NVIDIA’s GPU Technology Conference (GTC) is a renowned event where the brightest minds in artificial intelligence, accelerated computing, and countless other fields converge.
GTC 2024 promises to be a landmark edition, showcasing what’s possible with AI and showcasing groundbreaking innovations set to revolutionize various industries.
Key information
- Dates: March 18-21, 2024
- Location: San Jose Convention Center, San Jose, California (with virtual components)
- Website: NVIDIA GTC
What to expect from NVIDIA GTC 2024?
Get ready for a front-row seat to the future!
NVIDIA CEO Jensen Huang and other industry titans will deliver keynote addresses unveiling cutting-edge technologies, awe-inspiring advancements, and bold predictions for the future of AI.
Here is what we are expecting to see at NVIDIA GTC 2024:
More SKUs, GDDR7, and generative AI
Critics question whether NVIDIA’s roadmap focuses too much on frequent iterations rather than incremental value.
This is reminiscent of the AMD vs. Intel CPU market – choice abounds, fueling competition and continuous innovation.
It’s a similar story for AI. Model sizes are exploding, and the H100/H200 reigns supreme today. But with models doubling every few months and the human brain as the ultimate benchmark, we can expect a continued push for more powerful GPUs.
GDDR7, the new memory standard, could be a game-changer. While the L40S is a solid generative AI inference option, its GDDR6 memory could prove a bottleneck.
Expect NVIDIA to capitalize on GDDR7, perhaps with a new Lovelace GPU offering double the capacity and power for large LLM inference.
Perhaps a mobile GPU push?
The NVIDIA-MediaTek partnership was last year’s surprise. While initially focused on automotive, there’s potential for NVIDIA to enhance MediaTek’s mobile GPU capabilities. A move like this could signal a more aggressive play for the mobile market, potentially utilizing chiplet technology.
Inference & Edge in the spotlight
Last quarter, NVIDIA reported data center inference revenue hitting 40%. It’s likely they’re gunning for over 60% by late 2024. To get there, a more affordable platform is key, and an updated Lovelace GPU with GDDR7 could deliver that needed balance.
Edge AI, while taking a temporary backseat to the H100-driven data center surge, remains critical. The Jetson platform is still based on Ampere – an upgrade to Hopper or even Lovelace for a more price-sensitive edge market seems likely.
Automotive and robotics announcements should be a major focus.
Software-driven growth
NVIDIA hit a $1 billion software run rate, encompassing Enterprise AI and Omniverse. This software is a goldmine – sticky, valuable, and powerful way to monetize their massive GPU installed base.
Expect NVIDIA to double down on software growth – they’re starting at a strong base, and opportunities abound.
Don’t miss GTC 2024
While this is speculation, GTC 2024 is sure to reveal the specifics of NVIDIA’s roadmap to address these points. How much will it cost?
See the GTC 2024’s detailed pricing information below:
| Option | In-Person (4-Day Pass) | Virtual (4-Day Pass) | In-Person (Thursday, March 21 Only) | Virtual (Thursday, March 21 Only) |
| Conference + Training | $2,495 | $1,245 | None | None |
| Conference Only | $2,095 | $1,045 | None | None |
| Exhibits + Keynote | $745 | $75 | None | None |
| DLI Workshops (Sunday, March 17 Only) | $495 | $495 | $149 | $149 |
Keynotes, in-depth sessions, and the bustling exhibit hall will provide insight into how NVIDIA aims to solidify its AI leadership.
Stay tuned for what Jensen Huang and team have in store by following the GTC Session Catalog!
Featured image credit: NVIDIA.
