AMD bulldozer
- Thread starter omario
- Start date
Where to begin guys.
This has changed by entire specification for a new PC I was going to order within the next 2 months, obviously from here
Was going to purchase a laptop here 7 months ago, but I came to my senses and waited for AMD FX so I could build a new Desktop :winkiss: and return to the promised land HOWEVER
It appears to me that AMD has perhaps focussed too much on the idea of Multi-Threaded HPC and the more cores the better. The problem is there is insufficient software available to take advantage of multi-core processors and they have relied too much on this, as evident in the performance of the AMD FX.
It has very competitive multi-core performance when utilised with programs like MediaShow Espresso, Handbrake, and certain Adobe products which offers performance equal and indeed superior to the 2600K justifying it's pricing in between the 2500K and 2600K. It thrashes the 2500K in a lot of these benchmarks. The problem however is that AMD is relying on software to be compliant with multi-core hardware, even Windows 8 will supposedly bring an improvement to the performance of the AMD FX though likely not as much as they would like you to believe.
In the world we live in today there are few programs which actually fully utilise 4 or more cores, especially games to the non delight of pc gamers everywhere. The problem is the AMD FX's performance per core is worse than its predecessors due to its architecture design for multi-core applications. It consistently comes behind the X6-1100T and sometimes even the X4-980 in games and single thread or core processes. Even in 2 core processes it just does not have the muscle and considering this is the situation the CPU will find itself in predominantly it will make it very difficult to market to anyone but Media Enthusiasts and Designers who need multi-core operations.
Now I am not sure how the Opteron's will fare. You can clearly see in the Dhrystone and Whetstone benchmarks the capability and potential that justifies the price point, (though in reality this price point is not justified to poor actual performance). The multi core abilities may fare very well in the server market and indeed maybe equal to Intel's products, there is reason to believe they could even offer superior performance (especially price performance). Has AMD tried to sell the consumer market a Server chip?
Overall though, I am perplexed as are many to the strategy. The architecture is perhaps too reliant on multi-core software capability rendering its architecture fairly disadvantageous to competitors products especially in the consumer market where I feel people would have preferred a faster per core Quad Core instead of a slower per core Octo Core. This will be a hard sell indeed, it has just changed by specification for example, I will now have to go down the monopoly Intel route and maybe buy an AMD GPU just for ethics.
Finally I am wondering whether this strategy could seriously damage profitiability or perhaps even bankrupt the CPU Division. Hopefully the Opteron's will offer exceptional performance so AMD can capitalise on the server market (Where the most money is made) together with their relatively successful laptop and netbook range and survive.
On the plus side to a unlikely but potential bankruptcy, maybe a court order could bypass the x86 licencing agreement and allow big blue IBM to purchase one the CPU division :rockon:
What do you guys think?
P.S This is just my own opinion formed by looking at the reviews and architecture. Feel free to correct me If I have slipped up anywhere btw, knowledge is power.
This has changed by entire specification for a new PC I was going to order within the next 2 months, obviously from here
Was going to purchase a laptop here 7 months ago, but I came to my senses and waited for AMD FX so I could build a new Desktop :winkiss: and return to the promised land HOWEVER It appears to me that AMD has perhaps focussed too much on the idea of Multi-Threaded HPC and the more cores the better. The problem is there is insufficient software available to take advantage of multi-core processors and they have relied too much on this, as evident in the performance of the AMD FX.
It has very competitive multi-core performance when utilised with programs like MediaShow Espresso, Handbrake, and certain Adobe products which offers performance equal and indeed superior to the 2600K justifying it's pricing in between the 2500K and 2600K. It thrashes the 2500K in a lot of these benchmarks. The problem however is that AMD is relying on software to be compliant with multi-core hardware, even Windows 8 will supposedly bring an improvement to the performance of the AMD FX though likely not as much as they would like you to believe.
In the world we live in today there are few programs which actually fully utilise 4 or more cores, especially games to the non delight of pc gamers everywhere. The problem is the AMD FX's performance per core is worse than its predecessors due to its architecture design for multi-core applications. It consistently comes behind the X6-1100T and sometimes even the X4-980 in games and single thread or core processes. Even in 2 core processes it just does not have the muscle and considering this is the situation the CPU will find itself in predominantly it will make it very difficult to market to anyone but Media Enthusiasts and Designers who need multi-core operations.
Now I am not sure how the Opteron's will fare. You can clearly see in the Dhrystone and Whetstone benchmarks the capability and potential that justifies the price point, (though in reality this price point is not justified to poor actual performance). The multi core abilities may fare very well in the server market and indeed maybe equal to Intel's products, there is reason to believe they could even offer superior performance (especially price performance). Has AMD tried to sell the consumer market a Server chip?
Overall though, I am perplexed as are many to the strategy. The architecture is perhaps too reliant on multi-core software capability rendering its architecture fairly disadvantageous to competitors products especially in the consumer market where I feel people would have preferred a faster per core Quad Core instead of a slower per core Octo Core. This will be a hard sell indeed, it has just changed by specification for example, I will now have to go down the monopoly Intel route and maybe buy an AMD GPU just for ethics.
Finally I am wondering whether this strategy could seriously damage profitiability or perhaps even bankrupt the CPU Division. Hopefully the Opteron's will offer exceptional performance so AMD can capitalise on the server market (Where the most money is made) together with their relatively successful laptop and netbook range and survive.
On the plus side to a unlikely but potential bankruptcy, maybe a court order could bypass the x86 licencing agreement and allow big blue IBM to purchase one the CPU division :rockon:
What do you guys think?
P.S This is just my own opinion formed by looking at the reviews and architecture. Feel free to correct me If I have slipped up anywhere btw, knowledge is power.
Last edited:
Gorman
Author Level
Where to begin guys.
This has changed by entire specification for a new PC I was going to order within the next 2 months, obviously from here
It appears to me that AMD has perhaps focussed too much on the idea of Multi-Threaded HPC and the more cores the better. The problem is there is insufficient software available to take advantage of multi-core processors and they have relied too much on this, as evident in the performance of the AMD FX.
It has very competitive multi-core performance when utilised with programs like MediaShow Espresso, Handbrake, and certain Adobe products which offers performance equal and indeed superior to the 2600K justifying it's pricing in between the 2500K and 2600K. It thrashes the 2500K in a lot of these benchmarks. The problem however is that AMD is relying on software to be compliant with multi-core hardware, even Windows 8 will supposedly bring an improvement to the performance of the AMD FX though likely not as much as they would like you to believe.
In the world we live in today there are few programs which actually fully utilise 4 or more cores, especially games to the non delight of pc gamers everywhere. The problem is the AMD FX's performance per core is worse than its predecessors due to its architecture design for multi-core applications. It consistently comes behind the X6-1100T and sometimes even the X4-980 in games and single thread or core processes. Even in 2 core processes it just does not have the muscle and considering this is the situation the CPU will find itself in predominantly it will make it very difficult to market to anyone but Media Enthusiasts and Designers who need multi-core operations.
Now I am not sure how the Opteron's will fare. You can clearly see in the Dhrystone and Whetstone benchmarks the capability and potential that justifies the price point, (though in reality this price point is not justified to poor actual performance). The multi core abilities may fare very well in the server market and indeed maybe equal to Intel's products, there is reason to believe they could even offer superior performance (especially price performance). Has AMD tried to sell the consumer market a Server chip?
Overall though, I am perplexed as are many to the strategy. The architecture is perhaps too reliant on multi-core software capability rendering its architecture fairly disadvantageous to competitors products especially in the consumer market where I feel people would have preferred a faster per core Quad Core instead of a slower per core Octo Core. This will be a hard sell indeed, it has just changed by specification for example, I will now have to go down the monopoly Intel route and maybe buy an AMD GPU just for ethics.
Finally I am wondering whether this strategy could seriously damage profitiability or perhaps even bankrupt the CPU Division. Hopefully the Opteron's will offer exceptional performance so AMD can capitalise on the server market (Where the most money is made) together with their relatively successful laptop and netbook range and survive.
On the plus side to a unlikely but potential bankruptcy, maybe a court order could bypass the x86 licencing agreement and allow big blue IBM to purchase one the CPU division :rockon:
What do you guys think?
P.S This is just my own opinion formed by looking at the reviews and architecture. Feel free to correct me If I have slipped up anywhere btw, knowledge is power.
like i said, lol.
Yes, that is what I was thinking. In the long-term it is a good strategy, however there is an awful lot of potential earnings loss before that occurs. Also add this to the fact that when multi-core hardware like that begins to be utilised you better believe it that Intel will have something out then and it will likely be better - Haswell. Tbh, when a company probably spends more on R@D than another company receives in revenue you can assume the first company will win all future product battles unless they are extremely inefficient or get complacent and ditch the R@D - remember the Athlon it caught them napping. Tick Tock is too much though.
NMEBowen
Master Poster
Was that pish i heard that the 8 cores can be paired making it shear amazing running 4 core programs? I looking a fair bit into bulldozer and based my motherboard choice on its compatability as i saw that the sabertooth was designed primarily for this chip. i hope to sli my 580 and add a bulldozer chip aswell as a wc loop a topic i may ask about Gorman prices from you and parts i would have to send but is this chip going to be worth it?
That is fair enough. It is really down to choice on GPU's really. The products are similar. Like I said in one of my previous posts that appears to have now sprung up on the last page this long term strategy will be shown its flaws when Intel just wait for multi-core computing to mature then release Haswell.
well we have multi core processing, just not designed for 8 cores, that is why intel always outperform AMD, because they have the right amount of cores required, im really unsure whether to go for an 8 core bulldozer or i5 2500 because the 4 cores may not be enough in a few years but the AMD cores are clocked much lower
NMEBowen
Master Poster
3.6 base speed 4.2 clocked aint that slow tbh
Edit: not to metion the fact their is 4 core and 6 core versions that aint that slow and can multithread now so lets just say amd are back in the race for a bit their chips can compete a bit better with intel till the next step.
Edit: not to metion the fact their is 4 core and 6 core versions that aint that slow and can multithread now so lets just say amd are back in the race for a bit their chips can compete a bit better with intel till the next step.
Last edited:
It's like I said, we will have to wait for multi-core technology to mature further, perhaps enabling software to scale. I personally am going to go with the 2500K. Intel will have Haswell out by the time there is multi-core support for 4+ cores in mainstream computing. And AMD will still have the same problem as they will still have the same architecture. The FX-8150 matches the 2600K in applications designed for 4+ multi-core but it will not match Ivy Bridge or Haswell.
I guess I will have to go off and spec up. Might throw in a few questions tomorrow in the forums about a spec I am thinking of due to the annoyance of HDCP, PAP and soundcards.
I guess I will have to go off and spec up. Might throw in a few questions tomorrow in the forums about a spec I am thinking of due to the annoyance of HDCP, PAP and soundcards.
NMEBowen
Master Poster
Gorman
Author Level
Best post in the history of this forum.
