lol single core K10 anyoneQuote:
Originally Posted by Kuroimaho
anyone know how much AMD pays for a silicon ingot ?.....
Printable View
lol single core K10 anyoneQuote:
Originally Posted by Kuroimaho
anyone know how much AMD pays for a silicon ingot ?.....
good thinking. either that or the price will go up substantially. but there is no much room to go higher since the tri-core prices. all depends on how much garbage chips they get from wafer. and i hope this is only temporary, amd needs to offer true dualies with new process.Quote:
Originally Posted by Kuroimaho
If AMD runs out of broken chips that's good news for us. That would mean they no longer fab flawed K10 quads at all!!!! Nobody on the planet has a process that amazing so I doubt they will run out anytime soon.
This little chip could be a nice stopcap solution before the 45nm phenoms roll out :)
Quote:
Originally Posted by FlanK3r
Sad.Quote:
Originally Posted by techpowerup
so, you expected 2.3ghz to outperform 3.2? It's not like this is a new architecture...Quote:
Originally Posted by Periander6
I still wonder how "some people" could make a 2.3Ghz Phenom beat a 3.2ghz K8...
OC's?Quote:
Originally Posted by Shintai
And IPC wise it's better too. But stock, hardly to not. X2 6000+ is 'hard' enough.
Quote:
Originally Posted by Shintai
We don't. I'd say that it sometimes gets close to or exceeds the 6000+, but definitely not the 6400. Maybe if you stopped pushing people's words with insults attached you'd get taken barely seriously.
review X2 6500+ on expreview
http://www.expreview.com/img/review/kuma/kuma_cpu01.jpg
http://www.expreview.com/img/review/...ry_changed.png
CPU Z
http://www.expreview.com/img/review/..._cpuz_01_s.png
http://www.expreview.com/img/review/...puz_idle_s.png
load 5000+ system 215W and 6500+ 237W.
A OC, 30minutes wprime ok 1.44V:
http://www.expreview.com/img/review/...c_sp2004_s.jpg
1.45V
http://www.expreview.com/img/review/kuma/oc_limit.png
http://www.expreview.com/img/review/kuma/score.png
SO, it uses 22W more at 2.3ghz vs a K8 at 2.6? Seems unlikely, as the B3's are capable of 65W rating at 2.0ghz for 4 cores....
Look at the rest then. At 2.3Ghz Quadcore the TDP rating is what?Quote:
Originally Posted by ryboto
http://www.amd.com/us-en/Processors/...E15347,00.html
But again, Nice PR stunt to cheat customers. 6500 rating...5000 performance.
it is maybe only 1 cons for this CPU, performance is good
maybe now 65w TDP (new 9950 125W, 9750 new 95W, new 8450 65W)
Conroe has 4mb of cache and this Kuma has 3mb.
thats 25% less cache and 1-2% slower pre clock.
the Deneb should come to have about 512kb L2 x2 plus 6mb L3 thats 7mb's that's more then Wolfdale has (6mb)
if it where just added cache it would be 2-5% faster but where getting a bouns of added IPC to that it.
Looks like it's going to be a great 45nm chip for AMD.
And total difrenet architecture (core vs Agena) :)...Phenom, Deneb is more similar as NehalemQuote:
Originally Posted by demonkevy666
But look to LostCircuits, the B2 9600 only consumes 73W before the VRMs, that's the same speed, with 2 more cores.Quote:
Originally Posted by Shintai
It looks like the 6500 won't be a bad CPU for gaming compared to what AMD has out now, when overclocked at least. If you can get the 6500 to 3.3GHz, it will perform fine in games and easily outrun any Brisbane or the X2 6400+. If it ends up costing somewhere around ~$80 or so, then it won't be that bad of an option. If it costs $100 or more though, then it is a really bad deal in comparison to the E7200.
Whoa, 3.3 Ghz stable? I might get this if the price is right, it should match great with the SB750 DFI I wanna buy next month. :D
And I think they called it 6500+ because, at max OC, it's way faster than what a 6400 OC-ed can achieve.
Btw, this guy Shintai hates AMD or what's wrong with him?!
Perhaps he's a little blunt, but in the end all desktop derivatives of the 65nm Phenom core have been lousy CPUs compared to their Intel competitors. The 2.3 Kuma looks to be no exception given it will be outperformed by the E5200 while using significantly more power.Quote:
Originally Posted by azraeel101
Not even to talk about the 2.6Ghz K8 with lower power consumption.Quote:
Originally Posted by accord99
Performance/watt K10 loses to K8 in the dualcore shootout.
Like I said, it's strange, since K10 performance/Watt is technically greater than K8. Maybe the disabled cores are the issue?Quote:
Originally Posted by Shintai
Backlash only because if its name! Its not like a Tech Gimmick put there to mislead but Named to mislead. Intel named some of the slower Core processors Pentium while everyone knows Core is faster.Quote:
Originally Posted by ryboto
It boils down to simply scaling of two variables that have terms within them that are not interdependent nor scale 1:1, and the fact that power has an inherent offset -- leakage current. The overall affect is it will cause a different perf/watt ratio... example, if the IPC at a given clock improves 10%, but the circuit/architecture increases power consumption by 15%, then the Perf/Watt ratio will go down.Quote:
Originally Posted by ryboto
Another example is the preliminary Nehalem results, Anandtech leaked that power increased 10% over a similarly clocked Yorky, but performance was up 20-40% so Perf/Watt will go up. Logic/circuits added that improve IPC will also affect power consumption .... after reading around, the Nehalem design was contrained to a ratio of 2% perf to only 1% consumption increase, if it did not meet this goal then the feature was not included.
Not much to see here, and no reason to grab one of the last 65 nm CPU's now.
Sure, some people want AMD, but if you haven't already got a K10 then you might as well wait for 45 nm.