Hacker News Reader

Comment by ShakataGaNai | Hacker News Reader

I would wager that's because there isn't a lot of existing silicon that fits the bill. What COTS equipment is there that has all the CPU/Tensor horsepower these systems need... AND is reasonably power efficient AND is rated for a vehicle (wild temp extremes like -20F to 150F+, constant vibration, slams and impacts... and will keep working for 15 years).

Yea, Tesla has some. But they aren't sharing their secret sauce. You can't just throw a desktop computer in a car and expect it to survive for the duration. Ford et all aren't anywhere close to having "premium silicon".

So you're only option right now is to build your own. And hope maybe that you can sell/license your designs to others later and make bucks.

View on HN

NVIDIA orin series is the big one for tensor horsepower. Horizon robotics and Qualcomm also have competitive automotive packages.

They are all expensive, but less than the risk adjusted cost of developing a chip.

Isn't that risk balanced by a healthy reward of controlling their verticals and possible secret sauce?

And their chips give "1600 sparse INT8 TOPS" vs the Orin's "more than 1,000 INT8 TOPS" -- so comparable enough? And going forward they can tailor it to exactly what they want?

Mostly it costs hundreds of millions to develop a chip; it relies on volume to recover the cost.

NVIDIA also tailor their chips to customers. It's a more scalable platform than their marketing hints at... Not to mention that they also iterate fairly quickly.

So far anyway, being on a specialised architecture is a disadvantage; it's much easier to use the advances that come from research and competitors. Unless you really think that you are ahead of the completion, and can sell some fairly inflexible solution for a while.

Orin is Nvidia's last generation. Current gen is Thor at 1k TOPS. Rivian's announcement specifies TOPS at the module level. The actual chip is more like 800 and probably doubled. Throw two Thors on a similar board and you're looking at 2000 sparse int8 TOPS.

I've been involved with similar efforts on both sides before. Making your own hardware is not a clear cut win even if you hit timelines and performance. I wish them luck (not least because I might need a new job someday), but this is an incredibly difficult space.

Having to work with Qualcomm is enough reason to not buy Qualcomm

Reply on news.ycombinator.com