I think a few directed tests could prove weather the part being discussed is steamroller based or piledriver based. Since Streamroller features a split decoder, such directed tests would be wise to aim at depicting a Piledriver weaknesses there and comparing with the alleged Steamroller part.
The AMD Execution Thread [2007 - 2017]
- Thread starter overclocked_enthusiasm
- Start date
-
- Tags
- amd
- Status
I think a few directed tests could prove weather the part being discussed is steamroller based or piledriver based. Since Streamroller features a split decoder, such directed tests would be wise to aim at depicting a Piledriver weaknesses there and comparing with the alleged Steamroller part.
Even they wouldn't have any use for AMD.
AMD as a whole seems to be a very unattractive purchase target.
Their best shot at (pseudo-)acquisition seems to be, to be divvied up and sold for the parts.
Charlie think's it's game over for the big cores 
http://semiaccurate.com/2012/11/19/amd-kills-off-big-cores-kaveri-steamroller-and-excavator/
http://semiaccurate.com/2012/11/19/amd-kills-off-big-cores-kaveri-steamroller-and-excavator/
Makes complete sense to me. What really is the point in wasting money developing for a market that intel has had completely sewn up for years?
AMD needs small <100mm2 chips like Bobcat and Kabini to survive. They can't afford to have low margins on big chips. Why do that when they can have the same low margins on small chips, yet 3-4x the volume?
The key to taking on intel is about volume. AMD has the capacity to do massive damage to intel in the tablet/notebook space but they've never exploited it properly, constantly wasting their GF wafers on chips that are too big and have poor perf/mm.
They literally have the capacity to sell over 300 million chips in a year but have limited themselves to well under 100 million while trying to compete with intel in the high end space. Absolutely god-awful planning to be honest - Bulldozer should never had made it past the drawing board on it's size alone.
Kabini and its successor are likely to be their last throws of the dice so they might as well do it right and go all in before ARM gains a proper foothold.
AMD needs small <100mm2 chips like Bobcat and Kabini to survive. They can't afford to have low margins on big chips. Why do that when they can have the same low margins on small chips, yet 3-4x the volume?
The key to taking on intel is about volume. AMD has the capacity to do massive damage to intel in the tablet/notebook space but they've never exploited it properly, constantly wasting their GF wafers on chips that are too big and have poor perf/mm.
They literally have the capacity to sell over 300 million chips in a year but have limited themselves to well under 100 million while trying to compete with intel in the high end space. Absolutely god-awful planning to be honest - Bulldozer should never had made it past the drawing board on it's size alone.
Kabini and its successor are likely to be their last throws of the dice so they might as well do it right and go all in before ARM gains a proper foothold.
Is this scenario possible?
- AMD develops a fabless console CPU/APU for MS and SONY
- AMD makes CPU/APU for MS and SONY
- AMD disappear
- MS and SONY pay "royalties" to INTEL, so
- MS and SONY make CPU/APU in another foundry
In case of AMD bankrupcy, could INTEL stop production of CPU/APUs of MS and SONY, or going bad with both is too dangerous?
- AMD develops a fabless console CPU/APU for MS and SONY
- AMD makes CPU/APU for MS and SONY
- AMD disappear
- MS and SONY pay "royalties" to INTEL, so
- MS and SONY make CPU/APU in another foundry
In case of AMD bankrupcy, could INTEL stop production of CPU/APUs of MS and SONY, or going bad with both is too dangerous?
I think both Sony and MS are smart enough to have covered all their bases in case something happens to AMD. AMD hasn't been a healthy company for a while now.
More specifically:
[emphasis mine]
Does that mean Kaveri will be released in 2013 after all and previous rumors and leaked roadmaps were bogus? This is all very confusing.
itsmydamnation
Veteran
no it isn't, AMD hasn't said dick about Kaveri, all we had was one slide. Everything else is a rumor. Also i think AMD must of butt hurt charlie lately because he was been ultra negative almost but not quite to the same level as his NV hate.
The roadmap I think you're talking about:
http://www.extremetech.com/wp-content/uploads/2012/11/AMD-Desktop-Roadmap1.png
was somewhat ambiguous: it wasn't clear whether the time line extends all the way to the end of 2013 or just the 1st quarter.
chavvdarrr
Veteran
Any idea why AMD is losing GPU market share ?
People are not buying AMD GPUs.
It's never one thing with AMD, but lots of things.
Llano was totally osborned by Trinity these past few months. Brazos is still competitive but too old. The 7-series cards have been around for what feels like ages while the Nvidia cards feel new. Piledriver was an improvement but not enough.
I believe Trinity is selling well though, but honestly it's so cheap it should be. Scan has run out of stock of the 5800K twice this week.
It's a bad time of year to have an ageing portfolio or one that isn't quite good enough. The only place where they are probably winning (Trinity), it's a low margin product just starting it's ramp. Crap timing is just what they didn't need but it's not really a surprise.
Llano was totally osborned by Trinity these past few months. Brazos is still competitive but too old. The 7-series cards have been around for what feels like ages while the Nvidia cards feel new. Piledriver was an improvement but not enough.
I believe Trinity is selling well though, but honestly it's so cheap it should be. Scan has run out of stock of the 5800K twice this week.
It's a bad time of year to have an ageing portfolio or one that isn't quite good enough. The only place where they are probably winning (Trinity), it's a low margin product just starting it's ramp. Crap timing is just what they didn't need but it's not really a surprise.
Those aren't counted to graphics marketshare on those groups/categories.
In Q2/12 for the category/group where APU's are accounted for (meaning everything from integrated to discretes, mobiles, you name it) Intel had 62% share, AMD 22.7% share, nVidia 14.8% share, VIA/S3 0.5% share and Matrox 0.04 share
The biggest problem for them is they missed Windows 8. Trinity is the only thing out and its in a category that doesn't matter.
I don't see any tablet wins and very few laptops with amd chips.
Samsung can just hire whatever it wants from AMD, and wait for the carcass to hit the ground before picking up the patents it may crave...or wait for JP & friends to give Rory this really super duper mega consulting pro-tip (bro-tip?) to sell off the patent portfolio so that they live for another year and get closer to the ultimate winning strategy of making non-descript ARM chips and having super-optimal costs by outsourcing everything. IMHO anybody that's already involved in making MPUs / SOCs has no reason whatsoever to buy AMD or merge with them, the only ones that may care to put in the cash would be some investment fund that just wants to increase the portfolio / put money in high-tech.
AMD's credibility gap is so wide these days. Even if there were an illusion of competence, there is now the open question of whether AMD will even exist to keep many of its promises.
I wonder if J.P. Morgan is looking into how AMD can get whatever cash it can to pay of as much of the debt that it can and allow the board to bail with the best payday they can extract.
Could AMD ransom its CPU IP to Intel, just to keep any potential trolls from a big payday?
Can AMD sell the non-x86 IP to someone else?
The board of directors could then try to cash out after it runs through the dregs in the development pipeline.
I wonder if J.P. Morgan is looking into how AMD can get whatever cash it can to pay of as much of the debt that it can and allow the board to bail with the best payday they can extract.
Could AMD ransom its CPU IP to Intel, just to keep any potential trolls from a big payday?
Can AMD sell the non-x86 IP to someone else?
The board of directors could then try to cash out after it runs through the dregs in the development pipeline.
- Status
