Thursday, May 16th 2024
NVIDIA Blackwell GB200 Superchip to Cost up to 70,000 US Dollars
According to analysts at HSBC, NVIDIA's upcoming Blackwell GPUs for AI workloads are expected to carry premium pricing significantly higher than the company's current Hopper-based processors. The analysts estimate that NVIDIA's "entry-level" Blackwell GPU, the B100, will have an average selling price between $30,000 and $35,000 per chip. That's already on par with the flagship H100 GPU from the previous Hopper generation. But the real premium lies with the top-end GB200 "superchip" that combines a Grace CPU with two enhanced B200 GPUs. HSBC analysts peg pricing for this monster chip at a staggering $60,000 to $70,000 per unit. NVIDIA may opt to primarily sell complete servers powered by Blackwell rather than individual chips. The estimates suggest a fully-loaded GB200 NVL72 server with 72 GB200 Superchips could fetch around $3 million.
The sky-high pricing continues NVIDIA's aggressive strategy of charging a premium for its leading AI and accelerator hardware. With rivals like AMD and Intel still lagging in this space, NVIDIA can essentially name its price for now. The premium pricing reflects the massive performance uplift promised by Blackwell. A single GB200 Superchip is rated for five PetaFLOPs at TF32 of AI compute power with sparsity, a 5x increase over the H100's one PetaFLOP. Of course, actual street pricing will depend on volume and negotiating power. Hyperscalers like Amazon and Microsoft may secure significant discounts, while smaller players could pay even more than these eye-watering analyst projections. NVIDIA is betting that the industry's insatiable demand for more AI compute power will make these premium price tags palatable, at least for a while. But it's also raising the stakes for competitors to catch up quickly before losing too much ground.
Source:
Tae Kim (on X)
The sky-high pricing continues NVIDIA's aggressive strategy of charging a premium for its leading AI and accelerator hardware. With rivals like AMD and Intel still lagging in this space, NVIDIA can essentially name its price for now. The premium pricing reflects the massive performance uplift promised by Blackwell. A single GB200 Superchip is rated for five PetaFLOPs at TF32 of AI compute power with sparsity, a 5x increase over the H100's one PetaFLOP. Of course, actual street pricing will depend on volume and negotiating power. Hyperscalers like Amazon and Microsoft may secure significant discounts, while smaller players could pay even more than these eye-watering analyst projections. NVIDIA is betting that the industry's insatiable demand for more AI compute power will make these premium price tags palatable, at least for a while. But it's also raising the stakes for competitors to catch up quickly before losing too much ground.
30 Comments on NVIDIA Blackwell GB200 Superchip to Cost up to 70,000 US Dollars
If it serves a purpose it will add jobs. Theoretically yes. And yet, people still have jobs. Unemployment can always be brought down, especially and even in high income sectors. You can indeed cut jobs, and then you'll find them growing on another end of the business again, and you're screwed if you ignore that. For example, look at everything fintech. Now we haven't got banking around the corner IRL anymore, but we have towers full of idiots watching money go across the globe. Internet banking, sure, saved a lot of jobs, and then required a slew of new ones to keep it secure and up to date. And on top of internet banking spawned a whole new world of fintech apps, plus crypto. The trust in the digital transaction made that all possible (which in the case of crypto is unbelievable irony).en.wikipedia.org/wiki/Dot-com_bubble
Mhm.
Or are you saying AI is bigger than this ;)
"can be seen as similar to a number of other technology-inspired booms of the past, including railroads in the 1840s, automobiles in the early 20th century, radio in the 1920s, television in the 1940s, transistor electronics in the 1950s, computer time-sharing in the 1960s, and home computers and biotechnology in the 1980s"
Gosh its almost like every economical uptake or big invention, where there is a lot of free money, has its own little bubble doesn't it...
Like I said, most of what the AI does today has relatively low usefulness. But that will change in the future.
That's why we're seeing the consumer breed of AI. The checking is done by ourselves, or by artists making their 'art'. Anywhere else, its just automation.
What's far more likely to me is that (and we already see this) AI will add new jobs because new possibilities are opened. You don't throw away your old bathwater that easily. Once those new possibilities have become predictable enough to be acting upon, you've really just automated something.