Kann AMD mittelfristig noch mithalten?
- Ersteller Markstar
- Erstellt am
Markus Everson
Grand Admiral Special
Also ganz ehrlich - Leute deren IQ so deutlich einstellig ist das sie nichtmal in der Lage sind neben der Performance eines Brazos-Laptop / eines Ultrabook auch deren Preis (500€ <-> 1000€) zu berücksichtigen sollten doch wirklich nicht AMDs Zielgruppe sein.
[Edit]Wers kauft weil er sich davon Vorteile verspricht oder es schlicht heiß findet ordentliche Performance bei minimalem Gewicht zu haben ist hier nicht gemeint.
Zuletzt bearbeitet:
boxleitnerb
Vice Admiral Special
- Mitglied seit
- 27.01.2008
- Beiträge
- 517
- Renomée
- 5
Also ich hab einen Fileserver, da steckt ein E-350 drin. Zwecks Installation, Einrichten und Treibersuche hab ich mit dem auch etwas "gearbeitet". Ich muss ganz ehrlich sagen, dauerhaft wäre das für mich kein Vergnügen, daran aktiv was zu machen, trotz SSD. Dafür ist mir der Brazos zu langsam.
Mir persönlich würde es vorrangig ums "Gefühl" gehen, da reißt eine schnellere CPU schon was raus. Und wer sagt denn, dass man nicht auchmal was Intensiveres mit dem Ultrabook machen will? Für viele Kunden ist aber auch ein E-350 ausreichend, nur pauschalisieren würde ich das mitnichten.
Mir persönlich würde es vorrangig ums "Gefühl" gehen, da reißt eine schnellere CPU schon was raus. Und wer sagt denn, dass man nicht auchmal was Intensiveres mit dem Ultrabook machen will? Für viele Kunden ist aber auch ein E-350 ausreichend, nur pauschalisieren würde ich das mitnichten.
Warum Brazos? Man nehme einen 17W-Trinity. Viel besser, außerdem kann man dann mit dem Grafikargument kommen ("so dünn und trotzdem genug Leistung für Spiel xy").
Die Geräte werden eh relativ teuer, auch wenn billige AMD-Technik drinne ist. Mit dem Preisargument bei der CPU kann man da vmtl. nur wenig ausreichen, auch wenn die wohl immer noch klar billiger ist als die 17W-Sandy-Bridges, aber durch die nicht mehr so strengen Vorgaben was z. B. die maximal erlaubte Dicke etc. angeht kann man vmtl. mehr erreichen, da die Hersteller da nochmal sparen können.
Zuletzt bearbeitet:
y33H@
Admiral Special
- Mitglied seit
- 16.05.2011
- Beiträge
- 1.768
- Renomée
- 10
Erst mal auf den 17W-Trinity warten und gucken, was da am Ende über bleibt. Die CPU-Leistung wird sicherlich weit unter der eines IVB-Ultrabooks liegen, die GPU-Power gilt es gegenüber der HD 4000 abzuwarten. Bleibt wie üblich die Frage nach dem Preis und im Zweifel, da ich an meinem Subnotebook auch mal ältere Titel zocke, wird's eher was mit einer Radeon.
Klar, erstmal heißts abwarten. Aber im Vergleich zu Brazos dürfte er wohl die bessere Wahl sein. CPU-seitig muss man sich zwar erstmal abwarten, wieviel er dann noch mitbringt, und klar, gegen Ivy Bridge dürfte er dann kaum Chancen haben. Gegenüber Brazos wird man abwarten, aber auch da dürfte man einen Sprung sehen, ist natürlich mit das spekulativste.
Aber GPU-seitig ist er gegenüber Brazos auch im ULV-Segment wohl jetzt schon als klar besser einzustufen, allein das doppelte Speicherinterface macht ihn da interessanter. Aber auch viel mehr Einheiten, die man dann niedriger takten kann, sollten energieeffizienter sein.
LoRDxRaVeN
Grand Admiral Special
Bzgl. der CPU Leistung muss man mMn erst abwarten. Es ist noch nicht in Stein gemeiselt, dass Trinity bei ~17W TDP IVB "weit unterlegen" sein muss. Dass akutell der Prozess noch einiges an Spielraum hat, ist wohl unbestritten und weiters kann man kaum direkt von 95/125 W TDP Prozessoren auf 17W TDP Prozessoren -der nächsten Generation- schließen. Es besteht sehr wohl die Möglichkeit, dass die Energieeffizient bei Trinity um einiges besser dasteht und eben auch die Möglichkeit, dass die Energieeffizient bei niedrigeren Takten deutlich steigt.
Das einzige was man mit hoher Wahrscheinlichkeit prophezeihen kann ist, dass die IPC wohl kaum an IVB heranreichen wird - aber wie wir wissen, ist dies nur ein Teil der Rechnung...
Die Aussage ist somit mMn einer Glaskugel entnommen.
Das einzige was man mit hoher Wahrscheinlichkeit prophezeihen kann ist, dass die IPC wohl kaum an IVB heranreichen wird - aber wie wir wissen, ist dies nur ein Teil der Rechnung...
Die Aussage ist somit mMn einer Glaskugel entnommen.
Zuletzt bearbeitet:
y33H@
Admiral Special
- Mitglied seit
- 16.05.2011
- Beiträge
- 1.768
- Renomée
- 10
Das ist es meiner Ansicht nach sehr wohl: Intels schnellste SNB mit 17W kommt auf 2C/4T mit 1,8 bis 2,9 GHz - wie soll AMD das mit 17W erreichen? Der Core i7-2677M ist mit den Specs auf Höhe eines Bulldozers FX-4k mit deutlich über 2,0 GHz ... und IVB kommt als Core i7-3667U auf 2,0 bis 3,2 GHz bei Bumms pro Takt - das erreicht Trinity nie im Leben bei 17W TDP. Die Kluft wird eher größer denn keiner - selbst bei der IGP.LoRDxRaVeN schrieb:
Markus Everson
Grand Admiral Special
Früher im Jahr gabs doch diese Präsentation wo GCN vorgestellt wurde, richtig?
Ich glaube mich zu erinnern, das dort bei der Roadmap für die APU schon für 2012 Energiebalance - das Aufteilen des TDP-Budget auf CPU und GPU - als Topic stand.
Wenn ich das richtig verstanden habe und AMD es umsetzt, dann wird Trinity in jedem Fall bei reiner CPU Last deutlich Performance gegenüber einem Llano gleicher TDP gewinnen weil im Gegensatz zum letzteren keine starre Reservierung für die Grafik stattfindet.
.
EDIT :
.
Ich halte Deine 2C und erhöhe um 320 SP.
Schon lustig zwei Prozessoren miteinander zu vergleichen, die noch nicht auf dem Markt sind. Ich würde mal abwarten, was bei Ivys IGP am Schluss wirklich rauskommt...Bei AMD wirds auf jeden Fall gut, denn die haben schon bei Llano bewiesen, dass sie das können. Intel hat immer wieder gezeigt, dass es bei ihren GPUs Probleme gibt, nicht nur was die Treiber angeht.
Und bei der CPU.....Wenn man sich anschaut, was TDP mäßig jetzt schon im Serverbereich geht (Stichwort 4256 EE), dann wird klar, dass der Verbrauch von Bulldozer gerade bei niedrigerem Takt durchaus einigen Spielraum gegenüber K10.x bietet. Ein bisschen Optimierungspotential dürfte ja Piledriver auch noch haben und dann siehts doch gar nicht so schlecht aus. Ein Trinity muss auch nicht auf Augenhöhe mit Ivy Bridge liegen, aber er muss halt genug Leistung bieten, um für ein Ultrabook interessant zu sein.
Der größte Vorteil ist aber mMn auf Seiten der Hersteller zu sehen, die an AMD APUs auch deshalb ein Interesse haben dürften, um den Preisdruck auf Intel zu erhöhen. Für Intel ist das ganze Ultrabook-Zeugs ja nur dazu da, die Margen im Notebookbereich nicht zu weit sinken zu lassen bzw. sogar zu erhöhen.
Und bei der CPU.....Wenn man sich anschaut, was TDP mäßig jetzt schon im Serverbereich geht (Stichwort 4256 EE), dann wird klar, dass der Verbrauch von Bulldozer gerade bei niedrigerem Takt durchaus einigen Spielraum gegenüber K10.x bietet. Ein bisschen Optimierungspotential dürfte ja Piledriver auch noch haben und dann siehts doch gar nicht so schlecht aus. Ein Trinity muss auch nicht auf Augenhöhe mit Ivy Bridge liegen, aber er muss halt genug Leistung bieten, um für ein Ultrabook interessant zu sein.
Der größte Vorteil ist aber mMn auf Seiten der Hersteller zu sehen, die an AMD APUs auch deshalb ein Interesse haben dürften, um den Preisdruck auf Intel zu erhöhen. Für Intel ist das ganze Ultrabook-Zeugs ja nur dazu da, die Margen im Notebookbereich nicht zu weit sinken zu lassen bzw. sogar zu erhöhen.
Zuletzt bearbeitet:
AMD Q & A from WFG Management Discussion Series
Cody G. Acree, CFA
214.234.9475
cacree@williams-financial.com
Advanced Micro Devices (AMD, Buy, $5.56) Management Discussion Series Notes
Summary: During our call, it was obvious that in just the first few month’s of Rory Read’s tenure, that
he is already having a very positive impact on the attitude and motivation of the employee base. One
of the biggest early changes appears to us to be a focus of AMD at identifying segments where the
company truly has opportunities for leadership, rather than simply being a strong alternative to Intel
(INTC, Hold, $25.66). Rory, at least initially, also seems to be a much more visibly present CEO as
he is spending time interacting locally with employees, suppliers, customers, and partners, versus
prior management that appeared to lead remotely from the Austin offices. The health of the
computing market doesn’t appear to have changed significantly over the last few months with a
flattish Europe, a bit better U.S. holiday trends than feared, and the primary growth still being driven
by the emerging markets. Component and PC inventories appear lean, which could lead to
somewhat muted negative Q1 seasonality if North American holiday trends continue to improve and
is followed by a healthy Chinese New Year. We are encouraged that customers appear to be sticking
with AMD’s APUs despite the production issues at Llano. It doesn’t look as though there were major
order or program cancellations and steady progress is being made to correct production volumes. To
help with the yield problems, Rory has brought a team from IBM into GlobalFoundries (as GF is part
of the IBM Alliance) and has also brought in PDF, a yield consultancy group, to help identify issues.
These moves are helping to drive yield improvements although they are not yet optimal. However,
AMD met its 32nm supply commitments in both October and November. It appears that Bulldozer is
being well received by OEMs, with Dell (DELL, NR, $15.63) HP(HPQ, NR, $28.09), and IBM (IBM,
NR, $192.73) ramping servers this quarter. We expect these rollouts should begin to lift AMD’s
server share and are encouraged by AMD strategy to target specifically the segments of
virtualization, cloud, and high-performance, where the new architecture offers the most differentiation.
Overall, we believe AMD is well positioned to earn material share gains in servers and mobiles in
2012 and that these share gains should be solid drivers of margins and earnings, which is not
currently reflected in the company’s share price.
Questions & Answers:
1. Given the macroeconomic volatility, what is your view of the health of your major markets?
· The health of the computing market doesn’t appear to have changed significantly over
the last few months with Europe “uninspiring” not better but also not worse, a bit better
U.S. holiday trends than feared, and the primary growth still being driven by the emerging
markets. Chinese demand growth is not as rapid as it was last year, but it is still very healthy growth.
2. What is your view of channel and OEM inventory health?
· Over the last several months, everyone from suppliers to retailers have been working
to reduce inventory exposure and are now managing with very lean inventories.
· If holiday demand globally is better than many feared or economic activity improves
next year, there is likely some room for the channel and OEMs to restock a bit of
inventory.
3. What impact are you seeing from the Thailand floods?
· There is still not a consistent picture, but the impact doesn’t look to be as bad as some
of the most pessimistic estimates.
· With several prior natural disasters and uncertain economic outlooks, the supply chain
has become increasingly nimble and resilient, which is likely helping to shorten the
length of the impact.
· In Q4, AMD believes it is seeing only limited impact and that most companies are
managing through it. However, a broader impact is likely seen in Q1, although
estimates vary widely.
· For AMD, the company does earn about 40% of revenue from the channel, which is
likely to see the most shortage as OEMs are given preferential supply. Therefore, with
60% of AMD’s revenue from OEMs and HP being its largest customer, the AMD
impact may be somewhat less than the broader PC industry.
4. Coming into the quarter, OEMs appeared very cautious regarding the potential for much
holiday seasonality. With recent retail numbers looking a bit more positive, how could such a
scenario impact AMD over the next few months?
· If holiday sales are materially higher than expectations, then it is not too late for AMD
to see some benefit during the fourth quarter, however much of the holiday electronic
strength thus far has been in TVs and other electronics. While PC sales appear solid,
there is not yet evidence of a strong enough upside to cause channel restocking,
although lean inventories and even a descent holiday demand might help to lessen
Q1’s negative seasonality.
5. For last quarters call, Rory Read had just been with the firm a short while and was busy getting
up to speed operationally and starting to meet with customers. What have been the areas of
focus over the last quarter and what are some of the changes that he has affected?
· Rory has been focusing on getting up to speed on all areas and has been spending
time with customers, partners and the supply chain.
· It was apparent during the call, in our opinion that in just the first few month’s of Rory’s
tenure, that he is already having a very positive impact on the attitude and motivation
of the employee base.
· One of the biggest early changes appears to us to be a focus of AMD at identifying
segments where the company truly has opportunities for leadership, rather than simply
being a strong alternative to Intel (INTC, Hold, $25.24). These areas look to center
around AMD’s set of core competencies of low power, cloud/server, and serving the
needs of the emerging markets.
· He, at least initially, also seems to be a much more visibly present CEO as he is
spending time interacting locally with employees, suppliers, customers, and partners,
versus prior management that often appeared to lead remotely from the Austin offices.
· Given his technical and operational background, he looks to be hyper focused on
creating a discipline of execution, where historically AMD has had a spotty track
record.
· Rory has also supplemented his expertise with additional team members, such as
hiring Mark Papermaster to be CTO.
6. Have the yield issue on Llano continued to improve? Are there any major hurdles that are yet
to be crossed or is it now more of a linear improvement? When would you expect yields to
reach normal levels?
· Yield issues have continued to steadily improve although they are not yet at optimal
levels.
· One of the first things Rory did was put together a team from IBM, as GlobalFoundries
is a member of the IBM Alliance, and embed them in the foundry. He also brought in
PDF, a fab yield consultancy group, to help identify areas for improvement and
consistency.
· GlobalFoundries new CEO has also been very focused on correcting these issues.
· Improvements have been made such that AMD was able to meet its 32nm supply
commitments for both October and November.
7. Now that you’ve had a couple months post the yield issue announcement, what has been the
broad customers’ action? Investor fear has been that customer relationships may have been
damaged regarding overall adoption of the APU architecture.
· Customers were obviously annoyed and disappointed but do not appear to have made
any major change in their demand for AMD’s APUs. In the end, the customers need
the best processor for the best product and continue to see the benefit of high
performance integrated graphics priced below similar products from Intel.
· Further evidence that customers have not been moving away from AMD is that 28nm
design win momentum has not slowed.
8. You mentioned on our last call that Brazos and Llano had about 320 design wins together.
How far are you through the process of these wins pushing into production? Did the win
trajectory change much with the yield issues?
· The 320 is a good base for the life of these products, some refreshes will be added
and some designs may never make it to market. This volume compares to only about
70 wins that were typical on prior generations. Most of the Brazos designs are now in
the market, and Llano will continue over the next couple quarters, particularly as Llano
availability improves. As these make it to market, design wins for Llano’s successor,
Trinity, are being earned.
9. Could you talk about early customer reaction, feedback and adoption of Bulldozer? How is the
architectural change being received?
· Bulldozer, the Opteron 4200 and 6200, are being well accepted with 30 servers
ramping from Dell, HP, IBM, and Cray (CRAY,NR, $6.36). In the next few weeks,
AMD will also begin shipping to channel and motherboard partners.
· Server growth last quarter, was the first sequential growth in seven quarters and
AMD’s server share has fallen to 5%, which should begin to improve as OEMs are
rolling out platforms through Q4.
· AMD is not trying to address the entire server market with Bulldozer and therefore will
have only a small presence in mainstream servers for data management, but are
rather focusing on high performance computing, virtualizations, and cloud computing,
where the Bulldozer architecture offers the most benefit.
· For very high-end workloads Bulldozer can deliver up to 84% better performance and
in cloud or virtualization applications Bulldozer can save up 2/3rds floor space and
offers 2/3rds lower platform price.
10. Do you believe you are likely to have any tangible benefit from the delay of Romley?
· There may be some small benefits of AMD having the newest architecture available
before Romley ramps, but if the delay is only about a quarter, most can wait that long
to purchase.
11. Are you seeing any noticeable impact of your addition of server sales people and your more
focused strategy instead of trying to target all server segments? What are the odds of missing
important segments?
· Initially the additional people were brought in to ensure the successful launch of
Bulldozer. Now that this has occurred, the increased staff should help to accelerate
AMD’s server share.
· The risk of missing an important segment is not likely an issue when AMD only has
5% market share. The question is more of an issue if AMD’s share were well above
20%.
· AMD believes the combination of virtualization, cloud, and high performance
represents an opportunity to address about 40% of the total server market.
12. As the market has yet to see a demonstrable improvement in consumer or corporate
confidence, have you seen any changes in the pricing environment?
· Pricing is always competitive but no major changes are being seen.
· Historically AMD has seen more aggressive pricing when its market share approached
30% in a given segment.
· Intel’s pricing strategies are also under much more scrutiny, which may also help to
avoid historic price pressure.
13. If we assume that AMD is an eventual participant in the tablet/smartphone market, there is a
contingent out there that believes you may not have the R&D budget to effectively service this
and your core market. What is your view?
· For the next year or so, the tablet market is likely to continue to be dominated by
Apple (AAPL, NR, $393.11) as AMD doesn’t believe anything currently in the market
can truly change the market share landscape.
· AMD sees low power as one of is core competencies and believes the success of
Brazos demonstrates this. The recent restructuring and reduction in head count by
about 10% provides about $160 million that can be redirected to invest in new
opportunities.
· AMD is comfortable with its R&D budget as a percent of revenue and believes this can
actually decline modestly as revenue growth improves.
14. Can you talk about the competitive positioning of the Radeon family and what market share
trends have been like?
· Radeon demand continues to be strong and AMD believes its focus on DirectX 11 is a
key differentiator. The company continues to be able to introduce the industry’s
highest performing cards and is now demonstrating 28nm which will ship before the
end of the year.
Important Disclosures and Certifications
Williams Financial Group Capital Markets is the research division of WFG Investments, Inc.
Securities are offered through WFG Investments, Inc., member FINRA and SIPC.
This report is for informational purposes only and is not a solicitation to buy or sell securities from
any company mentioned herein. Further, the information contained in this report has been
obtained from sources believed to be reliable, but its accuracy and completeness is not
guaranteed and should not be the sole basis of any investment decision. The reports and ratings
are not personal recommendations for any particular investor or client and do not take into
account the financial needs, investment objectives or risk profile of any particular client.
Recommendations made in this report may not be suitable for all investors.
Cody G. Acree, CFA
214.234.9475
cacree@williams-financial.com
Advanced Micro Devices (AMD, Buy, $5.56) Management Discussion Series Notes
Summary: During our call, it was obvious that in just the first few month’s of Rory Read’s tenure, that
he is already having a very positive impact on the attitude and motivation of the employee base. One
of the biggest early changes appears to us to be a focus of AMD at identifying segments where the
company truly has opportunities for leadership, rather than simply being a strong alternative to Intel
(INTC, Hold, $25.66). Rory, at least initially, also seems to be a much more visibly present CEO as
he is spending time interacting locally with employees, suppliers, customers, and partners, versus
prior management that appeared to lead remotely from the Austin offices. The health of the
computing market doesn’t appear to have changed significantly over the last few months with a
flattish Europe, a bit better U.S. holiday trends than feared, and the primary growth still being driven
by the emerging markets. Component and PC inventories appear lean, which could lead to
somewhat muted negative Q1 seasonality if North American holiday trends continue to improve and
is followed by a healthy Chinese New Year. We are encouraged that customers appear to be sticking
with AMD’s APUs despite the production issues at Llano. It doesn’t look as though there were major
order or program cancellations and steady progress is being made to correct production volumes. To
help with the yield problems, Rory has brought a team from IBM into GlobalFoundries (as GF is part
of the IBM Alliance) and has also brought in PDF, a yield consultancy group, to help identify issues.
These moves are helping to drive yield improvements although they are not yet optimal. However,
AMD met its 32nm supply commitments in both October and November. It appears that Bulldozer is
being well received by OEMs, with Dell (DELL, NR, $15.63) HP(HPQ, NR, $28.09), and IBM (IBM,
NR, $192.73) ramping servers this quarter. We expect these rollouts should begin to lift AMD’s
server share and are encouraged by AMD strategy to target specifically the segments of
virtualization, cloud, and high-performance, where the new architecture offers the most differentiation.
Overall, we believe AMD is well positioned to earn material share gains in servers and mobiles in
2012 and that these share gains should be solid drivers of margins and earnings, which is not
currently reflected in the company’s share price.
Questions & Answers:
1. Given the macroeconomic volatility, what is your view of the health of your major markets?
· The health of the computing market doesn’t appear to have changed significantly over
the last few months with Europe “uninspiring” not better but also not worse, a bit better
U.S. holiday trends than feared, and the primary growth still being driven by the emerging
markets. Chinese demand growth is not as rapid as it was last year, but it is still very healthy growth.
2. What is your view of channel and OEM inventory health?
· Over the last several months, everyone from suppliers to retailers have been working
to reduce inventory exposure and are now managing with very lean inventories.
· If holiday demand globally is better than many feared or economic activity improves
next year, there is likely some room for the channel and OEMs to restock a bit of
inventory.
3. What impact are you seeing from the Thailand floods?
· There is still not a consistent picture, but the impact doesn’t look to be as bad as some
of the most pessimistic estimates.
· With several prior natural disasters and uncertain economic outlooks, the supply chain
has become increasingly nimble and resilient, which is likely helping to shorten the
length of the impact.
· In Q4, AMD believes it is seeing only limited impact and that most companies are
managing through it. However, a broader impact is likely seen in Q1, although
estimates vary widely.
· For AMD, the company does earn about 40% of revenue from the channel, which is
likely to see the most shortage as OEMs are given preferential supply. Therefore, with
60% of AMD’s revenue from OEMs and HP being its largest customer, the AMD
impact may be somewhat less than the broader PC industry.
4. Coming into the quarter, OEMs appeared very cautious regarding the potential for much
holiday seasonality. With recent retail numbers looking a bit more positive, how could such a
scenario impact AMD over the next few months?
· If holiday sales are materially higher than expectations, then it is not too late for AMD
to see some benefit during the fourth quarter, however much of the holiday electronic
strength thus far has been in TVs and other electronics. While PC sales appear solid,
there is not yet evidence of a strong enough upside to cause channel restocking,
although lean inventories and even a descent holiday demand might help to lessen
Q1’s negative seasonality.
5. For last quarters call, Rory Read had just been with the firm a short while and was busy getting
up to speed operationally and starting to meet with customers. What have been the areas of
focus over the last quarter and what are some of the changes that he has affected?
· Rory has been focusing on getting up to speed on all areas and has been spending
time with customers, partners and the supply chain.
· It was apparent during the call, in our opinion that in just the first few month’s of Rory’s
tenure, that he is already having a very positive impact on the attitude and motivation
of the employee base.
· One of the biggest early changes appears to us to be a focus of AMD at identifying
segments where the company truly has opportunities for leadership, rather than simply
being a strong alternative to Intel (INTC, Hold, $25.24). These areas look to center
around AMD’s set of core competencies of low power, cloud/server, and serving the
needs of the emerging markets.
· He, at least initially, also seems to be a much more visibly present CEO as he is
spending time interacting locally with employees, suppliers, customers, and partners,
versus prior management that often appeared to lead remotely from the Austin offices.
· Given his technical and operational background, he looks to be hyper focused on
creating a discipline of execution, where historically AMD has had a spotty track
record.
· Rory has also supplemented his expertise with additional team members, such as
hiring Mark Papermaster to be CTO.
6. Have the yield issue on Llano continued to improve? Are there any major hurdles that are yet
to be crossed or is it now more of a linear improvement? When would you expect yields to
reach normal levels?
· Yield issues have continued to steadily improve although they are not yet at optimal
levels.
· One of the first things Rory did was put together a team from IBM, as GlobalFoundries
is a member of the IBM Alliance, and embed them in the foundry. He also brought in
PDF, a fab yield consultancy group, to help identify areas for improvement and
consistency.
· GlobalFoundries new CEO has also been very focused on correcting these issues.
· Improvements have been made such that AMD was able to meet its 32nm supply
commitments for both October and November.
7. Now that you’ve had a couple months post the yield issue announcement, what has been the
broad customers’ action? Investor fear has been that customer relationships may have been
damaged regarding overall adoption of the APU architecture.
· Customers were obviously annoyed and disappointed but do not appear to have made
any major change in their demand for AMD’s APUs. In the end, the customers need
the best processor for the best product and continue to see the benefit of high
performance integrated graphics priced below similar products from Intel.
· Further evidence that customers have not been moving away from AMD is that 28nm
design win momentum has not slowed.
8. You mentioned on our last call that Brazos and Llano had about 320 design wins together.
How far are you through the process of these wins pushing into production? Did the win
trajectory change much with the yield issues?
· The 320 is a good base for the life of these products, some refreshes will be added
and some designs may never make it to market. This volume compares to only about
70 wins that were typical on prior generations. Most of the Brazos designs are now in
the market, and Llano will continue over the next couple quarters, particularly as Llano
availability improves. As these make it to market, design wins for Llano’s successor,
Trinity, are being earned.
9. Could you talk about early customer reaction, feedback and adoption of Bulldozer? How is the
architectural change being received?
· Bulldozer, the Opteron 4200 and 6200, are being well accepted with 30 servers
ramping from Dell, HP, IBM, and Cray (CRAY,NR, $6.36). In the next few weeks,
AMD will also begin shipping to channel and motherboard partners.
· Server growth last quarter, was the first sequential growth in seven quarters and
AMD’s server share has fallen to 5%, which should begin to improve as OEMs are
rolling out platforms through Q4.
· AMD is not trying to address the entire server market with Bulldozer and therefore will
have only a small presence in mainstream servers for data management, but are
rather focusing on high performance computing, virtualizations, and cloud computing,
where the Bulldozer architecture offers the most benefit.
· For very high-end workloads Bulldozer can deliver up to 84% better performance and
in cloud or virtualization applications Bulldozer can save up 2/3rds floor space and
offers 2/3rds lower platform price.
10. Do you believe you are likely to have any tangible benefit from the delay of Romley?
· There may be some small benefits of AMD having the newest architecture available
before Romley ramps, but if the delay is only about a quarter, most can wait that long
to purchase.
11. Are you seeing any noticeable impact of your addition of server sales people and your more
focused strategy instead of trying to target all server segments? What are the odds of missing
important segments?
· Initially the additional people were brought in to ensure the successful launch of
Bulldozer. Now that this has occurred, the increased staff should help to accelerate
AMD’s server share.
· The risk of missing an important segment is not likely an issue when AMD only has
5% market share. The question is more of an issue if AMD’s share were well above
20%.
· AMD believes the combination of virtualization, cloud, and high performance
represents an opportunity to address about 40% of the total server market.
12. As the market has yet to see a demonstrable improvement in consumer or corporate
confidence, have you seen any changes in the pricing environment?
· Pricing is always competitive but no major changes are being seen.
· Historically AMD has seen more aggressive pricing when its market share approached
30% in a given segment.
· Intel’s pricing strategies are also under much more scrutiny, which may also help to
avoid historic price pressure.
13. If we assume that AMD is an eventual participant in the tablet/smartphone market, there is a
contingent out there that believes you may not have the R&D budget to effectively service this
and your core market. What is your view?
· For the next year or so, the tablet market is likely to continue to be dominated by
Apple (AAPL, NR, $393.11) as AMD doesn’t believe anything currently in the market
can truly change the market share landscape.
· AMD sees low power as one of is core competencies and believes the success of
Brazos demonstrates this. The recent restructuring and reduction in head count by
about 10% provides about $160 million that can be redirected to invest in new
opportunities.
· AMD is comfortable with its R&D budget as a percent of revenue and believes this can
actually decline modestly as revenue growth improves.
14. Can you talk about the competitive positioning of the Radeon family and what market share
trends have been like?
· Radeon demand continues to be strong and AMD believes its focus on DirectX 11 is a
key differentiator. The company continues to be able to introduce the industry’s
highest performing cards and is now demonstrating 28nm which will ship before the
end of the year.
Important Disclosures and Certifications
Williams Financial Group Capital Markets is the research division of WFG Investments, Inc.
Securities are offered through WFG Investments, Inc., member FINRA and SIPC.
This report is for informational purposes only and is not a solicitation to buy or sell securities from
any company mentioned herein. Further, the information contained in this report has been
obtained from sources believed to be reliable, but its accuracy and completeness is not
guaranteed and should not be the sole basis of any investment decision. The reports and ratings
are not personal recommendations for any particular investor or client and do not take into
account the financial needs, investment objectives or risk profile of any particular client.
Recommendations made in this report may not be suitable for all investors.
Markus Everson
Grand Admiral Special
next couple quarters? Wann wurde der Call gehalten? Wann soll Trinity raus kommen?
FredD
Gesperrt
- Mitglied seit
- 25.01.2011
- Beiträge
- 2.472
- Renomée
- 43
Noch in der zweiten Hälfte der Maya-Apokalypse, was aber nicht so schlimm ist, weil der Zielmarkt der Maya im Anteil am Gesamtmarkt eher zu vernachlässigen ist.
deadohiosky
Gesperrt
- Mitglied seit
- 13.07.2011
- Beiträge
- 1.624
- Renomée
- 26
Falls richtig, würde das meine Annahme bestätigen, dass Llano nicht einfach ab Q/I entsorgt wird sondern noch ein klein wenig überleben könnte bis ins QII/III hinaus. Keine Produktpflege (außer den neueren K-Modellen), vllt noch ein paar Ausschussmodelle als Athlons um dann langsam auf Trinity umzuschwenken.
Der Call war offensichtlich im Dezember.
Heute erneut rory Call
Er sagte unter anderem, dass er keinen Sinn darin sieht den kleineren nodes hinterzujagen. Er will profitabler, innovative Produkte mit guter Ausbeute zuverlässig an die Kunden liefern.
Als Beispiel nannte er den Brazos erfolg.
Heute erneut rory Call
Er sagte unter anderem, dass er keinen Sinn darin sieht den kleineren nodes hinterzujagen. Er will profitabler, innovative Produkte mit guter Ausbeute zuverlässig an die Kunden liefern.
Als Beispiel nannte er den Brazos erfolg.
Markus Everson
Grand Admiral Special
Das könnte man als Bestätigung für das Canceln der 28nm APUs interpretieren.
Markus Everson
Grand Admiral Special
Klingt für mich deutlich nach Themenwechsel und damit nach GPU.
BavarianRealist
Grand Admiral Special
- Mitglied seit
- 06.02.2010
- Beiträge
- 3.358
- Renomée
- 80
Ich neige auch zu Deiner Sichtweise, zumal die Ankündigung der 28nm-APUs bei TSMC schon recht alt ist ("AMD wird 28-nm-APUs ebenfalls bei TSMC fertigen lassen" vom 17.11.10). Dass man am Ende nur entweder bei TSMC oder GF fertigt macht Sinn. Und von daher sollte man längst bei AMD erkannt haben, welcher Ansatz erfolgreicher läuft. So könnte man den GF-Ansatz gekillt haben, weil die TSMC-Lösung womöglich einfach so weit voraus war/ist?
Zuletzt bearbeitet:
Woerns
Grand Admiral Special
- Mitglied seit
- 05.02.2003
- Beiträge
- 2.983
- Renomée
- 232
Im Q&A Wohl eher kein Temenwechsel nicht, da gets um die Auswirkung aufs Geschäft durch die niedrigen Ausbeute der APUs und die Antwort darauf ist das die Kunden verärgert sind Jedoch nichts an ihrer Bestellung geändert haben da sie weiterhin eine Gute integrierte Grafik in den Produkten haben wollen.
Das 2. Argument dazu das die Lieferschwierigkeiten keine Auswirkung auf das Geschäft haben wird aufgeführt das die Kunden (dieselben!) nicht weniger Designes mit 28nm integrierter Grafik (chipsatzgrafik adee, somit bleibt nur APU) planen. Was meiner Meinung nach heisst, da kommt im APU-Bereich, was auf dem 28nm Node von irgeneier Fab zumindest irgendwann im nächsten jahr, und das sicher in weniger als einem Jahr pessimistisch gesehen.
Vorallem da bei allen anderen Fragen eigentlich immer beim angefragten Segment geblieben wird und nicht vom einen in das andere Gehüpft wird. Zudem hast du meine Frage nicht beantwortet.
Zu Trinity bisher hatte doch immer die Vorgängergeneration einen relativ grossen Marktanteil selbst nach dem Launch der neusten Generation, was soll man da rauslesen? Trinity kommt nächstes Jahr lauf offizieller mir zuletzt bekannter Roadmap und daran scheint sich nichts zu ändern.
Bisher hab ich nur von 40nm TSMC für Brazos 2.0 gelesen.
Wenn der im 1Q12 in 28nm käme, wäre es schon eine Überraschung für mich.
Da Brazos ein syntetisches Design ist, sollte dass doch relativ leicht machbar sein. Neue Prozessparameter in den Compiler, Knopf drücken, Masken erstellen lasssen.
Ach, was soll die Spekulatius, im Februar werden auch wir schlauer.
LoRDxRaVeN
Grand Admiral Special
Das ist doch völlig klar - somit kann man das "könnte" (usw.) bedenkenlos streichen. Selbst wenn Trinity tatsächlich (ich nenne es einfach mal) pünktlich gelauncht wird, wird Llano mit absoluter Sicherheit noch einige Monate in diversen Geräten zu finden sein.
Besonders im Mobil Markt ist dies einfach Fakt, da ja erst die Notebooks und Co. mit den Prozessoren bestückt werden müssen. Und wohl keine Firma könnte es sich leisten, 1000e Notebooks ohne Prozessor aber sonst fix fertig im Lager liegen zu lassen.
Nach der (Erst-)Auslieferung von Mobilprozessoren vergeht einfach zwangsweise etwas Zeit, in der der Vorgänger idR auch noch am Markt vertreten bleibt. AMD für sich "launcht" ja nicht die kompletten Notebooks, sondern eben nur die Prozessoren...
HelixUltra
Fleet Captain Special
- Mitglied seit
- 21.11.2008
- Beiträge
- 254
- Renomée
- 9
Kurze technische Frage: Warum kann man ein Design für einen Chip mit einem festgelegtem Verfahren zB 28nm oder mit/ohne SOI nicht bei zwei Herstellern also Glofo und TSMC gleichzeitig fertigen lassen?
Ich kenne zwar die Praxis nicht, aber stelle mir so ein Design wie ein Layout in Eagle oder CAD vor - quasi ein Baupläne nach dem jeder fertigen kann. Das wäre besser für die Verfügbarkeit und würde Probleme mMn reduzieren.
Freu mich auf fachliche Antworten
Ich kenne zwar die Praxis nicht, aber stelle mir so ein Design wie ein Layout in Eagle oder CAD vor - quasi ein Baupläne nach dem jeder fertigen kann. Das wäre besser für die Verfügbarkeit und würde Probleme mMn reduzieren.
Freu mich auf fachliche Antworten
Ähnliche Themen
