Fri, Dec 20, 2013, 11:59 PM EST - U.S. Markets closed

# EZchip Semiconductor Ltd. Message Board

• dubbyao dubbyao Apr 24, 2002 7:45 PM Flag

## Potential chip sales

I emailed Cisco and Juniper today, asking them how many high-end router units were sold in 2001.
Cisco has not gotten back to me yet, but Juniper replied, "Juniper sold 4519 chasis and 54,894 cards in 2001".
Since they had 34% of the high end market last year, it can be extrapolated that Cisco is roughly triple these numbers.

SortNewest  |  Oldest  |  Most Replied Expand all replies
• Great job; you are at the core of the potential of LNOP.
The only problem I see is this dam high tech. depression.

• This is the part I love. Please double check my work.

Juniper would need 54,984 chips for 54,894 cards (2001 sales figure). @\$1145 each=\$62,853,630 gross sales for EZ Chip.

We know that Juniper's market share is roughly one third of Cisco's so triple this number to get an estimate of Cisco's sales numbers. Which=164,682 chips, x \$1145 each= \$188,560,890 gross sales to EZ.
\$62,853,630
+ \$188,560,890
--------------
\$251,414,520 total gross sales for both.
x.60 margin (may be high)
--------------
\$150,848,712
x.57 % of EZ Chip that LNOP holds
--------------
\$85,983,765 gross net/7.3 million shares=\$11.77/share, x P/E of 20= \$235/share.
In a good market this p/e could easily be double. These numbers are staggering, but it looks high because of the low # of shares outstanding. Thats the benefit of a low # of S/O.

• Good thinking to collect this information. So what does that translate to in terms of EZ chips? How many per chasis and how many per card? Can multiple chips go into one card or chasis?
Also there are many other applications in which these chips can be used. Any idea which applications require the highest chip count?

• 2 Replies to jonsventnor1
• PS: I'm not an expert in the field and I'm presuming one EZ chip per card. I'm just trying to get a handle on the potential we have here.
By the way, Jonsventnor, I have enjoyed your posts for a long time. They have always been very helpful. Thanks so much.

• Network processor can be used for storage device, too. Then IBM, EMC, CPQ, SUNW, MCDT, STOR, EMLX, to name a few, may need EZChip. It may be used for optical switch/router and metro DWDM, then CIEN, JDSU, CORV, AVCI, SCMR, FDRY, EXTR, RSTN, ONIS will be potential customers. How about SONS, INRG, COSN? I am not a techy and am not sure, but it may be not entirely unrelated. Anyone, please enlighten us about this.

• Once again, excellent post dubbyao
I'm hoping to get some reasonable numbers for the total available market that EZ could supply. Listed below are the major applications with estimates provided by dubbyao for the number of chips needed for Juniper and Cisco. It would be most helpful for anyone who has information to provide it.
Who are the suppliers for these applications and what quantities of chips do they need per year?
Routers:
Juniper:
chasis: 4519
cards: 54894

Cisco:
chasis: 13500
cards: 165000

Switches:
Gateways:
Storage Devices:
3G base stations:
Servers:
Probes:
Analyzers:

Who are the major suppliers and what is the volume for each? Who knows?
Please fill in the blanks. I'll summarize if information is provided to me. Also please challenge any numbers that are presented so realistic #'s are achieved.
Thanks in advance to all who contribute.

• Those type of numbers assume that all router sales from Juniper & CSCO would be powered by NP1's... not too realistic.

However, what is AMAZING is that with only ONE 10K chip sale, LNOPs PE Dives from 800ish to 16is. ... accompanied by a large move up for the share price.

Conservativly, even if LNOP only sells 50,000 chips this year, revenues would have gone from basically ZERO to \$57.25 MILLION in about 7 months (thats only selling about 7,200 chips per MONTH). With 57.25 million in sales (multiplied by estimated 60% profit margin) gives you about 34.3 million in profit. Multiply that by the 57% ownership in EZChips by LNOP and you get about 19.6 million. Devide that by the number of outstanding shares (7.29 million) gives you a conservative estimate of \$2.68 PER SHARE of earnings by the end of the year.

Bear in mind, the profit margin could be 75 or 80% depending on the IBM contract and if they are picking up the fab costs or not. Also bear in mind that I would guess that any Tier one would be INITIALLY more hesitant to migrate teh heart of their cash cows untill the chip is run through their own "in-house" testing....

Once the ball starts rolling... I would definately say the PE should remain at least 45 to 65 after ramping revenue from Zero to 30, 40 or 50 million.

- A Long... Estatic about the posibilities, both the conservative ones and the optomistic ones.

• Sorry, the numbers for year 2000 and 2001 is hypothetical values.

• I am extremely enthusiatic about high speed communication in the future. I believe that eventually evry home will have fiberoptic cable connected beyond the cable/DSL, and we will have full screen full speed streaming video. This is possible with current technology, the only limit is demand and cost. We may see this happen faster than we imagine. Video on demand, Video conference, and ultra wide band wireless communication is around the corner.

• I believe that in the future communication is top of everything, and subsequently network processor is more on demand than CPU for PC. INTC is the king for CPU and who will be king for NP? EZchip is 2 generation ahead of any other NPs. I wish EZChip go beyond INTC in MC and technology. Even NP-1 is technically superior to Pentium 4 right?

• Hi everyone. Good post and detective work dubbyao.

I apologize to everyone for the long post, but I assure you every word and every link will be worthwile. Either you will learn something new, see it a different way, or totally reinforce everything you know and believe now. With that said....... here goes......

While I am bullish for the prospects of EZchip in the long run, I thought I'd just make a few clarifications about "forecasts" and "guesstimates".....

(1) If JNPR had roughly 34% (1/3rd) of the high-end market last year, then CSCO would not be "3x as big" ...... rather only *2x as big*.

54,894 (55K) x 3 = 164,682 (or roughly 165K) [ENTIRE MARKET 2001]

55K/165K = 1/3 [JNPR's share 2001]

The other 2/3 (110K) = CSCO and anyone else

Make sense? I hope so. And I'm not being sarcastic, and I am not short the stock (or long put options). I am looking to build a substantial position over time.

======================================

(2) EZchip is only partly owned by LNOP (between 57-70%, based on unverified posts)

Any revenue figures must be cut by that ratio first, IMHO.

======================================

The future still looks bright for EZchip and LNOP, but it is best to not get carried away IMHO. This creates unrealistic expectations and weak-handed owners in the short run.

Also, I believe we need to factor in some "unknown competition", like Bay Micro (and who knows what other private fabless chip company).
http://www.linleygroup.com/npu/wire042302.html
http://www.linleygroup.com/npu/index.html

Does anyone truly know what other chips are in "top secret R&D" everywhere else, that could pop up and take away EZchips major advantage?

Personally I'd rather know all of these things in advance, so I can determine when the stock is getting way ahead of itself in the next 12-18 months.

To me, it seems like JNPR and CSCO will be bidding for Bay Micro and EZchip (to compete against each other, if they don't have IN-house products on the way) with the VCs:
http://www.ezchip.com/html/ab_investors.html

Good luck to all.....