I’m trying to figure out $ANET’s potential for durable growth into the future; here is what i have so far:
How $ANET approaches durable growth:
1. Let your customers tell you what to build
- They’ve been doing this literally since inception; their business started as a response to a request from Google for a system that could handle what was considered at the time to be a near-impossible level of performance requirements.
- To this day, they have the world’s largest hyperscalers working essentially side-by-side with them, directly with $ANET Engineers to provide $ANET with the real-world requirements for the next generation of networking products.
- This gives them insight nobody else has regarding what the world’s highest-value networking customers want next! Can’t be bad for their durable growth prospects.
2. Go after the big margins, and let others tend to the commodity stuff.
- They don’t mess around in networks where commodity stuff will suffice: residential LAN, small datacenter LAN etc.
3. Growth via “adjacencies”: leverage their strength in large-scale Ethernet solutions into additional types of networks:
- The first opportunity they identified to provide differentiated products was cloud networking e.g. the “front-end” network in huge data centers.
- Next they figured out their low-latency products would stand out in financial applications; back in 2009 one-third of their customers were big Wall Street firms.
- They then went after “front-end” networks in large Enterprise data centers e.g. so-called “campus” networks.
Does $ANET have a durable growth future?
- They have determined that the exponential increase in GPU speed, and GPU cluster size, in AI stacks will drive high-dollar needs for “back-end” network products that are more manageable than Infiniband, and they are aggressively gunning for an Ethernet standard that can compete with Infiniband for latency/throughput yet retain the visibility, manageability etc. of ethernet.
- This is yet another example of $ANET identifying a high-value adjacency and gunning for it.
How about just plain ol’ greenfield growth potential?
- They are STILL taking market share from $CSCO at a good clip, and have a long way to go before they overtake $CSCO
- They estimate the whole networking field is growing at 13% CAGR into the forseeable future
- The number of customers using their SaaS offering (CloudVision) is growing at a healthy clip
An example of another kind of high-value network they could potentially decide to move into
- (pure speculation on my part)
- Check out a company called Calix $CALX
- $CALX has solutions that implement visibility, security, manageabilility etc. to the “backbone” networks that connect homes/businesses to ISPs, and ISPs to datacenters and other intermediaries like $NET
- $CALX is doing about $1B annually in revenue; maybe it’s just a matter of time before this attracts $ANET’s attention?
- …again, just PURE SPECULATION on my part, but my point is there are additional kinds of networks that $ANET could potentially move into.
What do you think…does $ANET have durable-growth prospects sufficient to the demands of this board?
Just so everyone knows. Technically it seems to me $ANET isn’t really appropriate for this board; I just read over their CC and they are guiding for 10-12% growth for 2024. Which I think is sandbagging, big-time, and several Analysts asked questions implying as much, but still. Not the growth we look for here.
Here are some things to watch, though, if you are interested in $ANET:
They are ALREADY getting traction with their EXISTING 400 and 800 gigabit Ethernet solutions in AI back-end networks.
Quote from the CC: “We expect both 400 and 800 gigabit Ethernet will emerge as important pilots for AI back-end GPU clusters. We are cautiously optimistic about achieving our AI revenue goal of at least $750 million in AI networking in 2025.”
They said during the CC that increased deployment of back-end GPU stuff MIGHT drive additional front-end-network spending
Aside from that, it’s business as usual while they steadily gain market share as they 've been doing over the years
And let’s not forget the possibility that Ultra Ethernet will eventually become a thing, and the possibility that $ANET will hit the ground running as soon as it is…
OK, so I’ll stop yammering about $ANET now unless/until some needle-moving thing happens.
Long $ANET 1.5%-ish
Thank you for sharing your thoughts on ANET. I’ve owned the stock since 2015, and haven’t added to it since 2018. It has grown into a significant position for me (20%), given the stock appreciation over that time.
Given it’s revenue guidance of 10-12% I agree it probably should not make the cut for this board. The one item that gives me pause is that the Gen AI trend could be a signficant tailwind for a few years. AWS, Microsoft, Google, Meta need to build out their data centers to provide Gen AI products. This build out has created demand for NVDIA’s chips and SMCI’s servers driving their stock appreciation. ANET is expensive compared to it’s peers (mainly Cisco) but it also has the better products. As the cloud titans build out their data centers, they will need networking gear and ANET has the best products on the market. Seems like a strong opportunity for future growth but ANET’s leader’s are only guiding for 10-12% growth. I’m still trying to make sense of that before deciding whether I should trim. I don’t plan to add giving it’s size in my portfolio already. If anyone has thoughts on how ANET’s future revenue guide connects or doesn’t connect to the data center build out for GEN AI, I’d love to hear your thoughts.
Stewaj1, if you haven’t read the transcript of the earnings call yet, you may want to. Analysts talk multiple times about “conservatism” in estimations but ANET held firm.
My take-away based on the call was that it was awfully challenging to predict this next year based on timing vs 2025 which is why they were being conservative. So maybe that networking spend will come in 2024, but maybe not until 2025 - but it will come in. Some money has already been spent and the need for networking equipment if very cyclical so 2024 may be a light year.
FWIW I have held this stock and it’s been a great earner for me, but I did trim half after earnings only because of the 10-12% guide. I don’t hold a full position of any stock that doesn’t expect to grow at least 20% a year (lower threshold than the board, but works for me) So why did I keep half? Because i believe ANET is a great LT hold and there is a chance that the money will come in this year and if not, I think 2025 is going to be awesome. I am happy to keep a half portion for the long haul but the rest of the money needs to go to stocks that can do heavy lifting this year.