spite of it. Nor, how certain things they did were counterbalanced or facilitated by other things.

Interviews are particularly fraught, due to a natural tendency towards self-serving narrative:
"Huang believes in walking away from businesses or sectors that have become commoditized, which is why Nvidia left the smartphone and tablet SoC market a few years ago."​
Or, perhaps because they just didn't do a very good job of competing, there. Their first 5+ years of trying to play in that market were a comedy of errors, during which many bridges were burnt. Towards the end, they sunk way too many resources in designing their own CPU cores, which I'm not sure ever delivered competitive performance. The only thing they really had going for them were their iGPUs.

However, based on some things they said around the time of their Maxwell and Pascal launches, it sounds as though their foray into mobile was very healthy for their dGPU business, since it really forced them to focus on energy efficiency in a way they hadn't previously done. That, in turn, enabled them to perform much better at scale, because they weren't as power- or thermally-limited. It might've even been behind their transition to tiled-rendering. Their SoC business also set the stage for their move into robotics and self-driving.

Something else that irked me:
"Another Huang deviates from established practices is by trusting intuition"​

Yeah, just ask their chip designers how much they rely on intuition. I don't buy it. I don't know how much of this is misdirection vs. Huang actually not knowing how their architects and design engineers do their jobs, but I can pretty much guarantee you they model everything and rely on data for the multitude of their decisions.

What I'd say stands out for me is their ability to learn and adapt. That, and the foresight behind pushing CUDA into educational institutions and investing big in AI were the real reasons for their current success. Not the management style, IMO.
 
Last edited:
Always be wary of over-interpreting N=1 case studies. You can't tell if they succeeded because of what's being highlighted, or in spite of it. Nor, how certain things they did were counterbalanced or facilitated by other things.
In this N=1 case study, nVIDIA, for better or worse, is the leader of the GPU / AI market

Or, perhaps because they just didn't do a very good job of competing, there. Their first 5+ years of trying to play in that market were a comedy of errors, during which many bridges were burnt. Towards the end, they sunk way too many resources in designing their own CPU cores, which I'm not sure ever delivered competitive performance. The only thing they really had going for them were their iGPUs.
The many bridges burnt has been a consistent thing with nVIDIA. Many companies and people within the Tech industry aren't fond of nVIDIA's behavior. The console market has a tendency to reflect this. Many companies are (One and Done) when dealing with nVIDIA.

- After Bump-Gate, Apple refuses to do business with nVIDIA.
- Sony did business with nVIDIA after their own internal GPU's failed and they needed one, so nVIDIA helped them out on the PS3, after that, Sony never went back.
- MS did business with nVIDIA on the original Xbox, after that, they were swapping to AMD afterwards with only 1x dabble in POWER for the CPU, and the rest using x86 from AMD afterwards.
- Nintendo is on it's current phase with nVIDIA, we'll see how long that relationship lasts.

Yeah, just ask their chip designers how much they rely on intuition. I don't buy it. I don't know how much of this is misdirection vs. Huang actually not knowing how their architects and design engineers do their jobs, but I can pretty much guarantee you they model everything and rely on data for the multitude of their decisions.
I don't expect Jensen Huang himself to know all the details that his numerous engineers are doing. He has to trust his employees to some degree and trust that they're using good data for their decision making.

What I'd say stands out for me is their ability to learn and adapt. That, and the foresight behind pushing CUDA into educational institutions and investing big in AI were the real reasons for their current success. Not the management style, IMO.
Actually, his management style resembles the OODA Loop IMO.
Whether or not he got it from John Boyd or came to it on his own is unknown.

Jensen Huang has stated in his interview that "He wanted a company that was smaller, not larger; you want a company as small as possible; not as large as possible; it needs to be as large as necessary to do the job well; but to be as small as possible"
RHVpREK.png
That's ALOT of head-count for nVIDIA just for Graphics & AI
 
Last edited:
  • Like
Reactions: artk2219
Don't forget that they bought a large player in datacenter networking and they build their own DGX systems. I'm sure a chunk of those people also work in robotics & self-driving.
But all those other divisions are small potatoes compared to their core 2-businesses.

DataCenter & Gaming.

Everything else falls by the wayside in comparison to the big 2 drivers.
P24slEX.png
 
  • Like
Reactions: artk2219
But all those other divisions are small potatoes compared to their core 2-businesses.

DataCenter & Gaming.

Everything else falls by the wayside in comparison to the big 2 drivers.
P24slEX.png
The core AI investment feeds all 3 businesses, really. The way people might be spread across the organization probably better reflects the future value of those businesses, more than the current revenue.
 
  • Like
Reactions: artk2219
From the Jensen Huang Interview:
Jensen Huang:
We never talk about marketshare in our company, and the reason for that is because:
why are you talking about I have 23% of MarketShare, and they have 27% MarketShare.
Why are you fighting people for MarketShare?
Because the whole concept of MarketShare says that there is a whole bunch of other people who are doing the same thing.
If they are doing the same thing, why are we doing it?
Why am I squandering the lives of these incredibly talented people to go do something that has been done.
Unless we enjoy the competition, which I tend not to.
So we tend not to go fight people for marketshare, fight people over markets that are comoditized.

Talk about speaking from a privileged position where you are the #1 player in the GPU market for gaming and DataCenter.

If his company wasn't in that position, and actually had to compete, he'd probably have a very different tone/attitude about it.

Not surprising that he doesn't like competition, he's been sitting on the top of the mountain for too long.

The core AI investment feeds all 3 businesses, really. The way people might be spread across the organization probably better reflects the future value of those businesses, more than the current revenue.
Given how Jensen has stated that he doesn't want to do "Commoditized Work" and is willing to leave sectors of business if it becomes "Too Commoditized", I wonder how much attachment he has to each of those other sectors.

There are plenty of people/competitors in:
- DataCenter & Networking
- Robotics
- Self Driving.

Jensen himself stated that he's more than happy to abandon those industries if there are too many players in there and they're doing "Commoditized Work".

So if AI becomes to "Commoditized", what will he do?

It's not like there are shortages of AI companies right now, there are literally hundreds around the world.

Plenty with their own silicon & software stack.
 
Last edited:
  • Like
Reactions: artk2219
Given how Jensen has stated that he doesn't want to do "Commoditized Work" and is willing to leave sectors of business if it becomes "Too Commoditized", I wonder how much attachment he has to each of those other sectors.

There are plenty of people/competitors in:
- DataCenter & Networking
- Robotics
- Self Driving.
Those really don't fit the definition of commodities. A commodity isn't defined by how many people are competing in the market, but rather by being something where you're basically reduced to competing on price, because the products are otherwise so indistinguishable and interchangeable. That means opportunities to add value are also extremely limited. I get why he doesn't want to be in a low-margin business with few opportunities for differentiation. Totally makes sense.

Noun​

commodity (countable and uncountable, plural commodities)

  1. ...
  2. ...
  3. ...
  4. (marketing) Undifferentiated goods characterized by a low profit margin, as distinguished from branded products.
    Although they were once in the forefront of consumer electronics, the calculators have become a mere commodity.
  5. ...
Source: https://en.wiktionary.org/wiki/commodity

So if AI becomes to "Commoditized", what will he do?
If the hardware gets commoditized (which doesn't seem likely to happen any time soon), I guess they'd just follow the value chain. They're investing quite a lot in the AI application domain, like self-driving and robotics, but also lots of other things.

It's not like there are shortages of AI companies right now, there are literally hundreds around the world.
Making hardware? None at Nvidia's level. Some are successful in their own niche, but most are just wannabees. The AI hardware industry will likely consolidate a lot, over the next couple years. Investors don't have unlimited patience, and if you can't make a go of it in the current climate, then there's probably not much hope for you.
 
Well, I could have told you they are incredibly shortsighted and run towards the next moneymaker without too much care about what used to be their core business.
Just look at the gaming GPU market, where the only "sensible" purchases nowadays are in the "I don't know where the money I spend comes from" price range.
 
I think Elon Musk is pretty much the same, maybe he took a page or two from Huang, or maybe it's just how modern management should be
Eh, I've heard that people at Tesla and SpaceX have found it very productive to have Elon being distracted by Twitter, over the past year. That's not a very strong endorsement of Elon's management style... nor is Twitter's financial performance.
 
  • Like
Reactions: TJ Hooker
I think Elon Musk is pretty much the same, maybe he took a page or two from Huang, or maybe it's just how modern management should be...
Hi Keng…I worked for Mubadala in Abu Dhabi and spend 14-years as an expat in the UAE including nearby Dubai! Mubadala is owned by Sheikh Mohamed bin Zayed, the current ruler of Abu Dhabi and President of the United Arab Emirates and chairman of the company. Sheik Mohamed also owns among other things the ‘Global Foundries’ with 14,000 employee’s now manufacturing chips for approximately 7-percent of the semiconductor industry. He would frequently visit our (his) major Abu Dhabi real estate properties including the US$1.6bn Cleveland Clinic located on Al Maryah Island. I gotten much unrestricted access to talking with him directly with hands-on discussions and onsite conditions compared to when I worked in NYC and Boston as a corporate SVP, where I would spend approximately 85% plus of my time writing reports and memos . Totally contrary while on the ground in the UAE, Kuala Lumpur and even Singapore. In those days writing no more than a few hundred memos over all during my many expat years. Face to face meetings, walks and relationships always maintained to be a big deal. Actually the only deal! That is also a big reason that specifically U.S executives once having worked overseas for any length have a exceedingly difficult time in ever returning to the island of the U.S and its way of its business life and how they are regarded. (Picture on the job on Al Maryah Island, Abu Dhabi.

a27d24_e3ddbac2da244b6cad369e2314d4f7fe~mv2.jpg
 
Last edited:
I find reading these comments more interesting and funny because most of them show lack of basic logical understanding of why Jensen said what he said here.
People with fancy degrees understand very little about what they know in actual practice. Successful people usually don't follow a rule book. Basic rules yes. But anyone in daily life has those.
Intuition certainly comes into play when you don't have a rigid system. Intuition plays off of wisdom gathered over time. Sure, other information is factored in but making the final decision sometimes comes down "how do we feel about doing this?".

Jensen, though had a vision a long time ago about what NVIDIA is tapping into and leading in now. Nothing comes out of a vacuum. Jensen never claimed that in this interview.

Seems many here just want to criticize and find fault. And most times those people either are jealous or never look in the mirror and see that big piece of wood in their own eye.

NVIDIA right now is winning. Live with it. Learn from it.
 
  • Like
Reactions: cyrusfox
People with fancy degrees understand very little about what they know in actual practice. Successful people usually don't follow a rule book. Basic rules yes. But anyone in daily life has those.
Intuition certainly comes into play when you don't have a rigid system. Intuition plays off of wisdom gathered over time. Sure, other information is factored in but making the final decision sometimes comes down "how do we feel about doing this?".
Do you have a source on that, or are you just going by this vaunted intuition?

The problem with intuition is that it exists to solve a certain set of problems that humans & other animals face, in the wild. There are well-known pitfalls to being over-reliant on intuition or a "gut feel". Scientists, engineers, accountants, and others can't simply rely on intuition. You probably don't want to ignore it, either, but it's not sufficient.

You need to choose the right action for the right reasons, not simply based on prior experience, because new situations won't always align perfectly with previous ones and you don't necessarily take the right lessons from every prior situation.

Jensen, though had a vision a long time ago about what NVIDIA is tapping into and leading in now. Nothing comes out of a vacuum. Jensen never claimed that in this interview.
Sure, he had visions of what might be possible and pursued some of those. However, it's a mistake to look backwards and assume the only path to the present was to repeat all the decisions of the past, or that different decisions wouldn't have yielded an even better outcome or earlier success. Nvidia was flirting with disaster, about 10 to 15 years ago. They made some costly mistakes and misjudgements. For a time, it looked like the company might be circling the drain.

Jensen is far from perfect and so is his stated methodology. Without more examples and an impartial observer watching how things at Nvidia actually work and how those same approaches worked in other companies, you can't know how accurate his version is and which aspects played into their success or were things that they succeeded in spite of. And, of the things which benefited them, was that more due to exceptional circumstances, or would they reliably produce the right outcome?

I think you can probably look at every wildly successful business leader and show how someone who behaved similarly either failed or met with middling performance, under a different set of circumstances. We like to believe in the myth that you can succeed if you just work hard and have the right methodology, but a lot of it comes down to being in the right place at the right time and dumb luck. I don't mean to suggest Nvidia just lucked into its success, as there's clearly been a tremendous amount of work and brainpower behind it. But, there's always an element of luck, and I happen to think luck might've played an outsized role in Jensen's success.

Seems many here just want to criticize and find fault. And most times those people either are jealous or never look in the mirror and see that big piece of wood in their own eye.
I'll admit that I listen to him speak and he just doesn't sound very smart, to me. That's not to say he's not shrewd and perhaps he indeed has a relatively good intuition. Maybe relying on it has truly worked out, for him. It's just not a very reliable strategy, IMO.

NVIDIA right now is winning. Live with it. Learn from it.
It's funny that you assume anyone's assessment of Jensen has anything to do with Nvidia. I own Nvidia products. I know people who work there. I'm glad for their success.
 
Last edited:
Status
Not open for further replies.