SI
SI
discoversearch

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.
Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?


To: Reginald Middleton who wrote (28743)2/24/1998 11:18:00 AM
From: AK2004  Respond to of 1572777
 
Regy
you still posting your distorted visions here, as I can see.

8 scenarios? And what is underlying process (You know what underlying process means - usually represented by tiny multi-factor SDEs). I would think it would take at least few hundred to get convergence on interest rates alone. And you managed to represent all that in 8 scenarios. Wow, now I am impressed :-))



To: Reginald Middleton who wrote (28743)2/24/1998 11:26:00 AM
From: Skipper  Read Replies (1) | Respond to of 1572777
 
I hesitate to throw my 2 cents worth in here since I am not a "pratitioner', and since I was handily spanked by AMD in 97, but the recent upward move this stock is showing speaks to the market's collective tendency to arrive at pricing that is driven from investor SENTIMENT,(that is they have finally solved their yield problems), not third order calculus equations that spit out a company's true "valuation"

Skipper



To: Reginald Middleton who wrote (28743)2/24/1998 2:29:00 PM
From: Petz  Respond to of 1572777
 
Reggie, re:<Under 8 out of ten scenarios, AMD is still no where near $45 per share>

Lets look one reasonable and one very pessimistic scenario on yields for K6's.

Lets say AMD only manages 5% yields for the 0.25 process for 1Q, 25% for 2Q and 50% for the rest of the year. For the 0.35 process we'll assume 40% yield in 1Q, increasing to 60% in 4Q.

Lets assume AMD processes 3000 wafers/week for 1Q, increasing to 5000 wafers/week by end of year.

Lets assume that 50% of wafer starts are for 0.25 process in 1Q, increasing to 80% in 4Q.

The K6 (0.35) has approximately 150 candidate die per wafer (162mm^2 with 23% wasted space)

The K6-3D and K6 (0.25) have at least 345 candidate die per wafer.

The ASP of the 0.35 K6's will be $110 declining to $80
The ASP of the 0.25 K6's and K6-3D's will be $200 declining to $140(very conservative)

Q PRODUCT WAFERS CAND_CPU GOOD A.S.P. REV
1.......K6(.35) 1500 x 13 2,925,000....1,170,000..$110...$129M
1.......K6(.25) 1500 x 13 6,727,500....336,375.....$200...$67M
2.......K6(.35) 1466 x 13 2,858,700....1,334,060..$100...$133M
2.......K6(.25) 2200 x 13 9,867,000....2,466,750..$180...$444M
3.......K6(.35) 1300 x 13 2,535,000....1,352,000..$ 90...$122M
3.......K6(.25) 3033 x 13 13,603,000..6,802,000..$160...$1,088M
4.......K6(.35) 1000 x 13 1,950,000....1,170,000..$ 80...$94M
4.......K6(.25) 4000 x 13 17,940,000..8,970,000..$140...$1,256M

TOTAL K6* REVENUES.......................................$ 3,330M

AMD had total sales of 2.36B in 1997, and of this only 0.683B was from the CPG group. If you add up the last column, you get a staggering 3.33 BILLION in sales from the K6 alone. In fact, even if 0.25 yield never exceeds the 25% level, and K6 prices decline as shown above, the K6 revenues would total 1.5 B, an increase of 817M over 1997. Roughly 50% of this would be profit since the only cost increase is about 200M for wafer production and 200M for packaging. The 25% maximum yield scenario results in a net profit for 1998 at AMD of $396M, or $1.67 per share (40% tax). $30 to $45 would be a reasonable valuation for this scenario, given what I haven't talked about, such as increasing domination of flash market, K7, etc.

If yields can reach 50% in the fourth quarter, the potential is staggering, an extra BILLION in profits or $5.40 a share in profits (45% tax). $21 is not reasonable for either of these scenarios.

Petz