AMD Q3 Result

[pablo87]

The year over year gross profit decline is shocking, from $494* to $239.

Now some would argue there is a one time $65 inventory writedown buried in the cogs but to me that is part and parcel of revenue and cogs management. Its not like the purchasing vp was high and bought the wrong stuff - every single quarter is like that - very high inventory levels and no ability to bring it down due to the WSA ( Imo, it would take ~6 quarters of $65 write offs to normalize inventory levels). Its how they roll - they dont look at what they need, they look at what Global Foundries needs.

By the way, its very easy to play around with this stuff ( write offs and writedowns), lose $100M one quarter, make $50 the next, etc...

Its nice they sold packaging & test. So now they'll have a PTSA to go with the WSA.

*Millions

 

[RampantAndroid]

The year over year gross profit decline is shocking, from $494* to $239.

Now some would argue there is a one time $65 inventory writedown buried in the cogs but to me that is part and parcel of revenue and cogs management. Its not like the purchasing vp was high and bought the wrong stuff - every single quarter is like that - very high inventory levels and no ability to bring it down due to the WSA ( Imo, it would take ~6 quarters of $65 write offs to normalize inventory levels). Its how they roll - they dont look at what they need, they look at what Global Foundries needs.

By the way, its very easy to play around with this stuff ( write offs and writedowns), lose $100M one quarter, make $50 the next, etc...

Its nice they sold packaging & test. So now they'll have a PTSA to go with the WSA.

*Millions

Aren't they obligated to purchase certain quantities from GF?

 

[ShintaiDK]

Its nice they sold packaging & test. So now they'll have a PTSA to go with the WSA.

It does look like a complete clone of the foundry spinoff with all its following disasters in terms of cost and flexibility.

 

[mrmt]

Chances are they would have incurred even larger penalties due to the horrendous WSA Agreement with GloFo than writing down oversupply of inventory.

Care to explain how paying a take-or-pay charge can be worse than taking goods for free and write it down?

 

[Phynaz]

The year over year gross profit decline is shocking, from $494* to $239.

Now some would argue there is a one time $65 inventory writedown buried in the cogs but to me that is part and parcel of revenue and cogs management. Its not like the purchasing vp was high and bought the wrong stuff - every single quarter is like that - very high inventory levels and no ability to bring it down due to the WSA ( Imo, it would take ~6 quarters of $65 write offs to normalize inventory levels). Its how they roll - they dont look at what they need, they look at what Global Foundries needs.

By the way, its very easy to play around with this stuff ( write offs and writedowns), lose $100M one quarter, make $50 the next, etc...

Its nice they sold packaging & test. So now they'll have a PTSA to go with the WSA.

*Millions

Agreed, AMD is sitting on 75 days of inventory vs 30 days for Intel. They need to get that under control or else there will be more writedowns in the not so distant future.

 

[Arachnotronic]

Chances are they would have incurred even larger penalties due to the horrendous WSA Agreement with GloFo than writing down oversupply of inventory. It's a miracle AMD was even able to survive the Bulldozer architecture disaster. It has now been more than 4 years since Bulldozer launched, which means for the last 4 years, Intel basically just had to show up. Technically, Intel could have gotten away just shrinking Sandy Bridge and raising clocks over the last 4 years and even the 9590 wouldn't have kept up. That's the scary part for Zen -- by the time it launches, Cannonlake will be around the corner and in that time Intel will have gone through SB, IVB, HW, BW, SKL and then CNL.

Yeah, that's why I'm not exactly aboard the hype train that a lot of the pro-AMD posters are on. AMD is literally competing with two of the world's best CPU core teams (Intel Converged Core Design Org. in Hillsboro, Oregon & Israel Design Center in Haifa) with virtually unlimited budget and easily the world's best process technology.

Zen should be an improvement from the really bad construction core line, but I doubt it will be anything close to Skylake in terms of power/performance. They might be able to get to Haswell-levels of performance, but I think they will need considerably higher power to do so which kind of wrecks it for everything but high-perf desktop (which Intel can easily counter by being more aggressive in which server dies it deigns to bring to the HEDT platform).

AMD really needs to get itself out of Intel's way, and I think they had said this was their intention a while ago, but they've gone full circle and are now talking about how a brand new, clean-slate architecture is going to help them be more competitive against Intel.

Good luck to AMD, but this looks like a losing long-term strategy.

 

[pablo87]

Aren't they obligated to purchase certain quantities from GF?

Yes they are, that's what I meant by Gloflo needs.

 

[RussianSensation]

YoY means Q4 2015 compared with Q4 2014.

I don't understand this comment -- how does it apply to anything I posted? I know what YoY means.

Thank you. I'm assuming Russian replyed to me, but I have him on ignore.

Good to know. :thumbsup:

Ok... RIP AMD. They fought well, but they moved to other sectores toó late

I wish you to make thousands of dollars on shorting the stock over the coming 5 years. :awe:

Care to explain how paying a take-or-pay charge can be worse than taking goods for free and write it down?

What does that even mean? AMD gets nothing for free.

Option 1 - You are obligated to purchase 10 seeds of apple trees whether you will sell 100 apples or not. If you only buy 9 seeds of apple trees, you get penalized $100 since your contract stipulates that you will pay a penalty unless you buy a minimum of 10 seeds.

Option 2 - You bought 10 seeds of apple trees as per the contract but you cannot sell 100 apples at regular prices. So you negotiate a lower price and sell the remaining apples at a reduced price. The difference between the normal price you were projecting to sell and the reduced price is the inventory write-down. The loss due to 'inventory' write-down is $70.

You save $30 against option #1.

"DEFINITION of 'Inventory Write-Off'
An accounting term for the formal recognition that a portion of a company's inventory no longer has value. An inventory write-off may be handled in the company's books by charging it to the cost of goods sold or by offsetting the obsolete inventory allowance. Most inventory write-offs are small, annual expenses; a large inventory write-off (such as one caused by a warehouse fire) may be categorized as a non-recurring loss. If inventory still has some value, it will be written down instead of written off.

Read more: Inventory Write-Off Definition | Investopedia http://www.investopedia.com/terms/i/inventory-write-off.asp#ixzz3ohlwL6AL

Do you think AMD will throw away $65 million of parts just like that? That's not what an Inventory Write Down means.

Yeah, that's why I'm not exactly aboard the hype train that a lot of the pro-AMD posters are on. AMD is literally competing with two of the world's best CPU core teams (Intel Converged Core Design Org. in Hillsboro, Oregon & Israel Design Center in Haifa) with virtually unlimited budget and easily the world's best process technology.

Zen should be an improvement from the really bad construction core line, but I doubt it will be anything close to Skylake in terms of power/performance. They might be able to get to Haswell-levels of performance, but I think they will need considerably higher power to do so which kind of wrecks it for everything but high-perf desktop (which Intel can easily counter by being more aggressive in which server dies it deigns to bring to the HEDT platform).

AMD really needs to get itself out of Intel's way, and I think they had said this was their intention a while ago, but they've gone full circle and are now talking about how a brand new, clean-slate architecture is going to help them be more competitive against Intel.

Good luck to AMD, but this looks like a losing long-term strategy.

I don't think anyone who understand business expects AMD to outperform Intel and/or NV long-term. However, if AMD can figure out a way to generation positive net free cash flow, it doesn't have to outright outperform Intel/NV.

If you think about it, you propose that AMD just gives up because you haven't provided an alternative strategy for them to pursue. So if you think they have no chance of ever making $ in CPUs, APUs or GPUs, what's the point of AMD existing at all?

 

[StinkyPinky]

Geez, are they even going to survive long enough to release their next desktop cpu's?

It will be grim if AMD go under.

 

[pablo87]

Agreed, AMD is sitting on 75 days of inventory vs 30 days for Intel. They need to get that under control or else there will be more writedowns in the not so distant future.

Under the current WSA, it seems they can't. Its been killing them for 5 years.

I'll never forget (i hope!) listening to an earnings call years ago and someone asked about ramping up Brazos production and filling backorders as the product was a smashing success. The then CFO gave this bizarre answer to the effect that we don't want to sell those. Turns out they were built at TSMC 40nm and it was screwing them up WSA wise. Talk about convoluted.

I still believe to this day that Dirk was fired because he questioned the WSA.

 

[Arachnotronic]

I don't think anyone who understand business expects AMD to outperform Intel and/or NV long-term. However, if AMD can figure out a way to generation positive net free cash flow, it doesn't have to outright outperform Intel/NV.

If you think about it, you propose that AMD just gives up because you haven't provided an alternative strategy for them to pursue. So if you think they have no chance of ever making $ in CPUs, APUs or GPUs, what's the point of AMD existing at all?

There are several alternative, (IMO) more reasonable strategies that AMD can pursue. They just require management and the BoD to put their egos aside and be realistic about the resources they have available to them and what markets they want to target.

 

[Tsavo]

Does anyone know if AMD can jettison the WSA? It's pretty clear that GloFlo is a dead man walking. All the stuff IBM paid them to take isn't going to help them, and they are going to get curb stomped on the next shrink.

 

[Tsavo]

I still believe to this day that Dirk was fired because he questioned the WSA.

Ruiz is the disaster that keeps on giving.

 

[Arachnotronic]

Does anyone know if AMD can jettison the WSA? It's pretty clear that GloFlo is a dead man walking. All the stuff IBM paid them to take isn't going to help them, and they are going to get curb stomped on the next shrink.

They can probably do so if they want to wipe out their shareholders by filing for bankruptcy. Bankruptcy is only an option when there are no other options left, and they seem to be doing a lot to make sure they stay solvent (as they should be!)

 

[Tsavo]

They can probably do so if they want to wipe out their shareholders by filing for bankruptcy. Bankruptcy is only an option when there are no other options left, and they seem to be doing a lot to make sure they stay solvent (as they should be!)

Bankruptcy is more of an option than a tombstone.

 

[Elixer]

I will wait for the transcript ...

Here is the transcript:
http://seekingalpha.com/article/357...u-on-q3-2015-results-earnings-call-transcript

 

[Arachnotronic]

I wish you to make thousands of dollars on shorting the stock over the coming 5 years. :awe:

Not that I think there is much value in dark zero's posts, but it's really not as simple as that.

Shorting is something that all but experienced investors should stay away from, especially as short-term moves can wipe you out even if you nail the long-term thesis if you're not very careful in how you manage the position.

For example, take UniPixel. If you shorted this thing at $21.97 in 2013, you would have gotten your teeth kicked in as it ran up to nearly $40 per share, esp. as borrowing costs on the thing were probably fairly high.

Obviously if you managed to hold through the pain of the further run up until now, you could have made a significant amount of money. But, really, for most investors the stress and work required to manage such a position would not be worth it.

It is far less stressful and, in this raging bull market, easier for most folks to make money being long quality companies than trying to short garbage ones.

Also, the nice thing about being long is that your upside is unlimited and your downside is limited to 100%. By straight up shorting a stock, your upside is limited to 100% and your downside is unlimited.

You could argue that betting on a decline thru long-term put options is a better way to do it, but there are limitations there too that make it not so straightforward.

tl;dr -- just because somebody thinks they spot a dying business doesn't necessarily mean they should call their broker and short as much as possible.

 

[Arachnotronic]

Bankruptcy is more of an option than a tombstone.

For shareholders, bankruptcy is effectively a tombstone.

 

[AtenRa]

Sampling or shipping for revenue? See sig.

Ok here it is,

Zen remains on track for availability in 2016 and we believe will return AMD to the mainstream server and high-end client market in a significant way in 2017 and beyond.

and

Lisa T. Su - President and CEOYes, Hans so let me talk overall about Zen and our roadmap in Zen, as specifically to your question. So as I said in prepared remarks Zen is on schedule for availability in 2016 and first full year of revenue ramp in 2017. As you know these microprocessor projects are multiyear projects so the architecture of the execution team very much in place, I think we’re pleased with the progress and we will continue to work hard to meet our objectives in that area.

So it seems that ZEN will be released in 2016 (Q4 2016 ??) with full year of revenue in 2017.

 

[Elixer]

we also tapped out multiple products in FinFET technologies across both of our foundry partners that are on track to enter production next year.

UMC & TSMC, or TSMC & Gloflo?

They increased CD revenue 12% sequentially primarily due to GPU and APU sales

I wonder what the breakdown is for Fury vs 3xx line?

Revenue increased 13% sequentially driven by strong semi-custom sales to support holiday game console demand. We are on track to set an annual record for semi-custom unit shipments in 2015.

Consoles are a huge part of their business, and is their lifeboat.

They also talk about medical communications... no idea what they are cooking there.
Talk about R-series SOC which will have significant performance per a watt for embedded designs, compared to their older stuff.

Zen is their ticket back to the mainstream server & high-end client markets, so, they bet the farm on Zen.
They also said again that Zen had a target of 40% IPC performance of Zen over our previous generation.

Fury 2 (or whatever the name) will be out in 2016, 2x performance per watt improvement.

They also say they have a "we've got a pretty good supplier agreement that we have put in place for the suppliers", so, no idiotic GloFlo type of agreement with the new partners.

 

[pablo87]

^ I'm not up to date on accounting rules but I think if net realizable value is lower than cost, they have to take a write down. Otherwise its "ok" accounting wise unless management ascertains in its sole discretion that there is impairment. Something like that.

Business wise its crazy. You want to sell the new higher margin, higher ASP product and to do this you want to move on as fast as possible from the old stuff. In other words, there is a real (non GAAP) cost to carrying too much inventory and it can be measured. This has been ingrained in silicon valley inventory mgmt practices for many many years but since the WSA, AMD hasn't been able to follow these.

 

[pablo87]

For shareholders, bankruptcy is effectively a tombstone.

FYI, Western Digital filed for Chapter 11 in early 90's, market cap was $135MM when they came out of it. They had terrible product (40mb 40ms stepper motor drives designed by Tandon) but good people. By filing, they got to start with a clean slate ( puerto rico ops gone, fab gone, etc...).

Whether shareholders were wiped out or not after filing I can't recall. But had they doubled down and purchased new shares for $3 (there was 45MM shares o/s post filing), they would have made a killing (I was one of those).

I don't have the same confidence in AMD mgmt though. Just by their actions of kicking the can down the road instead of biting the bullet like WD did.

 

[Elixer]

I don't have the same confidence in AMD mgmt though. Just by their actions of kicking the can down the road instead of biting the bullet like WD did.

You think they should file for Chapter 11 and restructure?
That would be a very huge undertaking, and would be a mess for the patents, and all the cross-licensing agreements that they have now.

It isn't a simple thing to just restructure like that, if it was, tons of companies would be doing it to get out of 'bad deals'. There are some pretty severe restrictions, and not all companies that have tried that route succeed in coming out it bankruptcy.

I forgot when that GloFlo contract ends, but, it seems to me that they could apply significant leverage for future purchases if GloFlo don't renegotiate the current contract. As in, TMSC & UMC for all future business plans, and GloFlo is out if they don't renegotiate new terms.

 

[Raftina]

Option 2 - You bought 10 seeds of apple trees as per the contract but you cannot sell 100 apples at regular prices. So you negotiate a lower price and sell the remaining apples at a reduced price. The difference between the normal price you were projecting to sell and the reduced price is the inventory write-down. The loss due to 'inventory' write-down is $70.

GAAP inventory is recorded at lower of cost or market. IFRS inventory is recorded at lower of cost or net realizable value. In either system, write-downs only occur if they are selling each unit at a loss.

Or, to put it another way, if a chip costs $10 to make, sells for $1,000, and costs $200 to sell, then under GAAP, the chip can lose 99% of its market value without any write-down. Under IFRS, there's still a massive 98.75% loss of market value before any write-down.

 

[ShintaiDK]

I wonder what the breakdown is for Fury vs 3xx line?

ASP being flat says it all.

They also say they have a "we've got a pretty good supplier agreement that we have put in place for the suppliers", so, no idiotic GloFlo type of agreement with the new partners.

They said the exact same thing about GloFo and the WSA.