AMD CEO talks of long-term turnaround

[Phynaz]

You don't get it. AMD warned that their sales for the just finished quarter will not meet expectations. Intel issued no such warning.

Therefore the issue lies with AMD.

The logic isn't that difficult.

 

[Fjodor2001]

You don't get it. AMD warned that their sales for the just finished quarter will not meet expectations. Intel issued no such warning.

Therefore the issue lies with AMD.

The logic isn't that difficult.

Intel issued a similar revenue warning just 3-4 months ago. Do you really think Intel's analysts are so crappy that they can't make a realistic estimate 3 months ahead, so they'd have to issue yet another warning again now!?

If you want to continue living in your fake Intel-fanboy world where the PC business is crap in general but all rosy and dandy only for Intel, then I'm afraid there's not much I can do. Ignorance is a bliss...

 

[dave_the_nerd]

PC sales have slowed down the last quarter. Effected my company as well.

 

[erunion]

. Do you really think Intel's analysts are so crappy that they can't make a realistic estimate 3 months ahead, so they'd have to issue yet another warning again now!?

You dont know seem to know how quarterly guidances work.

AMDs Q2 guidance was given 3 months ago at the end of Q1.

AMD has chosen to update their guidance shortly before their Q2 results because they now know that earnings won't fall within their guidance(which was a prediction).


Intel releasing a warning last quarter has nothing to do with Q2, its a completely new guidance.

 

[Fjodor2001]

You dont know seem to know how quarterly guidances work.

AMDs Q2 guidance was given 3 months ago at the end of Q1.

AMD has chosen to update their guidance shortly before their Q2 results because they now know that earnings won't fall within their guidance(which was a prediction).


Intel releasing a warning last quarter has nothing to do with Q2, its a completely new guidance.

At some point they have to issue a warning if things turn out worse than projected. AMD might have believed that things would turn out better than they would so they didn't issue a warning in Q1, but reality caught up with them.

But once a company issues a warning, they are not likely issue another one just next quarter. The analysts are not that poor after all, and things do not change that fast.

The reality is that PC sales have slowed down in general, and Intel just happened to issue their warning 3 months earlier than AMD.

 

[Arachnotronic]

At some point they have to issue a warning if things turn out worse than projected. AMD might have believed that things would turn out better than they would so they didn't issue a warning in Q1, but reality caught up with them.

But once a company issues a warning, they are not likely issue another one just next quarter. The analysts are not that poor after all, and things do not change that fast.

The reality is that PC sales have slowed down in general, and Intel just happened to issue their warning 3 months earlier than AMD.

Wow, just, no.

AMD's revenue guidance for Q2 already assumed a weak PC market. They did worse than already low expectations.

 

[Fjodor2001]

Wow, just, no.

AMD's revenue guidance for Q2 already assumed a weak PC market. They did worse than already low expectations.

Yes, that's why they issued a warning. Just like Intel did in Q1.

 

[ShintaiDK]

Yes, that's why they issued a warning. Just like Intel did in Q1.

You act like AMDs Q1 was on target.

In less than 3 months their shareprice went from 2.87 to currently 2.09.

 

[Phynaz]

You act like AMDs Q1 was on target.

In less than 3 months their shareprice went from 2.87 to currently 2.09.

As of today's market close, AMD shareholders now have negative equity.

 

[frozentundra123456]

Well to be fair, Intel is down more than 20% since the first of the year too, although obviously in absolute terms and compared to historical levels they are doing much better than AMD. AMD could make a nice turn around if Zen delivers, but as usual the hype train is already out of control. Lets see if they can deliver on time and if 14nm is all it is cracked up to be.

 

[DrMrLordX]

It's down in after-hours trading too. It's gonna be a rough ride until Sept/Oct 2016!

 

[mrmt]

It's down in after-hours trading too. It's gonna be a rough ride until Sept/Oct 2016!

If it makes until there. With sales crashing like that they won't make until Q316, let alone Q416 without further cuts and/or asset sales.

 

[shady28]

If it makes until there. With sales crashing like that they won't make until Q316, let alone Q416 without further cuts and/or asset sales.

True that. AMD had about $800M in the bank at the end of March after ending 2014 with just over $900M, and that's after they took in $60M from financing activities. They already have more debt than assets.

If their losses get back up to 2014 levels (-$500M\yr), they'll be pretty close to done by this time next year.

 

[B-Riz]

Can we get a sub-forum where all the arm chair experts predicting / hoping / wanting AMD's death can go and wallow?

 

[pablo87]

What I think is, the current structure is unworkable.

They need to spin off GPU ("ATI") (said that 3 yrs ago at another site). For many good reasons:

- gaming is a big consumer business and its only going to grow - its killing tv/cable in the youth market. 4K will accelerate this.

- their direct competition, Nvidia, can be beaten to wit HBM time to market.

- ATI has been short of resources, focus, management, you name it since the acquisition in 2007. They the BoD and mgmt robbed ATI to keep AMD afloat.

- HBM is a huge opportunity. Think GPU card on a single chip. Its also going to revolutionize the sub $100 market which has been bandwidth starved for a long time.

- raises money to restructure AMD. Considering Nvidia is worth $12B, a $2B pre-IPO valuation on ATI is not unreasonable. ( btw, you can go to Sedar.com to see all of ATI's old financials.)

**************

What's left of AMD would be restructured. They'd have ~$4.5B in cash and current assets against $3.3B in current liabilities and long term debt plus the big off balance sheet liability ( "WSA") of course so you'd think this could still be done outside the confines of Chapter 11.

Then you'd think they would cut sg&a to the bone and put all resources on Zen/x86 while still generating revenue from Console and existing APUs.

 

[jeff_rigby]

What I think is, the current structure is unworkable.

They need to spin off GPU ("ATI") (said that 3 yrs ago at another site). For many good reasons:

- gaming is a big consumer business and its only going to grow - its killing tv/cable in the youth market. 4K will accelerate this.

- their direct competition, Nvidia, can be beaten to wit HBM time to market.

- ATI has been short of resources, focus, management, you name it since the acquisition in 2007. They the BoD and mgmt robbed ATI to keep AMD afloat.

- HBM is a huge opportunity. Think GPU card on a single chip. Its also going to revolutionize the sub $100 market which has been bandwidth starved for a long time.

- raises money to restructure AMD. Considering Nvidia is worth $12B, a $2B pre-IPO valuation on ATI is not unreasonable. ( btw, you can go to Sedar.com to see all of ATI's old financials.)

**************

What's left of AMD would be restructured. They'd have ~$4.5B in cash and current assets against $3.3B in current liabilities and long term debt plus the big off balance sheet liability ( "WSA") of course so you'd think this could still be done outside the confines of Chapter 11.

Then you'd think they would cut sg&a to the bone and put all resources on Zen/x86 while still generating revenue from Console and existing APUs.

HBM and multiple power efficiencies added to the next node (14nm FinFET) and/or PD-SOI are going to allow a cheap multi-chip on interposer PC for the living room @ PS4 power levels. A dGPU is possible but the combination is outside the sweet spot of price and power/fan noise for the living room Skype-Connected Home-IoT-Vidipath PC. So:

1) Don't breakup AMD
2) A comeback is coming with Vidipath, ATSC 2 and 3 as well as the FCC DSTAC eliminating the cable card and allowing downloadable security.
3) Gaming, VR and AR require the power of a PC in the living room on that 4K TV that is under power regulations that don't permit it to do the same

 

[NTMBK]

HBM and multiple power efficiencies added to the next node (14nm FinFET) and/or PD-SOI are going to allow a cheap multi-chip on interposer PC for the living room @ PS4 power levels. A dGPU is possible but the combination is outside the sweet spot of price and power/fan noise for the living room Skype-Connected Home-IoT-Vidipath PC. So:

1) Don't breakup AMD
2) A comeback is coming with Vidipath, ATSC 2 and 3 as well as the FCC DSTAC eliminating the cable card and allowing downloadable security.
3) Gaming, VR and AR require the power of a PC in the living room on that 4K TV that is under power regulations that don't permit it to do the same

Betting the future of AMD on cable companies moving rapidly seems like a bad idea. Have you looked at a cable box recently?

 

[mrmt]

Betting the future of AMD on cable companies moving rapidly seems like a bad idea. Have you looked at a cable box recently?

Even if they wanted this ship has sailed. They ditched the cat cores, which would be ideal to this kind of venture.

 

[Dresdenboy]

Even if they wanted this ship has sailed. They ditched the cat cores, which would be ideal to this kind of venture.

Hmm.. how does XV compare to Puma+? Remember, one module is still 2 cores.

 

[mrmt]

Hmm.. how does XV compare to Puma+? Remember, one module is still 2 cores.

How does XV compares to Silvermont or ARM? It is still a low cost market after all, and XV chips aren't exactly cheap to design and manufacture.

 

[NTMBK]

How does XV compares to Silvermont or ARM? It is still a low cost market after all, and XV chips aren't exactly cheap to design and manufacture.

AMD can license ARM cores just as well as anyone else Even if Skybridge got canned, the work to make GCN play nicely with an ARM design is already done. I suspect their future embedded designs will replace cat cores with A72/53 (or whatever the latest equivalent is). And in fact, this is my guess for what ends up in the next Nintendo console.

 

[mrmt]

AMD can license ARM cores just as well as anyone else Even if Skybridge got canned, the work to make GCN play nicely with an ARM design is already done. I suspect their future embedded designs will replace cat cores with A72/53 (or whatever the latest equivalent is). And in fact, this is my guess for what ends up in the next Nintendo console.

I agree here, but then how would AMD get into tv boxes? This ARM chip wouldn't be engineered for lower costs, it would be standard after all, and AMD wouldn't be able to fetch a lower cost with their foundry partners for obvious reasons.

 

[NTMBK]

I agree here, but then how would AMD get into tv boxes? This ARM chip wouldn't be engineered for lower costs, it would be standard after all, and AMD wouldn't be able to fetch a lower cost with their foundry partners for obvious reasons.

I would agree with that. Given their graphics IP, they could potentially get in if the device was also meant to do gaming (perhaps running Android TV, like the NVidia Shield)... but I would honestly expect someone like Broadcomm to win the spot instead.

 

[pablo87]

HBM and multiple power efficiencies added to the next node (14nm FinFET) and/or PD-SOI are going to allow a cheap multi-chip on interposer PC for the living room @ PS4 power levels. A dGPU is possible but the combination is outside the sweet spot of price and power/fan noise for the living room Skype-Connected Home-IoT-Vidipath PC. So:

1) Don't breakup AMD
2) A comeback is coming with Vidipath, ATSC 2 and 3 as well as the FCC DSTAC eliminating the cable card and allowing downloadable security.
3) Gaming, VR and AR require the power of a PC in the living room on that 4K TV that is under power regulations that don't permit it to do the same

I'm not familiar with this tech but based on my observations of my local cable co., they haven't a clue. A lot of their base loss comes from their refusal to do a la carte though that would only delay the inevitable. Gaming is not only content, its YOUR content, and its social too - kids are playing games with their close relative even though they're thousands of miles apart.

Back in the day, we were designing video cards based on chips like 90C33, it was 64 bit 2mb, now you can see 128 bit 2GB so HBM fixes that in a very elegant way. $50 HBM video card (or better yet socketed chip) is possible. Edit: I'm always stupefied at the sheer qty of poor quality of video cards on Steam - clearly a consumer $$$ issue.

Also Integration has been going on for al 30 years since the days of C&T but its not straighforward anymore eg lets suck in these jelly beans and cost down. Recently we see that some vendors have abandoned trying to integrate the radio. With HBM the disintegration of video with CPU starts to make a lot sense thus ATI.

Certainly does for AMD - when we recently purchased that beautiful Intel powered ASUS Chromebook for $169, it made me realize that the DataCenter market can subsidize anything ( most companies have a very profitable side and the not so). Imagine if Intel sold that Gxxxx processor for $9 instead of $69, instead of subsidising tablets? Instant Chapter 11 for AMD. Obviously won't happen re Antitrust but what kind of company (talking AMD here) has to hide behind his poppa like this?

Honestly AMD would be better off to turn the table on Intel - stop selling x86, continue development and collect royalties from Intel - would be a lot more profitable with a lot less headaches.

 

[Phynaz]

I like the idea of spinning off graphics.