Go ask analysts, investors, and tech bros what kind of company NVIDIA is and they’ll immediately shoot back with “AI”. It’s definitely true that AI has fuelled the company’s meteoric rise to become the largest company in the world by market cap in recent years. However, NVIDIA’s expertise is in creating massive parallel processors, and it got its start in gaming. If it weren’t for PC gaming, there wouldn’t be an NVIDIA today.
Most of HotHardware‘s audience probably doesn’t need to be told that, as you, like us, were probably there twenty-five years ago for the launch of the original GeForce card. That’s right: it’s been twenty-five years since NVIDIA released the GeForce256 SDR version, to be followed later by the GeForce256 with DDR SDRAM onboard. The GeForce256 was the chip that prompted NVIDIA to coin the term “GPU,” as it was one of the first graphics processors that was anything more than just a dumb rasterizer.
While earlier GPUs had included some measure of additional processing capability onboard, it was the GeForce256 that brought hardware transform & lighting capability to consumer GPUs. This allowed the graphics card to take some of the load of geometry setup and vertex lighting calculations off of the CPU—a critical step forward when everyone was still using single-core processors, and mostly at sub-GHz speeds, too.
Games like Quake III Arena, Drakan: Order of the Flame, Homeworld, and Giants: Citizen Kabuto were among the first to make use of the integrated geometry and lighting hardware on the GeForce256, and it could really be a game-changer for users on the slower CPUs of the day. While a fast CPU could more than make up for the lack of such hardware, fast CPUs of the time could be unbelievably expensive. The 1.7 GHz Pentium 4 cost $777 on its release in 2000 for just the CPU alone; that would be equivalent to $1422 today—and you still had to buy pricey RDRAM memory, too.
