Notebook 03.13.12: Datacenter, Equities and HPC for Wall Street

I have been sans posts for the last week.  I did write a long post on the evolution of datacenter and broader network architecture.  I think it came in around 2.2k works, which is long for a blog.  I shared with a few colleagues and it stirred up some interesting responses and the consensus was to wait and post after integrating reviewer comments.  On a rainy Saturday I thought some bookkeeping was in order:

Tech Field Day Networking Event: I spent part of the past week listening to presentations from a number of companies.  I really did not see anything new, just more of the same.  I hope ONF 2012 is not the same as ONF 2011.  Presenters should be banned if they reuse an old presentation.  Back to TFD, I thought the event was interesting and clearly geared towards a technical audience; I am really disappointed in the vendors – not the TFD team.  Innovation seems dormant and no amount of science projects and Crossing the Chasm buttering will take the place of bold innovation.  I will take another look around on Monday in NYC, but ODM hardware + Openflow does not = innovation.

RIMM: What is there to say that I have not said before?  Prior posts are in the RIMM category.  Missing a product cycle really sucks.  I thought the other day that maybe they should just focus on the keyboard market for smartphones and uses who want to use BBM.  It looks like it will go the way of Palm.

APKT, BRCM and FNSR: I am recently long all three of these stocks.  If there is really an uptick in orders for optical stuff, I think inventories are lean and FNSR should work.  I like the product cycle at BRCM.  APKT is the biggest risk.  If NA CAPEX is better, APKT should work and I think there are a lot of lazy shorts in the stock right now.

Old Trading Instincts: Back and forth with a few buysiders during the past week has led me to trim overall gross exposure to equities.  It feels like correlations are breaking down and after the best Q1 since 1998, I feel better being less exposed to equities because theUS economy in Q1 2012 does not feel like theUS economy in Q1 1998 or even at the end ofClinton’s first term.

I am off tomorrow to NYC for the Flagg 2012 High Performance Computing Linux for Wall Street conference on Monday.  Looking forward to seeing many of you at the show on Monday, I will be in town for drinks Sunday night.


* It is all about the network stupid, because it is all about compute. *

** Comments are always welcome in the comments section or in private. ** 

Notebook 01.08.12: Cloud Providers, APKT, Macro and Finding Time to Think

Cloud Providers

It was just a month ago that I posted some thoughts on Cloud Providers being the CLECs of the twenty-tens and a few days into the New Year look what happens: a cloud provider M&A deal.  Here is summary of the deal.  I guess the cloud provider consolidation is under way.  It appears that the price per user for an acquisition in the cloud services market is $28k per user as of last week.  I calculated this price using the Internap/Voxel comp, I assumed 1,250 users and assumed they will get the earn out, but I should really take the time to run the Terremark and Savvis deal comps and maybe I will get to that this week.

Acme Packet (APKT)

APKT updated their 2011 guidance this past week to reflect a lower anticipated revenue result.  I have written positively about the company in the past.  Observing APKT over the past year reminded me of lessons I learned about scaling PE-VC companies.  There are real operational barriers when a company approaches and exceeds $75M a quarter.  The barriers to growth that emerge around $75M a quarter and into the next revenue band between $75-299M are quite real and often the business activity level required to execute on a 90 day cycle and transition between revenue bands is really challenging for leadership teams.  There are marked internal differences between a company executing around $75M a quarter and a company executing at $150M a quarter.  As companies organically approach the $75M level it is like a plane climbing to altitude and encountering turbulence.  The company has to fly through it and it is going to be bumpy.  Leadership teams can (1) go it alone, (2) buy something or (3) find some help by combining efforts with a larger company.

Scaling venture backed startups is a topic I wrote about years ago and had given a thought to as part of five PE-VC backed startups in my business career.  Here is an excerpt from a paper I had written in 2006 and updated a few times into 2007:

“When all the numbers were entered into the model, companies were placed into four revenue bands.  These bands were created from my experience on the operating side of the business.  The revenue bands were chosen because within each band, the complexity of the business is different and the scale of a company’s infrastructure changes from band to band.

  • Tier 1 Companies: >$2B Quarters
  • Tier 2 Companies: $300M-$1.99B Quarters
  • Tier 3 Companies: $75-$299M Quarters
  • Tier 4 Companies: <$75M Quarters

I am looking at technology companies in terms of revenues (i.e. market share) because this is how CEOs look at their companies.  EPS and operating margins are important, but leadership teams in technology companies think in terms of top line revenues and market share.  This is in part due to the cycle of creation/destruction within the technology industry, but also because revenue growth means the possibility of an equity event.  There are a few additional reasons why I put companies into revenue tiers.  If you accept the World is Flat hypothesis or some variation of the premise that competing in global markets is important, then it is important to consider that only a few companies can compete in all geographic markets in their target technology verticals.

The chart below [not included in this 2012 blog post] illustrates total revenue (i.e. market share) for each tier on the left axis and number of competitors per tier on the right axis.  This chart shows that 56 competitors are playing for 12.89% of the market and nine competitors control 87.11% of the market. [The 65 companies in the model are networking/comm equipment companies.]

Moving between revenue tiers is very difficult.  It is the movement between these tiers that produces investment value.  To move between tiers requires a (i) technology paradigm shift to be occurring in the broader market or a specific technology sector or the (ii) expansion of the size of an existing market.  An example of a paradigm shift was the adoption of the client/server model and an example of the expansion of an existing market is the Telecom Act of the 1996 and the creation of ~300 CLECs, ~1300 ISPs and ~15 IXCs.

The objective of this paper is to look at the companies in the market and ask: (1) Are any companies moving between tiers, (2) are their any consolidation candidates and (3) are there any macro technology changes that will enable companies to cross tiers?

As a starting point, the following page shows a graph of the total revenues from 1990 through mid 2007.  The broad technology drivers in the market are identified as well as a selection of key events.  The term “turning point” is taken from Carlota Perez’s book and used in her context to describe the behavioral economic aspects of technology adoption.”

By reviewing my prior thinking on the subject of APKT, it makes me think that APKT would be an attractive M&A candidate for a company like CSCO.

Macro Market Thoughts

This past week a few friends and I were discussing the various market outlooks for 2012.  We were having all sorts of debates on risk on and risk off.  Then there was one email from buy side friend that succinctly summarized the back and forth email thread.  Here is the quote “The troubling thing for me is every time the market rallies – prices of commodities (especially oil) go up more than the market, which inflicts pain on the economy.  It is like QE2.  If the market rally means commodity price go up (risk on) – it means economy gets hurt and there is also inflation.  Inflation makes it impossible to launch QE3.  Ergo seems like there is implicit cap on the ability of market to rally because of this Risk On thing and what happens to commodity prices.”

Finding Time to Think

I am making a commitment in 2012 to find time to think about problems and not be distracted by the tyranny of the present.  This means I am not tweeting, I am not on Facebook, but I am reading and taking time to think about problems and plans.  Here is a link to an interesting post on creativity and problem solving.  If you are going to watch the video and just want to view the thinking portion then start at the 19 min mark.  I fully admit that I use the computer as a tool.  I write and analyze data sets on my computers and I use programs to organize data sets, create communications and interact, but I try to guard against time spent doing unimportant and non-urgent tasks.  Ideally what I want to be working on and thinking about should be important, but not urgent subject matter.  That is one of the reasons I like to write and my writing manifests into books, blog posts and if you ask the people I work with occasional long emails.  For me, writing is a method to framing my thoughts and my computer is a tool for this process that I must guard against becoming a distraction and creative inhibitor.


* It is all about the network stupid, because it is all about compute. *

** Comments are always welcome in the comments section or in private. **


October 2011 Earnings 1.0

Here is the start of my October 2011 earnings thread.  I am already a bit late having been positive on GOOG from the past, I did tweet it last week.  A few reactions so far:

GOOG: Sticking with my GOOG call from April, not much to update.

INTC: I have been positive on INTC based on my web 3.0 thesis.  It will be interesting to see if results were strong enough to break INTC out of the ten year down channel on a weekly basis.  Need a close above $22, but a close above $23.16 on a weekly basis would be a strong signal as that is the 38.2% Fibb going back to 2002.  I pitched this stock and thesis to a PM in NYC in September and he laughed at me.  Who is laughing now?

AAPL: I still like the product cycles and would be a buyer as it settles.  It is all about product cycles.

JNPR: JNPR was defiantly on the mind when I was thinking about my web 3.0 thesis and Moore’s Law exhaustion.  I read about fifteen reports on JNPR this morning, which is probably ten too many.  I would start by saying I think the company is in the midst of (i) product, (ii) market and (iii) leadership transitions.  These three forces need to harmonize for the company to go on a run and I do not see that happening for some time.

INFN: The company is making a bet on 100G and that they have the ability to scale their business to be a supplier in size to tier 1 service providers.  One cannot criticize the company for not taking on difficult endeavors.

APKT: Waiting on the official results after the pre neg which I posted about here.

YHOO: I think this stock is un-investible.  Negative product cycles and poor leadership; are there any other questions?

PWAV: I posted some thoughts on that last night.

MACRO: The market is really difficult to deal with on a daily and weekly basis.  It reminds me so much of 2008, even though I see people on CNBC who say it is not 2008.  I have posted on this in the past, but I will add when the market swings on reports from news outlets like Reuters, CNBC, WSJ and government officials and central bankers, we might as well appoint a Committee of Public Safety and start the Reign of Terror.

I will endeavor to post additional updates to this tread over the next few days.

Stocks of interest in the next few days:

10.19: RVBD, WDC

10.20: NOK, ERIC, T, STX, MSFT

10.21: GE, VZ


* It is all about the network stupid, because it is all about compute. *

** Comments are always welcome in the comments section or in private.  Just hover over the Gravatar image and click for email. ** 

Part II Cisco: It is all about the network stupid

This section is a modified email of an exchange I had with a CTO of networking startup.  I have removed any specific references to the company and focused on what I think are the emerging trends in networking and this is where Cisco needs to be focusing. I am going to start with how value has been created in the networking segment post 2000.

When I look at a handful of IPOs in recent years (RVBD, INFN, DDUP, BBND, STAR, MITL, BSFT, APKT, etc) the parallels to the 1990s are clear.  I looked at all of these companies while on the buy side.  I became very familiar with the management teams, operating characteristics of the companies, what really worked and what did not work.

– RVBD: What did they do?  Put a drive in a network element to cache content and improve the compute function at the edge of the network because the CIFS protocol was too chatty and the network did not have an enough bandwidth/capacity to overcome the software coding limitations.  So the content was moved deeper into the network.  In the case of RVBD, this was not really innovative.  I was on a engineering team that coded various run length compression algos into bridges and routers in the late 1980s and put PBX functions in an ATM based CPE box in the 1990s.  What was innovative about RVBD was the mechanism used to identify the content frequently called by remote users and move that content locally to improve the local compute cycles and then sync the data between the storage points.  The key point to focus on is compute.

– FFIV: In my view, the game change for FFIV came after the 2001 crash.  They opened their platform (Big IP) up to 3rd party apps with a developer’s kit, API, simulators, tools etc.  They then let the community of users share these apps.  When you talk to big banks (I spent a lot of time with CIOs and developers of the top ~10 banks over the past few years) you find out they have 10s of thousands of apps.  FFIV organically created a community of users who became very loyal because of this community.  This community also became compelling to new users.  Lesson learned was opening Big IP platform to market to create a community of application users.  BTW…applications are about compute.  Note…JNPR is trying to do this with JUNOS, but it is not quite the same.  JNPR is more partner OS related than end-user related.

– INFN: I am going to say very little about INFN.  I will just summarize by saying that the commoditization of the optical component supply chain happened far faster than they estimated and when they brought a system level product to market the price disruption was far less than expected.  Hence…a lesson…processor, memory, capacities tend to be commoditized on a horrible curve of buy more for less.  When processing, capacity, memory are sliding on the curve of more for less, this means software will just consume the compute cycles.  If you doubt me, there is a reason Apple developed their own processors.

– DDUP: I think these guys wrote the book on how to build an innovative, disruptive business in a market structure with large dominant companies.  Lesson here is that operational software that removes the need to spend huge amounts of CAPEX on storage arrays and increases the effective utilization of sunk cost is a winner.   Why would a customer what to effective move data from tier 1 storage to other tiers?  Answer is to allow compute cycles and new apps to use the top tier.

– STAR: They made a really smart bet that EVDO would have much longer theta than the incumbents where willing to assume.  Hence, they were the only real innovator of software on a HW platform.  If you have ever seen an ST-40 up close it takes up a whole rack of power in a CO.  This thing has more silicon in than a beach and runs hot, hot, hot – yet it has software margins.  How?  When you looked at what STAR sold it was a software MCP with all the management tools built on a custom hardware platform.

– APKT: I wrote a blog post on APKT the other day so that will stand as my place holder.

Network Trends I Believe:

Processors, memory, storage and bandwidth are elements that you can buy more for less over time.  The companies that create extraordinary value find ways to arbitrage the gap when one or more of these elements become dislocated.  Look at RVBD; when drives became cheap enough and large enough it made economic sense to fix the transaction cost of CIFS for SQL servers by caching the content.  The ultimate need is compute, but the problem was an inefficient, poorly designed protocol and not enough network capacity that was cheap enough to fix the protocol problem with the network.  DDUP is another example.

Akamai, what did they do?  Built a better way to call for and process content.  They cached content which means they moved the compute function deeper in the network to overcome network bandwidth/capacity limitations with their routing algorithms. What problem did AKAM really address?  The answer is compute.  They made compute (consumption of web content) easier because the network could not handle the traffic load so they distributed the compute function by moving the content deeper into the network, replicating it and finding more efficient routing paths to call for the content.

Virtualization and all the commodity cloud networking trends are forcing applications and data to be partitioned and replicated across multiple data centers.  Data sizes will continue to grow as we move from click streams, to scientific data, to user audio, photo, and video collections.  Processing or the compute function of these trends will really run in parallel on thousands of machines.  All of this is leading to the enablement of scale out networking (to use the Google term) which is…

  • Scalable interconnection bandwidth
  • Aggregate bandwidth = # hosts × host NIC capacity
  • Economies of scale: Price/port constant with number of hosts, Must leverage commodity merchant silicon
  • Anything anywhere: Don’t let the network limit benefits of virtualization
  • Network Management: Modular design that avoids actively managing 100’s-1000’s network elements

If you believe in this compute trend and that optics/bandwidth will follow the curves of processing, memory then the network is out of sync with the other elements.  The need becomes TB and PB metro and access networks that evolve to EB networks in ~10 years (a guess on my part).  Virtualization should really be a positive driver for network equipment suppliers because of replication and accessibility of data for compute.  CSCO had the correct thesis, it is all about the network – the problem is they got to big and lost focus on the network. That created opportunities for companies like RVBD, FFIV, APKT, DDUP, to really hurt them in market silos that CSCO cannot seem to get the A team to focus on.

What Do Networking Products Look like in the Future?

If we assume the network trends discussed above are true, then networking companies need to be focused on applications and compute to be a strong value creator – instead of a hardware box supplier.  As part of this thesis here are five key elements:

  • Back end analytics: Virtualization and content evolution from short form to long form means that data processing will dominate the future. The network has to enable that function. By the way, that function is called compute.  Enable that function to occur and you are a winner.  This means networking companies need to sell tools in a software form that changes the value proposition from a box seller to compute enabler.  I would be looking to add analytics and processing software tools on the platform.
  • Modular software
  • User Definable Platform Tools
  • Ability to Embed Value Add Apps into the Network Element or Call for Virtualized Apps in the Network
  • Security and Policy Tools

It is all about the network stupid, because it is all about compute.


Acme Packet and Innovation Leadership

Acme Packet (APKT) reported earnings on 04.26.11.  The stock was off ~600 bps in reaction to the numbers.  By 1800 hours it was only off ~200 bps.  The primary driver of the stock reaction was the high multiple of the stock and the guidance of Y/Y growth of ~35%.  This was raised from 30% when the company reported Q4 FY2010 numbers.  I suspect that this growth rate will be raised two more times in 2011.  We know that CAPEX was off to a slow start in Q1 2011.  For entertainment purposes I put a chart at the bottom that still has the original channel I drew for the stock in early 2010.

The purpose of this blog is not really to talk about CSCO, because I plan to write a deeper analysis of CSCO in the coming days, but rather to talk about innovation.  As a primer, I suggest that you watch this video (part 1 and part 2) from Andy Bechtolsheim on innovation, why it fails and why it works. There are some interesting corollaries to APKT and CSCO.

When I was on the buy side and I started looking at APKT, I understood the concept of a SIP translator built on an appliance for service mediation.  If you have no idea what I just wrote then think: build a language translator that allows people calling from different locations, speaking a different language to communicate.  Next, substitute machines for the people in the visualization.  Now you understand the APKT business plan.  What you do not understand is (1) why they have a ~$5B market cap for a company turning out $70M quarters, (2) how innovation meets luck and (3) the similarities to CSCO.

Let me start with #2.  APKT got lucky.  I define luck as: where planning meets opportunity.  The APKT team made the assumption that the proliferation of billions of IP enabled devices will use thousands of networks to communicate with each other and these networks will need devices on the borders to translate the languages between the networks.  This is called session border control and it is a software function that resides on an appliance.  The technology assumption was correct; all the company needed was an event to make the market.  Luckily for APKT a large number of investment banks (IB) massively over inflated the residential as well as the commercial real estate markets using leverage, derivatives and synthetic financial products.  I am pretty confident that the global financial crisis was not in the APKT business plan in the years leading up to 2009.

When the global credit crisis hit in 2008, it changed the technology buying process.  APKT became a winner and CSCO not so much a winner.  What really happened was the buying decision process and decision makers for technology provided by APKT changed.  CFOs started showing up in the decision making process and they wanted to buy technology that had a longer shelf life.  They wanted to break from the past and acquire technology that had longer deployable life span in the network.  You can see this in the revenue line in mid-2009.  Quarterly revenues start to move significantly higher after ten quarters of a relatively tight range.  This is when the market began to broadly adopt their technology and eight years of work began to pay off.

Q1 Q2 Q3 Q4
2011 $74.0
2010 $51.0 $53.3 $56.6 $70.2
2009 $31.0 $32.9 $36.3 $41.3
2008 $31.7 $25.7 $28.4 $30.6
2007 $25.1 $27.0 $29.6 $31.4
2006 $22.3 $23.7

Turning to #1 the reason APKT has a $5B market cap is the revenue growth, operating metrics of the business model and in large part the potential size of the market.  Investors like growth stories and they like growth stories with an operating margin of 39.7%, but they really like a growth story that says the evolution from old school PSTN networks to IP/packet networks will take years and that APKT will play into this sea change transition for years to come.

Looking at #3 and the primary reason I wrote this post was the similarities I see in APKT with CSCO from the early days (late 1980s and early 1990s).  CSCO was building routers for years before many people had heard of the company and years before their market exploded.  CSCO was able to hone their routing code; test it, try it out, and acquire field experience.  Routing and network control plane code is not optimized in the lab.  It is needs field testing and real world experience.  Things happen in production networks that do not happen in lab networks.  CSCO had a huge developmental lead on competitors and this was a significant advantage when their market exploded.

When I look at APKT I see that same developmental lead in their product set and they have a innovation lead in real production networks.  It will be interesting to see if other companies try to put a competitive solution set into their product lines.  The router vendors tried to do this to compete against RVBD and FFIV and failed.  This is going to be part of a future post, but for now I look at the developmental lead (i.e. years) that APKT has in their market, the potential for that market to grow strongly and I believe that APKT has a high probability to increase in value over the lower probability that the company is done growing, competitors will catch up and the market will reverse or slow adoption rates.