It seems like it has been an eternity since Intel made their first major dive into the low-cost CPU market. In 1998, at the Intel Developer Forum, Intel CEO Andy Grove outlined what would be a major step for the company, the idea of applying a single processor core design to multiple product segments in the market. In doing this, Intel effectively created what they liked to call the “Basic PCs,” which, at the time, referred to PCs that ranged from $800 - $1200 in price and provided “a base level of functionality with limited expandability to meet the core needs and affordability requirements common to many new home and business users.”
It was on that same day, March 4, 1998, that Intel released a new brand name for their upcoming CPU that was designed for this “Basic PC” market segment, the Celeron. A little over a month later, Intel made the official introduction of the Celeron CPU at 266MHz for a price under $200.
The 66MHz FSB Celeron made its introduction on the same day that the Pentium II gained 100MHz FSB support provided by the 440BX chipset for the 350 and 400MHz parts. At the time, little attention was paid to the performance benefits provided by the 100MHz FSB frequency over the Celeron’s and older Pentium II’s 66MHz FSB speed.
By combining the fact that the Celeron used the “slower” 66MHz FSB as well as the fact that the first Celerons had no L2 cache, Intel attempted to truly differentiate the Celeron from its Pentium II brother. Intel did not want anyone using the Celeron as the basis for a high performance system as that was what the Pentium II was for and the profit margins on the Pentium II were greater than they were on the Celeron.
This original Celeron was a success among hardware enthusiasts because the 266MHz Celeron could easily make it up to 400MHz by simply increasing the FSB frequency from 66MHz to 100MHz. At 400MHz, the Celeron offered gaming performance virtually equivalent to that of the Pentium II 400, and thus, the Celeron became the ideal gaming solution.
With no L2 cache, the Celeron suffered incredibly in business/office application benchmarks and while the majority of gamers were enjoying Intel’s extremely overclockable Celeron solution, the Celeron came out looking very under par since most magazines use a combination of business and office applications to measure performance. By removing the L2 cache from the Celeron’s SEPP card, Intel could cut costs while establishing a clear performance difference between their low-end (Celeron) and their high-end (Pentium II) chips.
In spite of the fact that business application performance isn’t that big of a deal, Intel was forced to phase out the cacheless Celerons after hitting 300MHz. It was in August of 1998 that Intel introduced the Celeron A at 300 and 333MHz. The Celeron A improved on the weaknesses of the original Celeron by integrating 128KB of L2 cache running at core speed onto the die of the processor itself. This helped the business performance of the Celeron tremendously without affecting the price point of the processor since the Celeron’s 0.25-micron process allowed for the 128KB of L2 cache to be integrated onto the Celeron’s die without significantly increasing the cost of the CPU.
The Celeron A was introduced 19 months ago, and it has been with us ever since -- until today.