Pentium D805 Vs Pentium D820

tvl

Distinguished
Jan 4, 2006
58
0
18,630
I saw were someone was asking about a Pentium D 805 - 2.66 GHz and got somewhat curious.

I am currently runnning a Pentium D 820 - 2.8 GHz. Is this basically the same chip, other than the D820 has a slightly faster processor speed?

What was Intel's first Pentium D named ........ I actually thought it was the 820?

Just curious!!
 

ltcommander_data

Distinguished
Dec 16, 2004
997
0
18,980
The 820D will probably more than a bit faster since it has the advantage of not only extra clock speed but more importantly a 800MHz FSB vs. a 533MHz one.

The first Pentium D was the 820. However, Intel has a lot of Pentium Ds left over that couldn't clock to a 800MHz FSB. There's really nothing else wrong with them. So they're just selling them on a lower 533MHz FSB to clear inventory.
 

ltcommander_data

Distinguished
Dec 16, 2004
997
0
18,980
Introducing a budget dual-core was probably an important motivation as well as a side benefit, but the primary reason was probably still to clear defective inventory. Intel could have just as easily priced the 820D at the 805D's price level if they just wanted to introduce a budget processor because Smithfields are on clearance anyways.

What I found strange was that Intel is introducing new 90nm Prescott 2Ms with the new 12x multiplier. The 65nm Cedar Mills already have the lower 12x multiplier which is valuable for EIST. I just don't understand why Intel would bother going to the trouble of adding a new feature and creating a new stepping to the already replaced 90nm products. Allowing the 6x0s to downclock to 2.4GHz also takes away a selling point for the 6x1 line.

Intel announced the 12x multiplier Prescott 2Ms in tray form in December about the same time the first 65nm part the 955EE was launched"
http://developer.intel.com/design/pcn/Processors/D0105693.pdf

And announced transitioning the boxed Prescott 2Ms in February
http://developer.intel.com/design/pcn/Processors/D0105884.pdf

Since Smithfield is based on Prescott 2M, Intel could have probably have added the 12x multiplier to the line too if it wanted to.

On another note, I really wish Intel would transition the Celeron D to a 667MHz FSB. I understand why they didn't do it before since it wasn't related to the existing 400MHz, 533MHz, 800MHz or 1066MHz FSBs. However, with the 1333MHz FSB finalized they just need to cut it in half. The new 65nm 512k cache Celeron Ds look to be pretty good and are great overclockers, but with a 667MHz FSB they'd give the AM2 Semprons great competition. Celeron Ds on 667Mhz, high-end single cores and low-end dual cores on 800MHz, mainstream dual cores on 1066MHz, and the Extreme Edition on 1333Mhz should work out nicely as a product matrix.