That could turn this year, at least in terms of product sales, but there is a strong headwind blowing in the sector and the company's history makes us to give a pass on this one.
The company is an innovator in memory technology, from the 2018 10-K:
We provide high-performance modular memory subsystems to customers in diverse industries that require enterprise and storage class memory solutions to empower critical business decisions. We have a history of introducing disruptive new products, such as one of the first load reduced dual in-line memory modules ("LRDIMM") based on our distributed buffer architecture, which has been adopted by the industry for DDR4 LRDIMM. We were also one of the first to bring NAND flash memory ("NAND flash") to the memory channel with our NVvault non-volatile dual in-line memory modules ("NVDIMM") using software-intensive controllers and merging dynamic random access memory integrated circuits (“DRAM ICs” or "DRAM") and NAND flash to solve data bottleneck and data retention challenges encountered in high-performance computing environments. We recently introduced a new generation of storage class memory products called HybriDIMM to address the growing need for real-time analytics in Big Data applications, in-memory databases, high performance computing and advanced data storage solutions. We also resell NAND flash, DRAM products and other component products to end-customers that are not reached in the distribution models of the component manufacturers, including storage customers, appliance customers, system builders and cloud and datacenter customers.
The company derives revenues from licensing its IP and from producing products at its facility in China. How is that going?
Not really good. While the company has been able to more than triple its revenue in the memory up-cycle, they haven't been able to produce a profit during that whole term, which has us fear the worst now that the memory cycle has turned.
The company holds over 100 patents covering memory modules, SSDs and high bandwidth memory and other related products and it continues to add to these with 13 new patents granted in the past two years and dozens more in the pipeline.
The company is involved in multiple patent infringement cases which are a considerable drain on resources. Is there actually any hope this will subside, or at least lessen? Here are the main cases:
- Netlist 623 and 907 patents: ITC (International Trade Commission) against Hynix
- Netlist 912 patent: PTAB (U.S. Patent and Trial and Appeal Board) against Google (GOOG).
- Netlist 295 patent: Smart Modular against Netlist.
To start with the latter as it is the easiest, both companies jointly agreed to drop claims against one another, which has settled that case. The 912 case with Google is also likely to end with a favorable outcome for Netlist as Google just lost with the PTAB rejecting Google's three main arguments for invalidity.
Google is not one to give up though, they (with some other companies who have dropped out in previous stages) have pursued this for a decade, and still have 60 days to file an appeal. But the prospects have brightened here.
The most complex case is the 623 patent as that involved multiple courts, some of them ruling favorably, but other ruling against the company:
- ITC against Hynix has a favorable background from two recent wins, one in February with the ITC denying Hynix efforts to go directly to a final ruling without a trial and a favorable Markman order from the ITC last year.
- Netlist lost its case against Hynix (and Hewlett Packard) in Germany in January but the company is appealing the decision.
If the company wins, SK Hynix will be barred from importing its RDIMM (patent 623) and LRDIMM (patent 907) products in the US. The initial determination is set for June 14 this year.
Whatever the outcome, we are worried about the extent these cases are a drain on resources for the company the size of Netlist. There is ample evidence of this (Q4CC):
We ended of 2018 with cash and cash equivalents and restricted cash of $16.7 million, compared to $19.4 million at the end of the third quarter, a net cash burn of $2.7 million, of which $2.3 million was used to pay down legal expenses.
Almost all of their cash burn comes from litigation expenses, that's not good, but at least there is the perspective that this is going to decrease this year, from the Q4CC:
We do expect it to go down Q1 and then further decreased after the ITC trial.
Management also argues it is going to take a more proactive stance
Their main products are:
- NVMe SSDs
- HybriDIMM storage class memory
- NV vault
NVMe SSDs are in a hyper-growth (40%+) phase and the company might be gaining traction here (Q4CC):
central to the continuing effort to grow our product business are the new lines of enterprise and NVMe SSDs that we introduced last year. NVMe SSDs are the fastest growing part of the enterprise SSD market, with a 40% compound annual growth rate. These recently introduced SSDs have been very well received in the market with more than 55 customers' samplings, of which 15 qualifications are now complete and 12 more are in process.
They specified their expectations here during the Q4CC:
We believe that, our plan calls for us to exceed $10 million in sales of SSDs this year. I think we will see some evidence of that traction in starting next quarter. We -- as I mentioned, we have some 50 customers in the pipeline, a dozen or more qualifications completed and another dozen in process.
Probably the most innovative products are the company's HybriDIMM, which are part of NVDIMM products. These are non-volatile memory products (part DRAM, part flash) that slot in the fastest bus, that for DRAM (DIMM). You can find a really good primer here on the different approaches, there are at least four (NVDIMM-N, NVDIMM-F, NVDIMM-X and the upcoming NVDIMM-P).
This is the technology that was central to a deal with Samsung (10-K):
we partnered with Samsung Electronics Co., Ltd. (“Samsung”) in November 2015 through a joint development and license agreement (“JDLA”) to jointly develop new storage class memory technologies, including a standardized product interface for NVDIMM-P memory modules, in order to facilitate broad industry adoption of this new technology. We believe Samsung represents an important strategic partner with a high level of technical capability in memory that can facilitate bringing our HybriDIMM technology to market.
Here is The Register describing it in 2016:
DRAM gives Netlist's Hybrid DIMMs a cache veneer, providing what looks like DRAM to applications but is really persistent NAND underneath, cheaper than DRAM and lots of it.
HybriDIMM is the result of combining Netlist's HyperVault technology with Samsung DRAM and NAND – a Netlist project that began in November last year.
The idea is to use NAND as a DRAM substitute, masking its slowness with predictive software called PreSight, that loads data into DRAM from NAND in anticipation of it being needed.
There was a lot of optimism back in 2015 as that deal $23M licensing deal looked to be a break-through, here is SA contributor The Smallcap Authority back in November 2015:
In this partnership, NLST and Samsung will produce a new class of NVDIMM-P memory solutions. While the stock is up 50% on the news, we believe the long-term implications of the deal are far more significant than that. The most significant part of this deal for investors is that it validates NLST LRDIMM technology and proves that they own the IP for this transformational technology going forward. While the company has been saying this for some time, investors have been waiting to see some traction in the market.
Well, investors had to wait for an additional 3+ years but it indeed looks like some versions of it are finally readying for commercial production (10-K):
We publicly demonstrated a HybriDIMM prototype in August 2016 and we sampled HybriDIMM to select customers in the second half of 2017. We are now working with certain customers to transition to volume production.
Netlist has two products that fall under this approach, the HybriDIMM and their NVvault. Management claims the HybriDIMM module is unique, from the 10-K:
we believe HybriDIMM is the industry’s first storage class memory product capable of operating in existing Intel x86 servers without BIOS and hardware changes. HybriDIMM unifies DRAM and NAND flash in a plug-and-play module, delivering terabyte storage capacities operating at DRAM-like nanosecond memory speeds
With respect to NVvault (10-K):
NVvault DDR4 NVDIMM (“NV4”). NV4 is an NVDIMM-N that provides data acceleration and protection in a JEDEC standard DDR4 interface. It is designed to be integrated into industry standard server or storage solutions.
But quite frankly, while these products might be unique in a very narrow sense, it seems that the competition is achieving much the same via similar routes, here is the above linked article from The Register again:
Gen 2 HybriDIMM will add to this and be configurable to 1TB NAND + 32GB DRAM per DIMM/2400 MTS 2DPC for Purley processors. It will have both Linux and Windows support.
This is broadly similar to Diablo's Memory1 technology, which currently has 128GB DDR4-format DIMMs available, with 256GB ones coming, enabling up to 4TB of "memory" in a 2-socket server.
Although it has to be also said that the article mentions a number of performance claims from Netlist itself:
- Unites DRAM and NAND flash on LRDIMMs.
- Lowers memory costs by up to 80 per cent vs traditional DRAM.
- With buffering technology, improves access times to near-memory speed and 1000x vs PCIe SSD.
- Is 100 per cent compatible with current x86 servers.
- Uses PreSight predictive algorithms to pre-fetch and stage required data.
- Enables the scaling of "memory" to storage-level capacities in terabytes.
But it is a little worrying that the primer we linked above does mention a number of approaches and companies, but not Netlist. So despite the deal with Samsung, it looks like the company has found a really interesting approach in combining non-volatile flash and DRAM for the much faster DRAM bus, but they won't have the field all to themselves.
The upshot is that the company has interesting products (and IP) that will generate revenues and got important validation through the Samsung deal, but the optimism that this was something fairly revolutionary has subsided considerably.
Not really a pretty picture, from the 10-K:
This is not a pretty picture, 10% gross margin is disappointing, even if it's a little better in non-GAAP (product gross margin was 14.3% in Q4).
Cash burn isn't as bad as two years ago but it is still solidly negative. Much of it comes from the legal cost though (Q4CC):
we had about $2.3 million in Q4 of total OpEx without legal, of which $2 million is cash based. We brought in $1.3 million from product margin. So you're looking at less than $1 million cash burn.
The company still has $16.6M in cash but they also have $17.3M in notes outstanding and the amount of dilution is really quite worrying:
Management doesn't give specific revenue or earnings guidance, but nevertheless there are some worrying developments, from the Q4CC (our emphasis):
we do not guide given the current pace of bookings as of today, we do anticipate a flattish or slightly down first quarter revenue performance compared to last quarter due to the industry wide decrease in demand and resulting lower ASPs from what we see as conservative purchasing behavior by our customer base.
Management was a little more specific when asked about their Samsung branded products (Q4CC, our emphasis):
Given the massive decline in the DRAM and NAND quarter-to-quarter I think we're expecting about a 50% price decline year-on-year. But from last quarter alone, I think we're already seeing anywhere from 20% to 30% price decline. So if the volumes were the same, I think you just have a natural decline. As Gail mentioned, I think we're looking at a flat to a slightly down quarter, which means we have -- we are generating more volume, but I don't think enough to make up for the decline in the ASPs of memory and SSD products.
We have to say a 50% y/y price decline, that's rather concerning.
Despite a history of losses and cash burn, the shares are actually not all that cheap which surprised us a little.
The company's NVE SSDs, hybriDIMM and NVvault products might finally start to generate meaningful revenues this year and there is the prospect of declining litigation cost.
Unfortunately for the company, this is occurring against a background of steep expected (50%) memory price declines. The end result is that the company is unlikely to become profitable or generate positive cash, in fact that situation could very well worsen, rather than improve despite the progress in the product roadmap.
Given the history of losses, cash bleed and dilution, we're not optimistic, to put it mildly.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.