NVIDIA's A Beast...And They're Only Getting Bigger, Better and FASTER!
The exponential rise in computing power is off the charts and is only going to increase by leaps and bounds.
So, I’m kind of a tech geek. I love all things technology, especially in the field of computers. The advent of classical computing has allotted us many privileges and capabilities on multiple levels, from industrial to commercial to personal use. I usually try my best to keep up with some of the keynotes that tech companies present. One of (if not the biggest) would be NVIDIA. I’d have to say that they are the leaders in chips (both CPU and GPU’s) that are used in many of the servers around the world.
Another area that NVIDIA seems to have the market cornered is in the filed of A.I. when it comes to processing speed and power. NVIDIA’s H100 A.I. LLM super chip was released in 2023. It was 4X’s faster than its predecessor, the A100 chip. The H100 catapulted NVIDIA to a multi-trillion dollar company…would have been nice to have purchased a healthy serving of this company in its infancy, right?
Well, as if that wasn’t huge, NVIDIA just released their new GPU chip called the Blackwell B200 GPU and GB200 “superchip.” It boasts a whopping 208 Billion transistors (that’s INSANE!) and offers up to 20 petaflops of FP4 horsepower from its 208 billion transistors.
Let’s just say that NVIDIA’s new advancement in GPU and CPU construction is a massive game changer in speed and for LLM’s (large language models)…A.I.
Here’s the article for all to read. Little known fact that what powers CHAT GPT is none other than NVIDIA hardware.
Though this was not geared towards gamers, NVIDIA did say that most likely these chips or similar ones with massive processing power will be released for gaming systems as well…again, game changer on multiple levels.
Let me see if I can put this into perspective as the article does a great job of doing. It says “In total, Nvidia says one of these racks can support a 27-trillion parameter model. GPT-4 is rumored to be around a 1.7-trillion parameter model.” Did you catch that? CHAT-GPT4 is said to be around a 1.7 trillion parameter model…these new racks from NVIDIA already have a 27-trillion parameter model. That’s a 15.9X’s increase in less than a year and a half.
Folks, the rate at which processing power is increasing is exponential to say the least. This is only increasing the speed of which deception will flow (no pun intended). Technology is a good thing, but did you ever ask yourself if it’s really all in the name of “efficiency”?
I believe that we are witnessing the system that will already be in place once the body of Christ is taken out of this world in the rapture.
I will be talking about technology on a very consistent and regular basis in the coming weeks and months, and should the Lord tarry, years to come…though I don’t believe that to be the case.
Stay tuned. Especially for a “wedding” that most are unaware is coming. No, I’m not referring to that of Jesus and His Bride…but, rather to the marriage of Hardware and Software.
Don’t lose heart. Stay informed. And, most importantly, in the Word of God!
Pablo.
ARTICLE: NVIDIA Reveals Blackwell B200
Thanks for sharing this article. I'm looking forward to all your upcoming messages on technology!