Technically speaking the acquisition is nearly neutral. AMD spent some cash to acquire an asset, which means the books simultaneously reflect the loss of cash and the gain of an asset.
In practice AMD actually came out ahead on the Seamicro acquisition: ignoring what they directly paid for Seamicro, they had 6mil in acquisition costs (that is, costs for lawyers and such), but realized a tax benefit of 36mil, meaning they came out 30mil ahead.
It falls under the current period and other periods. It impacts the balance sheet more, as it is capital expenditure, and doesn't have a significant impact on the income statement, where you see profit or losses.
Rather than being reflected as an expense for the current year, the cost is amortised over the life of the acquisition. They pay $334m now, and if it had a life of 10 years or so, you would might see an expense of $33.4m/yr, give or take, reflecting the decline in value of the purchase over time.
They are good now. They now compete solely like nVidia. No Fabs. No losses from fabs. Let the Arabs in Abu Dhabi take the losses at GloFo. They should have done this long ago.
What this really proves is how difficult and expensive it is to continually upgrade Fabs to new process nodes. The fact that Intel does it so well, now, is simply a testament to their ability to learn from their own past. Too bad they will never offer to manufacture other companies designs at their Fabs. Could be interesting.
And, despite many people calling NVIDIA's demise over the past couple of years, it's never been remotely close to happening.
It's great to have fabs... if it didn't cost so much to upgrade them and work your way towards a mature process, that is. As AMD don't have to pay GloFo for failures, the only thing AMD can lose money on is not getting enough product back after aforementioned failures. I'm definitely simplifying it, though.
Fab 8's completion should provide a decent boost to 28nm production during the second half of this year, should AMD decide to sign up for it.
LOL I may have to ignore the ignore; according to Charlie, apparently AMD has signed a deal with GloFo to produce Kaveri on their 28nm process for release early 2013.
Hey, he's often right on the mark. Forecasted Kepler would be the winner vs GCN 3 months ahead of launch, the only site where I read that, and lo and behold... http://semiaccurate.com/2012/01/19/nvidia-kepler-v... Or was there a story here about that as well? Reading the comments of that article now in retrospect is kind of funny.
"No losses from fabs." Well, perhaps not directly but AMD lose money when the foundries don't get the process ready in time. They lose money when they have to cancel gate-first designs at GloFo and start over with gate-last at TSMC. Against nVidia that's probably not a big deal as they're in the same boat but versus Intel it's a very big deal. Not only because of cost but also because of power efficiency which means longer battery life and cheaper data centers. For once Intel's schedule has slipped a bit too letting AMD get off easy in Q1 but this coming week I expect Ivy Bridge. And comparing AMDs roadmap with Intel's tick-tock roadmap, it's not going to get easier...
I think you are wrong, and here is why...CONSUMERS! you see consumers don't read benchmarks, they don't care about logos, all they know is those AMD netbooks and laptops are REALLY affordable while letting them do light gaming and more importantly letting them enjoy HD video with zero stutter or lag. I mean you can get a nice AMD Brazos netbook in the $350-$430 range (I lucked out and hit a sale and got an E350 for $350 with an 8Gb of RAM upgrade AND a carrying case) that they can not only carry but pop an HDMI cable into and make it into an HTPC you can carry around.
The two big screw ups AMD did IMHO was killing Thuban (which made a great affordable desktop line and thanks to turning off cores/cache could cover the ENTIRE Phenom II and Athlon line while giving those willing to core unlock a chance at a free upgrade) and the killing of the Brazos update Krishna but knowing what we now know about GloFlo it was probably smart to kill it and go for an update of Brazos with TSMC instead.
Walk into any Walmart or other B&M store and you'll see Fusion stickers everywhere and talking to friends that work at these shops they sell like hotcakes. I think they have a real winner with the Bobcat design and if they can make say a 4 or 6 core version with a more midrange GPU this will further cement that big chunk of the consumer market because bobcats are cheap to make and run well. I think we can all agree that Bulldozer is a failure but if they can get the new GPU, which they are trading VLIW for vector based BTW, they should be able to salvage the design. You see the weak spot of the BD/PD design is two integers but only one FP unit per pair, but with vector basing they should be able to use the built in GPU as a super powered FP unit.
So I'd say the future looks bright for AMD, especially once they are completely done with GloFlo as anything but one contractor of many. Sure they won't be beating Intel but as Jobs said you don't HAVE to kill the giant to carve out a profitable niche and frankly AMD does make a good value for consumers. Who doesn't like netbooks that get 6+ hours, plays 1080p, never gets hot, and is affordable to boot?
When they release a cpu that is a whopping 3 times slower than intel at superpi, is it any surprise they are not making money? They should be bankrupt after pulling that stunt.
Yes because as we all know.. Super Pi is the be all end all of computer productivity or gaming apps...... I guess Intel deserved to go bankrupt over the p4?
Uhm, somehow it didn't work other way round, did it? Oh, and on top of being SLOWER and more power hungry, Intel's netburst CPUs were also more expensive.
Whether you want to call these loses Gap or Non-Gap or related to or caused by GloFlo or not is meaningless. The bottom line is that the company had to pay out a lot of money and lost a lot of money.
The fact is that these expenses represent loses to the company due to bad business decisions which were made in the past and are reflected in the current.
And frankly, given the poor state of AMD's current product offerings, one has to wonder if these financial burdens won't be the straw that killed the camel. Every dime spent on this and on legal costs represents a dime the company can not spend on R&D.
Anyone remember AMD laying off 10 percent of its workforce awhile back? Think it had nothing to do with this? ? ? You probably also believe in Santa.
Uhh, AMD has successful product offerings in every CPU and GPU segment except for high-end desktop CPUs, which is quite tiny (although quite profitable). Have you looked in a Best Buy flyer at the desktops and laptops lately?
AMD really is similar to Nvidia now - now fabless and diversifying into different markets as traditional CPU and GPU businesses gradually change.
We need the competition, and I think their products are better than ever. Honestly 7, 10 years ago I didn't use AMD because the third party chipsets were unacceptable IMO. Now? They seem to be as rock solid as Intel, so I've been recommending their stuff left and right.
On the low end they're close to being the only company that can give you a balanced system, and as far as I'm concerned they're still perfectly reasonable at the mid range too.
I doubt Intel's chips would be half as powerful or fast if not for AMD constantly improving too. I've got my first AMD system (that actually works right at least...3rd party chipsets always gave me problems and ended up being replaced for Intel), and I'm seriously considering a Trinity notebook in a few months.
1.) The Article fails to write about the growth in sales for AMD in a seasonally down quarter. AMD actually sold more processors that last quarter of last year as first quarter is a seasonally weak quarter. INTEL;s sales were down 7-10% as well. So AMD on contrary did very well.
2.) The article says " In any case, not counting their problems with GloFo, AMD’s non-GAAP net income for Q1’12 would have been $92M, which is a small but notable increase over their non-GAAP net income of $56M in Q1’11" Is an increase from $56 to $92 a "SMALL increase". Dude its more like its nearly doubled.
3.) $703 charge is not all cash related impact and its also not solely related to GloFlo. Its a total charge for GloFlo and Sea Micro acquisition.
4.) The article says "All things considered Bulldozer doesn’t seem to have lit a fire under AMD, and the lack of new APUs to replace Brazos isn’t helping, but at the very least AMD is holding their ground." Now its seems its true that Bulldozer is not making much money for AMD but they did mention that sales of Bulldozer doubled from last quarter. So it could be assumed that its faring a bit better and the market adoption is continuing, albeit slowly and at not much profit to AMD. Also Brazos is doing extremely well for AMD. Have you seen any of the Best buys, Amazons, Tiger direct etc websites. E and C processors are really selling well and doing well for AMD, so why would they be in a hurry to replace those. Also Brazos 2 is already shipping now.
I don't mean to be rude or anything but feel that as a JOURNALIST, complete information were not provided which should be the aim of any article. A reader would likely think that AMD is losing money as it was 4 years ago but the truth is that they are actually doing better and making some money. Do you know that AMD has reduced its debt from over $6B after ATI acquisition to close to $2B.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
27 Comments
Back to Article
Bull Dog - Friday, April 20, 2012 - link
Thanks for taking the time to distill down what the numbers really mean.fredisdead - Monday, April 23, 2012 - link
check amds stock price lately, or search on this term ...'interlagos design win'
anand, LOL
silverblue - Friday, April 20, 2012 - link
Wouldn't this be responsible for a good proportion of the loss, or will it fall under another period?Ryan Smith - Friday, April 20, 2012 - link
Technically speaking the acquisition is nearly neutral. AMD spent some cash to acquire an asset, which means the books simultaneously reflect the loss of cash and the gain of an asset.In practice AMD actually came out ahead on the Seamicro acquisition: ignoring what they directly paid for Seamicro, they had 6mil in acquisition costs (that is, costs for lawyers and such), but realized a tax benefit of 36mil, meaning they came out 30mil ahead.
Lonyo - Friday, April 20, 2012 - link
It falls under the current period and other periods. It impacts the balance sheet more, as it is capital expenditure, and doesn't have a significant impact on the income statement, where you see profit or losses.Rather than being reflected as an expense for the current year, the cost is amortised over the life of the acquisition. They pay $334m now, and if it had a life of 10 years or so, you would might see an expense of $33.4m/yr, give or take, reflecting the decline in value of the purchase over time.
silverblue - Friday, April 20, 2012 - link
Many thanks to you both; I must confess that I know little about financial dealings. :)shriganesh - Friday, April 20, 2012 - link
Ryan! I thought your were a GPU and graphics specialist! But now you write articles as though you are a financial analyst! Well done and great job!Wreckage - Friday, April 20, 2012 - link
Unless AMD finds a buyer soon, I imagine bankruptcy is on the horizon.BSMonitor - Friday, April 20, 2012 - link
They are good now. They now compete solely like nVidia. No Fabs. No losses from fabs. Let the Arabs in Abu Dhabi take the losses at GloFo. They should have done this long ago.What this really proves is how difficult and expensive it is to continually upgrade Fabs to new process nodes. The fact that Intel does it so well, now, is simply a testament to their ability to learn from their own past. Too bad they will never offer to manufacture other companies designs at their Fabs. Could be interesting.
silverblue - Friday, April 20, 2012 - link
And, despite many people calling NVIDIA's demise over the past couple of years, it's never been remotely close to happening.It's great to have fabs... if it didn't cost so much to upgrade them and work your way towards a mature process, that is. As AMD don't have to pay GloFo for failures, the only thing AMD can lose money on is not getting enough product back after aforementioned failures. I'm definitely simplifying it, though.
Fab 8's completion should provide a decent boost to 28nm production during the second half of this year, should AMD decide to sign up for it.
silverblue - Friday, April 20, 2012 - link
Hehe ignore the last part; AMD aren't exactly going to go for something they've paid to get out of.silverblue - Friday, April 20, 2012 - link
LOL I may have to ignore the ignore; according to Charlie, apparently AMD has signed a deal with GloFo to produce Kaveri on their 28nm process for release early 2013.http://semiaccurate.com/2012/04/20/where-will-amd-...
JarredWalton - Friday, April 20, 2012 - link
Charlie? Here's your salt shaker.... ;-)Spoelie - Friday, April 20, 2012 - link
Hey, he's often right on the mark. Forecasted Kepler would be the winner vs GCN 3 months ahead of launch, the only site where I read that, and lo and behold...http://semiaccurate.com/2012/01/19/nvidia-kepler-v...
Or was there a story here about that as well? Reading the comments of that article now in retrospect is kind of funny.
Kjella - Saturday, April 21, 2012 - link
"No losses from fabs." Well, perhaps not directly but AMD lose money when the foundries don't get the process ready in time. They lose money when they have to cancel gate-first designs at GloFo and start over with gate-last at TSMC. Against nVidia that's probably not a big deal as they're in the same boat but versus Intel it's a very big deal. Not only because of cost but also because of power efficiency which means longer battery life and cheaper data centers. For once Intel's schedule has slipped a bit too letting AMD get off easy in Q1 but this coming week I expect Ivy Bridge. And comparing AMDs roadmap with Intel's tick-tock roadmap, it's not going to get easier...bassbeast - Thursday, April 26, 2012 - link
I think you are wrong, and here is why...CONSUMERS! you see consumers don't read benchmarks, they don't care about logos, all they know is those AMD netbooks and laptops are REALLY affordable while letting them do light gaming and more importantly letting them enjoy HD video with zero stutter or lag. I mean you can get a nice AMD Brazos netbook in the $350-$430 range (I lucked out and hit a sale and got an E350 for $350 with an 8Gb of RAM upgrade AND a carrying case) that they can not only carry but pop an HDMI cable into and make it into an HTPC you can carry around.The two big screw ups AMD did IMHO was killing Thuban (which made a great affordable desktop line and thanks to turning off cores/cache could cover the ENTIRE Phenom II and Athlon line while giving those willing to core unlock a chance at a free upgrade) and the killing of the Brazos update Krishna but knowing what we now know about GloFlo it was probably smart to kill it and go for an update of Brazos with TSMC instead.
Walk into any Walmart or other B&M store and you'll see Fusion stickers everywhere and talking to friends that work at these shops they sell like hotcakes. I think they have a real winner with the Bobcat design and if they can make say a 4 or 6 core version with a more midrange GPU this will further cement that big chunk of the consumer market because bobcats are cheap to make and run well. I think we can all agree that Bulldozer is a failure but if they can get the new GPU, which they are trading VLIW for vector based BTW, they should be able to salvage the design. You see the weak spot of the BD/PD design is two integers but only one FP unit per pair, but with vector basing they should be able to use the built in GPU as a super powered FP unit.
So I'd say the future looks bright for AMD, especially once they are completely done with GloFlo as anything but one contractor of many. Sure they won't be beating Intel but as Jobs said you don't HAVE to kill the giant to carve out a profitable niche and frankly AMD does make a good value for consumers. Who doesn't like netbooks that get 6+ hours, plays 1080p, never gets hot, and is affordable to boot?
Wreckage - Friday, April 27, 2012 - link
Tablets are starting to eat into the netbook market and ARM netbooks are on the way. So that segment will be shrinking for AMD.medi01 - Monday, April 30, 2012 - link
Except he was talking notebooks, not netbooks. And as it stands now, tablets are toys (and yes, I own one) you can't compare them to notebooks.Shadowmaster625 - Friday, April 20, 2012 - link
When they release a cpu that is a whopping 3 times slower than intel at superpi, is it any surprise they are not making money? They should be bankrupt after pulling that stunt.Mathos - Friday, April 20, 2012 - link
Yes because as we all know.. Super Pi is the be all end all of computer productivity or gaming apps...... I guess Intel deserved to go bankrupt over the p4?medi01 - Monday, April 30, 2012 - link
Uhm, somehow it didn't work other way round, did it? Oh, and on top of being SLOWER and more power hungry, Intel's netburst CPUs were also more expensive.Beenthere - Friday, April 20, 2012 - link
Without the GF contract buy out AMD would have netted ~$200 million so not a bad quarter at all.Southernsharky - Friday, April 20, 2012 - link
Whether you want to call these loses Gap or Non-Gap or related to or caused by GloFlo or not is meaningless. The bottom line is that the company had to pay out a lot of money and lost a lot of money.The fact is that these expenses represent loses to the company due to bad business decisions which were made in the past and are reflected in the current.
And frankly, given the poor state of AMD's current product offerings, one has to wonder if these financial burdens won't be the straw that killed the camel. Every dime spent on this and on legal costs represents a dime the company can not spend on R&D.
Anyone remember AMD laying off 10 percent of its workforce awhile back? Think it had nothing to do with this? ? ? You probably also believe in Santa.
Taft12 - Monday, April 23, 2012 - link
Uhh, AMD has successful product offerings in every CPU and GPU segment except for high-end desktop CPUs, which is quite tiny (although quite profitable). Have you looked in a Best Buy flyer at the desktops and laptops lately?AMD really is similar to Nvidia now - now fabless and diversifying into different markets as traditional CPU and GPU businesses gradually change.
CeriseCogburn - Monday, April 30, 2012 - link
And amd just lost almost a billion dollars.Wolfpup - Friday, April 20, 2012 - link
We need the competition, and I think their products are better than ever. Honestly 7, 10 years ago I didn't use AMD because the third party chipsets were unacceptable IMO. Now? They seem to be as rock solid as Intel, so I've been recommending their stuff left and right.On the low end they're close to being the only company that can give you a balanced system, and as far as I'm concerned they're still perfectly reasonable at the mid range too.
I doubt Intel's chips would be half as powerful or fast if not for AMD constantly improving too. I've got my first AMD system (that actually works right at least...3rd party chipsets always gave me problems and ended up being replaced for Intel), and I'm seriously considering a Trinity notebook in a few months.
tonny b - Sunday, April 22, 2012 - link
There are few Discrepancies in the article.1.) The Article fails to write about the growth in sales for AMD in a seasonally down quarter. AMD actually sold more processors that last quarter of last year as first quarter is a seasonally weak quarter. INTEL;s sales were down 7-10% as well. So AMD on contrary did very well.
2.) The article says " In any case, not counting their problems with GloFo, AMD’s non-GAAP net income for Q1’12 would have been $92M, which is a small but notable increase over their non-GAAP net income of $56M in Q1’11" Is an increase from $56 to $92 a "SMALL increase". Dude its more like its nearly doubled.
3.) $703 charge is not all cash related impact and its also not solely related to GloFlo. Its a total charge for GloFlo and Sea Micro acquisition.
4.) The article says "All things considered Bulldozer doesn’t seem to have lit a fire under AMD, and the lack of new APUs to replace Brazos isn’t helping, but at the very least AMD is holding their ground." Now its seems its true that Bulldozer is not making much money for AMD but they did mention that sales of Bulldozer doubled from last quarter. So it could be assumed that its faring a bit better and the market adoption is continuing, albeit slowly and at not much profit to AMD. Also Brazos is doing extremely well for AMD. Have you seen any of the Best buys, Amazons, Tiger direct etc websites. E and C processors are really selling well and doing well for AMD, so why would they be in a hurry to replace those. Also Brazos 2 is already shipping now.
I don't mean to be rude or anything but feel that as a JOURNALIST, complete information were not provided which should be the aim of any article. A reader would likely think that AMD is losing money as it was 4 years ago but the truth is that they are actually doing better and making some money. Do you know that AMD has reduced its debt from over $6B after ATI acquisition to close to $2B.
None the less thanks for the article.