...and that the only way AMD could persist is to successfully compete in low volume, high margin, high end chips?
I sure do.
And then a few things happened. Jim Keller, the lead architect of Zen, up and quits without a word long before Zen was done. Pascal benchmarks leak and Pascal's launch priority over Polaris was established. AMD's spin-meister Roy Taylor, does a 180 turn around on Polaris, dropping the claim that it will be competitive at the high end and a high performance part, instead claiming that it will be a volume lower end part. The pricing for Polaris launches at just $200 a card, making it the cheapest new AMD GPU product line ever.
Obvious conclusion? GF/Samsung 14nm manufacturing sucks. Zen architecture sucks. Polaris architecture sucks. Trying to design new products with a skeleton crew of bottom of the barrel employees, against much better staffed, better funded competitors sucks. Lisa Su's leadership and vision suck. 2016 and 2017 are going to suck hard for AMD investors and fan boys.
...the discount bargain bins at Micro Center and Fry's.
While Pascal owns VR, just like that. Poof. AMD's VR dreams gone.
Nah. I'd say the multiple downgrades, some of them from firms formerly positive on AMD, have more to do with the collapse. The smartest analysts on Wall Street don't buy AMD's management story nor its Bernie-Sanders-like math. No profits for at least 2 years, if ever again.
That's less than Lisa Su wasted on her micro-server project that was jettisoned last year as a total loss.
And the shares tumble another 4% on no news a cooler heads and detailed analyses of AMD's report trickle in, along with massive downgrades from firms who say "we don't believe AMD management; the numbers don't add up; there will be no profits this year and probably all next year too".
Because anyone who follows AMD knows the pump and dump when they see it.
Sure the 12000 Intel employees being laid off represent Intel's second tier of employees, and are being let go due to out-of-date skills, but even Intel's second best employees are three or four rungs up the competence ladder than AMD's very best talent. AMD could fire everyone and replace them with these Intel cast offs, and have 4-5x more competence and productivity. Makes you think...
After all, there are janitors at Intel that could do a better job as CEO than Lisa Su.
Memory Read Benchmark:
Core i7 6700k = 30174 MB/s
Zen Summit Ridge = 27165 MB/s (-10%)
Vishera (a 2012 product CRUSGED by Haswell Core i5!) FX-8350 = 26518 MB/s
Zambezi (Sandy Bridge era!) FX-6100 = 26456 MB/s
FP/Ray Trace Benchmark:
Core i7 5820 (a 2014 Haswell chip!) = 7300 KRay/s
Zen Summit Ridge = 5198 KRay/s (-29% - OUCH!)
Opteron 6274 (ancient Q1 2012 product!) = 5991 KRay/s
Shockingly the benchmarks reveal that AMD has made little if any progress improving either performance or efficiency. Both Summit Ridge and Opteron 6274 have 8 cores. A 4 year old Opetron beats Summit ridge core for core. That ancient Interlagos core was built on 32nm process! Zen is 14nm FinFET. The new chip is actually worse than the ancient one! HA HA HA HA! But the fan boys say, Zen is more efficient! WRONG! Opteron 6274 has a TDP of 115W and Summit Ridge a very similar 95W!
No wonder Jim Keller quit and ran away! Bulldozer 32nm is FASTER and as EFFICIENT as Zen! HA HA HA HA!
Summit Ridge looses to a Skylake with half as many cores and worse still, to Haswell with 25% fewer cores!
AMD may as well put its head between its legs and kiss its #$%$ good-bye!
HA HA HA HA!
AMD thinks consumers are stupid.
It's AMD that's stupid if it thinks it can peddle the current garbage against mobile Pascal.
Reportedly, the new Radeon R9 M470X will be a rebranded Radeon R9 M385X, which is based on the old Bonaire XT GPU, and features 896 stream processors, 56 TMUs and 16 ROPs.
AMD GPUs Using HBM2 Won’t Launch Anytime Soon Thanks to High Costs
The word on the street is that HBM2 is currently in a very limited supply, which is the reason behind making things expensive.
AMD can't afford to use it!
Will you get pummeled tomorrow when AMD misses, gives #$%$ guidance and triggers the dump tomorrow. It'll be in the 1s by Friday. I just sold my 3800 shares and am now short 2000 shares. It's hard to short AMD there are so many wanting to short it today.
Good luck to everyone. I'll be back to gloat tomorrow and hear your woes.
AMD is going to take hot manmeat up its backside from Stacy tomorrow.
Last earnings call: "We ain't got much business."
This coming earnings call: "We still ain't got much business."
Q2 2016 earnings call: "Nope, still no business."
Q3 2016 earnings call: "Nuh uh. Polaris shipped. It sucks. No business."
Q4 2016 earnings call: "Zen shipped. It sucks. Still no business, boss."
AMD Q1 2018 price target: $0.47
"The current short interest as a percentage of the float for Advanced Micro Devices is very high at 17%. That means that out of the 646.51 million shares in the tradable float, 110.34 million shares are sold short."
"If you're long biased I would avoid Advanced Micro Devices. Look for short-biased trades after disappointing earnings and guidance fail to trigger a breakout. The shares should drop back below some near-term support levels at $2.61 a share to its 50-day moving average of $2.43. If we get that move, then this stock will set up to re-test or possibly take out its next major support levels at its 200-day moving average of $2.16 to around $2 a share or even lower potentially."
Digitimes sources claim that AMD CPU and GPU market share has hit a new low in Q1 2016.
"Demand for standalone graphics cards has already been sluggish given a weak PC market, and the market for standalone graphics cards continues to decline. In the already-shrinking market, AMD's rival Nvidia has continued to eat away at AMD's market share in the quarter."
"In the PC processor market, AMD's gap behind Intel has widened further in the quarter. And it remains doubtful that AMD's next-generation Zen architecture will help the company make a comeback."
Ouch, just ouch.
That's certainly one way to mitigate yield and quality problems; artificially reduce demand by raising price. AMD is just trying to save a little face.
I can't wait for earnings! AMD is going to get loaded onto the train, sent to the work camp, worked to exhaustion and then burnt to ashes.
Liar. He rates AMD a HOLD. Not a Buy or Strong Buy. Hold. That's a neutral rating. It's bad to have coverage initiated at Hold. It says the firm feels the shares are fairly valued at present levels for likely near term results. In other words, he doesn't think it will go down a lot, but you certainly shouldn't be buying it at this level.
Besides, nVidia will double by this time next year. AMD will be in the 2s still. Smart money is buying NVDA.