AMD and the death of the crypto-mining GPU

(AMD stock and its options are extremely liquid as AMD is one of the top 10 most traded stocks.)

Previously, I’ve written about how AMD’s profitability has been inflated by the cryptocurrency bubble.  This will likely come to an end in the coming quarters as the market for video cards will shrink thanks to the emergence of ASICs for top cryptocurrencies.  We are already seeing the effects of this as prices of video cards continue to normalize at the retail level.  From PC Part Picker:

Most investment bank analysts anticipate that AMD’s earnings will go to the moon.  While that type of behaviour may help their employer earn underwriting profits if AMD chooses their employer for an equity raise, these optimistic projectons are unlikely to materialize.

I/B/E/S estimates for AMD, courtesy of Sentieo.

In my opinion, AMD is a compelling short as its earnings will likely shrink rather than grow in the coming quarters.

The AMD bear raid from Viceroy + CTS-Labs + YLZ, volume V

It turns out that I missed other parts of CTS’ old website from January 2018.  The Management Team section from January has a different story than the one from March.

Secondly, I should own up to some of my mistakes in covering CTS Labs.

1. AMD has confirmed the vulnerabilities.  I originally thought that perhaps (A) CTS Labs was rehashing known issues or (B) they didn’t have the technical chops to find security vulnerabilities.  I was wrong.  AMD’s story differs from CTS Labs in other respects though.  CTS Labs has suggested that the vulnerabilities may take months to fix.  AMD states: “AMD is working on PSP firmware updates that we plan to release in the coming weeks.”  *AMD’s blog post is not entirely clear about when fixes for the Promontory chipset will be available.
2. Apparently CTS Labs (via its PR consultant) is claiming that Yaron Luk-Zilberman (YLZ) no longer actively manages NineWells Capital.  It’s unclear to me as to when that occurred since YLZ’s name appears on a March 8, 2018 SEC filing for NineWells.  Regardless, it wouldn’t be accurate for me to equate YLZ with NineWells.

CTS Labs is a lie (The AMD bear raid from Viceroy + CTS-Labs + NineWells, volume IV)

While CTS’ predecessor Flexagrid is registered with the Israeli government, I could not find a business registration for CTS Labs.  This is because CTS Labs and Flexagrid are the ‘same’ company.  According to the archive.org history of the CTS Labs website, Flexagrid “became” Catenoid Security which implicitly became CTS-Labs.  Flexagrid was founded in 2013.  The charlatans have been claiming that CTS Labs was founded in January 2017, which is a lie.

Less importantly:

1. The Business Wire press release has a 302 temporary redirect on it that makes the press release impossible to read (live link, archive.org link).  Effectively, Business Wire has retracted the press release.  (EDIT 2:48PM: Well I was wrong.  The redirect is no longer in place.)
2. In 2016, Uri Farkas registered a number of domains that seem like SEO spam.  The websites are filled with Amazon affiliate links.  The domain registrations suggest that Uri and Ilia Luk-Zilberman (of CTS Labs) were involved in a 2016 venture involving low-quality cookie cutter websites.  This is the kind of startup business an unemployed person might try if they can’t get a job in the IT security field.  The affiliate marketing game requires a completely different skillset compared to IT security and is easier to get into (even teenagers can do it).  Uri and/or Ilia are a little sleazy because two or more of the websites are written under fake names.

The AMD bear raid from Viceroy + CTS-Labs + NineWells, volume III

Two developments:

1. It looks like Fraser John Perring has told different facts to Reuters and Vice Motherboard.  Reuters says that Perring received the report on Monday March 12.  Vice Motherboard says that Perring received the report last week, which would be March 4 to 10.
2. CTS Labs has released a “Clarification About the Recent Vulnerabilities“.  It turns out that all 4 named vulnerabilities are only relevant AFTER an attacker has gained control of the system (“An attacker would only need to be able to run an EXE with local admin privileges on the machine”).  I don’t see how the severity of these vulnerabilities would differ much from BIOS rootkit attacks that have existed for years.  For example, the Vault 7 documents released by Wikileaks revealed that the CIA developed malware (‘evil software’) to spy on computers.  In 2017, Intel Security released a patch to detect such software embedded into the BIOS.  Viceroy Research did not write an obituary on Intel.
   • Earlier speculation from myself and others on CTS’ research seems to be misguided.  CTS Labs is not rehashing vulnerabilities that work without having control of the system (e.g. BadUSB).  CTS Labs is presumably alleging novel flaws (although they don’t seem to be actually claiming that their discoveries are novel).

The AMD bear raid from Viceroy + CTS-Labs + NineWells, volume II

“It looks like the IT security world has hit a new low.”
-Linus Torvalds, creator of Linux (via Google+)

Linus Torvalds has basically summarized the whole situation: clickbait media sites (e.g. CNET, Tom’s Hardware, Gizmodo, Vice, The Hacker News) breathlessly report on security vulnerabilities without critical thinking or fact checking.  The security industry takes advantage of that by making exaggerated claims and being attention whores.  On CTS’ report, Linus states: “I refuse to link to that garbage. But yes, it looks more like stock manipulation than a security advisory to me.”

Thankfully there are some journalists trying to do real journalism.  (I know the industry is dying but I’d like to thank the journalists out there who are upholding their journalistic integrity.)  In comments to these journalists, charlatans like Fraser John Perring and Yaron Luk-Zilberman have been quite disingenuous.  The short and distort campaign has been getting more bizarre.

The bear raid from Viceroy Research + CTS-Labs, and the NineWells Capital Management connection

CTS-Labs has come out with a “research” piece on AMD processors.  The reader might be misled into thinking that the “white paper” reveals previously undisclosed security vulnerabilities.  However, the CTS-Labs disclaimer (archive.org) states that the CTS report “summarizes security vulnerabilities, but purposefully does not provide a complete description of such vulnerabilities […]”.  So from my reading of the so-called “white paper”, CTS isn’t actually revealing previously unknown security flaws.

This isn’t like Muddy Waters’ work on St. Jude, where Muddy Waters alleged security flaws with St. Jude’s pacemakers; St. Jude has since recalled pacemakers to fix security vulnerabilities (Zdnet, FDA).  The difference between Muddy Waters and CTS is that Muddy Waters did actual research to find previously unknown problems with the company’s products.  Now if CTS actually did find a novel security vulnerability, then I would apologize.  However, the CTS report does not clearly articulate what’s a rehash of previously known security issues and what isn’t.

EDIT (3/21/2018):  Correction: the bugs are real.  Their severity has been overstated as they only work on systems that have already been compromised.  AMD says that fixes will be available within weeks while CTS Labs is still claiming that it will take months.  Status of fixes for the Promontory chipset are less clear at the moment.

Please don’t fall for the Viceroy Research / CTS Labs bear raid on AMD

Viceroy has come out with a “research” report on AMD based on the work of CTS Labs.  Just read CTS Lab’s disclaimer in its white paper PDF or on its website (archive.org).

1. CTS Labs doesn’t provide a complete description of security vulnerabilities.  So… I don’t see any original research here.
2. CTS Labs advises visitors that they have “an economic interest in the performance of the securities of the companies whose products are the subject of our reports”.  This is unusual for a IT security company.

Despite being short AMD (see my posts tagged AMD), I disagree with the ethics of what Viceroy is doing.

Blockchain is a useless technology

Blockchain, a way of implementing a distributed ledger (distributed record-keeping), is a novel technology with little real-world practicality.  The original Bitcoin white paper published back in October 31, 2008 spurred little interest in distributed ledgers.  The distributed ledger was ignored for years until Bitcoin started receiving mainstream attention and a few years had passed.

I simply couldn’t find much evidence that distributed ledgers are useful for any real-world applications (other than speculative asset bubbles).  Once you understand that blockchains are bad at solving real-world problems, then you will understand why Bitcoin will fail.  The blockchain imposes limitations that makes Bitcoin a bad version of something that has been tried in the past: e-gold (description here and Wired profile here).

A company’s stance on blockchain can also serve as a test of a company’s management.  In my view, companies pushing blockchain technology (e.g. IBM, Microsoft, Intel, Oracle) are disconnected from customers’ actual needs and have mediocre management.  Companies that don’t talk about blockchain (e.g. Facebook, Amazon, Google, Apple) are more likely to produce sensible technology that will work in the real world.

Betting against the cryptocurrency bubble via AMD

The cryptocurrency craze has infected the real world economy, driving up the prices of GPUs (graphics chips used for playing 3-D computer games that are also really good at crypto calculations).  Mining Ethereum with GPUs has become an increasingly profitable endeavour largely because the price of Ethereum went up about 170 times from $8.24 at the beginning of 2017 to a peak of around $1400.  Mid and high-end GPUs are selling out everywhere, retailing for 2 to 3 times their suggested retail price.  PC Part Picker has some good data on market pricing of GPUs such as the Radeon RX 570.

Note that computer hardware normally depreciates over time, roughly halving in price every 1.5 to 2 years due to Moore’s Law.  Instead of depreciating, most GPUs have appreciated wildly thanks to rising Ethereum prices.

If the crypto mining market collapses (due to a broad collapse in cryptocurrency prices), I am willing to speculate that AMD’s valuation would better reflects its difficulties in generating profits.  Historically AMD has never been a particularly profitable company, losing $7.82B of the $8.34B in capital raised.  In YE2017 AMD earned $62M before taxes, buoyed by aggressive accounting (perhaps $40M-97M+?) and unusually high GPU demand (perhaps a few hundred million?).  If AMD returns to its money-losing ways, its shares ought to trade closer to its $0.611B book value (plus the value of its x86 license) rather than its current $10.83B market cap.

Heterogeneous computing / GPGPU (INTC, AMD, NVDA)

Heterogeneous computing refers to computing systems that use a mix of different types of computational units.  The type of heterogeneous computing that I am most interested in is rise of the general purpose graphics processing unit (GPGPU).  The industry trend has been to use the GPU graphics chip for purposes other than graphics.  Innovations in the GPGPU could cause some changes to the landscape between Intel, AMD, and Nvidia.  I believe that heterogenous computing will grow from a small niche into a larger one.