Discussion Forums > Technology
Your view on AMD's Bulldozer
kureshii:
--- Quote from: TMRNetShark on October 13, 2011, 12:36:33 PM ---They are releasing a water cooled bundle of the FX chip in Japan. I wonder how people will like the CPU over an Intel chip. As of right now, only AM3+ boards will fit it, I guess, right?
--- End quote ---
Intel's right on that too:
http://www.legitreviews.com/article/1707/1/
Although it'll only come with SB-E. Seems they went with the same OEM, unsurprisingly.
http://www.legitreviews.com/news/11637/
kitamesume:
even if they could reduce the TDP it would cause them to lower in performance as well(if it doesnt then tell me how they do their wicked magic XD ), lowering the performance would make it even less viable as theres other options.
not only that, alot servers are placed on rackmounts which are known to be compact, putting in a full blown reactor in their would cause it to overheat which would further reduce the performance.
so theres alot of problems to tackle right now, the 8core BD are already running at around 200watts, pressing it another 2cores would further increase it by 50watts? reducing the clock speed would just set back the advantage of adding another set of cores...
TMRNetShark:
--- Quote from: kitamesume on October 13, 2011, 01:01:28 PM ---even if they could reduce the TDP it would cause them to lower in performance as well(if it doesnt then tell me how they do their wicked magic XD ), lowering the performance would make it even less viable as theres other options.
not only that, alot servers are placed on rackmounts which are known to be compact, putting in a full blown reactor in their would cause it to overheat which would further reduce the performance.
so theres alot of problems to tackle right now, the 8core BD are already running at around 200watts, pressing it another 2cores would further increase it by 50watts? reducing the clock speed would just set back the advantage of adding another set of cores...
--- End quote ---
So the Bulldozers will be in the 2.0 GHz-3.0 GHz bracket. XD Hey, remember kids... Pentium 4's had their limits. Thanks to multi-cores, we got crazy CPUs that take all this POWER and not do much with it.
Lupin:
--- Quote from: kitamesume on October 13, 2011, 01:01:28 PM ---even if they could reduce the TDP it would cause them to lower in performance as well(if it doesnt then tell me how they do their wicked magic XD ), lowering the performance would make it even less viable as theres other options.
not only that, alot servers are placed on rackmounts which are known to be compact, putting in a full blown reactor in their would cause it to overheat which would further reduce the performance.
so theres alot of problems to tackle right now, the 8core BD are already running at around 200watts, pressing it another 2cores would further increase it by 50watts? reducing the clock speed would just set back the advantage of adding another set of cores...
--- End quote ---
I already mentioned it:
--- Quote from: Lupin on October 13, 2011, 12:31:16 PM ---Workloads are vastly different as well so mediocre desktop performance doesn't immediately translate to mediocre server performance.
--- End quote ---
There are workloads were core count matters more than frequency. For example, in a virtualization workload, you can cram in more services into one server as there are more cores to operate on them. Different services have different utilization; they do not tax all the cores everytime and may fit into the thermal budget. HPC workloads tend to use custom code, compiled and optimized for the processor used. Throughput matters more on server/HPC workloads
--- Quote from: kureshii on October 13, 2011, 11:27:06 AM --- (click to show/hide)http://www.xbitlabs.com/news/cpu/display/20111012151052_AMD_Aims_to_Improve_Bulldozer_Performance_per_Watt_by_Up_to_50_by_2014.html
And we worry more and more about their future ... Remember what we saw from Bulldozer at its best?
They're currently about 50% behind i7-2600K in PPW for heavily multithreaded applications. a 50% improvement in that might help Bulldozer catch up with it ... in 2014. Are we going to see AMD in their Pentium 4 doldrums?
--- End quote ---
Anand's benchmark tool version is ancient (1342). The new version is of that benchmark is also old (1913). The current version of x264 is 2085 (from x264.nl) released about 3 weeks ago. I'm interested in how BD performs in the latest version.
kitamesume:
good point, then again if that would've been the case, i wonder why'd they even released it in the desktop segment, it runs awkwardly horrible on usual apps, and compared with the older models it looks more "meh".
if what you'd say that the newer benches are more efficient that would mean the competition's score would increase as well, or are you saying the newer version is more bulldozer biased? or theres simply a hidden trick on running bulldozer that the older version lacks thats hindering bulldozer's performance?
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version