To: gnuman who wrote (42481 ) 5/18/2000 10:43:00 AM From: jim kelley Read Replies (2) | Respond to of 93625
Hey! I was asking for other opinions not my own. <g> "Remember that the Jan-Mar production was 8-9 million parts and included a royalty credit for Samsung (the main producer of RDRAM) that was to run out during that production." -----Samsung----Toshiba Jan--1-----------1 Feb--2-----------1 Mar--2-----------1 Leaving out contributions from any other manufacturers we get about 8 M parts from Samsung and Toshiba. Toshiba seems to be providing all the PS-II parts and maybe providing some of the PC parts. Samsung said it was going to 2 M parts per month in January. "For Apr-Jun production I think we can safely say that we'll do double that, 16-18 million and Samsung would be paying royalties on all that production." This assumption looks a little low. I get ~ 20 million parts for the Apr-Jun qtr. "Remember that last Q production was based on 1-2 million parts and no royalties from Samsung due to their royalty credit." Actually, my recollection is that RAMBUS said 1-2 million parts but Asiabiztech had it at 3.5 M parts including PS-II. The difference appears to be Samsung was subtracted out of Rambus numbers because they were working off a prepaid royalty credit. " During that Q., total royalties hit a record $3.5 million (I get 2.63 M in recorded royalties - this does not include the Samsung prepaid royalties) (about 14.6 cents a share) WITH NO! RDRAM ROYALTIES!! So you can think of the $3.5mil as an other than RDRAM royalties that will continue to increase every Q." Actually the 3.5 million in royalties was probably due to PS-II controllers and memory. PS-II is shipping .5 M units per month (average) with 2 memory chips and a controller. That is about 1M memory components per month for the PS-II by itself. DELL began shipping RDRAM systems in late January. "For June we'd have: 8.5m parts x $26/part x .02 x .7 = $3.1m from RDRAM royalties alone or 13 cents a share. Now remember that a portion of that is Samsung royalty credit. I think we can safely say that even with half the royalties taken care of the credit, Rambus should easily do 20 cents a share for the quarter, beating by 4cents (or 25%) the estimates. That's 150% earnings growth from last year of 8 cents." I get royalties at 2.63 (PS-II) + (5.5 parts x $26/part x .02 )/2 = 2.63 + 1.43 = 4.06 Then, adding in contract fees: 9.32 + 4.06 = 13.38 So my earning number is lower than what he is calculating using his own assumed $26 per Pc memory component and Rambus statement regarding prepaid royalties. "For the Sep quarter, that is where the fun begins. With 17m parts we have: 17m x $26 x .02 x .7 = $6.2m/23.9m shares = 26cents. With this much royalties, there will be no question that Rambus will BLOW AWAY the estimates by a huge amount. I can't see how on earth it can be less than 30c a share (that's 50% above estimates and 200% growth from last year of 10c). Remember that this is JUST!!! RDRAM royalties. It does not include any of the $3.2m royalties earned during this past Q from other non-RDRAM parts, and it does not include increase production of 5% royalty of ASICs and RACs that matches the increased production of RDRAMs. Neither does it include revenue from contracts, networking & communications area, consulting, HDTV that is coming on line at the end of 2000, etc.." Here he gets away from me totally with unaccountable parts. It is clear, however, that RDRAM production should be reaching 24 M parts by September quarter not the 17 M that he seems to have come up with. Conclusion: There is enough uncertainty in the royalty scheme and the shipment volumes as well as prepaid royalty issues to make calculations of the EPS uncertain. It will take a few more quarters data to be able to predict EPS with reasonable accuracy.