Revenues for Q3 will be flat to down for Q3, coming in below the +2-8% increase promised. More importantly, guidance for Q1 will flat to down relative to Q4. That will be enough to send AMD down 35-45% following earnings.
TSMC let the cat out of the bag in December, advising that AMD had scaled back console chip orders on "seasonally weakening demand". Got it? Demand for console chips in the 1H of 2014 will be lower than it was in the last half of 2013.
Since launch, Sony and Microsoft combined have managed to sell 7.2 million next gen consoles. Chips for those consoles were supplied by AMD in Q3 and Q4 of 2014.
In Q3 AMD reported that its graphical products division revenues more than doubled "due in large part" to sales of next generation console chips that AMD classifies as graphics product rather than computing products for reporting purposes. Revenue in Q3 for the graphical products division was $671 million.
"Graphics and Visual Solutions segment revenue was $671 million, more than double last quarter’s level, driven by shipment of game console, semi-custom SoCs."
If more than half of the revenues for that division were based on console chip sales to Sony and Microsoft, that could mean $340 million to $380 million. For sake of argument let's pick something in the middle; $365 million.
Certain AMD analysts believe that AMD's ASPs on console chips are around $100 each, and that it's margin on those chips is just 15%, yielding $15 per chip in income. If the ASP estimate is correct, it means that in Q3 AMD sold 3.65 million console chips to Sony and Microsoft.
That means that in Q4 it sold only another 3.55 million units to make up the balance of the 7.2 million consoles shipped by the end of the year. Got that? AMD sold slightly fewer console chips in Q4 than it did in Q3. Think about that for a minute.
Now, with that in mind, think about the TSMC announcement - that orders for 1H 2014 from AMD for console chips have been cut back on seasonally weakening demand. Makes sense. More "toys" are sold in Q3 and Q4 for Christmas than during any other time of year.
So AMD will sell fewer console chips in Q1 and Q2 than in Q4. Revenues are declining.
In the 2012-2013 time frame AMD lost all of its Apple business (it lost iMac and MacBook Pro GPU placement) and managed to hang on to only one single placement which it already had, the low volume Mac Pro.
AMD supplied fewer console chips in Q4 than it did in Q3, too, so gaming revenue will be flat to down compared to last quarter while Apple revenue is now the tiniest fraction of what it was previously when AMD had all of Apple's GPU buisness, including the high margin, high volume iMac and MacBook Pro business.
Gaming is spent. Apple represents less not more business, historically. You're an imbecile.
Well said. "Unbelievable" is exactly right. Much like last year's presentation where we were told that 2013 would see AMD Tehmash tablets everywhere, the pledges this year were equally improbable, and given that the pre-earnings run up appears to be stalling, apparently once again no one believes AMD.
It's not so much a "chip market recovery" as resumed growth in certain segments where MU's NAND flash are in demand. Those are tablets, smart phones and automotives. AMD has zero presence in any of those segments, though its competitors, Apple, Intel, Samsug, QCOM and nVidia are strong in one or more of the rebounding sectors driving MU's stock price appreciation.
No. However Q3 wasn't a "blow out". Guidance for Q4 was weak. There was clear evidence of Q3 channel stuffing. You know, the exact same kind of launch quarter channel stuffing AMD is about to be sued for? So flatish revenues after a single quarter of revenue growth says all kinds of bad things about the company's business prospects. Notice they haven't come out to raise guidance for Q4. Under Red FD were they going to surprise materially to the upside, they would have to pre-announce. This means that they'll come in somewhere in line with the feeble guidance given at the end of Q3.
Then there's all kinds of evidence that console APU sales are in freefall. TSMC told us that AMD cut back orders for console chips citing weakening demand for the chips. Console sales data shows demand in the post-preorder phase to be considerably less than the pre-order demand. The money AMD makes on console chips is a joke. It takes 2 console chip sales to replace just one lost low end laptop APU or desktop APU or CPU sale in income. In the last week before Christmas we saw slow selling Wii U outsell even slower selling XBox One. Wii U is a weak seller. That makes XBox one a pitiful seller. Now that console chip sales are cooling off rapidly, AMD has nothing to replace them with. Instead we see a barrage of silly AMD marketing cherry picking benchmarks and ignoring massive vices like massive watt sucking and feeble CPU performance. They do this kind of stuff all the time. It never works. The market recognizes AMD products for the garbage they are.
Then there's the debt. AMD is swimming in it. AMD is still burning cash. It has just a billion dollars left. It owes $200 million to GF in Q1 for renegging on AMD's purchase obligations last year. Then in 2015 it owes a big check for $2 billion. Now how is a company that is barely making a profit and still burning up cash going to come up with two billion dollars in twelve months to pay off its debt? It's not.
Fantards are wrong this quarter too.
The dipchits ran it up to the mid-4s in Q3 only to take a dollar buttreaming following earnings.
They've now run it up to the high 3s and are about to get another coal-chute expansion following Q4 earnings.
AMD will come in light. Worse still, on weakness in console APUs, it will be forced to guide down for Q1. This coupled with the $200 million penalty payment due to GF in Q1 will mean red ink and losses in Q1 for AMD. The return of losses for the company will drop it to the 2s for certain.
AMD's historical GPU margins are half that of their CPUs and APUs for PCs and laptops. 20% versus 40%.
GPUs for AMD are a very low margin business, actually. Before AMD lumped console APUs into the Graphics Product division, that division was solely GPUs and we have unquestionable data on margins for GPUs versus CPUs and APUs which at the time were in the other product division.
From the Q2 2013 CC transcript:
"Graphics and visual solution segment revenue was $320 million, down 5% compared to the prior quarter, primarily due to lower game console royalties. Graphics and visual solution segment operating income was breakeven compared to $16 million in the prior quarter."
Operating margins on GPUs are around 20% compared to 40% for CPUs and APUs for PCs, laptops and servers.
Nice try fantard. It's not your fault that your mother and father were also brother and sister.
Actually they don't call ARM Cortex the "Atom Smasher" at all anymore. That term was cooked up with reference to the old Atom CPU's architecture. The new Atom Avoton and Rangely chips are far ahead of the fastest current and next generation ARM designs, and more energy efficient as well. Most importantly of course, they are plug and play x86 allowing them to seamlessly replace older less efficient x86 systems, especially power hog, performance wimp Opteron based systems, instantaneously. All the software that runs on Xeon or Opteron, running 95% of all data centers today, runs on the new Atoms. None of it runs on ARM architecture currently and as we saw when Intel killed RISC with x86, the switching costs for an architectural shift are enormous. It worked for Intel because Intel chips were much cheaper than the RISC chips they replaced, and Linux was much cheaper than proprietary flavors of UNIX. There's no such value proposition supporting an ARM transition. It will cost you more in both the sort run and long run to migrate to ARM since Intel now offers an x86 server CPU that is both more efficient and more powerful than ARM server CPUs.
Here's what Leah Frakes, one of the principal pro-AMD analysts, has to say about Avoton versus ARM:
"The Atom C2000 is Powerful and More Energy-Efficient than ARM-based Server CPU."
"The Atom C2000 processors are 100% compatible with Google's existing x86 server configuration. A seamless transition can be done without the risks that entails from migrating old servers to new 64-Bit ARM-based processors. Google can save itself from a lot of headaches and money by just retaining Intel as its main processor supplier."
Intel killed the RISC guys. It has now killed AMD. It will kill ARM tomorrow. Intel is a killer. Don't be a victim. Avoid investing in roadkill like AMD.
"AMD has always been good about making promises. Its current generation Kabini and Temash were promised to take the world by storm, only to see a moderate uptick in sales in Q2 before the company resumed its market share decline in Q3. The big fear at this point is that AMD's PC chip business will become collateral damage in Intel's battle with the top ARM vendors since Intel is focusing on the low cost/low power space more feverishly than ever before. It's not going to be easy, that's for sure."
No sooner did the words (i.e., that Intel wasn't AMD's competition any more; ARM (tablets) were) leave perma-bumbler Rory Read's mouth than Intel massively and aggressively began developing products to eliminate AMD from its last remaining segment, the low end consumer market. You might say that Rory, or rather his empty head and big mouth, created Bay Trail.
With the CEO on record as saying ARM, not Intel, is the competition at the low end, Read signed AMD's death warrant in shareholders' blood. Bay Trail will have 20% of the tablet market by the end of the first half of 2014. It will have displaced ARM chips from dozens of high profile Android devices, as well as monopolized the entire Windows tablet ecosystem.
With tablets killing off low end consumer PCs and laptops, Intel's totally superior Core i5 and Core i7 dominating mid and high end PCs and laptops, and Intel owning 96% of the server CPU segment, there's little left for AMD. INTC to AMD: "All your base are belong to us!"
Long time AMD pumper rantusmaximus (aka fusionthread) was a big fan of both OCZ and AMD. He pumped OCZ on the AMD board and AMD on the OCZ board under at least a half dozen aliases. He was famous for claiming to be "averaging down" when either of his two favorite stocks took a nosedive.
The poor chump was an OCZ employee and an AMD fan boy. How f--ked up is that? Today it was announced that all of OCZ's assets would be sold out of Chapter 11 to Toshiba for a paltry $35 million. He was sacked by OCZ long ago when their troubles started, but his pitiful legacy should be studied by other AMD fantards.
Shareholders of OCZ will get...NOTHING! Not a penny. 100% loss of investment. Don't kid yourselves. AMD is every bit as much of a house of cards as OCZ was. You really do stand a good chance of losing every penny.