LIKE AN APPLE: Microsoft will design their own ARM chip based on an ARM Mali graphics core. WinRT will run on a dedicated Microsoft Windows Processor (No more Qualcomm, Nvidia fragmentation). TBA '13.
When Jen's boss -- the nvidia Board of Directors -- asks him how the tegra is doing,
does Jen show them a graph of the tegra's quarterly losses,
Logarithmic charts -- it's like Magic, right boys?
Makes everything look not so bad.
ah.... absolutely positively cannot owe up to the $180M loss last Q due to Tegra... don't blame you. As a long, must be pretty tough to see those losses balloon out of control. Good luck, you'll need it.
Where is it? Where do you see it? You've mentioned the 10-K, then after I pointed out the actual figure listed for the CPB segment's operating loss, I got weird replies acting as if page 28 of the 10-K provides the $180M figure you're quoting, but it doesn't. Nor does any section of the 10-K, unless you've found a page that my copy is missing.
So again, where is the figure from? Made up? You've mocked me as though I haven't gone through the most recent 10-K... but I'd challenge you to provide any source that actually supports your claim that Tegra lost $180M last quarter. I do notice you slightly rewording your original claim to sound a bit less inaccurate, but nonetheless it simply isn't true.
Good luck to you as well. It'd be great if we could debate the information that's actually in the 10-K, or any real objective info -- rather than a consistent reference to a made up "$180M loss last Q due to Tegra", which is false and makes it hard to engage in discussion of the actual data.
And, honestly, it just seems odd to me that you wouldn't use the real numbers, since they seem to support your premise well enough.... CPB had an operating loss just under $90M last quarter in spite of ~$180M in sales. The segment was definitely far from profitable. Shorts seem very confident that the accelerated sales in Q3 (everyone seems to agree CPB will be much higher Q3/Q2) and thus, if your argument is generally valid (though still false on specifics) then we should see Q3 CPB losses grow at a similarly accelerated pace, in Q3, right?
$312,158,000 __ CPB (Tegra) revenue for the 6 months ending July 29, 2012
($191,641,000) __ CPB (Tegra) Losses for the 6 months ending July 29, 2012
Does that mean that's Jen's "let's make commodity ARM chips" Tegra folly
actually Lost 100% of their $312,158,000 of CPB (Tegra) sales
-- PLUS they Lost another ($191,641,000) more?
My my my.
That's an Uncommonly Unprofitable business.
If nvidia is losing such a Horrendous amount on money
on making and then selling those Tegra chips well below their cost,
why on Earth don't they just Kill the Tegra?
That's what I would do, Master.
but the R&D for the CBP segment includes Tegra and Icera, and Tegra is the line that will product Denver. Part of the Icera R&D is towards developing discrete baseband products, too.
And to be clear, all of the "best" accounts are confirming that there is in fact no line in the 10K that attests to the claim that Tegra lost $180 over the quarter, right?
Where do you see that specific figure? I do believe that Tegra is losing money currently. Jen-Hsun Huang has hinted on CC's in the past that CPB needs to do ~$1B annually to be profitable -- which I think implies they need to sell ~40M SoCs annually from the Tegra line. Q3 should show a lot more progress getting there than Q2 did, considering that Q2 barely captured the limited launch of the Nexus 7.
the problem is not SF&A, the problem, once again, the run-away R&D spending in the Tegra division. Spending that has no relation to the reachable TAM and which is not even showing any returns - where is Wayne?!!
"If selling Tegra at scale actually hurts Nvidia's profitability"
that is NOT a problem, T3 has enough gross manufacturing margins. The problem is the run-away R&D spendings that is doing the heavy duty damage, like $180M loss in the Tegra division last Q when sales went up $30M+.
See this specifically seems to be inaccurate. Where do you come up with the "$180M loss in the Tegra division last Q" ... it's not what the 10-K says. Where do you see R&D broken down by business segment? Again, it's no in the 10-K
but sales this Q should go up an incremental $70-90M, perhaps more. If either Nexus 7 or Surface RT become hits into the holiday season, Q4 could show a similar gain.
One key point is that with higher volume design wins in production this quarter, SG&A as a percent of CPB revenue should be much lower -- which should help the margins.
That "$900M in guaranteed future payments from Intel" of which you speak?
You DO realize, I hope, that Intel's LAST annual payment of $300 million is coming up
really quick. January 15, 2013 to be exact.
Then there's only three (3) little annual payments of just $200 million left.
January 15, 2014 / January 15, 2015 / and January 15, 2016.
Then Intel is Done paying for the use of ALL of nvidia's Existing patents,
and ALL Future patents that nvidia receives thru March 31, 2017.
With No additional Intel payments to be made to nvidia -- like Forever.
And Intel will make rapacious use of those nvidia patents, until the very last one expires.
Probably killing-off a few more of nvidia's businesses.
Like the Tesla.
While the $200 million annual handout from Intel might Look like a big deal
in your pocket or mine, or even whosunw's pocket -- and she's filthy Rich --
it really isn't.
Oh no no no.
Not with Jen's CPB (Tegra) Division on track to LOSE over $400 Million per year.
The Tegra HAS INDEED lost ($191,641,000.00) for the six (6) months ended
July 29, 2012 -- right boys?
It sure has.
And the Tegra losses are Growing Exponentially, quarter after quarter after quarter.
So losing ($400,000,000.00) PER YEAR will be a walk-in-the-park for the Tegra.
In short, the Tegra will be Losing TWICE (2X) the amount of the annual Intel handout.
Yes, just blowing the cash.
TWICE as fast as it's coming in.
And blowing it Badly.
Did I mention?
While the $200 million annual handout from Intel might Look like a big deal in
your pocket or mine, or even whosunw's pocket -- and she's filthy Rich --
it really isn't.
Not with The Tegra around.
So $3.26B in net cash currently, and $900M in guaranteed payments -- like I said. Take the NPV of the future payments if you'd like -- it still leaves an enterprise value less than $4B, for a company doing about $850M in FCFO. I don't think the valuation makes any sense, and certainly don't understand the short thesis at this valuation.
cash in company's coffers is worth a whole lot less of cash in your or my pocket. Reasoning is simple and nvda is a good example of it. The company can easily waste that cash, like nvda is blowing through hundreds of millions at a time 'investing' in a market, which when you strip out the unreachable part like aapl and samsung, is really a very small market. That reachable TAM will never suupport such an overblown R&D program.
nvda is the only company losing money in app processors. mediatech, qcom, brcm, aapl, samsung are all making money with their SoC programs. nvda's program does not have a chance.
Not sure why you keep harping on BOM cost. BOM cost is only relevant for loss calculations when they have to dump their initial built because of no sell through. It is also relevant when msft is charging their msft tax.
It's about market positioning. msft has decided to not fight the challengers heads on, abandon the mass market, and concentrate on their version of 'ultrabooks'. And there goes the franchise.
No more wintel tax... go, el goog GO!