Tesla AMD Partnership Rumor Shows A Chink In Nvidia’s Armor – NVIDIA Corporation (NASDAQ:NVDA)

Nvidia (NVDA) is a terrific Company. The Company’s CEO, Jen-Hsun Huang, in our view, rates to be one of the most effective CEOs in high tech industry. But, a solid company with a great CEO does not make for a sound investment if the stock price is in the stratosphere.

In the recent past, NVDA has been a story stock with analysts falling all over themselves to paint increasingly rosier prospects for the Company. However, we believe the Nvidia investment thesis is stretched pretty thin with the onset of potent competition.

CNBC Story Raises Flags

This week’s CNBC story about Tesla (TSLA) working with Advanced Micro Devices (AMD) gives investors a preview of what is ahead for Nvidia stock as competition materializes.

To be sure, there has been a bit of controversy about the story and CNBC amended the initial story since it originally reported it on Wednesday. While the initial story indicated that Tesla was a customer of Globalfoundries, that no longer appears to be the case. When the dust settled, a close reading of the publicly available information boils down the narrative to the following items:

  • CNBC source suggests that Tesla has already gotten first silicon on its AI chip and the unit is currently undergoing testing.
  • Tesla is in an IP deal with AMD for the AI chip development

It has been long known that Tesla was working on its own silicon so the former news item is akin to a project status update. However, it is the latter item that has the market excited and has taken a toll on Nvidia stock.

Sizing The Potential Lost Opportunity

While we have no details on the deal, the rumored deal could have some or all of the following components:

  • AMD may be doing a semicustom design for Tesla.
  • AMD may have licensed its GPGPU technology, very likely Vega or a variant, to Tesla.
  • AMD may have licensed its Infinity Fabric interconnect technology to Tesla to enable Tesla chips to connect with other Tesla or AMD chips.

If Tesla apparently is already testing the silicon, it is possible that Tesla may be able to ramp this chip starting mid-2018. Assuming that Tesla can use AMD IP starting mid-2018, the worst case sales hit for Nvidia for 2018 may be a loss of 150,000 units of what otherwise would have been Nvidia business. 2019 sales hit could be 300 or 400 thousand units or higher. Over the lifetime of the chip, it is conceivable that Nvidia will lose the opportunity to sell a million Dive PX type units to Tesla (assuming Tesla is in business that long).

At a million units, assuming ASP per unit between $100 to $200, the range on potential revenue contribution from this project to Nvidia could have been in the range of $100M to $200M. Sizeable, but, not a material hit to Nvidia.

The Impact Goes Farther Than Tesla

It is never a good thing to lose a large high profile customer and there can be a massive reputational loss to Nvidia here. Any deal with Tesla is likely to offer AMD:

  • Significant reputational benefit to AMD for winning a socket against Nvidia.
  • A validation of AMD’s GPGPU IP. Vote of confidence from a high profile name in the auto industry may bring positive tidings to AMD such as design wins at other auto manufacturers.
  • A perception among customers that switching from Nvidia to AMD could be economical and not that difficult.

In other words, while the immediate opportunity loss may be limited, an AMD win at Tesla can be a harbinger of bad news ahead for Nvidia.

Whether the CNBC story turns out to be true or not, we continue to believe that AMD will now increasingly take market share from Nvidia. As such, in spite of teething problems, we believe that AMD has now become competitive with Nvidia in the GPGPU space. Also, when coupled with the strong Epyc server platform, we expect AMD to continue to start taking share from Nvidia.

The potential for market share loss does not stop with AMD. Nvidia also faces challenges from new entrants in the space. In addition to semiconductor players such as Intel, verticals such as Google (GOOG) (NASDAQ:GOOGL) and Apple (AAPL) building their own solutions, there has also been a massive VC investment boom in the machine learning semiconductor space.

The VC gold rush speaks to the attractiveness of emerging artificial intelligence and machine learning applications. Each and every company in the space are attempting to find superior solutions for high volume applications. It is impossible for Nvidia to defend its turf in an increasingly diverse application space. With billions of dollars invested, it is fair to say that some of the emerging chip companies will carve out pieces of the machine learning market from Nvidia.

Stock Reaction & Implications

Nvidia stock fell about $6 or over 3% since the CNBC story broke on Wednesday. The stock reaction demonstrates the jitteriness of Nvidia investors to competitive news.

The Nvidia storybook valuation is built on the Company having an unassailable lead over competition and a strong moat in software infrastructure such as CUDA. This potential loss of Tesla socket to AMD shows that this moat may be overrated. This loss, if it checks out, will show that the boosters’ claims about the barrier to entry for competition in the ML/AI space, especially against GPGPU peer AMD, are overstated. Over the coming months and quarters, we can expect AMD, Intel (INTC), and several chip startups to make significant headway in a market that Nvidia currently owns.

We believe this is the new reality for Nvidia. From now on, we expect a steady drum beat of competition winning sockets in the artificial intelligence and machine learning space. And, with each such competitor win, the shine on Nvidia’s story will reduce a bit until one day when the story fails to resonate.

From a quarterly results viewpoint, we continue to expect Nvidia growth to moderate substantially within the next couple of quarters.

We believe Nvidia, while a great company, is at a significant risk of valuation compression.

Our view of Nvidia: Sell.

Before it is here, it is on the Renewable Energy Insights subscriber platform. For timely and in-depth research and analysis of solar, wind, battery, and autonomous vehicle industry stocks and developing news, please consider subscribing to our Renewable Energy Insights platform.

Disclosure: I am/we are long AMD, GOOG, GOOGL.

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.

Tesla Wins ‘World’s Greatest Drag Race’ – Tesla Motors (NASDAQ:TSLA)

Rethink Technology business briefs for September 21, 2017.

A Model S P100D wins the quarter mile against . . . just about everyone

Source: Motor Trend

Although I can’t recommend Tesla (TSLA) as an investment at the moment, I’m still a shameless fan of the company and its products, and happy to report good news when it’s available. Motor Trend recently held what it called the “world’s greatest drag race” on the landing strip at Vandenberg Air Force Base in California. And a Tesla Model S P100D with Ludicrous mode won the race.

The Model S’s competitors included some potent and fairly exotic machines such as the Ferrari 488 GTB, Mercedes AMG GT R, Aston Martin DB 11, and McLaren 570GT. The P100D often wins such drag races, where its electric motor torque pushes it to the quarter mile finish first.

But at longer distances, the Tesla usually falls behind, since it doesn’t have the top speed of the exotics. But it was still fun to see the competition trail in the wake of the mighty P100D. And races such as this demonstrate the future of the performance sedan: electric, all-wheel drive. Whatever may befall Tesla and Musk, even its detractors have to admit that Tesla has shown us the future.

Tesla’s reported custom AI chip: that’s what Keller does

Source: wccftech

Yesterday, CNBC reported that Tesla is working with Advanced Micro Devices (AMD) on a custom AI chip. Jim Keller is reported to be leading a team of “more than 50 employees” for the project. Tesla reportedly has received samples of the new chips and is now testing them.

GlobalFoundries, the fabricator that does most of AMD’s work, reportedly made the parts. CNBC says that GloFo’s CEO Sanjay Jha mentioned the work being done for Tesla, although this was seemingly denied later on.

Reuters subsequently reported an email statement from the company:

Tesla has not committed to working with us on any autonomous driving technology or product.

Of course, there wouldn’t be a commitment at this point, but this statement isn’t a denial that some work has been done. Ever since Jim Keller was hired away from AMD in January 2016, rumors have swirled around him that he was designing a custom AI chip to power Tesla’s self-driving cars. Jim is a microprocessor architect. Designing chips is what he does. It’s probably not plausible to assume that Tesla hired him for any other reason.

However, designing microprocessors from scratch is a huge, billion-dollar undertaking, something that non-technical business writers may not appreciate. My take on the rumored Tesla chip is that a collaboration with AMD was always the plan. Even with Keller and his staff, many of whom came over from AMD, Tesla wouldn’t have the resources to go it alone.

So I think it’s likely that Tesla hired AMD to design a “semi-custom” chip along the lines of the console chips. This is the somewhat mysterious “third semi-custom design win” often referred to in AMD conference calls.

Keller’s staff are overseeing the design effort and providing design input. One of those who Keller brought over to Tesla, David Glasco, is listed on his LinkedIn resume as System Architecture Lead at Tesla. But he never left the Austin, Texas, area where AMD is located.

As to the composition of the chip, many have been assuming that it contains custom ARM CPU cores. I actually think this is unlikely for a number of reasons. AMD doesn’t really have the skills to design a true custom ARM core.

But the most important consideration driving this development process for Tesla is the desire to find a lower cost solution than what NVIDIA (NVDA) has to offer. Capitalizing on the development of Ryzen and Vega seems like a good way to do that.

So my take is that the chip is probably a combination of one or more Ryzen “Zeppelin” slices, combined via the Infinity Fabric with a Vega GPU and possibly a custom ASIC for hardware tensor processing. This would take advantage of AMD’s development of Infinity Fabric for EPYC and Threadripper, and make the resultant device relatively low cost to fabricate.

Trip Chowdhry begs to differ

Barron’s reports that Trip Chowdhry of Global Equities regards the CNBC report as “100% false.” In a research note, he writes:

Comprehensive view is that TSLA and NVDA have currently a 5-year contract in place, which was renewed just recently. AMD is not a Player. AMD is DoD (Dead-on-Departure) in DML (Deep Machine Learning) Workloads. Google threw AMD out of its GPU Training Cluster, as AMD had extremely poor performance on GOOGL TensorFlow framework. Later, as a courtesy to AMD, GOOGL redeployed the AMD GPU’s for only VDI (Virtual Desktop Infrastructure). So far we have attended no less than 60 DML (Deep Machine Learning) conferences…we have not seen even a single benchmark of any DML Framework running on AMD GPU’s for production workloads. Investors optimism is completely misplaced that AMD will become a significant DML player. There is only going to be one GPU player NVDA, just like there is only one CPU Player INTC ….rest all the players will be in the others category, which will be about 10% of the market.

Insofar as AMD’s disadvantage in AI on its GPUs is concerned, I don’t doubt that Chowdhry is correct. But I doubt that would stop Keller from pursuing this project. And even though there might be a contractual commitment to NVIDIA for some period of time, projects like this require a significant amount of time. Tesla may simply be looking forward to the next generation of devices post-NVIDIA.

The area where AMD is weakest compared to NVIDIA, in software support for machine learning, is precisely why this effort may be on a multi-year development track. Getting to the point of having a hardware platform is just the start. Now the real work begins to develop the software.

While I disagree with Chowdhry on the reality of the effort, that doesn’t mean I think it’s a good idea. I’ve written previously that Tesla’s autonomous vehicle effort appears to be in disarray. Now the development of a separate hardware platform and the concomitant software effort seems like grasping at straws.

Tesla hasn’t been able to make much progress on the current NVIDIA derived platform, which is half of a Drive PX 2, and I believe, inadequate to support full self-driving capability. So Tesla has decided to go off in a completely different direction. I believe that Tesla would have been better served devoting the money and resources from the AMD effort to solving the problems it has with the current system, whatever those are.

That just wasn’t going to happen once Keller landed on the scene.

Nvidia is part of the Rethink Technology Portfolio and is a recommended buy.

Disclosure: I am/we are long NVDA.

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.

Tesla to Develop Own Chip—Will NVIDIA Suffer?

In what is seen as a blow to NVIDIA Corp. (NVDA) Tesla Inc. (TSLA) is reportedly testing its own semiconductor for self-driving cars.

Citing a source familiar with the matter, CNBC reported that the green car manufacturer has already received and is currently testing samples of its own chip. The move is aimed at reducing its reliance on NVIDIA and other companies as it continues to manufacture vehicles. CNBC noted Tesla is working with Advanced Micro Devices Inc. (AMD) and will build its own chip on the intellectual property of one of NVIDIA’s biggest rivals. The chip is aimed at getting the company closer to its vision to deliver a completely self-driving car by 2019. The internal chip program is being headed by Jim Keller, who has been in charge of the Autopilot unit since June. CNBC reported that more than 50 people are working on the profect. (See also: NVIDIA Could Make $1B From Tesla’s Self-Driving Decree: Analyst.)

Are the Reports True?

News of Tesla’s move was pressuring NVIDIA’s stock in early morning trading with shares down 2.4% or $4.41 to $181.43 a share. Meanwhile, shares of AMD were up 2.04% or $0.28 to $14.02 a share. But at least one bull, RBC Capital Markets, told investors not to overreact and to “ignore the noise” about the CNBC report. In a research note to clients, RBC’s Mitch Steves reiterated his outperform rating and $205 price target, arguing that if the report proves to be true and NVIDIA is no longer the only provider of chips to Tesla, it will still remain a leader in the market. The analyst pointed out that neither company has confirmed the news. (See also: AMD Is a Sell, Positioned to Crash 70%: Citigroup.)

“We view this announcement as similar to dynamics we anticipate to occur in the data center: AMD winning some share while NVDA remains as the de-facto standard and market share leader,” wrote the analyst in the note covered by Barron’s. “We view the announcement as a near-term concern that should not impact the story over the next several years. AMD will likely gain some share both in auto and [data center] but we believe Nvidia will remain as the market share leader.” What’s more, the analyst said he thinks Tesla will continue to use NVIDIA’s semiconductors for artificial intelligence workloads while AMD chips will be deployed for specific jobs. That way Tesla isn’t 100% reliant on NVIDIA, noted the analyst. The main read from the potential move is ultimately a positive according to Steves: “the value of AI chips is continuing to increase which is causing companies to look in some instances to dual source.”

What an AMD deal with Tesla would mean for Nvidia and Intel

AMD is cutting into Nvidia’s early lead in artificial intelligence for cars.

Jim Keller, a former AMD

AMD, -2.00%

chip designer, now works at Tesla

TSLA, -2.13%

(Keller also previously worked at Apple

AAPL, -1.39%

) The news is that Tesla is working with AMD on a new custom artificial intelligence chip at the expense of Nvidia

NVDA, -2.36%

To be sure, a Reuters report said there’s no formal agreement.

Still, let’s explore the dynamics among AMD, Nvidia and Intel

INTC, +0.40%


Please click here for an intraday annotated chart of AMD. The most important information for investors to glean from the chart is that the VUD indicator stays strongly green as AMD’s stock price traces an explosive move up. The VUD indicator is the most sensitive measure of supply and demand in real time. Typically when AMD moves up on news, the VUD indicator first becomes green and then turns orange. Green indicates net buying; orange indicates net selling. This happens because as the momo (momentum) crowd runs up the stock, the “smart money” typically sells into the strength.

Read: The YouTube channels for investors to watch now

Please note from the chart that during the latest AMD price move, the VUD indicator stayed green throughout. There was no selling by the smart money into the strength, as has been the case in the past.

Ask Arora: Nigam Arora answers your questions about investing in stocks, ETFs, bonds, gold and silver, oil and currencies. Have a question? Send it to Nigam Arora.

Not a surprise

At The Arora Report, the news is not a surprise. It was anticipated for the following reasons:

• It is a good practice for companies to not become totally dependent on chips from one company. Companies often seek to diversify.

• It is only a matter of time before Tesla will be competing with electric cars from BMW

BMWYY, +0.59%


DDAIF, +0.24%


GM, +0.58%


F, +0.17%

and Toyota

TM, -0.19%

It behooves Tesla to develop its own intellectual property. Here Tesla is using AMD intellectual property to design a custom chip. In the process, Tesla will build significant intellectual property of its own.

• AMD had a relationship with an ex-staffer now at Tesla.

Financial impact

There is no significant financial impact on AMD and Nvidia.


Stocks often move on sentiment. The news boosts sentiment on AMD. The sentiment on Nvidia should be dented, but expect the effect to be mild as Wall Street analysts rush to defend Nvidia. Many Wall Street analysts have “buy” ratings on Nvidia, hence they are predisposed to defending Nvidia.

Bad day for Intel

The simple fact that, in spite of its significant efforts, Intel was not able to get in Tesla but AMD did, makes for a bad day at Intel.

Should you buy AMD, Nvidia or Intel?

The best way for investors to decide on which stocks to buy is to focus on potential risk-adjusted returns.

In plain English, it means returns in excess of those commensurate with the risk taken. For example, Nvidia could go to $250 but it could also fall to $90. As such, risk in Nvidia is fairly high. In contrast, the probability of getting the same return as Nvidia in Intel is lower but the risk is also lower. Intel also pays a nice dividend.

Nigam Arora on Tesla: Is the stock too expensive? Try these alternatives

The Arora Report provides clear signals based on risk-adjusted potential on these stocks to buy, sell or hold including buy zones, target zones, stop zones and appropriate quantities. All of the stocks mentioned here move on the news, and sentiment can shift on a dime. For this reason, it is important for these positions to be reviewed regularly, investors to stay nimble and have protective measures in place. The objective of investors should be to consistently make money and avoid the yo-yo of making money and then losing it.

Disclosure: Subscribers to The Arora Report may have positions in the securities mentioned in this article or may take positions at any time. All recommended positions are reviewed daily at The Arora Report.

Nigam Arora is an investor, engineer and nuclear physicist by background, has founded two Inc. 500 fastest-growing companies, is the developer of the adaptive ZYX Global Multi Asset Allocation Model and the ZYX Change Method to profit from change in trading and investing. He is the founder of The Arora Report, which publishes four newsletters. Nigam can be reached at Nigam@TheAroraReport.com.

Tesla Worker Says Timing of Firing Denied Him Lucrative Shares

A former Tesla Inc. factory worker filed a lawsuit alleging the automaker fired him a day before his one-year anniversary, denying him hundreds of thousands of dollars worth of stock options that he claims should have vested.

Stephen Platt — who said he began working as a machinist at Tesla’s Fremont, California, factory on Aug. 27, 2012 — was fired on Aug. 26, 2013, at the end of his shift despite having been told the previous month he’d be getting a raise for his performance, according to the complaint filed in California Superior Court in Oakland last month.

Platt had been offered 2,500 shares of Tesla common stock when he accepted the job, a quarter of which were slated to vest 12 months after the first day of his employment. Platt alleges he had completed exactly a year of work on the date of his termination and should have been allowed to purchase the shares.

A spokesman for Tesla didn’t immediately respond to requests for comment on the lawsuit.

Under the terms of his employment agreement, Platt would have been able to purchase his vested shares for $27.37 apiece after a year of work, according to the complaint. The day he was terminated, Tesla closed at $164.22 in New York. Had Platt’s options vested, he would have been able to acquire 625 shares at a fraction of the market price — shares that would now be worth more than $240,000 at Monday’s settlement.

“Tesla is cheating its employees out of stock options that they are entitled to, and they are worth a significant amount of money,” Yosef Peretz, Platt’s San Francisco-based attorney, said in an interview. “It’s a straight-up breach of contract case. The employment agreement says you vest after 12 months, he completed 12 months, and he should get his stock options.”

The plaintiff is seeking class action for all employees who joined the company under the same terms, which the suit estimates to be at least 200 former workers. Peretz also represented Tesla co-founder Martin Eberhard in his defamation suit against Chief Executive Officer Elon Musk, which was settled in 2009.

Palo Alto, California-based Tesla, which makes electric vehicles and energy storage devices, has ballooned in both size and market value since its June 2010 initial public offering. It posted its first quarterly profit during the year Platt was employed and has since become the largest U.S. automaker by market capitalization.

Many Silicon Valley startups give employees shares that fully vest over a four-year period, with equity in fast-growing companies a critical part of compensation packages. This spring, cyber-security startup Tanium Inc. was roiled by allegations that its CEO kept a list of workers who were close to cashing in their options and firing them before they could do so.

As the company ramps up production of its new Model 3, some workers at the Fremont factory have draw attention to wages and working conditions and have said that Tesla needs a union. Platt said in the lawsuit that he experienced breathing difficulties due to inadequate ventilation in the area where he worked, though an occupational health physician cleared him to work without restrictions in the month of his dismissal.

The case is Platt v Tesla Motors Inc., RG17873032, California Superior Court, Alameda County (Oakland).

Tesla: As CounterGate Ends, APGate Emerges – Tesla Motors (NASDAQ:TSLA)

Tesla (NASDAQ:TSLA) likes to make problems real hard. For instance, while Tesla has lots of cameras, it doesn’t even try to have pairs of same-spec (same lens, same field of view) cameras pointing the same way. They all have different fields of view, so a problem that’s already hard (reliably re-creating the world in 3D from camera vision alone) is made harder still.

A while ago, I covered the emerging CounterGate saga in my articles titled “A New Tesla Scandal Is Brewing” and “Tesla’s Countergate II – The Sequel.” In that saga, Tesla tried to covertly nerf customers’ car performance. Tesla did this, quite obviously, to limit possible warranty claims down the road stemming from the usage of the cars’ full power.

The CounterGate scandal had two acts:

  • First Tesla tried to limit the power on Performance models after a certain number of full-power uses. Then, having had that come to light, it relented saying it would no longer limit power.
  • However, it still did – just in another fashion. The second act was thus Tesla trying to limit the use of full power to certain restrictive instances (launch mode usage).

Well, this scandal is now over. Tesla got sued by one of its customers. Tesla still tried (includes news of the lawsuit, attempted NDA resolution, etc) to use its old trick of offering to fix only that customer’s car and then putting him under a NDA (Non-Disclosure Agreement). The customer did not accept the solution, and finally Tesla quit and simply accepted to restore the performance to all its customers’ cars (as demanded by the heroic owner). So that was the end of the CounterGate scandal.

With this scandal being over, today I am going to talk about yet another emerging scandal. That will be the APGate scandal.

APGate, The Origins

The APGate scandal begins in October 2016, with Tesla’s introduction of HW2.0 (Hardware 2.0) to implement AutoPilot 2.0. This hardware, as per Tesla, was ready to provide full self-driving.

A bit of context. Some might already be qualifying what this “full self-driving” ultimately means. Elon is quoted as it being Level 5, so fully autonomous under all conditions. But what Elon says and what Tesla sells are sometimes different things, so what does Tesla’s order page actually say? Here’s a screen grab (highlight is mine):

Notice the following: There are two instances (when the car seeks parking, and when it is used in a supposed Tesla Network) where the car self-drives while empty. This removes all doubt. When empty, the car needs to be able to face any possible circumstance. What this means, is that the level being implicitly promised is SAE Level 5 self-driving. This claim was present at all times starting in October 2016.

Now, Tesla doesn’t have self-driving technology. So Tesla is selling something it neither has, nor can be certain of having in the future. Tesla never promised any timeframe for having such a technology, but did mislead customers both with videos of Teslas supposedly self-driving, and promising an autonomous trip from California to New York during 2017– which I already explained wouldn’t prove anything.

Furthermore, testing data shows Tesla lags severely behind other companies seeking to deliver self-driving technology. This lag isn’t surprising. Tesla started late, and is trying to solve a more difficult problem, because while self-driving leaders use a combination of LIDAR+cameras+radar, Tesla chose to leave LIDAR out of its hardware suite (given the present cost).

In removing LIDAR, Tesla removed the most reliable way to currently detect the environment around the car. A car equipped with LIDAR can get a high degree of certainty regarding all relevant objects around it (even before recognizing those objects). A car equipped with cameras, can’t. It sees the world, but that’s different from actually detecting and measuring the distance to all objects around it.

Tesla, obviously, is trying to recreate the information given directly by LIDAR using cameras and computation. This task has been researched, and is possible to be performed (and indeed, used in some instances). However, the reliability in performing this task isn’t necessarily high enough for self-driving duties. The problems emerge both because some objects won’t be reliably detected, and because objects will be detected which don’t exist or aren’t in the car’s path (false positives).

In self-driving, false positives are a large problem. A self-driving car needs to turn conservative when faced with a possible solid object it might hit. If this object is “imaginary” (a false positive), the car will brake – sometimes brake hard – for something that isn’t there. In many instances, this will create a safety hazard for other cars in the road. Unexpected braking will be met by slower reaction times from human drivers, and crashes resulting from it are to be expected.

Now, on the issue of false positives, we already have at least two reasons to believe they’re affecting Tesla:

  • The fact that AP2 cars often brake for underpasses and trees (among other flaws) when using TACC (Traffic Aware Cruise Control) or AutoPilot (TACC+lane keeping).
  • And the fact that Teslas, when subjected to pedal misapplication, can happily run from a standstill into a building (see my prior Tesla article for examples).

In the first case, the car is seeing a “false positive” object in its path, and not ignoring it. Thus, it unexpectedly brakes. In the second case, the car is certainly getting a large radar signature from the building. However, in trying to ignore all those false positives which happen elsewhere, the car ignores the giant building straight ahead and plows into it.

Now, both of these cases also highlight something else: They highlight that at present, Tesla hasn’t yet solved the problem of using cameras to reliably map the surroundings. Had Tesla done that, and neither of those cases would have happened.

That Tesla hasn’t yet solved how to fully detect its environment using cameras, tells us just how far from FSD Tesla still is. Detecting the environment isn’t the end point of a self-driving effort. Instead, it’s the starting point. When using technologies like LIDAR or Flash LIDAR, the very first thing a self-driving developer gets is a reliable 3D picture of its (solid) environment. Using this picture alone, a startup self-driving effort would neither brake for overhead signage, nor accelerate into giant buildings from a standstill.

It’s not a coincidence that Google cars haven’t yet hit any other cars. Even when Google’s car got in an accident where it was at fault, the Google car was hit (scraped) by a bus, not the other way around. This happens because the Google car has a 3D picture of its solid environment, and thus can avoid hitting (but not being hit) literally anything almost by definition (there are exceptions with transparencies, small objects, etc).

Now, with this description up to here, we already have a good clue about where Tesla stands on FSD, and how far it is from actually delivering on its promises. It’s already a scandal that Tesla is selling something it cannot deliver on any specific schedule. Indeed, Tesla is selling something it might not be able to deliver using its present hardware suite at all. It’s also rather interesting that even Elon Musk himself puts the timeframe to Level 5 about 2 years away (counting from late April 2017…), and Elon is usually late. It’s interesting because Tesla cars often get sold on 2-year and 3-year leases.

But the scandal I’m bringing you today is rather different. Instead, the scandal has to do with the fact that all the FSD promises above were made on a hardware suite (HW2.0), which Tesla has already obsoleted by launching HW2.5. “Oh but that was already known,” you’ll say. Indeed it was, Electrek broke the story a month ago.

At the time, to soothe fears, Tesla said it would upgrade HW2.0 customers to the improved computer on HW2.5, if need be (Source: Electrek article linked above):

“However, we still expect to achieve full self-driving capability with safety more than twice as good as the average human driver without making any hardware changes to HW 2.0. If this does not turn out to be the case, which we think is highly unlikely, we will upgrade customers to the 2.5 computer at no cost.

Then, more recently, HW2.5 customers got notice that their AEB (Automatic Emergency Braking) feature would stop working for a few weeks, as it re-calibrates the new hardware.

It was the AEB going out which got me digging a little. You see, computers, even more powerful computers built on the same architecture, don’t need any re-calibration. So something else had to be amiss. And indeed it was.

At the time of the HW2.5 introduction, Tesla downplayed the changes HW2.5 brought (Source: Electrek article linked above):

“The internal name HW 2.5 is an overstatement, and instead it should be called something more like HW 2.1. This hardware set has some added computing and wiring redundancy, which very slightly improves reliability, but it does not have an additional Pascal GPU.”

With this description, it seemed that all that was changed was a bit of cabling for redundancy and reliability, and then the computer unit which could be upgraded for free down the line (as explained by Tesla in the first quote). But this is misleading, because (as discovered by enterprising Tesla owners):

  • Redundancy itself might be a regulatory requirement for FSD in the future.
  • And most importantly, wiring and the computer were not the only changes. Along with those changes, Tesla also added a new radar. Previously Tesla used a Bosch MRREvo14f, and now Tesla has changed this radar for a more capable Continental unit. This new radar, is the reason why HW2.5 cars now need re-calibration. And once the re-calibration ends, a surprise will emerge: the new radar is more capable, so these cars will perform better than the HW2.0 cars already.

As a result of these changes, the car is already getting more capable computers and sensors. Also, still more hardware changes still are oncoming. For instance, the Model 3, using mostly the same HW2.5 suite, already includes an interior camera. It won’t stop there.

The official Tesla line is that the original hardware will still be capable of the promises. However, faster and more capable hardware would always make for an easier FSD problem. As a result, at the very least a FSD solution would arrive on the more capable hardware first.

Now, given what we saw regarding the state of Tesla’s FSD development (both from testing data and from the fact it doesn’t have a reliable 3D view of the world), for now neither hardware solution (HW2.0 or HW2.5) is likely to attain FSD. But Tesla continuing to add to its hardware solution will, at some point, fully obsolete HW2.0 (and even HW2.5). In the meantime, tens of thousands of cars will be carrying that obsolete hardware together with a promise that it will someday be enough for FSD Level 5.

This development is an even larger liability than CounterGate ever was. This touches every car Tesla sold since October 2016. This touches even the cars which did not pay for the FSD option, because the capability to have FSD, even if bought later, could arguably have been at the core of the buyer’s decision. In the end, what’s happening to Tesla with the HW2.0 APGate, is no different from what happened with Volkswagen (OTCPK:VLKAY) and the emissions scandal. VW (and Tesla) sold cars based on capabilities and specs they did not have. And VW ultimately had to pay compensation and even buy back cars from owners.

A Side Note

As expected, Tesla has been on a tremendous discounting drive to improve Q3 2017 deliveries. This discounting drive included up to $30,000-$40,000 discounts on Model S P90DLs, P100Ds, as well as other discounts across the range.

There are also customers which managed to buy new 90Ds and P90Ds with leases as low as $600 per month. To put this in context, a new base Model S75 RWD, costing $69,500 in cash, on the cheapest terms possible (36 months, 10,000 miles/year) goes for $790/month.

This discounting wave also introduced something that’s remarkable. You see, Tesla discounts “discontinued” models more. Thing is, many of discounts were on “discontinued” 90Ds which were coming new off the line, being built during the quarter!

Tesla guided for a 20% drop in automotive gross margins (from 25% to 20%) for Q3 2017. Tesla put it down to starting up Model 3 production. However, it’s evident that a lot of the gross margin drop is coming both from including more equipment in base models, and extreme discounting (as predicted).


The CounterGate, so many times denied by Tesla fans, was not only true but has been solved by Tesla capitulating on it. As a result of that scandal, Tesla ended up spending some of its best customers’ goodwill. It will also be eating higher warranty claims in the future, because of the now unrestricted use of those cars’ performance.

The new APGate stands to be an even larger scandal. Already Autopilot hardware is evolving beyond the initial specification, promised to be enough to attain self-driving SAE Level 5. Ultimately, and with this continuing hardware improvement process, HW2.0 (and even HW2.5) will be fully obsoleted.

At that point, Tesla will have created a gigantic liability for itself, since it sold tens of thousands of cars which it said would be able to perform to SAE Level 5 self-driving duties, and this will never be attained on the original hardware. Neither will the hardware be retrofit-able, as what’s being changed goes well beyond just the CPU box.

On top of that, this development will see further customer goodwill being wasted even before the lack of FSD capability realization takes place. For instance, right now AP2 customers have cars which don’t perform to AP1 levels, even though such parity was promised for as early as December 2016. Moreover, with the new hardware improvements it’s highly likely that Autopilot features on HW2.5 cars will quickly surpass those on HW2.0 while HW2.0 cars don’t yet get their promises fulfilled. This will be more evident on TACC, because of the new radar. It’s my estimate that the new radar will quickly reduce the false positives which lead to severe braking out of the blue on HW2.0 cars. At that point current customers will still be stuck with the worse-performing TACC on their “FSD cars” while HW2.5 customers will no longer have to endure it.

Tesla’s most important asset is the brand it built. Yet, with these scandals and unfulfilled promises it’s burning a bit of that brand every day. This is already visible in Tesla Motor Club’s forums, where many former Tesla fans are growing every more skeptical. Tesla is slowly destroying its most valuable asset.

Disclosure: I am/we are short TSLA.

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.

NTSB Finds Tesla Autopilot Partly to Blame for Fatal Crash

The U.S. National Transportation Safety Board has concluded that the crash that killed the driver of a 2015 Tesla Model S electric sedan in Florida last year was at least partly due to the limitations of “system safeguards” on the vehicle’s Autopilot semiautonomous feature.

According to Reuters, NTSB chairman Robert Sumwalt said: “Tesla allowed the driver to use the system outside of the environment for which it was designed and the system gave far too much leeway to the driver to divert his attention.”

Autopilot is designed to control the steering and speed of a vehicle driving on a highway with exit and entrance ramps, well-defined medians and clear lane markings. Since it’s not intended to have full self-driving capability, the system alerts the driver repeatedly with visual and audible warnings to pay attention and keep his or her hands on the steering wheel.

But in January, both the NTSB and the National Highway Transportation Safety Administration determined that Joshua Brown, the driver of the Model S, had set the vehicle’s cruise control at 74 mph (higher than the 65-mph limit), was not driving on a controlled-access highway and ignored the system’s warnings to remain alert.

So when a semitruck turned left across the path of Brown’s vehicle, the Autopilot system failed to respond because it’s not designed to detect crossing traffic, and the driver did not apply the brakes or otherwise take control. As a result, the Model S crashed into the side of the truck, killing Brown instantly.

At the time, NHTSA concluded that the vehicle had no defects and that Autopilot had performed as designed. And NTSB attributed the crash to driver error.

Now, however, NTSB says that Autopilot’s “operational design” was at least a contributing factor to the crash because, as configured at the time, it allowed drivers to keep their hands off the steering wheel and otherwise let their attention wander from the road for extended periods of time. In other words, drivers can override or ignore warnings from the system, putting them at risk for collisions.

NTSB has devised a number of recommendations for automakers developing partially autonomous vehicles. These include going beyond simple alerts to ensure driver engagement, blocking the use of a self-driving system beyond the limits of its design, and making sure these systems are only used on specific types of roads.

Tesla responded that it would evaluate the agency’s recommendations and “will also continue to be extremely clear with current and potential customers that Autopilot is not a fully self-driving technology and drivers need to remain attentive at all times.”

Tesla has continuously updated Autopilot since its introduction. For example, the latest version doesn’t just give warnings; it will shut off completely if the driver doesn’t take control of the wheel.

Although the Tesla Autopilot crash prompted continued NTSB scrutiny, the agency stressed that its recommendations apply to other automakers as well. It specifically mentioned Audi, BMW, Infiniti, Mercedes-Benz and Volvo, suggesting that their semiautonomous systems should also receive upgraded warnings and features that prevent drivers from using them improperly.

There’s More Crossover Between Tesla and SpaceX Than Most Realize


Tesla and SpaceX CEO Elon Musk


One of the most respected Apple analysts, Gene Munster, now covers Tesla.

In a fascinating Bloomberg* report, Munster explained, “In this race to disrupt the world with both electric cars and autonomy as well as space, you don’t really work for Tesla or SpaceX. You just work for Elon Musk. You have the most wicked-smart people who can feed off of each other all working for Elon, and he can call on them to help crack various problems.”

Another Wall Street analyst with Robert W. Baird, Ben Kallo, explains: “SpaceX can contribute to what Tesla is doing. There’s a lot of crossover, and it gives Tesla a complete advantage over other automakers.”

Bloomberg also reports that analysts like Morgan Stanley’s Adam Jonas can’t help but wonder if a company that ultimately plans to build and launch its own satellites might have a leg up in the race for driverless cars, which will have to be connected to a vast wireless network. It’s an edge that other carmakers like General Motors Co. and Toyota Motor Corp. won’t have.

Elon Musk also addressed the unique synergies between SpaceX and Tesla on a recent earnings call:

“That’s cross-fertilization of knowledge from the rocket and space industry to auto back and forth, as I think it’s really been quite valuable.”

What’s a good example of this cross-company collaboration between SpaceX and Tesla? It turns out that engineers at Tesla found a quality problem earlier this summer with a cast aluminum auto part that was taking hours to diagnose and fix. They were stumped, so they called in the rocket scientists — literally.

*This article comes to us courtesy of EVANNEX (which also makes aftermarket Tesla accessories). Authored by Matt Pressman.

Tesla’s engineers reached out to their counterparts at SpaceX, who recommended the use of ultrasound sensors to isolate the problem. The solution saved Tesla about eight hours of work per car, an eternity on an assembly line aiming to ramp up to mass-market volumes for Tesla’s new Model 3. A Tesla spokesperson explained, “Given that Tesla and SpaceX are totally non-competitive and have a similar first-principles approach to problem-solving, employees at one company are occasionally able to share ideas that help the other.”

Other synergies exist between the two companies. Software was developed jointly by SpaceX and Tesla to manage massive amounts of data at both companies. And both cars and rockets have to stay trim and light to get where they’re going, making material science another key area where the two companies collaborate. Therefore, the materials teams at both Space X and Tesla often hold joint meetings to discuss materials issues.


Faux movie art of Elon Musk, CEO of SpaceX / Tesla and his on-again-off-again girlfriend, Amber Heard (Source: oneksy)

“As an industrial community — whether it’s aerospace or automotive — everyone is grappling with increasing data management and the search for stronger, lighter, cheaper materials,” Luigi Peluso, an aerospace, and defense consultant at AlixPartners explained. “People who can master those skills can play in either domain pretty fluidly.”

Tesla has more than 33,000 employees and SpaceX has roughly 6,000 — giving Musk a vast talent pool to draw from in order gain a competitive advantage in each of their respective industries.  A Tesla spokesperson elaborated, “This [collaboration] hasn’t been a major thing, but it’s still always nice to be helpful, especially given the shared respect for each company’s mission… It’s not unusual to see people at Tesla gathered around their computers to cheer on SpaceX launches, and lots of SpaceX employees drive Teslas.”


*Source: Bloomberg

*Editor’s Note: EVANNEX, which also sells aftermarket gear for Teslas, has kindly allowed us to share some of its content with our readers. Our thanks go out to EVANNEX, Check out the site here.

Are legislators penalizing Tesla? | Guest Perspectives

California’s last big auto manufacturer, a General Motors-Toyota joint venture, closed its Fremont factory in 2010. This year, Toyota is vacating its North American headquarters in suburban Los Angeles as it decamps to Texas.

Yet in the final week of this legislative session, lawmakers inserted language that takes aim at Tesla Motors, the upstart that builds its electric vehicles in the Fremont factory that Toyota and GM abandoned.

As legislative language goes, it’s mild. But it suggests that an automaker that runs afoul of the fine print could lose out on $140 million in rebates issued to consumers who want to help fight climate change by purchasing zero-emission vehicles.

The amendment directs the Air Resources Board to work with the Labor and Workforce Development Agency to “develop procedures for certifying” that companies that make autos that qualify for rebates are “fair and responsible in the treatment of their workers.”

Further, the Legislature’s “intent” is that the state labor secretary will “certify (auto) manufacturers as fair and responsible in the treatment of their workers” before companies’ vehicles qualify for the rebate program.

Legislators are within their rights to tie rebates to state-only standards. But the language, part of broader legislation, Assembly Bill 134, to divvy up $1.5 billion in cap-and-trade revenue, emerged in the final week of the legislative session, rarely a good sign. Why not include it in separate legislation, subject to full legislative review?

Although the bill is aimed at Tesla, the Alliance of Automobile Manufacturers, which represents Ford, GM, Toyota, Fiat and others, opposes the language. The Global Automakers, which represents Nissan, called the amendment “counterproductive to building a sustainable market for zero emission vehicles.”

Counterproductive or not, Gov. Jerry Brown, the internationally recognized champion of the fight against climate change, blessed the deal, The Sacramento Bee’s Jim Miller reported.

We support treating workers fairly and responsibly. We believe Tesla probably would gain market share if it attained labor peace. But what exactly constitutes “fair and responsible” treatment? Why limit the standard to car makers? Why not include manufacturers of zero-emission buses, and bio-digesters, and any company that receives cap-and-trade revenue?

And why should lawmakers step into the middle of a labor dispute with Tesla?

If the language becomes law, California would need to enforce it equally, whether zero-emission vehicles are made in Fremont, or Michigan, or right-to-work states, or foreign countries. How state regulators here would enforce workplace standards in far-flung plants remains to be seen.

As is their right, Tesla and its founder Elon Musk are fighting the United Autoworkers’ organizing effort. The National Labor Relations Board detailed Tesla’s rough tactics in a complaint two weeks ago.

Most definitely, Musk rubs some officials here wrong. His company benefited mightily from California subsidies and from consumers’ green attitudes, but Musk located his battery factory in Nevada three years ago.

Nonetheless, Tesla employs 10,000 Californians in Fremont, unlike Toyota, GM, Ford, Nissan and all the others without factories here. It’s as if lawmakers are penalizing Tesla for operating in this state. They wouldn’t do that. Would they?

NVIDIA Tesla V100 tested: near unbelievable GPU power

AMD might have just launched their new Vega GPU architecture with a slew of Vega-based products (Radeon Vega, Radeon RX Vega, Radeon Pro WX, and Radeon Instinct) but the real king is NVIDIA’s now months-old Volta GPU architecture.




We don’t hear much about NVIDIA’s Volta GPU architecture because it’s still a while out from finding its way into consumer GeForce graphics cards, but the supercomputer/AI/deep learning markets are now receiving their new Volta-based Tesla V100 accelerators which means… BENCHMARK TIME!




First off, let’s look at the difference between the previous-gen Pascal-based Tesla P100 and the new Volta-based Tesla V100. Starting off with 12x more deep learning training performance, with 10 TFLOPs on P100 up to a freakin’ is-it-real 120 TFLOPs of ‘DL training’ on V100.


NVIDIA has some huge memory bandwidth numbers on Tesla V100 as well, with 900GB/sec available – up from 720GB/sec on Tesla P100. NVLINK 2.0 is also featured, throwing the internal bandwidth up from 160GB/sec to a huge 300GB/sec (1.9x) while we have 10MB of L1 cache, up from 1.3MB on Tesla P100 (7.7x increase).


The new NVIDIA Tesla V100 has been tested on single-core Geekbench 4 compute tests, with an out-of-this-world score of 743,537… the next one close to that is the P100-based system with just 320,031 in comparison.


Even HP’s impressive Z8 G4 workstation PC is only capable of 278,706 points, and that system rocks 9 x PCIe slots with Quadro GP100 cards inside.


All-in-all, NVIDIA’s new Tesla V100 is a compute MONSTER and nothing else on the market begins to compete. AMD is radically behind here, until their new Vega-based Radeon Instinct graphics cards begin shipping – NVIDIA continues to reign supreme.