INTC Message Board - Msg: 1285438

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 : INTC

INTC 36.78

+2.7%

Nov 26 3:59 PM EST

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next

To: Larry Loeb who wrote (355)	4/28/1997 3:11:00 PM
From: Paul Engel	of 990

Larry - Re :"According to today's WSJ, AMD is producing chips using a .30 mu process as compared to INTC's .25 micron process. Is this consistent with your understanding?" This is what AMD said at the K6 Launch meeting - I was there and took notes: The K6 is made on a "0.35 micron process" using polysilicon gate lengths of 0.30 micron. The process also uses 5 layers of metal for interconnect and a LI (Local Interconnect) at the gate/Source-Drain Level" The process also employs C4 solder bump technology for external connections. The subtlety of 0.30 vs. 0.35 is as follows - The minimum feature size of the process was set up to be 0.35 microns. AMD could only achieve the performance of the K6 by tweaking this down to 0.30 FOR ONE LAYER ONLY - POLYSILICON. Note - Intel's Pentium II does the same - the poly gates are tweaked down to 0.28 microns to get the desired performance. The remaining layers (AMD and Intel) have feature sizes that are 0.35 microns or larger, depending upon the layer in question. AMD's 0.35 micron process is more agressive than Intel's 0.35 micron process in terms of packing density with one extra metal layer (5 vs. 4), a local interconnect layer (none for Intel) and the use of C4 solder bumps. HOWEVER, Intel's process is much more robust - as evidenced by the speeds (233 MHz, 266 MHz and 300 MHZ) of Pentium II devices that they can make on this process (0.35 micron). AMD appears to have real difficulty making 233 MHz devices. They have to run their 233 MHz device at 3.2 volts Vcc in order to achieve the 233 MHz speed, whereas the process was targeted for 2.9 volts, which the 166 and 200 MHz device use. Intel's Pentium II runs a t a nominal 2.8 volts. In a recent discussion by Bob Jecmen of Intel, he indicated Intel's 0.25 micron process would use 5 metal layers and C4 technology. At that point, Intel should be very close to AMD (except for the local interconnect level). Intel hasn't offically announced their 0.25 micron process. I suspect that they will make an announcement this quarter for a high speed Pentium MMX (200 and 233 MHZ) targeted at notebooks - with 1.8 volts Vcc for greatly reduced power consumption. Shipment of these devices will probably be early in Q3 - although "SAMPLES" are already in the hands of customers. Pentium II devices based on the 0.25 micron process will probably be announced for delivery in late Q3 or early Q4 - again at lower power dissipation than the 0.35 micron parts. Always keep in mind that the number of devices produced on a given process is perhaps more important than the overall process parameters. The best way to judge this is to look at K6 shipment rates vs. Pentium II shipment rates. Overall yields impact the die cost - a high yielding process can produce more and cheaper die than a low yielding, agressive process with a "smaller" die size. All those extra layers in the AMD process are not free - they come at a price! We may have to wait until AMD and Intel announce their June quarter results to get this information. Paul

Report TOU Violation

Share This Post

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next