Bulldozer
- Thread starter StrangleHold
- Start date
maroon1
New Member
Quick question
Just wanted to know whether there is going to be drop in prices in Intel CPUs?
The FX-8120 seems to low in cost compared to what it gives.
Why would intel drop the price when they have faster CPU ?
It is other way around. AMD should drop the price of bulldozer to compete with SB
i5 2500 beats FX-8120 in almost all benchmarks. It even beat FX-8150 in a lot of benchmarks
Not to mention that sandy bridge has lower power consumption.
Last edited:
Ankur
Active Member
I just saw the benchmarks and 2500k seems to beat them.
Hmm. . . I'm not sure which CPU to go for.
claptonman
New Member
http://blogs.amd.com/play/2011/10/13/our-take-on-amd-fx/
At least they responded to everyone. I'm still not buying, so until they prove this:
"This is a new architecture. Compilers have recently been updated, and programs have just started exploring the new instructions like XOP and FMA4 (two new instructions first supported by the AMD FX CPU) to speed up many applications, especially when compared to our older generation."
is truth, then no dice.
At least they responded to everyone. I'm still not buying, so until they prove this:
"This is a new architecture. Compilers have recently been updated, and programs have just started exploring the new instructions like XOP and FMA4 (two new instructions first supported by the AMD FX CPU) to speed up many applications, especially when compared to our older generation."
is truth, then no dice.
jonnyp11
New Member
well it is true to a degree, i said this before and will say it again, on the anandtech review they contacted amd and found that the win7 scheduler doesn't know how to properly use the bd modules so it doesn't get to turbo or get efficient spreading of the tasks over the modules, while the win8 dev preview does have the scheduling and in the 4 test they ran the bd improved its scores from 4 to 10% over the originals. then if the tasks themselves can't make use of some of the other features that would also hinder the performance, we can't truly see this processor for what it is when half of the things that make it faster can't be used. that is the problem with introducing newer technology. it's like (not sure if this is a good analogy but) putting a crapo transmission on a bugatti veyron.
Seems most or maybe all these benchmarks that came out on the 12/13th were still using a pre released ES models. Not retail models. Dont know how much difference there is in the ES and released/final model. But seems pretty bogas to me.
jonnyp11
New Member
i still like this though for all the stuff saying it sucks (from passmark)
Intel Core i7 995X @ 3.60GHz 10,945 NA
Intel Core i7 990X @ 3.47GHz 10,927 $989.99*
Intel Core i7 980X @ 3.33GHz 10,603 $1,021.43**
Intel Core i7 980 @ 3.33GHz 10,247 $599.49**
Intel Xeon W3690 @ 3.47GHz 10,227 $1,079.99*
Intel Core i7-2600K @ 3.40GHz 9,976 $427.31**
Intel Xeon X5690 @ 3.47GHz 9,943 NA
Intel Core i7 970 @ 3.20GHz 9,939 $559.99*
Intel Xeon X5680 @ 3.33GHz 9,838 $1,625.00*
Intel Xeon W3680 @ 3.33GHz 9,833 $599.99*
Intel Xeon X5670 @ 2.93GHz 9,214 $1,442.01*
Intel Xeon E31275 @ 3.40GHz 9,126 NA
Intel Core i7-2960XM @ 2.70GHz 9,084 NA
Intel Xeon E31270 @ 3.40GHz 9,084 NA
Intel Core i7-2600 @ 3.40GHz 8,955 $299.99**
Intel Xeon E31280 @ 3.50GHz 8,724 NA
Intel Xeon E31245 @ 3.30GHz 8,697 NA
AMD FX-8150 Eight-Core 8,681 NA
Intel Core i7 995X @ 3.60GHz 10,945 NA
Intel Core i7 990X @ 3.47GHz 10,927 $989.99*
Intel Core i7 980X @ 3.33GHz 10,603 $1,021.43**
Intel Core i7 980 @ 3.33GHz 10,247 $599.49**
Intel Xeon W3690 @ 3.47GHz 10,227 $1,079.99*
Intel Core i7-2600K @ 3.40GHz 9,976 $427.31**
Intel Xeon X5690 @ 3.47GHz 9,943 NA
Intel Core i7 970 @ 3.20GHz 9,939 $559.99*
Intel Xeon X5680 @ 3.33GHz 9,838 $1,625.00*
Intel Xeon W3680 @ 3.33GHz 9,833 $599.99*
Intel Xeon X5670 @ 2.93GHz 9,214 $1,442.01*
Intel Xeon E31275 @ 3.40GHz 9,126 NA
Intel Core i7-2960XM @ 2.70GHz 9,084 NA
Intel Xeon E31270 @ 3.40GHz 9,084 NA
Intel Core i7-2600 @ 3.40GHz 8,955 $299.99**
Intel Xeon E31280 @ 3.50GHz 8,724 NA
Intel Xeon E31245 @ 3.30GHz 8,697 NA
AMD FX-8150 Eight-Core 8,681 NA
Well I take back some of what I said about it. Not as good as I wanted, but better then I initially thought.
Legit is using a retail and it doesnt really do that bad.
http://www.legitreviews.com/article/1741/1/
Legit is using a retail and it doesnt really do that bad.
http://www.legitreviews.com/article/1741/1/
Troncoso
VIP Member
I'm still waiting for more than one sample from passmark. A single person can easily skew the tests and throw the numbers off the actual value. Once they get a few on there and come up with an average, we'll get a better idea of how it stacks up
linkin
VIP Member
http://translate.google.com/transla...2+translate&hl=en&biw=1472&bih=848&prmd=imvns
http://www.xtremesystems.org/forums...ew-(4)-!exclusive!-Excuse-for-1-Threaded-Perf
Rather interesting. Seems an updated BIOS or software patch is needed to show true performance.
http://www.xtremesystems.org/forums...ew-(4)-!exclusive!-Excuse-for-1-Threaded-Perf
Rather interesting. Seems an updated BIOS or software patch is needed to show true performance.
jonnyp11
New Member
where was it doubled? and also what was the point of disabling half the cores, to demonstrate the benefits of the shared resources? i mean when they disabled the other resources it gave the remaining ones access to more resources (mainly caches) so obviously it will do more than the 2 module 4 core, although seeing the benchmarked benefits is interesting.
Windows doesnt know how to distribute threads to Bulldozer/Zambezi yet. It will have higher performance if say, its running 4 threads if each thread is run one module each. Instead on all 4 on 2 modules. Windows just distributes threads on any open core.
kennebell347
Member
I bit the bullet and went Intel. Asrock P67 Extreme4 Gen3 motherboard and i5 2500k. I cant rely on AMD working out the issues.
spynoodle
Active Member
No wonder single core performance sucks: The whole point of a module was to provide a combination of great singe-threaded and multithreaded performance. If the threads are basically being distributed by a mentally disabled monkey (Windows 7), then the whole point is ruined. Really, Microsoft, really? I wonder if Bulldozer performs better in Linux.