Intel Corporation (INTC) Message Board

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Intel Corporation (INTC)

INTC 35.40

-0.4%

10:41 AM EST

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next

To: Elmer who wrote (133664)	4/30/2001 11:31:42 AM
From: pgerassi	Read Replies (1) of 186894

Dear Elmer: In case you hadn't heard, I am on a 1425MHz Tbird KT133A system (150x9.5) with 256MB SDRAM at default voltage of 1.75V. It out runs a 2+GHz P4 (15x140) in the real world. It would outrun any 2GHz P4 at (20x100) except for sheer bandwidth only synthetic benchmarks. Intel shows that it can't sell it by cutting the price in half. Intel shows that it needs more voltage to get to 1.7GHz. It also raised the minimum voltage which only means that yields at the lower minimum voltage is not enough for saleable speeds. By the act Intel showed that P3 could only go two steps (866 to 933 to 1000) as the third step failed. Applying same to P4, 1.5 to 1.6 to 1.7 on .18u. John P. showed that 1.9G is iffy due to current restraints of the socket 423. At least a resocketing (478?) is required for continuing the power (both electrical, speed, and performance). Performance of P4 when used with PC600 shows that P4 becomes slower than P3 and thus far below that of DDR Athlon. Single channel PC800 shows that even using PC1600 DDRDRAM would perform very poorly, slower than even Celeron. And you think Intel will sell this? P4 has the same problems like the 300 Celeron had. It needs more L1 cache which means a three cycle one. A 8KB 2 cycle L1 instruction cache just doesn't cut it. Intel needs to add transistors to P4 to add both decoders and L1 to balance the pipeline as the current front end is just not enough to feed the back end. The trace cache just does not add enough to performance in typical apps. That is what points to a redesign, or if you prefer, a rework of the core. I believe Intel realizes this and will do something towards this when going to .13u. The problem is that marketing has to make do until it happens which leads to the much hated PR BS (FUD)(AMD unfortunately does BS too). Pete

Report TOU Violation

Share This Post

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next