Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations wOOdy-Soft on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Intel Celeron

Status
Not open for further replies.

tinkertech

Technical User
Oct 29, 2002
285
US
Is the Intel Celeron Processor a 32-bit devise?

If at first you don't succeed, reboot!
 
Celeron is a generic term to describe a group of processors made by Intel that are not fully a PII, PIII, or a P4, but based on the same core. They are built on the same cores as their full-blown counterparts, but may have reduced Speed, Cache Size, or some other feature.

I have a Tulatin Core PIII Based processor at home I am still using. It was basically a PIII, but it only had a 256k size L2 Cache and ran at 1.2 Gig. However, there are at least 2 cores of P4 Celeron. Some have a 400Mhz FSB and 128k Cache. Some have a 533Mhz FSB and 256k Cache. Some 533Mhz FSB processors are Northwood or may be the newer cores that have more pins for socket T. So you see you have to know what version of P4 Celeron you are buying and what socket type it is designed for. I would guess that Intel will come out with a 800Mhz FSB Celeron some day.

If you do not like my post feel free to point out your opinion or my errors.
 
Are you asking if Celerons are 32 bit as opposed to the newer 64-bit? Yes they are, i think ALL celerons are.
 
Why are they developing 64-Bit Celerons?! What a waste of time. Just make slower P4 64-Bit CPUs like AMD did (you can get the AMD64 3000+ which is about £100 - entry level).
 
This goes back to the days of the 486 SX

When making semiconductors, not all of them come out perfect. Until they get the process just right, the yield will not be very high. There is a bit a voodoo to maximizing chip yeilds... it is sorta like tweaking a recipe to make the best cake.

Anyway, way back in the day when Intel started to put the Math CoProcessor on the same chip as the CPU they had crummy yields on these 486 DX chips because an error in either the CPU part or the Math part of the chip would make the entire thing useless.

Say you have a 75% yield on both parts... the yield of totally good chips might be only 50% 0.75 x 0.75 = 0.5

So what do they do? They can sell the ones with a good CPU but a faulty Math CoProcessor as a 486 SX ... all they have to do is disable the faulty piece and then they can sell the partially working chip as a different kind of chip.

Skip ahead to the modern era... If the chip has a bug in half of the onboard cache then just disable that part and sell it under a different name.
 
If Intel follows its normal line of reasoning they come out with the more expensive server CPUs and then they come out with Desktop/Mobile CPUs and then they make celerons after the they go get done developing at least one more and possible 2 or 3 more core generations. Never for the current technology.

If you do not like my post feel free to point out your opinion or my errors.
 
It was just a hypothetical example but yeah, it isnt uncommon to get less than 75% yield during the ramp up.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top