New Celeron D Core gets a Speed Boost

qtothemax writes "The new Celeron core was released on the 25th. The processor, using Intel's new model number naming convention, looks to be quite a bit faster than the old core. The new core is based on the 90nm Prescott, which offers respectable performance, compared to the very slow Northwood based Celeron. It features a 256kB L2 cache, and a 533mhz FSB. Looks like Prescott's longer pipeline is more then offset by the better branch prediction and most importantly the doubled cache when it comes to the smaller cached Celeron. This Celeron may be able to compete with AMD's offerings based on more then name brand alone. Reviews and benchmarks are at Anandtech. I couldn't find any other good reviews, as budget chips rarely generate much excitement."

Comments

Celeron 2.6GHz

[strictnein]

What Anandtech's review really seems to show is what an absolute piece of shit the 2.6GHz celeron was. In most of the benchmarks it was beat by the 1.6GHz Duron for fuck sakes. It was also beaten by a P4 1.8GHz, which wasn't too suprising, and even an AMD Athlon 1700+ (which runs at 1.47GHz - we're talking a 1.13GHz gap here).

Of course, last time a celeron interested me was when the good old Abit BP6 board was out.

Re:Core

[strictnein]

I'm not familiar with Intel's current family, but I seem to remember that Celerons were based one on the P2

The first ones were based on the P2. Then they based them on the P3. And then the P4. And now this one is based on a newer P4. As any intelligent manufacturer would do, their cheaper product line is simply based upon older versions of their more expensive product line.

Celeron 2.6GHz (better oveclocked)

[IYagami]

You can find a very good review at

http://www.xbitlabs.com/articles/cpu/display/cel er on-d.html

. They show that a Celeron D overclocked to 3.8 Ghz (yes, really) can outperform even a Pentium4 3.2E (Ok, only sometimes ;-) )

Sorry about my english

Re:Market Statistics

[TheRaven64]

The original 300MHz Celerons were very overclockable. Intel had a much higher yield than they expected, and most of them could run at 450MHz with no extra cooling. The same thing happened with the AXIA T-Birds (Athlons), where the 1GHz version could be pushed to 1.33GHz (again, with no extra cooling). The yields on 90nm chips are such that this kind of thing probably won't happen again for a while.

Re:celeron's are terrible

[Bishop]

Start reading here or here, and be enlightened.

From a December 2003 article:

When we can find a 1.6GHz Duron for just over half the price of a 2.6GHz Celeron and get better performance consistently in almost every test we ran, the choice is clear.