LRDIMMs future and end-user risk factors

LRDIMMs validated by JEDEC without licensing relevant IP

LRDIMMs are based on a single-sourced “iMB” buffer chipset from Inphi (more on that below).

Inphi does not hold the relevant IP and LRDIMM memory modules also lack the relevant IP backing.

Please see these articles for background info:

DISCLAIMER: I am not a lawyer, nor am I qualified to give legal advice – what follows is my own understanding – readers should do their own research before they arrive at a conclusion.

https://ddr3memory.wordpress.com/2012/06/04/examining-patent-docs-at-uspto/
Examining patent docs at USPTO
June 4, 2012

https://ddr3memory.wordpress.com/2012/05/30/legal-issues-with-lrdimms-repeating-metaram-2/
LRDIMMs similarities with MetaRAM
May 30, 2012

https://ddr3memory.wordpress.com/2012/05/27/what-are-ibm-hcdimms-and-hp-hdimms/
What are IBM HCDIMMs and HP HDIMMs ?
May 27, 2012

https://ddr3memory.wordpress.com/2012/05/24/intels-need-for-lrdimms-on-roadmap-to-ddr4/
Intel’s need for LRDIMMs on roadmap to DDR4
May 24, 2012

https://ddr3memory.wordpress.com/2012/05/24/the-need-for-high-memory-loading-and-its-impact-on-bandwidth/
The need for high memory loading and it’s impact on bandwidth
May 24, 2012

https://ddr3memory.wordpress.com/2012/05/24/lrdimm-buffer-chipset-makers/
LRDIMM buffer chipset makers
May 24, 2012

NLST ‘537 and ‘274 patent reexamination re-validation

Netlist patent ‘537 and ‘274 reexamination process at the USPTO has led to their validation with all claims intact.

If you look at the statistics of reexamination proceedings (where a significant number of patents have most or all claims reversed), this is a very strong signal of the strength of NLST IP.

Reading the reexamination docs

Further, the reexamination documents provide illustrative reading for those wanting to evaluate the future of LRDIMMs.

Inphi’s “iMB” buffer chipset is used by memory module makers like Samsung and Hynix to construct LRDIMM memory modules.

However, Inphi has very little IP in this area. They were relying on prior art of other companies to rebut NLST IP on load reduction and rank multiplication.

NOTE: it is load reduction which allows 32GB LRDIMMs to work at acceptable speeds when 32GB RDIMMs (4-rank) face abysmal speed slowdowns at 3 DPC and 2 DPC.

Yet each and every one of that prior art that was presented by Inphi was examined in the reexamination proceedings and all claims of the Netlist IP were re-validated by the USPTO. Inphi will appeal to BPAI (Board of Patent Appeals and Interferences). However Inphi has the burden of invalidating each and every USPTO argument for the validity of every claim of the Netlist IP – which is a great burden to overcome.

Netlist vs. Inphi

Inphi employed patent reexams as a dilatory strategy to stay Netlist vs. Inphi. When this case resumes, Inphi will have very little to stand on.

What will this mean for the future of LRDIMMs ?

Inphi currently the only maker of LRDIMM buffer chipsets

Inphi is currently the ONLY provider of LRDIMM buffer chipsets.

Of the top 3 buffer chipset makers:

– IDTI has prudently scaled back on the rhetoric over the course of a few quarters (as evidenced by their enthusiasm for LRDIMMs during conference call comments).

– IDTI has postponed LRDIMMs to end of 2012 (i.e. skipping Romley and targeting the Ivy Bridge series according to their conference call).

– Texas Instruments has not been interested in LRDIMMs – possibly related to settlement in Netlist vs. Texas Instruments a couple of years ago (it seems Texas Instruments may have been the original leaker of NLST IP to JEDEC).

– Only Inphi has persisted with LRDIMMs and are currently the only supplier of buffer chipsets for LRDIMMs.

Outcome of adverse ruling in Netlist vs. Inphi – lack of replacement parts

If Inphi LRDIMMs are seen as “infringing product” it will result in damages (or treble damages for wilful infringement). Payment of damages could be limited to Inphi, but could be expanded to memory module makers who have used the LRDIMM “iMB” buffer chipset for profit – since Netlist manufactures not just buffer chipsets, but also makes complete memory modules. In this way, Samsung and the other memory module makers may be liable for the sale of infringing product (complete LRDIMM memory modules).

However, for end-users what is far more important is the future of LRDIMMs that they have ALREADY purchased and that they will need in the future as replacement parts.

When LRDIMMs are used, they have to be used EXCLUSIVELY on ALL the DIMM slots that are going to be populated. This is because LRDIMMs are not interoperable with RDIMMs or HCDIMMs.

There is no other replacement product for LRDIMMs except the products produced by Inphi.

If Inphi is barred from the sale of “iMB” buffer chipsets, there may be NO replacement LRDIMMs available in the future (to replace faulty parts).

For this reason, LRDIMMs could become a “dead end” product in the market.

Changing language in Inphi’s SEC filing about infringement

An earlier IPHI SEC filing has some risk factors and info on the lawsuits and the impact it may have on IPHI’s memory module (buffer chipsets) business:

http://www.secinfo.com/d14D5a.q2jnb.htm
INPHI Corp · 424B4 · On 4/1/11

For example, Netlist, Inc. filed suit against us in the United States District Court, Central District of California, in September 2009, alleging that our iMB ™ and certain other memory module components infringe three of Netlist’s patents. For more details, see “Business—Legal Proceedings.”

Infringement claims also could harm our relationships with our customers or distributors and might deter future customers from doing business with us. We do not know whether we will prevail in these proceedings given the complex technical issues and inherent uncertainties in intellectual property litigation. If any pending or future proceedings result in an adverse outcome, we could be required to:

* cease the manufacture, use or sale of the infringing products, processes or technology;

* pay substantial damages for infringement;

* expend significant resources to develop non-infringing products, processes or technology, which may not be successful;

* license technology from the third-party claiming infringement, which license may not be available on commercially reasonable terms, or at all;

* cross-license our technology to a competitor to resolve an infringement claim, which could weaken our ability to compete with that competitor; or pay substantial damages to our customers or end users to discontinue their use of or to replace infringing technology sold to them with non-infringing technology, if available.

Any of the foregoing results could have a material adverse effect on our business, financial condition and results of operations.

Oddly enough this boilerplate info is missing from the most recent Inphi SEC filings:

http://www.secinfo.com/d14D5a.p6xdw.htm#1stPage
INPHI Corp · 10-Q · For 3/31/12
Filed On 5/9/12 2:32pm ET · SEC File 1-34942 · Accession Number 1193125-12-221626

Evidently Inphi now thinks it no longer needs to warn shareholders about the risks of infringement.

Risk factors

This has been an examination of the worst-case risk associated with the use of LRDIMMs.

In practice, that risk might be mitigated by the licensing of Netlist IP.

However is this risk factor known to most users of memory ?

Hedging your options

Currently all the OEMs are supporting LRDIMMs.

However IBM and HP have hedged their bets and are offering IBM HCDIMM and HP HDIMMs (based on Netlist HyperCloud memory).

The HyperCloud product is superior to the LRDIMMs.

See the article:

https://ddr3memory.wordpress.com/2012/05/27/what-are-ibm-hcdimms-and-hp-hdimms/
What are IBM HCDIMMs and HP HDIMMs ?
May 27, 2012

And DDR4 is copying HyperCloud in even greater detail than LRDIMMs were doing.

Advertisements

12 Comments

Filed under Uncategorized

12 responses to “LRDIMMs future and end-user risk factors

  1. Pingback: JEDEC fiddles with DDR4 while LRDIMM burns | ddr3memory

  2. Pingback: Why are LRDIMMs single-sourced by Inphi ? | ddr3memory

  3. Pingback: LRDIMM buffer chipset makers | ddr3memory

  4. Pingback: Non-viability of 32GB RDIMMs | ddr3memory

  5. Pingback: Memory buying guide – when to use RDIMMs ? | ddr3memory

  6. Pingback: Is Montage another MetaRAM ? | ddr3memory

  7. Pingback: Memory buying guide – including 1.35V memory for Romley | ddr3memory

  8. Pingback: Financial institutions retreat from Romley LRDIMM story | ddr3memory

  9. Pingback: Infographic – memory buying guide for Romley 2-socket servers | ddr3memory

  10. Pingback: Examining LRDIMMs | ddr3memory

  11. Pingback: HyperCloud to own the 32GB market ? | ddr3memory

  12. Pingback: Awaiting 32GB HCDIMMs | ddr3memory

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s