Re:An Interview with Nvidia CEO Jensen Huang About AI’s iPho...
by Dalewyn on Sunday March 26, 2023 @02:43AM
(#35311734)
Apple's business is squarely with the frontend of computing: The desktops, laptops, phones, tablets, and even watches that your mom and pop, artists, designers, and engineers use in daily life.
I don't see what Apple stands to gain from getting into the enterprise market, other than simple diversification of their portfolio.
Re:An Interview with Nvidia CEO Jensen Huang About AI’s iPho...
by lwhi on Sunday March 26, 2023 @04:11AM
(#35312252)
>I don't see what Apple stands to gain from getting into the enterprise market, other than simple diversification of their portfolio.
They'd stand to win a market they have no foothold in and profit.
Re:An Interview with Nvidia CEO Jensen Huang About AI’s iPho...
by Dalewyn on Sunday March 26, 2023 @05:29AM
(#35312637)
Like I said, I can see diversification of portfolio as something to gain.
On the other hand, Apple's non-existence in the enterprise space isn't due to a lack of trying. They've been there and done that already.
Re:An Interview with Nvidia CEO Jensen Huang About AI’s iPho...
by lwhi on Sunday March 26, 2023 @12:45PM
(#35316507)
Do you mean tried and failed?
Re:An Interview with Nvidia CEO Jensen Huang About AI’s iPho...
by smoldesu on Sunday March 26, 2023 @03:37PM
(#35318647)
Apple explicitly failed in the enterprise space because they didn't care. They built a very well-decorated walled-garden, but it's not what enterprise customers wanted. Apple wanted to sell UNIX, the customers wanted to buy servers. When the dust settled, Apple made no attempt to respond to customer demands. They smothered the product with a pillow and told their enterprise partners to pound sand or buy a Trash Can Mac.
...but, if core size is your jam (for whatever reason), keep an eye on Nvidia's Grace CPU. It's their stab at a datacenter-scale ARM SOC, and it should be releasing before EOY. Then there's the Ampere offerings that already have acceleration for PyTorch, ONNX and Tensorflow, along with Graviton for general-purpose efficiency... there's a lot of low-profile ARM cores in the datacenter today.
A good start for Apple would be updating the rackmount Mac Pro with an 80 core Double Ultra chip, but even that feels fairly pedestrian next to the 144-core-complex Grace is teasing. I'm sure it sounds silly to the readers of this website, but I genuinely don't think Apple is up to the task of competing in the datacenter. Obviously so on the software side, but arguably not even on the hardware front either.