Discussion in 'News' started by btarunr, Sep 24, 2011.
Who has a Xeon?
Not enough is known about performance to determine who trumps who yet.
im sure that 8 core intel chips trumps 16 core AMD.
end of story
That just proves your ignorance.
This I guess is proof why for all intents and purposes AMD is actually doing very badly in the processor front, compared to the video card front.
With their GPUs, they're able to compare their next-gen with the competition's (Nvidia's) next-gen, or more obviously their current-gen is comparable to the competition's current-gen.
In the CPUs, they're essentially only making their next-gen comparable to the current-gen of the competition (Intel).
Bulldozer is meant to compete with SB, Piledriver is meant to compete with IB. What's the issue?
the issue is with AMD is unable to compete with SB
at least in the raw core performance
And Bulldozer was released more than 6 months ago?
Although Nvidia wasn't quick as well with the 400 and 500 series, the "time-delay" between their and AMD's respective "current-gens" aren't as far apart as Bulldozer is to Sandy Bridge.
And as an additional point, with AMD comparing the FX 8150 to the Core i7-980X would be akin to Nvidia comparing the GTX 480 to the HD 4870X2, which they didn't do, unless you considering "HD 5870" as corresponding to "HD 4870X2"...
Hyper threading is basically a virtual "core" So when i see a BD chip going head to head to a 6 core and 6 virtual "core" (12 cores total) 980X i think that pretty dam impressive and for quarter of the price.
You argue like you want to go back to AMD so bad....
Don't worry it hardly shows...
And that matters how? I don't care what's inside the CPU as long as it performs. It's stupid to say that what is inside the CPU is more important than how the CPU performs.
No, my argument is that 2600k will sometime perform like the cheaper 2500k. So is it more about a broken architecture or more about well threaded programs?
BD has 8 integers, 4 FPUs.
2600k has extra decoders to help with threading (if I remember correctly).
It's about getting more threaded performance for the die space. In either case the better the program (important part here) is written for multi threading, the better performing the 2600k and the 8 core BD will be over the 2500k.
When it comes to performance over 2500k, then BD wins every time (if the price is correct which I'm not sure it is).
While I agree on this (the performance part not so much the virtual core part), it also proves my point of multi thread processors need programs that take advantage of them or you end up with a CPU (AMD or Intel) that performs like a cheaper version.
The virtual cores don't really do a whole lot of work, the processor is still has 6 physical cores to do the work, and they can only do the work of 6 physical cores. No matter what it appears to the OS.
The BD chip doesn't go head to head with the 980x, it barely touches the 980x in multithreaded apps, and the 980x is a generation old. The BD really goes head to head with the 2600K for pretty much the same price.
If you hadn't noticed, I do own an AMD rig(actually several, just only one in the 4 primary machines I use). And in the next few weeks I'll be replacing another one of my Intel rigs with an AMD because for what it is used for, AMD is a better buy.
Really, because you pretty clear said:
Your lack of knowledge on how HT works, and how closely the 2600K performs to the 2500K really shows. Even in the best coded multi-threaded apps, the 2600K only gets a relatively small boost thanks to HT.
Than why are you still arguing about how "bad" AMD 8 cores is?
If the price & perform is real , it will be a good CPU for the price IHO no matter how many cores it has
If the 2600k is a 4 core then why doesn't it perform exactly like a 2500k? Because of extra hardware enabled so it can handle more threads.
Sorry for poorly chosen words. My point was pointing out that saying that the 2600k is just a 4 core CPU is misleading since it can handle 8 threads. It is more then just a 4 core (4 core plus extra hardware) otherwise it would perform like the 2500k since it could only handle 4 threads like the 2500k.
I prefer to compare performance based on number of threads a CPU can handle. Cores are misleading (I think) because you leave out the ability of some CPUs to handle more threads (therefore) and allow them to perform better in certain circumstances.
If a program only does 4 threads, then both BD and 2600k look bad from an architecture and price standpoint (compared to a 2500k). The more threads the program can do, the greater the potential for BD and 2600k to distance themselves from the 2500k and provide a better price/performance. The closer the program relies on more actual "core hardware" the better the BD architecture should distance itself from SB. At this point, it really depends on how well the program is written to take advantage of the architecture. Lean either way (BD or SB) and one will perform better then the other.
If anything, I hope this will drive down Intel's 32nm Westmere prices so I can move up to a 6 core 1366 CPU for even less. That's my hope.
I think AMD marketing its 4 modules as 8 cores is really biting them in the butt when it comes to some people.
Going by all the information AMD has released they really are just dual threaded cores with enough extra non shared hardware that AMD felt it suitable to market as separate cores thus some people think each "core" should be better than they are.
What makes me laugh though is that when it comes to average consumers the whole 8 core thing will go down well and lend it's self great to marketing use so as far as mass market goes it will probably be a good thing but it's the people who should know better that just can't seam to accept the architecture for what it is.
Then whats the point of hyper threading then???
Far as i know a virtual core is about 25% of a normal cores performance, it mimics the core.
Well we dont know that with 100% as none of the tests show that, only against he 2600K and it wins half losses half.
We need some better benches to see how it realy all turns out. So far im happy been a AMD user i can now get a CPU upgrade to be on par with others from the other camp for cheaper, win win for me.
FX 8150 vs. i7-990X, FX 8150 performs well: Boohoo, 980X is slower than SB
FX 8150 vs. i7-2600K (SB), FX 8150 performs well: Boohoo, that's an unfair 8 core vs. 4 core comparison
so to get what some people are saying straight: is bulldozer using AMD's version of hyper threading, or are they all 'real' cores?
i think thats what i've seen people argue about a few times now
OK, imagine the lulz if AMD marketed Bulldozer as an SMT quad-core chip (4c/8t) instead. Bulldozer would have still gone on to beat i7-980X (6c/12t) in multi-threaded tests, and Intel would run for cover.
They are indeed a real core (to be exact, an integer core. AMD has moved away from a "real core" marketing slogan and what we see as a "real core" right now).
4 module, each module has 2 integer core, and each integer core has its own L1 data cache, integer scheduler and integer datapath, while sharing L2 and L3 cache between integer core in one module.
This is the best die shot I have on BD architecture:
Fixed that for you.
Neither. It's the difference between SMT and CMT.
Bulldozer(2011) => Piledriver(2012) => Steamroller(2013)
Haswell(2013) -> Haswell(2013) -> Haswell(2013)
and 8-cores Sandy Bridge-EP vs 16-core Interlagos
Interlagos wins the price and power brackets at being $1400~ below and below 150~ watts while maintaining Intel's $5000 and above performance bracket
Server market is much more diluted than the Desktop market
Also, these benchmarks/demos weren't done by AMD...just telling
Buahahahaha, you have got to be joking right. Or am i blind? In which multithreaded test did BD beat I7-980X(which is more than a year old now)? Your comments are a joke. Mean no disrespect, but seriously, wtf.
Oh and lets not forget, that AMD also uses higher clock speeds, so that means more cores with higher clocks, for less performance. Well done AMD.:shadedshu
Well, at least AMD has gotten one gen closer to Intel.
BTW: 2 all those that say 2600K outperforms 980X (let alone 990X) - i think it's on a contrary: don't remember @ which sites, but in most cases 990X & 980X outperform the afformentioned CPU stock & OC'd; in games & synthetic benchies i think it's better than 2600K excluding apps that specifically desingend for SB if there's any. No fanboy s*** on my side, but this Westmere 980/990X is still heck of CPU compared to SB; not that 2500/2600K's are bad: if i had enough ca$h i would buy one BD-based gaming rig & one SB/SB-E-based rig & than pit one gainst the other to see which beats the f*** out of the other.
Regardless of what i said above, these news completely acceptable by me, since this is how i hoped the BD (this FX-8150) will fair out. Now the only thing left to know is how it'll run when paired with C5F, Sabertooth 990FX or any other 990FX/X-based mobo, 2x4GB DDR3 1600MHz RAM, 2xGTX 4xx/5xx or 2xHD 6xxx, 1kW+ PSU & fast HDD or SSD & tested here. If you'll pit it (FX-8150) gainst both 990X & 2600K - ace.
P.S. Is there a CPU-based Photoshop benchie to test multithreading? I might check this one out once i'll buy this CPU; probably end of October/middle of November.
Separate names with a comma.