That doesn't mean I can't feel sorry for them.Quote:
Originally Posted by Hell Hound
Printable View
That doesn't mean I can't feel sorry for them.Quote:
Originally Posted by Hell Hound
:clap::clap::rofl:Quote:
Originally Posted by Solus Corvus
Its actualy kinda wierd how far it seems hardware has come in the last 4 years or so. and then to watch my kids playing Crysis on those old machines and it actualy looks not that bad (for Low settings) with smooth FPS realy makes me question just how far have we realy come.
It feels like a huge leap for me. But then again, I came from a FX-51 to a QX6700.
But honestly, Donnie is my favourite. :ROTF: And actually, I don't mean it in negative way. :) I'd grant him a beer or two if I was able to. :yepp:Quote:
Originally Posted by savantu
Yes you're my favorite too but I wouldn't want to get in trouble for buying a minor a beer:rofl: Anyway, lighten up it's alright:up:Quote:
Originally Posted by Calmatory
I just tested K10 B2 L3 Cache Latency with RMMA giving an additional 200 MHz boost at the integrated northbridge. So far i achieved an average latency of 16,89ns (random access) for the L3 Cache. I will share the data if you want.Quote:
Originally Posted by savantu
We cannot draw a line to Deneb, but my guess its that Denebs iNB will be clocked higher then Agena. So i think AMDs 45nm silicon will be close to Nehalems latency score of the L3 Cache.
I doubt.
http://www.behardware.com/art/imprimer/682/Quote:
Originally Posted by Boschwanza
Are these figures wrong?Quote:
Latencies, associativity and dissipation
The measured latencies of the K10’s caches show 3 cycles for the L1, 15 for the L2 and between 30 and 45 for the L3. These values merit some explanation.
L1 latency has not changed compared to the K8, and this is for the best because it contributed to the latter’s good performances. The L2 figure is good news following the "surprise" of the 65 nm version Athlon 64 (Brisbane core) whose latency underwent an increase that was truly detrimental to performances. Luckily, the K10 has been spared and L2 latency reflects more the value that we observed on the Athlon 64 90 nm (Windsor core).
Memory latency has also slightly increased from the presence of L3 cache due to memory access which occurs from the moment when each cache level has not responded positively to a request. To be sure, it takes a little more time to access memory, but we shouldn’t conclude from this that the addition of a cache level slows memory access. It’s even the opposite because the presence of L3 significantly reduces the number of memory accesses.
I'm guessing the difference is in cycles versus time (in nanoseconds).
but even at 3400 still not able to really outperform the stock 3500+ :DQuote:
Originally Posted by Donnie27
:yepp:Quote:
Originally Posted by xVeinx
At same clock frequency, less cycles is always faster. More cycles & faster means higher frequency, and thus can't be compared one-to-one in terms of IPC.Quote:
Originally Posted by xVeinx
exactly.Quote:
Originally Posted by duploxxx
you were able to oc the netburst cpus like hell, but especially for games this wasn't enough to outperform a lower clocked k8.
seriously, to debate about whether k8 was really faster than any netburst cpu is nonsense. we know that k8 > netburst since these cpus got released. no need to question half a decade of benchmark results rofl... :rofl:
even the p4 server cpus are slower than any older opteron. we're still running a lot of server xeons (server p4s) in some of our blades at work, and so we do k8-opterons. and even in spi a dualsocket singlecore opteron at 2,4ghz beats a dualsocket singlecore xeon at 3,6ghz.
(yes, in my spare time at work i do indeed bench all of our non-production windows-servers :D)
I think what you're seeing is the prefetchers at work.What block size did you use ?Quote:
Originally Posted by Boschwanza
Digit life obtained very different results.
http://www.digit-life.com/articles3/...ma-phenom.html ( look in the middle )
why are you guys bickering about p4 vs k8? k8 was to p4 as conroe is to k8 right now, no competition. I think you guys should focus more on the deneb discussion, I for one am very interested in how the latency of the l3 cache will affect deneb's performance compared to nehalem (of course we have to take into account nehalem is the superior design from the start)
hahaha, now that is extreme!! lol :up:Quote:
Originally Posted by RaZz!
Because they always have. Anything showing AMD having any advantadge over Intel, even in the past, can't be allowed. Think twice before you listen to some of these folks, you can't trust anyone with an agenda. :rolleyes:Quote:
Originally Posted by AliG
On some apps yes, on others no it couldn't. Problems start when Fans try to say either would out perform the other all the time. That ain't happening. As I've been jumped for saying before, it was app dependant on which was faster. I liked my 3500+ better than my OC 2.6C for games, easy choice. I moved the 2.6C to the HTPC where it fit better than the 3500+.Quote:
Originally Posted by duploxxx
honestly donnie, stop kidding yourself, same way a 3.4ghz k8 can barely touch an e6700, a 3.4ghz p4 could barely touch a 2.5ghz k8, end of story.
Just drop the subject already, both architectures are long extinct and no longer matter that much anymore, now its all about k10.5 and core 2/nehalem
No, not exactly and not even close=P Netburst overclocking weren't just wasted faster cycles, only someone fooling themselves or blind believes something that lame. That's exactly what I meant in my reply to Duploxx. I had both and still have the 3500+ powering my wife's computer right now.Quote:
Originally Posted by RaZz!
I'd tried to sell the 3500+ and the 2.6C, the person testing both picked the 2.6C. When I asked her why, she said the P4 felt smoother LOL!:rofl: No. honestly she said she trusted Intel more. I was glad she didn't pick the 3500+, I liked it better:up:
Sorry, I'd rather be unpopular and tell the truth than lie to make you and others feel better. I absolutely stand by what I said. 3500+ was better at games and running some apps faster one at a time. Multi-Tasking was faster on the 2.6C, Majx-, Nero, Creative, Roxio and many other Multimedia apps ran faster on the 2.6C one at a time and two at a time=PQuote:
Originally Posted by AliG
He is right though. Just as Phenom can beat Core 2 in certain apps but loses in most, back then P4 could beat K8 in certain apps. If I was building a computer to do media encoding back then I would have picked P4, but since I was a gamer I went K8. Processors handle different tasks at different speeds, then as now.Quote:
Originally Posted by AliG
It's not as if processor performance is a swimming match where the first to the line is the winner. It's more like a decathlon where each user picks their own events.
But I agree that this is quite a bit of a tangent. I want to know how Deneb/Shanghai stacks up against Nehalem. Speculating is fun, but we will have to wait for benches to know for sure.
QFT! And very well put! If more folks understood this, there'd be fewer Flame Fests.Quote:
Originally Posted by Solus Corvus
Deneb official specs at the moment are a NB speed of 2.2 on the first released stepping. Can change offcourse, so can actual launch speed, but will be higher then current but stay in same thermal for early release.Quote:
Originally Posted by Boschwanza
then start to check your own posts, its always you that brings up the old p4-3500+ comparison, multi tasking with more computers :D , aan price of the long due x2 3800 against current pricing....Quote:
Originally Posted by MrMojoZ
not to mention your QFT quotes (+1000 pieces) and the unknown multi-quote button.....