ADVANCED MICRO DEVICES AMD
January 22, 2011 - 10:52pm EST by
YCOMBINATOR
2011 2012
Price: 7.54 EPS $0.64 $0.00
Shares Out. (in M): 682 P/E 12.0x 0.0x
Market Cap (in M): 5,140 P/FCF NM 0.0x
Net Debt (in M): 700 EBIT 848 0
TEV: 5,810 TEV/EBIT 6.8x 0.0x

Sign up for free guest access to view investment idea with a 45 days delay.

 

Description

KEY TAKEAWAYS

 

  • After being essentially noncompetitive with Intel for the past four years, Advanced Micro Devices' new x86 server CPUs slated for release in the second quarter of 2011 should help the firm recapture lost market share. We believe the stock has the potential for substantial appreciation as AMD's results improve due to market share gains.
  • The last time AMD fielded a very competitive server CPU, it captured 20 points of server market share from 2003 to 2006 and the stock increased from the midsingle digits to $40.
  • AMD's new chips, code-named Interlagos (16-core) and Valencia (8-core), are based on Bulldozer, the company's first new architecture since the original Opteron in 2003. AMD walked a down much different architectural path than Intel with Bulldozer, which will allow it to fit more cores within a given silicon area and power consumption budget. Servers with many more cores are attractive to IT departments running virtualized datacenters.
  • We estimate the x86 server CPU market to be roughly $9 billion annually. Intel currently owns about 95% of the market, which means AMD has very little share left to lose.
  • We estimate that if AMD were to capture an incremental $225 million of quarterly server revenue (a $900 million annual run rate, or 10 points of market share) at a 65% gross margin, the revenue would flow through at about a 55% operating margin. All other things equal, AMD's operating margins would increase from their current 8% to 14%.
  • AMD currently trades at about 7 times EV/EBITDA. Since spinning off its chip fabrication plants to GlobalFoundries and winning a $1.25 billion antitrust settlement with Intel, AMD is now a much less capital-intensive business with a clean balance sheet. We believe the shares are attractively priced ahead of the Bulldozer introduction in mid-2011, which could double AMD's earnings.

 

AMD AND INTEL: OPPORTUNITY ABOUNDS

As the 10-year stock price chart of AMD shows, the market is almost comically bad at pricing in step changes in relative product quality between Intel and AMD, even when newly released chips from both companies have been fully benchmarked and reviewed. We'll briefly review the past decade of the Intel/AMD rivalry to highlight the opportunity available to investors who pay attention.

Chart at http://2.bp.blogspot.com/_BbvwQAOOzxw/TNhP5yrET5I/AAAAAAAAAkQ/pkXNpNV7DIs/s1600/amd-intc.JPG  

 

2003: AMD Introduces the Opteron

There are many similarities between the server CPU market in 2002 and 2010. Much like today, in 2002, Intel dominated the market with 90%-plus share and AMD was relegated to picking up scraps. However, in early 2003, AMD introduced the original Opteron, which handily took the performance crown away from Intel and eventually captured roughly 25% of the server market. Again much like today, one reason the Opteron was successful was that AMD took a much different architectural approach than Intel.

Intel

AMD

Outcome

64-bit product was Itanium, which was a non-x86 instruction set.

Developed x64 as 64-bit extensions to the industry standard x86 instruction set.

Industry, including Intel, standardized on AMD's x64. Itanium effectively dead.

NetBurst architecture (Pentium 4) focused on high clock rate (Ghz). Resulted in a hot, power-hungry chip that did less work per cycle than Opteron.

The Opteron architecture focused on IPC (instructions per cycle), which made the Opteron a more powerful and energy-efficient chip.

Opteron went on to capture 25% of the server CPU market. Intel killed NetBurst and adopted the IPC approach for its Core 2 architecture.

When the Opteron was introduced and benchmarked in April 2003, AMD's share price was languishing in the midsingle digits even though it was clear that AMD would have Intel beat for the upcoming processor generation. The following chart published on the AnandTech website on April 23, 2003 (http://www.anandtech.com/show/1099/5), showed the Opteron 244 outperforming the Intel Xeon by 13% in a CPU-bound benchmark, even though the Xeon was clocked a full gigahertz higher.

The market eventually caught on as AMD racked up market share and the stock peaked at more than $40 in early 2006.

 

2006: Intel Introduces the Core 2

We argue that one reason the Intel/AMD trade is so lucrative is that investors tend to extrapolate current processor generation trends far into the future. However, each processor generation represents an opportunity for step changes in relative product quality between Intel and AMD. After the NetBurst and Itanium debacle, Intel executed an amazing turnaround. The company scrapped its processor roadmap and refocused on an architecture developed by Intel's mobile group in Israel. That architecture was the basis for Intel's Pentium M notebook chips, and its descendant was released in mid-2006 as the Core 2 family of chips.

The Core 2 architecture was a massive step change improvement in the two critical metrics of price/performance and performance/watt. Benchmarks showing that Core 2 outperformed AMD by 25%-30% began appearing online in March 2006 when AMD shares were trading in the mid-$30s (e.g., http://www.anandtech.com/show/1966/1). Once the details of Core 2 were known, AMD shares were an undeniable short. Core 2, along with its successor Nehalem, allowed Intel to use its traditional advantages of scale and process technology to recapture market share and crush AMD's margins.

 

AMD'S BULLDOZER ARCHITECTURE

AMD has been noncompetitive in server CPUs during the past few years, but we expect that will change with the introduction of chips based on the Bulldozer architecture in mid-2011. Bulldozer will be AMD's first new architecture since the original Opteron in 2003, and much like the Opteron, AMD has again taken a very different architectural path than Intel.

We emphasize that we do not expect Bulldozer to outperform Intel's 2011 server chips to the degree that the Opteron or Core 2 did when they were introduced. We simply believe the market is not pricing in any measure of success for Bulldozer, which, given our current knowledge of the architecture, is too pessimistic. AMD has very little server market share left to lose, and this is a case where AMD will feel better once it stops getting repeatedly punched in the face by Intel.

 

Intel Is a Ferrari, AMD Is a 16-Passenger Van

AMD's key architectural decision with Bulldozer is to share certain chip circuitry such as the floating point unit among two integer units, which allows AMD to fit more cores within a specified silicon area and power consumption budget. Each of the two integer units in a Bulldozer module will appear as a separate core to the operating system. By comparison, Intel's chips sport cores with much less shared circuitry and a 1:1 ratio of integer and floating point units. 

Bulldozer chips are likely to face off against chips based on Intel's new Sandy Bridge architecture. Given what we currently know, we believe Sandy Bridge will win hands down against Bulldozer in single-threaded processor performance. However, AMD is focused on overall chip throughput, rather than pure performance. There are certain workloads where Bulldozer's greater core count and throughput will be an advantage. To use an analogy, Sandy Bridge is a finely tuned Ferrari, while Bulldozer is 16-passenger van. Sandy Bridge will get you where you need to go quicker, but Bulldozer will let you bring along 15 of your closest friends for the ride. Sometimes you just need the van.

 

The Virtualized Datacenter

One area where Bulldozer's high core count will be an advantage is virtualized environments. A majority of enterprise application workloads are integer based (thus we do not believe Bulldozer will lose much by sharing a floating point unit among two integer units) and IT administrators tend to follow a one-virtual-machine-per-core rule. Every physical server entails a fixed cost in terms of datacenter space and power consumption. Thus, the more cores that can be packed into a physical server, the more virtual machines that server can host, with the upshot that fewer physical servers will be needed. For example, a server with four CPU sockets can be outfitted with four 16-core AMD Interlagos Bulldozer chips for 64 cores versus four 8-core Intel Xeon chips for 32 cores. A physical server with Bulldozer chips can potentially host twice as many virtual machines as one with Intel chips (assuming memory and I/O are also scaled appropriately).

Virtualization is a secular growth trend in technology, and we expect Bulldozer to be especially attractive for enterprises looking to consolidate their servers with virtualization.

 

POTENTIAL FINANCIAL IMPACT OF BULLDOZER

Server CPUs sell at much higher prices than desktop/notebook CPUs, and accordingly have much higher margins. We estimate that Intel's server CPU gross margins are 70%-75%. In modeling the potential financial impact of Bulldozer, we make three key assumptions: (1) AMD is able to capture an incremental 10 points of server market share, resulting in an additional $225 million of quarterly revenue ($900 million annual run rate); (2) AMD realizes 65% gross margins on Bulldozer server CPUs; and (3) AMD's operating costs stay relatively flat, given the high fixed cost nature of the business. As shown in the following table, all other things equal, an incremental $225 million flowing through at a 65% gross margin should allow AMD to approximately double its operating income.

 

Hypothetical AMD Quarterly Results

Table at http://2.bp.blogspot.com/_BbvwQAOOzxw/TNhRo_QwhrI/AAAAAAAAAkg/1mRdm2DHBjM/s1600/amdtable2.GIF  

 

WE BELIEVE AMD IS A TIMELY LONG IDEA AHEAD OF BULLDOZER

After spinning off its capital-intensive chip fabrication plants to GlobalFoundries and fortifying its balance sheet with a $1.25 billion antitrust settlement from Intel, AMD is well positioned to reap the financial benefits of its Bulldozer introduction. AMD shares currently trade at ~$7, or about 7 times EV/EBITDA and a 50% discount to our $15 fair value estimate. Investors have historically missed the important inflection points in the Intel/AMD story, and we believe Bulldozer is one of these inflection points. AMD shares offer significant return potential over the next 24 months as the market prices in much better financial performance from the company.
 
UPDATE
 
This was my application idea that was originally submitted in early November 2010. Here are some key updates:
 
1) AMD fired CEO Dirk Meyer, which killed short-term momentum after the Citi upgrade b/c of uncertainty. Whether the firing was justified is something upon which reasonable people can disagree, but it was certainly poorly timed ahead of the 2011 product refresh for AMD. However, mitigating that is the fact that AMD's roll out is coming along nicely. AMD's low-power Brazos ("Bobcat" cores) platform was introduced at CES with numerous design wins (1 million units shipped).
 
2) Llano, AMD's mainstream APU (combined CPU and GPU) for notebooks and desktops is sampling in high volume currently and on track for a Q2 launch. Llano's CPU will be slower than Intel's Sandy Bridge, but the GPU performance will absolutely crush Sandy Bridge's GPU. The desktop version of Bulldozer is also sampling and on schedule for a Q2 launch.
 
3) I focused the write-up on AMD's server opportunity, but the success of Brazos, Llano, and Bulldozer are fairly independent events. The desktop and notebook CPU markets are about $9 billion and $18 billion, respectively. Thus, if AMD can, say, gain 10 points of share in the notebook market with Llano and its successors (where it is currently almost non-existent), that is a ~$800 million improvement in EBIT (assuming 55% GMs for notebook chips). With the full product refresh, there are many ways to win with AMD.
 
4) Early Bulldozer performance leaks are promising (http://www.fudzilla.com/processors/item/21512-bulldozer-to-come-close-to-core-i7). If this is indeed true, the 16-core Bulldozer will likely take the performance crown away from Intel, until at least Q4 2011 when 8 core Sandy Bridge chips are launched.

Catalyst

Over the next 24-36 months, we believe market share gains and improved financial results will be catalysts.

Putting 2011 on the Street's radar: AMD's refresh of its entire product line is still flying under the Street's radar. We believe getting the 2011 story out to the Street may be a positive catalyst given the strength of the new products. AMD's new product line is centered on two new architectures and the integration of the graphics processor (GPU) on the same silicon die. The first new architecture is "Bobcat" for low power consumption applications such as netbooks and thin-and-light notebooks. The second new architecture is "Bulldozer" for servers and high-performance desktops.

In addition to Bulldozer, AMD's Bobcat based products also look very compelling. Bobcat is targeted toward netbooks and thin-and-light notebooks, a market where AMD has been completely absent. The Ontario and Zecate chips featuring Bobcat CPU cores and ATI-based GPU will have Intel's (INTC) Atom beat on every performance metric. These chips, had they been available, would have been a perfect fit for Apple's (AAPL) new MacBook Air.

Clarity around launch dates: One risk to AMD's 2011 story is the manufacturing ramp of these new chips at both GlobalFoundries and TSMC (TSM). If AMD provides detail on the progress of the ramp that helps to reduce uncertainty around ship dates, this would be a positive catalyst.

Performance data: AMD was rightly been coy about performance data to avoid giving Intel a target, but based on our research we believe both Bobcat and Bulldozer will offer significant performance improvements over AMD's existing product line. We are very confident that Bobcat will outperform Atom and AMD's integrated GPU will outperform Intel's graphics. If AMD releases more performance data that helps to fill in the question marks in the table above, we believe it will be a positive catalyst. 
    sort by   Expand   New

    Description

    KEY TAKEAWAYS

     

    • After being essentially noncompetitive with Intel for the past four years, Advanced Micro Devices' new x86 server CPUs slated for release in the second quarter of 2011 should help the firm recapture lost market share. We believe the stock has the potential for substantial appreciation as AMD's results improve due to market share gains.
    • The last time AMD fielded a very competitive server CPU, it captured 20 points of server market share from 2003 to 2006 and the stock increased from the midsingle digits to $40.
    • AMD's new chips, code-named Interlagos (16-core) and Valencia (8-core), are based on Bulldozer, the company's first new architecture since the original Opteron in 2003. AMD walked a down much different architectural path than Intel with Bulldozer, which will allow it to fit more cores within a given silicon area and power consumption budget. Servers with many more cores are attractive to IT departments running virtualized datacenters.
    • We estimate the x86 server CPU market to be roughly $9 billion annually. Intel currently owns about 95% of the market, which means AMD has very little share left to lose.
    • We estimate that if AMD were to capture an incremental $225 million of quarterly server revenue (a $900 million annual run rate, or 10 points of market share) at a 65% gross margin, the revenue would flow through at about a 55% operating margin. All other things equal, AMD's operating margins would increase from their current 8% to 14%.
    • AMD currently trades at about 7 times EV/EBITDA. Since spinning off its chip fabrication plants to GlobalFoundries and winning a $1.25 billion antitrust settlement with Intel, AMD is now a much less capital-intensive business with a clean balance sheet. We believe the shares are attractively priced ahead of the Bulldozer introduction in mid-2011, which could double AMD's earnings.

     

    AMD AND INTEL: OPPORTUNITY ABOUNDS

    As the 10-year stock price chart of AMD shows, the market is almost comically bad at pricing in step changes in relative product quality between Intel and AMD, even when newly released chips from both companies have been fully benchmarked and reviewed. We'll briefly review the past decade of the Intel/AMD rivalry to highlight the opportunity available to investors who pay attention.

    Chart at http://2.bp.blogspot.com/_BbvwQAOOzxw/TNhP5yrET5I/AAAAAAAAAkQ/pkXNpNV7DIs/s1600/amd-intc.JPG  

     

    2003: AMD Introduces the Opteron

    There are many similarities between the server CPU market in 2002 and 2010. Much like today, in 2002, Intel dominated the market with 90%-plus share and AMD was relegated to picking up scraps. However, in early 2003, AMD introduced the original Opteron, which handily took the performance crown away from Intel and eventually captured roughly 25% of the server market. Again much like today, one reason the Opteron was successful was that AMD took a much different architectural approach than Intel.

    Intel

    AMD

    Outcome

    64-bit product was Itanium, which was a non-x86 instruction set.

    Developed x64 as 64-bit extensions to the industry standard x86 instruction set.

    Industry, including Intel, standardized on AMD's x64. Itanium effectively dead.

    NetBurst architecture (Pentium 4) focused on high clock rate (Ghz). Resulted in a hot, power-hungry chip that did less work per cycle than Opteron.

    The Opteron architecture focused on IPC (instructions per cycle), which made the Opteron a more powerful and energy-efficient chip.

    Opteron went on to capture 25% of the server CPU market. Intel killed NetBurst and adopted the IPC approach for its Core 2 architecture.

    When the Opteron was introduced and benchmarked in April 2003, AMD's share price was languishing in the midsingle digits even though it was clear that AMD would have Intel beat for the upcoming processor generation. The following chart published on the AnandTech website on April 23, 2003 (http://www.anandtech.com/show/1099/5), showed the Opteron 244 outperforming the Intel Xeon by 13% in a CPU-bound benchmark, even though the Xeon was clocked a full gigahertz higher.

    The market eventually caught on as AMD racked up market share and the stock peaked at more than $40 in early 2006.

     

    2006: Intel Introduces the Core 2

    We argue that one reason the Intel/AMD trade is so lucrative is that investors tend to extrapolate current processor generation trends far into the future. However, each processor generation represents an opportunity for step changes in relative product quality between Intel and AMD. After the NetBurst and Itanium debacle, Intel executed an amazing turnaround. The company scrapped its processor roadmap and refocused on an architecture developed by Intel's mobile group in Israel. That architecture was the basis for Intel's Pentium M notebook chips, and its descendant was released in mid-2006 as the Core 2 family of chips.

    The Core 2 architecture was a massive step change improvement in the two critical metrics of price/performance and performance/watt. Benchmarks showing that Core 2 outperformed AMD by 25%-30% began appearing online in March 2006 when AMD shares were trading in the mid-$30s (e.g., http://www.anandtech.com/show/1966/1). Once the details of Core 2 were known, AMD shares were an undeniable short. Core 2, along with its successor Nehalem, allowed Intel to use its traditional advantages of scale and process technology to recapture market share and crush AMD's margins.

     

    AMD'S BULLDOZER ARCHITECTURE

    AMD has been noncompetitive in server CPUs during the past few years, but we expect that will change with the introduction of chips based on the Bulldozer architecture in mid-2011. Bulldozer will be AMD's first new architecture since the original Opteron in 2003, and much like the Opteron, AMD has again taken a very different architectural path than Intel.

    We emphasize that we do not expect Bulldozer to outperform Intel's 2011 server chips to the degree that the Opteron or Core 2 did when they were introduced. We simply believe the market is not pricing in any measure of success for Bulldozer, which, given our current knowledge of the architecture, is too pessimistic. AMD has very little server market share left to lose, and this is a case where AMD will feel better once it stops getting repeatedly punched in the face by Intel.

     

    Intel Is a Ferrari, AMD Is a 16-Passenger Van

    AMD's key architectural decision with Bulldozer is to share certain chip circuitry such as the floating point unit among two integer units, which allows AMD to fit more cores within a specified silicon area and power consumption budget. Each of the two integer units in a Bulldozer module will appear as a separate core to the operating system. By comparison, Intel's chips sport cores with much less shared circuitry and a 1:1 ratio of integer and floating point units. 

    Bulldozer chips are likely to face off against chips based on Intel's new Sandy Bridge architecture. Given what we currently know, we believe Sandy Bridge will win hands down against Bulldozer in single-threaded processor performance. However, AMD is focused on overall chip throughput, rather than pure performance. There are certain workloads where Bulldozer's greater core count and throughput will be an advantage. To use an analogy, Sandy Bridge is a finely tuned Ferrari, while Bulldozer is 16-passenger van. Sandy Bridge will get you where you need to go quicker, but Bulldozer will let you bring along 15 of your closest friends for the ride. Sometimes you just need the van.

     

    The Virtualized Datacenter

    One area where Bulldozer's high core count will be an advantage is virtualized environments. A majority of enterprise application workloads are integer based (thus we do not believe Bulldozer will lose much by sharing a floating point unit among two integer units) and IT administrators tend to follow a one-virtual-machine-per-core rule. Every physical server entails a fixed cost in terms of datacenter space and power consumption. Thus, the more cores that can be packed into a physical server, the more virtual machines that server can host, with the upshot that fewer physical servers will be needed. For example, a server with four CPU sockets can be outfitted with four 16-core AMD Interlagos Bulldozer chips for 64 cores versus four 8-core Intel Xeon chips for 32 cores. A physical server with Bulldozer chips can potentially host twice as many virtual machines as one with Intel chips (assuming memory and I/O are also scaled appropriately).

    Virtualization is a secular growth trend in technology, and we expect Bulldozer to be especially attractive for enterprises looking to consolidate their servers with virtualization.

     

    POTENTIAL FINANCIAL IMPACT OF BULLDOZER

    Server CPUs sell at much higher prices than desktop/notebook CPUs, and accordingly have much higher margins. We estimate that Intel's server CPU gross margins are 70%-75%. In modeling the potential financial impact of Bulldozer, we make three key assumptions: (1) AMD is able to capture an incremental 10 points of server market share, resulting in an additional $225 million of quarterly revenue ($900 million annual run rate); (2) AMD realizes 65% gross margins on Bulldozer server CPUs; and (3) AMD's operating costs stay relatively flat, given the high fixed cost nature of the business. As shown in the following table, all other things equal, an incremental $225 million flowing through at a 65% gross margin should allow AMD to approximately double its operating income.

     

    Hypothetical AMD Quarterly Results

    Table at http://2.bp.blogspot.com/_BbvwQAOOzxw/TNhRo_QwhrI/AAAAAAAAAkg/1mRdm2DHBjM/s1600/amdtable2.GIF  

     

    WE BELIEVE AMD IS A TIMELY LONG IDEA AHEAD OF BULLDOZER

    After spinning off its capital-intensive chip fabrication plants to GlobalFoundries and fortifying its balance sheet with a $1.25 billion antitrust settlement from Intel, AMD is well positioned to reap the financial benefits of its Bulldozer introduction. AMD shares currently trade at ~$7, or about 7 times EV/EBITDA and a 50% discount to our $15 fair value estimate. Investors have historically missed the important inflection points in the Intel/AMD story, and we believe Bulldozer is one of these inflection points. AMD shares offer significant return potential over the next 24 months as the market prices in much better financial performance from the company.
     
    UPDATE
     
    This was my application idea that was originally submitted in early November 2010. Here are some key updates:
     
    1) AMD fired CEO Dirk Meyer, which killed short-term momentum after the Citi upgrade b/c of uncertainty. Whether the firing was justified is something upon which reasonable people can disagree, but it was certainly poorly timed ahead of the 2011 product refresh for AMD. However, mitigating that is the fact that AMD's roll out is coming along nicely. AMD's low-power Brazos ("Bobcat" cores) platform was introduced at CES with numerous design wins (1 million units shipped).
     
    2) Llano, AMD's mainstream APU (combined CPU and GPU) for notebooks and desktops is sampling in high volume currently and on track for a Q2 launch. Llano's CPU will be slower than Intel's Sandy Bridge, but the GPU performance will absolutely crush Sandy Bridge's GPU. The desktop version of Bulldozer is also sampling and on schedule for a Q2 launch.
     
    3) I focused the write-up on AMD's server opportunity, but the success of Brazos, Llano, and Bulldozer are fairly independent events. The desktop and notebook CPU markets are about $9 billion and $18 billion, respectively. Thus, if AMD can, say, gain 10 points of share in the notebook market with Llano and its successors (where it is currently almost non-existent), that is a ~$800 million improvement in EBIT (assuming 55% GMs for notebook chips). With the full product refresh, there are many ways to win with AMD.
     
    4) Early Bulldozer performance leaks are promising (http://www.fudzilla.com/processors/item/21512-bulldozer-to-come-close-to-core-i7). If this is indeed true, the 16-core Bulldozer will likely take the performance crown away from Intel, until at least Q4 2011 when 8 core Sandy Bridge chips are launched.

    Catalyst

    Over the next 24-36 months, we believe market share gains and improved financial results will be catalysts.

    Putting 2011 on the Street's radar: AMD's refresh of its entire product line is still flying under the Street's radar. We believe getting the 2011 story out to the Street may be a positive catalyst given the strength of the new products. AMD's new product line is centered on two new architectures and the integration of the graphics processor (GPU) on the same silicon die. The first new architecture is "Bobcat" for low power consumption applications such as netbooks and thin-and-light notebooks. The second new architecture is "Bulldozer" for servers and high-performance desktops.

    In addition to Bulldozer, AMD's Bobcat based products also look very compelling. Bobcat is targeted toward netbooks and thin-and-light notebooks, a market where AMD has been completely absent. The Ontario and Zecate chips featuring Bobcat CPU cores and ATI-based GPU will have Intel's (INTC) Atom beat on every performance metric. These chips, had they been available, would have been a perfect fit for Apple's (AAPL) new MacBook Air.

    Clarity around launch dates: One risk to AMD's 2011 story is the manufacturing ramp of these new chips at both GlobalFoundries and TSMC (TSM). If AMD provides detail on the progress of the ramp that helps to reduce uncertainty around ship dates, this would be a positive catalyst.

    Performance data: AMD was rightly been coy about performance data to avoid giving Intel a target, but based on our research we believe both Bobcat and Bulldozer will offer significant performance improvements over AMD's existing product line. We are very confident that Bobcat will outperform Atom and AMD's integrated GPU will outperform Intel's graphics. If AMD releases more performance data that helps to fill in the question marks in the table above, we believe it will be a positive catalyst. 

    Messages


    SubjectMore Bulldozer performance leaks
    Entry01/24/2011 10:04 PM
    MemberYCOMBINATOR
     
    Apparent leak of a NDA performance slide. The chart shows the 8 core desktop version of Bulldozer outperforming an Intel Core i7 950 by 50% on some multi-threaded applications (taking advantage of Bulldozer's 8 cores). If that is the case, then Bulldozer should handily take down current 4 core Sandy Bridge chips was well.
     
    We'll get more information as more leaks come out, but early indications are that Bulldozer will be more competitive with Sandy Bridge than even I expected when I first wrote the AMD pitch.

    SubjectAny thought on the earnings released today?
    Entry01/25/2011 11:08 AM
    Membermitc567
    What you expected?

    SubjectRE: Questions
    Entry02/01/2011 11:30 AM
    MemberYCOMBINATOR
    Hi Utah, Sorry for the delay in replying.
     
    To be clear, I don't expect Bulldozer to beat Intel's offerings to the same degree that the original Opteron did. AMD has just been so bad since 2006 that it's hard to do any worse in the server market. Even if AMD gains my hypothetical 10 points of server share, it's still 85% INTC, 15% AMD.
     
    CPU architectures have a multi-year development cycle, so when key architectural decisions are made, it can take years for the other player to catch up (it took INTC 3 years from 2003-2006 to come back against the original Opteron and that was literally turning on a dime). AMD is taking a very different architectural path than INTC with Bulldozer (more cores in a given silicon area), so if this approach is successful, INTC can't respond for some time. Sandy Bridge server chips are due Q4'11 and after that is Ivy Bridge (process shrink on Sandy Bridge). INTC's next new architecture is Haswell expected in 1H'2013 and is already under development. AMD could make hay for a few years.
     
    Yes, I've seen GS's reports. I don't know what to say other than I think he's wrong.
     
    As for Q4, nothing unexpected. Quarterly results won't carry much signal until more of the new products are rolled out.

    SubjectBrazos review on Anandtech, Notebookreview
    Entry02/02/2011 10:09 AM
    MemberYCOMBINATOR
    Review of AMD's Brazos platform (Bobcat APU) for ultramobile and low-power applications.
     
     
    "For AMD, Brazos has to be exciting. The company finally has a value offering that it doesn't have to discount heavily to sell. Brazos does very well against Atom on absolute performance, die size and price. The E-350 isn't the most powerful Fusion APU we'll meet, but it's a great way to introduce the family."
     
     
    Conclusion
    At the end of the day, it's hard not to like the new HP Pavilion dm1z. HP and AMD resolved all of the main complaints that people had with the previous generation of AMD-based ultraportbales: Poor battery life, high temperatures, and only adequate performance.

    The new AMD Fusion APU technology really delivers some impressive results in terms of HD video playback, gaming, and even adds some "snappiness" in a basic Windows environment. At the time of this writing (January 2011) consumers will have a hard time finding an Intel-based ultraportable notebook or netbook that delivers better real world performance and battery life across the board for the same price ... and that's saying a lot.

    Several other companies announced other notebooks and netbook alternatives at CES 2011 that use the same AMD E-350 APU, so it's hard to say if the HP Pavilion dm1 is the best of the bunch, but it certainly is better than any netbook we've reviewed to date.


    SubjectDell introduces server with 96 CPU cores
    Entry02/14/2011 05:35 PM
    MemberYCOMBINATOR
    "The company rejects Intel chips for the server, choosing AMD because its chips provide better performance per watt. The server is designed to run scientific or math applications and can also be used in cloud and virtualized computing environments."
     
     
    This is for the existing 12 core Opterons. The story gets much better with Bulldozer b/c Bulldozer will be a much better chip with even more cores (16 cores for the Interlagos version). These high core counts are very well suited for virtualized environments.
     
    Rumors of AMD as a takeout candidate by Dell, which seems nutty to me. I don't think much about M&A, but I will say that the company I think AMD is most valuable to as an acquisition is Apple.

    Subjectrecent takeover speculation
    Entry02/20/2011 02:05 PM
    Membertyler939
    Ycombinator, do you think there is anything to recent speculation, if you can even call it that, of qualcomm or Samsung (or even IBM for that matter) acquiring AMD (see the following link, as referenced in this week's Barrons)?
    http://www.tomshardware.com/news/amd-athlon-amd-opteron-advanced-micro-devices,12236.html

    SubjectHP and the threat from GPUs
    Entry02/22/2011 11:10 PM
    Membertyler939
    I'm still digesting the HP earnings, but it looks like the miss was mostly in PC sales.  I know that HP sells some AMD-based machines.  I wonder if anyone has any thoughts on whether we should read anything through from HP's report to AMD?
     
    Also, a more technological speculation, given the massive parallel processing capabilities of graphics cards, is multithreading in the CPU really the technology of the future for threadable applications?  I just wonder whether having the Ferrari CPU and a Greyhound bus (e.g., a series of NVIDIA cards running the CUDA architecture) isn't the wave of the future for servers and other high-end systems.

    SubjectRE: HP and the threat from GPUs
    Entry02/24/2011 01:35 PM
    MemberYCOMBINATOR
    Hi tyler,
     
    Sorry for the delay in replying. Traveling through London this week.
     
    1) I don't buy the takeover speculation. Can't see why QCOM would want x86 -- I think they would prefer to move ARM upmarket. Samsung? Maybe, if they want to empire build to take on INTC. 
     
    2) If the world heads toward GPU compute, I can't think of a better positioned company than AMD since they can offer both high performance CPU and GPU on a single piece of silicon. We're a long way from GPU compute though since the application support is not there and will take a long, long time to build.
     
    3) The AMD thesis is based on market share gains, so I wouldn't read much into HPQ's earnings right now. Their earnings later in the year when more of AMD's new products are actually shipping. A hot PC market will help, but as long as things don't fall off a cliff 2008-style, AMD's new products will help take share.

    SubjectAMD expanding Bobcat APU production
    Entry02/24/2011 01:38 PM
    MemberYCOMBINATOR
    From DigiTimes:
     
    AMD has recently expanded its 40nm chip orders to foundry house Taiwan Semiconductor Manufacturing Company (TSMC) and is hoping to secure all new capacity from TSMC's 12-inch fab, Fab 14's phase 4 facilities to solve the shortage issue, according to a Chinese-language Commercial Times report.

    Since strong demand for AMD's new 40nm Zacate and Ontario CPUs has caused supply for the two processors to tighten in both Europe and the US, plus AMD is preparing to push the APUs into China in March, the company has been aggressive resolving the shortage issue as well as building its inventory levels, the paper added.


    SubjectAMD pits Llano against Sandy Bridge (with video!)
    Entry03/02/2011 09:32 AM
    MemberYCOMBINATOR
     
    "AMD has posted a rather interesting Youtube clip comparing its upcoming Llano APU to an Intel Sandy Bridge processor.

    The test pitted a Llano-based laptop against an Intel Core i7-2630QM rig in a couple of real life tests and, obviously, AMD came out on top. AMD ran some 3D and office apps side by side, proving that Llano could easily outperform Sandy Bridge in Final Fantasy, with not stuttering or dropped frames.

    AMD's power consumption figures were also better, staying under 48W, while Intel jumped back and forth between 45W and 55W. In a spreadsheet test, and the game running in the background, both processor were equally matched, but when the testers piled on some video and post processing, Llano pulled ahead. In addition, power consumption was still under 48W for the AMD, while Sandy went over 70W. AMD then launched a 3D modeling application, only to see Sandy Bridge struggle to cope, while Llano managed to deal with all four workloads.

    The test rigs were closely matched in terms of hardware. The AMD testbed was based on an A8-3510MX APU with HD 6620M graphics and the A70M Fusion Controller Hub (FCH). Intel sported a 2GHz Core i7-2630QM, HD 3000 graphics and H67 chipset. Both machines had 4GB of DDR3 and the exact same 128GB SSD.

    AMD's Llano clearly has a lot to offer, although it probably won't need to deal with such heavy workloads in real life and Intel will still dominate the high-end. However, Llano could easily hit the sweet spot and score plenty of design wins in the much broader mainstream market. In addition, Llano is expected to end up somewhat cheaper than comparable Intel Sandy Bridge processors and AMD might have a pretty good chance of unsettling Intel in the mid range mobile market." 

    Subjecthistorical R&D, intel vs. AMD
    Entry03/07/2011 02:57 PM
    Membersurf1680
    Intel has been spending 5x as much money on R&D as AMD for the last decade.  AMD spins off it's foundry  (potentially making it harder to control mfg costs?).  I understand that AMD could win a battle or two in the next few cycles of chipsets or whatever, but how, long term, will Intel not dominate?  Intel's valuation & balance sheet don't look half bad, either.   
     
     

    SubjectRE: historical R&D, intel vs. AMD
    Entry03/07/2011 06:32 PM
    MemberYCOMBINATOR
    Intel will always dominate the CPU market for all the reasons you mention (R&D spend, more advanced manufacturing, etc...). AMD is not a long term holding -- it's about picking the right point in the product vs. valuation cycle to own the stock (see 2003)
     
    I've been both long and short AMD over the last decade. I imagine if AMD's upcoming product cycle is successful, investors get overexcited about AMD's prospects, and Intel is about to hit back with new products (see 2006), I'll be writing up AMD as a short.

    Subjectycom, can you suggest a short to pair
    Entry03/09/2011 03:19 PM
    Membertyler939
    Is there a short in tech that you would recommend against being long AMD.  INTC looks to cheap to short, even if the thesis plays out.  Is there another short you can recommend to get rid of some of the tech exposure?

    SubjectRE: ycom, can you suggest a short to pair
    Entry03/13/2011 11:36 AM
    MemberYCOMBINATOR
    Hi tyler, so after thinking about it for a bit, I'll suggest NVDA.
     
    1) GPU business will be challenged as AMD and INTC integrate better GPUs with their CPUs.
     
    2) NVDA will try to compensate by going to the high-performance computing (HPC) market with their GPUs and into the smartphone/mobile device/tablet market w/ Tegra.
     
    2a) The HPC market might turn out alright, but it's not big in terms of size.
     
    2b) NVDA is just another ARM-based chipmaker in the mobile market, competing against QCOM, BRCM, MRVL, etc... The stock ran up on the mobile/tablet hype (and has come back down quite a bit), but I see the mobile chip space as being competitive.
     
    Also, to the extent you believe there is a "tablet bubble", NVDA may be a good short. The question is whether the iPad is like the iPod (category-killer) or iPhone (viable alternatives like Android). I'm leaning toward the former, where people don't buy a tablet, they buy an iPad.

    Subjectycombinator, I appreciate the updates
    Entry03/15/2011 12:31 AM
    Membertyler939
    Would you mind giving your thoughts on how you believe AMD will be affected by the situation in Japan? Thanks.

    SubjectRE: ycombinator, I appreciate the updates
    Entry04/04/2011 06:54 PM
    MemberYCOMBINATOR
    Hi Tyler, sorry didn't see your msg until now.
     
    There's a lot of PC-related production in Japan, so an impairment of the supply of DRAM or laptop batteries would not be good for the PC industry for whatever duration it takes to replace that supply. From your msgs, it appears you run long/short, so this is risk you can hedge out.
     
    The AMD thesis really is just share gains based on new and improved products. If the PC market is flat, but AMD gains 10 points of share, I think it'll work out just fine for us.

    Subject10-core Xeon, 22nm, and Apple
    Entry04/05/2011 01:00 PM
    Membertyler939
    YCOMBINATOR, Intel launched it's 10-core Xeon line (E7, formerly Westbridge EX).  Does this influence your views on the advantages of extra cores (i.e., now you can get 40 Intel cores instead of 32 to compete with AMD's 64 cores)?  Also, where does AMD stand versus IBM in terms of 22nm fabrication?
    Also, I know you mentioned Apple as a potential acquirer of AMD.  What would Apple's strategic rationale for that acquisition be?

    SubjectRE: 10-core Xeon, 22nm, and Apple
    Entry04/05/2011 01:50 PM
    Membertyler939
    I know you never pitched M&A as part of your thesis.  You mentioned a while back though that they would be most valuable to Apple.  I assume there would be severe antitrust issues if Apple intended to restrict AMD from selling to others, so what would be the rationale?  Thanks.

    SubjectRE: RE: 10-core Xeon, 22nm, and Apple
    Entry04/05/2011 04:47 PM
    MemberYCOMBINATOR
    To address your other two questions:
     
    1) Did you mean AMD vs INTC on the 22nm transition, or did you really mean IBM and their Power series of chips?
     
    2) As for AAPL, I agree there are anti-trust issues, but let's just imagine they don't exist for this hypothetical. Steve Jobs/AAPL understands the value of owning the full soup-to-nuts hardware/software stack. That's why AAPL bought PA Semi/Intrinsity and now has their own custom ARM chip for the iPhone/iPad. Also why AAPL owns a stake in Imagination Technologies (along w/ INTC). Silicon is a key way to differentiate your products. Buying AMD and perhaps keeping some chips AAPL only, say for a very specialized product like the MacBook Air would give AAPL an advantage in the PC market. The MacBook Air uses a old circa 2008 Core 2 Duo chip w/ NVDA graphics b/c INTC's graphics suck. Stick an AMD Fusion chip in there and voila great CPU/GPU performance at low power... and no one else can have it.
     
    AAPL also can't be too happy with INTC wanting to compete in the tablet/smartphone market. Imaginary acquisition aside, I think there will be an AMD Fusion chip inside a Mac in the not too distant future.

    SubjectRE: AMD 5/31 event, Llano & Bulldozer pricing
    Entry05/23/2011 05:48 PM
    MemberYCOMBINATOR
    The pricing for Llano and Bulldozer in the Digitimes article tells me two things:
     
    1) AMD's margins should expand nicely. Accordingly to Mercury Research, AMD sold 7mm mobile chips in Q1 at an ASP of $37 (compared to INTC @ $90) and 11mm desktop chips at an ASP of $54 (INTC @ $90).
     
    AMD's current mobile line-up tops out at $75, so Llano priced at $70-$170 for an est. 8-9mm units in 2011 should move the ASP up nicely.
     
    Bulldozer priced at $220-$320 lets AMD play in the high-end desktop market it's been completely locked out of since 2006. AMD's most expensive desktop chip now, the Phenom II X6, tops out at $165.
     
    2) If AMD is willing to price inline with INTC, then Bulldozer must be bringing the goods. The top end Sandy Bridge chip, Core i7 2600K, is priced at $310. When AMD can't match INTC on performance, it tries to make up for it by being cheap. It appears AMD is not going the cheap alternative to INTC route with Bulldozer. And if Bulldozer can take it to Sandy Bridge in the desktop space, then things look good for the server market as well.

    SubjectRE: Llano in da house, part 2
    Entry06/14/2011 10:50 PM
    Membercasper719

    SubjectRE: RE: RE: Llano in da house, part 2
    Entry06/15/2011 11:18 AM
    MemberDen1200
    From the Anandtech website on the Llano

    "Some people will try to tell you that AMD will sell you more CPU cores than Intel for a lower price, but unlike desktop parts, mobile Llano cores don't clock high enough to consistently outperform dual-core Intel processors. Even in heavily-threaded benchmarks where quad-core CPUs can shine, dual-core i5 processors are still typically 30% faster than the A8-3500M. Instead of selling you more CPU cores for less money, what AMD is now selling is substantially better graphics for less money. Home theater enthusiasts might find a use for such parts as well, but really the purpose of GPUs is simple: they're for playing games. Until and unless GPGPU can take off and provide some killer apps, businesses and non-gaming folk alike will be better served by Intel's processors-unless you want to save $100 to $200."

     

    I did some more reading on another blog, call cpuwars. Here is an interesting link.

    http://www.cpu-wars.com/2011/06/intel-starts-talking-about-8nm-node.html It is mostly how Intel continues to push the envelope. Remember that AMD just made it on the 32nm platform, while Intel is already on track to go to 22nm. Well Intel is already talking 14nm in two years and 8nm in 3.5 years. The article said it would also put Intel at a big advantage over its main competitors, mainly IBM. Interesting AMD isn't mentioned here.



    SubjectRE: Llano in da house, part 2
    Entry06/15/2011 12:15 PM
    MemberYCOMBINATOR
    1) Saving $100-$200 with way better GPU performance is a pretty good value proposition, no? OEMs are eating Llano up b/c it combines three chips (CPU, GPU, and northbridge) into one package -- lower power consumption and lower cost.
     
    Llano is based on the old Stars CPU core, so of course Sandy Bridge is going to kick its ass in CPU performance. But for a large portion of the population, CPU performance is not the bottleneck anymore. Sandy Bridge will run your browser a nanosecond faster, but Llano will play your HD video better. Llano solves more customer problems than Intel at this point, unless you're running massive Monte Carlo simulations on your laptop.
     
    From the same Anandtech review: "AMD has focused this round up upgrades more on reducing power consumption rather than increasing performance, and that's a perfectly reasonable approach for a mobile CPU. Most of us probably aren't doing 3D rendering, CAD/CAM, or unassisted video transcoding on our laptops anyway."
     
    2) Intel will always have a process advantage over AMD. The AMD thesis isn't based on AMD (GlobalFoundries actually) beating Intel in process technology, it's based on AMD no longer having really sucky products.

    Subjectjust some challenging questions
    Entry06/29/2011 04:18 PM
    Membertyler939
    YCOMB, thanks for the great updates.
    Why do you think the stock has performed so poorly recently relative to Intel?
    Do you have any thoughts on the CEO search?  If AMD is in the midst of a turnaround, why are they having trouble attracting a CEO?t
    IHS iSuppli said that Intel's share of the processor market increased from 82.6% in the first quarter from 81.0% in the fourth quarter, while AMD's share fell fell from 10.1% to 10.9% (which was still an increase in absolute terms, since the overall market grew by 20%).  This was despite the Sandy Bridge issues that emerged earlier this year.  What do you make of this?
    Susquehanna downgraded semiconductors generally (they didn't single AMD out), saying their meetings in Asia suggest that OEMs are not going to "build aggressively for the holiday season."  Do you agree that we may see muted seasonality this year?
    Are there any upcoming events we should be watching for now that Llano is in the market?
    Is there anything else that came up recently that we should be thinking about that isn't addressed in your writeup or the comments?
    I hope these questions don't sound too hostile, but it seems as though your thesis has been playing out as you predicted (e.g., good reviews on the new processors), but I'm puzzled by the stock's weakness.  I'm just looking for answers, so these questions really are meant to be a friendly inquiry.

    SubjectRE: just some challenging questions
    Entry07/05/2011 11:47 AM
    MemberYCOMBINATOR
    Hi tyler, those are all fair questions.
     
    1) INTC is just not going to have the volatility that AMD does. It's big, safe for approved lists, cheap, and pays a nice dividend. As for AMD's absolute poor performance, who knows? If the market occasionally doesn't act in a puzzling manner, then we couldn't do what we do as investors. I have a friend who presented at the Value Investor Conference in Omaha this year. He pitched AMD -- the reaction was crickets. Most investors I talk to give me the same reaction. I traded the last two cycles, long in 2003 and short in 2006, and it was just as puzzling that the market would not price in either the positive or negative implications of new products until the market share numbers are plain to see.
     
    2) The CEO has definitely taken much longer than I expected, but it makes sense why Hurd or Gelsinger turned them down (fortunately in the case of Hurd, and Gelsinger wants to run EMC).
     
    3) I wouldn't have expected any positive share movements for AMD in Q1, regardless of any INTC hiccup with Sandy Bridge, because AMD's product line-up in Q1 still sucked with the exception of the Bobcat APUs. Q2 should be more of the same, and in Q3 we'll finally see volume shipments of Llano. Digitimes had some expected mix numbers through Q1 2012: http://www.digitimes.com/news/a20110629PD228.html 
     
    4) I can't offer much insight into how the holiday season may go. The public comments by the big Asian OEMs are that they still expect a good holiday season, but Quanta and Compal have both recently lowered their shipment forecasts. This seems like a hedgeable macro risk to me though, and share gains should help AMD unless the entire PC market falls apart.
     
    5) Next up after Llano is Bulldozer, which is scheduled for introduction sometime in Q3 for both servers and desktops.
     
    I've been buying more stock at recent prices. The downside is fairly limited (AMD doesn't have blow up potential), 2 of 3 new products have launched and are doing well, and these are huge markets where small share shifts go a long way toward moving AMD's bottom line.

    SubjectSome Llano math for Q3
    Entry07/05/2011 02:12 PM
    MemberYCOMBINATOR
    From the various Digitimes reports, we have AMD's internal estimates of 1.5mm mobile Llano shipments in July with a total of 8-9mm units for all of 2011. So let's call it 4.5mm (1.5mm/month) for Q3.
     
    According to Mercury Research, AMD sold 7mm mobile CPUs in Q1 at a $37 ASP for a total of $260mm in revenue. Half of that 7mm was Bobcat based APUs targeted toward netbooks. INTC dominated the mobile market with 46mm units at a $90 ASP for $4.15b of revenue in Q1. Of those 46mm, 34mm units were in the mainstream mobile segment with a $108 ASP. The segment where Llano makes AMD relevant and competitive again.
     
    Assuming 4.5mm mobile Llano units in Q3 at a $80 ASP (it could be higher, INTC is at $108) and holding everything else constant from Q1 (Bobcat stays at 3.5mm units @ $33 ASP, no credit for Llano desktop chips, Q1 is seasonally weak, etc...), just the mobile Llano shipments will add an incremental $215mm to AMD's mobile revenues in Q3 (compared to $967mm total CPU revenue in Q1 according to Mercury).
     
    4.5mm mobile Llano * $80 ASP = $360mm
    3.5mm Bobcat * $33 ASP = $115mm
    Total 8mm mobile units * $60 ASP = $475mm

    SubjectYCOMBINATOR, what is the "mobile market"?
    Entry07/08/2011 12:15 PM
    Membertyler939
    YCOMBINATOR, on 6/15 (comment 30), you criticized Bergman's interview for focusing on laptops instead of the "mobile market".  Specifically, you said " AMD is finally relevant again in the $18 billion mobile market with Llano and Bergman gets sucked into talking about tablets."   How exactly are you defining the "mobile market"?  Are you jsut talking about notebooks?  Regarding sub-laptop form factors, as I read Siefert's statements, AMD has basically given up on challenging AMR's dominance in smartphones, and is basically waiting for tablets to reach the capabilities of today's notebooks before making a push in that market, while right now all they are doing is working on reducing power consumption in their current product line to anticipate that day.  When a CEO says the opportunity is on servers and clients, and that he doesn't want to spend a lot on R&D, that doesn't sound much like AMD becoming relevant to me.  Is it correct that tablets and smartphones are not part of AMD's addressable market?  Thanks in advance.

    Subjecttyler
    Entry07/08/2011 01:31 PM
    MemberYCOMBINATOR

    I should be more precise now that mobile might mean phones/tablets. Traditionally, when INTC and AMD referred to mobile it means laptops.

    Yes, tablets and smartphones are not part of AMD's addressable market. I'd rather have them focus on the very large laptop, desktop, and server markets than go chasing the tablet/smartphone market.


    SubjectRE: It's black! No, it's white!
    Entry07/12/2011 11:42 AM
    Membertyler939
    What numbers are you expecting today for earnings and guidance, and what levels do you think would move the stock significantly?

    SubjectRE: It's black! No, it's white!
    Entry07/22/2011 10:55 AM
    MemberYCOMBINATOR
    Okay, the sell-side is too easy to beat up on and we're all wrong at some point, but I can't resist.
     
    If you recall the JMP analyst downgraded AMD last week on the "failure" of AMD's APUs. Today, we get this...
     
    "On the other hand, analyst Alex Gauna of JMP Securities offered a more downbeat view of AMD shares. He reiterated a market underperform rating for AMD, saying in a note, "We caution that AMD does not have a particularly good track record of forecasting.""
     
    Thanks for your excellent forecasting though Alex, it made my Aug call buying experience very enjoyable.

    SubjectHP announcement and its effect on AMD
    Entry08/22/2011 07:03 PM
    Membertyler939
    AMD seems to be holding up will over the last few days in a down market.  Do you think this might be related to HP's announcement regarding seeking strategic alternatives for the Personal Systems Group?  For example, might this somehow result in someone bidding for AMD?

    SubjectSo is Rory the right guy?
    Entry08/25/2011 09:47 AM
    Membertyler939
    Any thoughts appreciated.

    SubjectRE: So... close... to... Apple...
    Entry06/08/2012 07:31 AM
    Memberghostly
    Are you still following this Ycombinator?  Trinity looks great for the value laptop / emerging market laptop segment but you wouldn't think so looking at AMD's stock price.  Granted the server chips have been disappointing, but that is probably insignificant next to Trinity and its successors.  Have there been manufacturing issues or other problems that could get in the way of Trinity's success?
      Back to top