• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD Bulldozer Finally!

I thought that the llano issues didn't apply to Bulldozer. Bulldozer is a CPU and GF/AMD have been building CPUs on SOI for years. Llano is a CPU/GPU, and GPUs haven't been built on SOI before, so that's where the trouble lay.
 
Last edited:
drunkenmaster said:
Trinity, huge gpu and much better yields, so what the heck is wrong with Llano?
Click to expand...

Well, if it's to do with SOI then it could be that they have learnt from Llano and built a better die for SOI. That means that they need to get Trinity out sharpish, both for GF and AMD's sakes. Plus, I'm betting everyone around here would be interested in seeing it. :)
 
Its certainly possible, I'm suspicious that a bad Fermi like decision was made, when 32nm turned out to be very different to expected, rather than do a full respin, base layer, 6 month delay, revalidation and losing a lot of time, they did a top layer respin to get things as good as possible get it out, hope yields were passable and limp along to Trinity.

That still might be the right decision, for a GPU less so, validation takes no time at all, the delay would be less long, and power is a huge issue on cores that are well, 2.5x the core size of a Llano. Also if Llano was delayed 6-8 months, and trinity is due in Jan/Feb, considering the cost to retape out and time spent with engineers working on that rather than trinity/bulldozer/piledriver/etc, then you're really just delaying everything else, to get a "good" Llano out, that will have only 2-3 months to make money.

I said a while ago that Bulldozer/Llano could well end up a bit cack, with all the effort being pushed into Piledriver/trinity, why fix Llano Bulldozer when the next cores are SO close, the time is better spent on the next parts.

SOI and GPU, certainly possible, Llano has a lot of surprisingly new bits, Sandybridge-e is all but identical to Sandybridge normal in 98% of the core, but the other 2%, new interconnect, pci-e 3, bunch more sata is responsible for the delay. Llano has a lot of GPU/CPu interconnects, and lots of new cool bandwidth saving gpu/memory tech going on(but mostly unused right now).

Its perfectly possible they learnt a lot from Llano but rather than make Llano mk2 with higher yields and better clocks, they just dumped all those tweaks into Trinity, I said 5 months ago that is probably the best plan of action, hurt short term, but means next year could be awesome for AMD. Or fix the first bulldozers/Llano, delay the better parts by 6 months to do so, and make very little extra short term, and have a crappy next year.
 
IF its really coming out on the 12th october wont the nda or whatever its called be lifted soon and we can see some real benchmarks?
 
tick-tock, time is passing and maybe AMD should learn from Intel and not combine major architecture and process shrinks into the same product release.

Surely Bulldozer now ranks worse than Fermi in the botched release league?
 
Martini1991 said:
That'll be no doubt the backlash from Rick.
Click to expand...

Except it had literally and entirely nothing to do with Rick at all, well done.

AMD posted their new updated forecast which was down on previous estimates, though not by much and considering lack of Llano supply and lack of Bulldozer supply, they actually did pretty well.

The US tech market is going up or down by 4-5% every few days, the market is stupidly volatile, its likely part of the 5-10% drop in these kinds of situations is bog standard, the market being what it is right now, swinging 5-10% each way every couple weeks, if one stock does bad, it drags everyone else down a few percent, which add's on to what AMD lost.
 
555BUK said:
tick-tock, time is passing and maybe AMD should learn from Intel and not combine major architecture and process shrinks into the same product release.

Surely Bulldozer now ranks worse than Fermi in the botched release league?
Click to expand...

Nah I've not seen a BD CPU turn up made out of woodscrews and bits of cut up PCB boards yet, with the guy holding it saying it was a working model. :)
 
drunkenmaster said:
Except it had literally and entirely nothing to do with Rick at all, well done.

AMD posted their new updated forecast which was down on previous estimates, though not by much and considering lack of Llano supply and lack of Bulldozer supply, they actually did pretty well.

The US tech market is going up or down by 4-5% every few days, the market is stupidly volatile, its likely part of the 5-10% drop in these kinds of situations is bog standard, the market being what it is right now, swinging 5-10% each way every couple weeks, if one stock does bad, it drags everyone else down a few percent, which add's on to what AMD lost.
Click to expand...

How much did Intel or Nvidia drop today then?
 
Martini1991 said:
How much did Intel or Nvidia drop today then?
Click to expand...

Why today, the market hasn't been open "today", yesterday Nvidia went from almost 14 to almost 12, Tuesday opened at about 14.35 and its lowest is around 12.5, can't see exactly from small graph, so around 15% lost in the past few days.

AMD have lost about 20%, INtel went from 23 on Tuesday to around, 21.8 yesterday.

However different companies are seen differently, Intel are rock solid for the tech industry and the current climate, Nvidia, netflix, AMD and a dozen others are VERY volatile.

52 week low high for Intel is 18.77/23.96, for Nvidia its 10.38/26.17, for AMD its 5.11/9.58.

So Intel over the past year has seen $5 change on highs of 23, or less than 25%, AMD is seen its value from max to min roughly half, Nvidia has had change of roughly 60% change in its market value.

Yeah, AMD and Nvidia are crazy volatile and no AMD haven't lost much more than Nvidia since Tuesday.

Netflix have 52 week low/high of 107/304, massive massive volatility, and its very much part of their downfall that caused Nvidia/AMD to lose so much in the past few days, not always but often whenever AMD or Nvidia have a new statement, more/less profits, those two and Netflix tend to move, fast.

Intel haven't got a huge amount of growth, they are just huge, traders aren't interested, day trading is a non starter with Intel, while day trading and short term trading on AMD/Nvidia/Ntfx can be highly profitable. Netflix fell 49% in September, alone, and thats why Nvidia/AMD have been doing VERY badly in the past week. Without Netflix crashing, Nvidia would probably have done nothing this month and AMD would have taken a 5% hit max for the marginally weak forecast(realistically 2-4% revenue drop due to Glofo capacity issues), and would likely have rebounded within a few days.

Incidentally, most investers don't want stocks that stay still, Intel aren't going to make you big money, AMD, Nvidia, and my god Netflix might. If you knew the forecast wasn't great, you'd have made a killing selling a week ago, and rebuying in probably a week or two from now(maybe now). ;)

Now is the time Netflix is either going to get into serious trouble, or probably rebound hard in the next few months, unless the EU goes down hard and takes most of the world economy with it, which is very possible.

Rick left for one reason, to be CEO somewhere else, he'd likely have stayed at AMD as CEO had they offered it too him, Bergman, the old CEO, its hard to judge performance, idea's men, technicians aren't always the best for business or scheduals, or not being in huge debt, sometimes you need an idea's man alone and the lemmings follow(Steve Jobs). Will Bergman/DIrk meyers architecture and plans be great in the next few years but sucked for 3-4 years(both had been in their positions since 2008 I think), maybe, would their architecture first, business second have cost AMD billions more in lost profits/stocks/sales/market share, who knows. Since 2008 AMD have only had one success ATi, and that happened before Dirk Meyer was CEO.
 
Last edited:
Share price gains / falls for the past 12 months are as follows.

AMD
Sept 30th 2010 - 7.11
Sept 30th 2011 - 5.11
loss - 28%

NVidia
Sept 30th 2010 - 11.68
Sept 30th 2011 - 12.79
gain - 9.5%

Intel
Sept 30th 2010 - 19.2
Sept 30th 2011 - 21.9
gain - 14%

It seems that whilst AMD is falling, both of it's arch rivals have been climbing. If AMD had any good news whatsoever, they would have made it public by now.

edit: Just checked the rating's agemcies where NVidia sentiment has been rising, whilst AMD is falling deeper into the avoid bucket.
 
Last edited:
Yes, companies peak at different times, well done, you worked it out. Randomly ignoring that AMD only went up $3 throughout the year from 7.11, and Nvidia went up to $26, and all the way back down.

I gave a thorough look to back up what I said, pretending Nvidia have been solidly gaining, is laughable.

Nvidia's peak came at the end of their best quarter when they gave out a positive earnings report, they hit 26, they've gone down from 26 since then, but you want to ignore that.

For the record as per usual Martini was being negative, and getting it wrong, by all means be negative but get it right ;)

THe drop wasn't because of Bergman, AMd dropped 20% while Nvidia dropped 15% in the past few days. I was giving high/low for the year to talk about and prove the volatility not the direction either company is going, so why the need to respond to that with which direction the companies were going when no one was actually discussing that, I don't know.

Also what good news would it "make public" by now if it had any, in regards to what, what are you on about?

AMD's chips have been delayed, the major reason for this is confirmed by EVERYONE, and that is Glofo having issues with 32nm and capacity. If you watched the market, you'd know everyone over reacts to EVERYTHING. Like for instance Nvidia going from 11.68, to 26.17, back to 12, in a year, the peak was short, in one quarter, based on an earnings result, which.... didn't stick.

AMD/Nvidia have these peaks every year, because they are volatile companys, Intel doesn't have such peaks because it is a rock solid, will always be around company.

Nvidia price at 26.17 was nuts, utterly unrealistic, thats what happens in a good quarter when everyone overbuys and over reacts to a simple earnings report, thats part of whats happening to netflixs. All three companies have had good quarters with big gains that come back down to earth, and all three companies have had bad quarters, with big losses, that come back up.
 
Last edited:
555BUK said:
Share price gains / falls for the past 12 months are as follows.

AMD
Sept 30th 2010 - 7.11
Sept 30th 2011 - 5.11
loss - 28%

NVidia
Sept 30th 2010 - 11.68
Sept 30th 2011 - 12.79
gain - 9.5%

Intel
Sept 30th 2010 - 19.2
Sept 30th 2011 - 21.9
gain - 14%

It seems that whilst AMD is falling, both of it's arch rivals have been climbing. If AMD had any good news whatsoever, they would have made it public by now.

edit: Just checked the rating's agemcies where NVidia sentiment has been rising, whilst AMD is falling deeper into the avoid bucket.
Click to expand...

Hardly a surprise. Bobcat's market is been taken over by ARM/Apple, BD is going to be ~6 months late and their foundry, GF, can't even make the simpler version of HKMG 32nm work(gate first).
 
Owenb said:
Hardly a surprise. Bobcat's market is been taken over by ARM/Apple, BD is going to be ~6 months late and their foundry, GF, can't even make the simpler version of HKMG 32nm work(gate first).
Click to expand...

What, Bobcat isn't being taken over by ARM, bobcat isn't a tablet chip, nor a smartphone chip, and was never meant to be, Atom is whats being killed, badly, by bobcat, in netbooks, which have rightly evapourated, the netbook bubble burst, and the tablet bubble will burst soon enough(if it hasn't already).

BD is 6 months late? Sandybridge -e is already 7 months late and getting worse, its going to launch without more sata, without pci-e 3 working and without something else it was supposed to have. BD was never supposed to launch till Q3, that was the target for the previous 18 months or so, in the final 2 months AMD thought they could bring it forward because frankly, its ready and Glofo were supposed to be, but, they weren't. Even then it was supposed to be May/June, which is barely Q2, its only 5 months late as it is.

Nvidia's 680gtx is set to be 6 months later than the 7970, Ivybridge has been pushed back 6 months, Haswell has been pushed back 6 months.

But AMD are terrible, while Intel are great despite having worse delays.

Delays suck, AMD shouldn't have gotten over excited and brought the date forward, but they did, thats quite probably part of the reason Bergman is gone. But Intel have had the same/worse delays and no one bats an eyelid.

You know Cederview, the new Intel Atom launched, its way late to 32nm, its supposed to have a dx10.1 gpu, they missed out on dx10.1, couldn't get that working, missed out on dx10 aswell.... it came out in 2006, Intel STILL can't get dx10 working right, its launched, and they won't have a final non beta dx9 driver till November, and December for the 64bit driver...... thats how late Intel are again.
 
You must log in or register to reply here.
Back
Top Bottom