NVDA: IBM Touts Partnership at conference


Big Blue Touts Partnership with NVIDIA at IBM Think

Emphasizing the key role GPUs will play as businesses race to turn raw data into intelligence, IBM invited NVIDIA CEO Jensen Huang to appear on stage Tuesday at its massive Think conference, in Las Vegas.

Using GPUs, IBM is able to use fewer machines to harness data 46x faster on a key benchmark, John Kelly, IBM’s senior vice president of cognitive solutions and IBM Research, said at the gathering of more than 40,000 technologists, consultants, developers and business leaders…

…“I believe every company in the world will have high-performance computers or supercomputers inside,” Huang said. “Instead of engineers doing coding, you will have supercomputers learning, and whoever learns the fastest will get to market soonest.

Companies that don’t use AI will be outcompeted and left behind. NVDA is the arms dealer for AI.

“I cannot think of a single industry that is not only going to be not only impacted, but completely transformed,” Kelly said.



<<<Using GPUs, IBM is able to use fewer machines to harness data 46x faster on a key benchmark,>>>

This is consistent with what Nvidia often says, “the more GPUs a customer buys the more money the customer saves.”

Meaning of course that GPUs are much more efficient computing model than CPUs and thereby you need fewer computers and ergo spending less money to get greater results.

Another example of this at work.


“The team used a terabyte-scale dataset to predict clicks on advertisements in 91.5 seconds, compared to the previous record of 70 minutes, set in February of last year.”

That’s not just beating the previous record, that is utterly destroying it. In just about a year. That kind of improvement should be very valuable. Nvidia is cleaerly still in the disruption phase of its business cycle.



I have now listened to Jensen Hwang’s presentation at the IBM conference. One thing that he said stuck out for me. He said that NVDA is using its own technology inside the company. I had speculated about this in some posts last year. The first clue was how fast then report earnings results after the end of each quarter’s close. The confirmation came during the last earnings call when the CFO said that NVDA uses its own technology inside the company. So what did CEO Hwang say the other day?

He said that NVDA has about 700 supercomputers that it is using inside the company. These computers has some of the most powerful supercomputers in the world. He said that we expects all companies to use supercomputers inside their organizations. He said that companies that use this technology can get to market faster. If you get to market faster then you can outcompete the competition and you can stay ahead of the competition. NVDA is clearly staying ahead of AMD. It has for years. NVDA iterates its products very fast. Volta came out last year. When will the next, more powerful version come out? What will NVDA announce at GTC next week. Software is writing software. NVDA can design and develop new products with the aid of their AI supercomputers. How many SW engineers do these 700 computers “add” to NVDA’s ranks. NVDA has about 11,000 - 12,000 employees. Intel has more than 100,000. NVDA is faster, more nimble, and more efficient. NVDA is making the world’s best brains in a box, and it is not just selling these brains but using them to make their company stronger and more competitive. NVDA is definitely the smartest company in the world. I think in this case the smartest company will get more of the marbles. Great stuff and I’m really looking forward to GTC next week.



Software is writing software. NVDA can design and develop new products with the aid of their AI supercomputers.

Nvidia has 6 letters, Skynet has 6 letters…coincidence? I think not!