By advantage of its relentless pursuit of ever quicker, ever extra highly effective GPUs, Jen-Hsun Huang has claimed that Nvidia, over the previous 20 years, has pushed the “price of computing down by a million instances”.
Once you take a look at the rising prices of contemporary graphics playing cards in contrast with their forebears, that is perhaps laborious to fathom. It positive seems to be like the price of a GPU has simply been steadily rising to most of us after we take a look at the objects of our silicon needs. However whenever you take a look at simply what the graphics chips of right this moment are able to, the extent of uncooked computational energy on the disposal of even a lowly RTX 4060 would have appeared borderline legendary 20 years again.
A GeForce 6800 Extremely from 2005 delivered a whopping 6.4 GFLOPS, whereas the underside of the Ada Lovelace era comes with 15,100 GFLOPS of processing grunt. That is a complete world of distinction from a $499 card of 20 years in the past versus a $299 GPU of right this moment.
And that is not even a card anyplace close to the highest of the stack, nor near what you may get from Nvidia’s strongest enterprise GPUs.
Jen-Hsun, at right this moment’s morning-after-keynote Q&A session, in contrast what Nvidia has carried out in creating extra highly effective graphics silicon, pushing down the relative value of GPU computational energy, to the impression of Moore’s Legislation.
“The rationale Moore’s Legislation was so vital within the historical past of the chip is that it drove down computing prices,” Huang remarks. “In the middle of the final 20 years we have pushed the marginal price of computing down by a million instances.
“A lot that machine studying turned logical: ‘simply have the pc go determine it out.'”
Principally, there’s a lot computational energy accessible for such comparatively little money that you simply may as properly simply begin throwing it at AI to unravel all our issues. Or, , draw us an image of a gold fish whenever you completely, positively simply have to have a freshly generated image of a fish.
There is no getting away from it, the graphics card, and its constituent element, the GPU, have turn into an important items of silicon in our fashionable time. There’s additionally no getting away from the truth that Nvidia is accountable for a few of the most vital silicon of our time, nonetheless you are feeling in regards to the rise and rise of synthetic intelligence and its potential impression on the world and humanity.
Does Jen-Hsun’s maths add up? I do not know, he did not present his workings, however what’s true is that for the reason that start of the GPU as we all know it, the cost-to-performance ratio has solely been going in a single path.
