Too bad AMD didn't update Thuban to a real 8 core 32nm processor. Give it a triple channel memory controller, more L3 cache (12MB+), faster L3 cache speed (3Ghz+) and a target speed of 4+GHz at launch.
Printable View
Too bad AMD didn't update Thuban to a real 8 core 32nm processor. Give it a triple channel memory controller, more L3 cache (12MB+), faster L3 cache speed (3Ghz+) and a target speed of 4+GHz at launch.
Maybe I'm the stupid one here...Quote:
Originally Posted by Manicdan
Llano's IPC difference vs Propus is neglegible. It's worth about 100 Mhz...or roughly 3%. Sure, up to 5% in some apps...too bad we are 30% behind Sandy with "STARS". As far as AMD's efficiency falling with more cores? Yeah. Thuban is a slower CPU in single thread than Deneb (partially due to less cache)...but look it up.
I leave you with a long string of quotes from chew in this thread that were posted in reaction to BD's leaked performance, ending with a response to freeloader:
By the way, VR-Zone Cinebench leak has shown ~5x scaling over all cores from single thread (not one whole module-core like the patent design states)
Regardless of what is behind NDA and what is not, between what we have for current leak and what chew* has said, I expect scaling from 5.0 to 5.5x on retail chips.
Quote:
Originally Posted by chew*
Quote:
Originally Posted by chew*
*insert flank3r post about how X8 would be a multithreading beast*Quote:
Originally Posted by chew*
...I've been claiming far before bulldozer talk how well "STARS" does multithreaded, if multithread was AMD's focus they should have stayed with it in my opinion. On 45nm, it was very competetive with Lynnfield and Bloomfield multithreaded.
Quote:
Originally Posted by chew*
Quote:
Originally Posted by chew*
Quote:
Originally Posted by chew*
Quote:
Originally Posted by chew*
Quote:
Originally Posted by chew*
Quote:
Originally Posted by chew*
Quote:
Originally Posted by chew*
*insert more arguements about how :banana::banana::banana::banana:ty BD is and its not 4 cores blah blah, cinebench sucks it sucks blah blah*Quote:
Originally Posted by chew*
Quote:
Originally Posted by chew*
...some example talk, because PR said it was true and the website said it was true, and AMD has a habit of eitherQuote:
Originally Posted by informal
A. Not talking to public
B. Talking to public (not us enthusiasts directly...like Simon) in horrible ways, misleading info, cheesy presentations, how Denebs with disabled cores are a "True Dual Core Design?"
Quote:
Originally Posted by chew*
I'd give that CPU a launch speed of 3 Ghz. New processes, GF lagging behind and seemingly sucking at yields looks like trouble with new process.Quote:
Originally Posted by freeloader
As far as 12MB of cache...triple channel memory controller, two more cores...about 45-50% more die space (rough estimate, larger DDR3 controller + reorder cores for 3+3+2 and AMD's horrible cache density equalling 15% of Thubans die (6MB)) on a 40% shrink so you are looking at 5% larger than thuban around 350-375 mm^2? Leaves no room at all to grow, even if they could pull that off.
BeepBeep, I admit that I have no clue about what was your intention with the last post. Summary of quotes plus some vague estimate how much die area would X8 K10 have? Who cares about 32nm X8 K10? It's not going to happen so why discuss it? And as for Bulldozer ,there are so many unknowns that it's not even funny. We have no real/official benchmark on final platform, no official launch date, no shipping date for Zambezi and no official price. We do have some bits and pieces here and there,some painting good some bad picture. That's all.
Thats not an upgrade, thats an entirely new processor architectureQuote:
Originally Posted by freeloader
It really should have been three or four posts...and I seem to add on to my thoughts so a lot of information does get jumbled up.Quote:
Originally Posted by informal
Ignore my jumbled response to manicdan + freeloader on "stars" arch to 8 cores on 32nm and the effects it would have on performance + die space. (in freeloaders case, 8 core with 12MB L3 + tri channel memory) would be a larger die on 32nm than thuban is on 45nm.
As far as Bulldozer:
Yes you are right, no "real" official benchmarks.
But there are two peices of information available:
chew*'s words about how the architecture is more like a 4 core with HT + leaked unofficial benchmarks that show multi-threading performance in several cases lower than Thuban, making some people assume that IPC is lower than Deneb. People are looking at these results and dividing by 7 to 8 for "theoretical leaked single thread performance" and talking about how IPC is lower. Maybe it is(!)
...chew* states that running a program as simple as pi on two cores "one module" brings better SINGLE thread performance, goes back to what I predicted a month ago as soon as he said the architecture works more like a 4 core with HT the first time.
Several weeks ago, my prediction was this:
Higher IPC in 4 threads or less
Lower IPC in more than four threads, but better scaling than HyperThreading.
Hopefully this post is a little easier to understand ;)
EDIT:
Cleaned up the first post a bit if you'd like to take another stab at it.
I think his point is: Read between the lines of the post's from one who has had vast experience with AMD chips and hands on with BD. Read the clues!! Has come to the same conclusion as me..... BD wont be the monster everyone is hoping for.Quote:
Originally Posted by informal
I also believe the events this summer, rumors and delays point to the same thing.
Disclaimer: ^ My personal opinion, hope I am wrong and the 5 years of BD planning nets a real winner for the enthusiast .
People in the AMD crowd seem to be way too hung up on multi-threaded performance and the fact that AMD calls this an 8 core CPU(even though they've been crying about single thread but pretending like more cores makes up for it, hence "why didn't they make X8 32nm").Quote:
Originally Posted by Naturecannon
Because leaked benchmarks show the 8 core CPU performing near 2500K level multithreaded, people are disappointed. However, chew* has stated multiple times the CPU internally is a 4 core with 8 threads. People have simply given him hell for making such statements. The funny thing here to me though, is that he has Bulldozer and they do not.
The goal here in my opinion was not to add more cores, but to get single thread performance up. If they can find a way to make CMT work well (1 thread per module) and that does in fact speed up single thread performance, then I see absolutely nothing wrong with this architecture. If you guys haven't noticed, Sandy cores are huge...Llano cores are not. Bulldozer modules are huge...Llano cores are not. 1 Sandy Core = 2 threads...1 BD module = 2 threads.
Don't call it HT, in relation to AMD, please. HyperThreading is a brandname of Intel for its implementations of SMT.Quote:
Originally Posted by BeepBeep2
So, CMT is a kind of Multi Threading. A quite different one to SMT. Otherways both are "physical".
I agree that it's a 4-core + CMT, from engineering POV. Altough, if it performs comparably to a true 8-core of earlier uarch's, then it can be valid to call it a 8-core.
But we have known the processor would act like a 4 core + advanced hyper threading for what, 2 years now? Maybe 3? Everyone got caught up in the whole 8 core thing but we have known for a LONG time now that this architecture that uses cluster multi threading was just a more advanced and better form of HT. It was talked about extensively years ago by AMD. Slides were shown explaining the 4 types of multi-threading in a core and why AMD hadn't done the same thing as Intel and what AMD would do and was trying to implement in Bulldozer for its multi-threaded cores.Quote:
Originally Posted by BeepBeep2
Just 5,95 pts in Cinebench: http://www.donanimhaber.com/islemci/...onuclari_5.htm
To calculate single thread performance one should look at JF's posts. One thread = 100% two threads =180%, this is probably top scaling so probably lower most often, maybe 40-60%. If BD IPC is less than 11% higher than Thuban it will not increase IPC per two threads (111% of Thuban + 88% of thuban) best case scenario. In non optimal scenarios it would need from 25-50% higher IPC than Thuban to equal it per two threads (run within one module of course).
Edit: The purpose of CMT is probably to get a more consistent scaling with the second thread per core than SMT. I also believe that the figure 80% is more likely to happen with a server workload than a desktop workload. That would also explain why the desktop test are lacking in performance compared to Thuban in MT.
shopblt was only 1 source, there were 4-5 other shops, which back-uped the lineup of 3 models.Quote:
Originally Posted by Mats
Now concerning your Interlagos example, that does not fit, because it is another segment. The retail segment of server CPUs is very, very small. If there is demand for an Opteron, than people buy a complete machine+service and don't do a "DIY-build" of the server in their basement.
Therefore, most shops do not have any Opterons, some only have a few. Furthermore, there is no other shop which could back up that fact.
Anyhow, we have confirmation now:
Attachment 120445
People who says that AMD wouldn't be able to add two cores to Phenom II and rise frequencies at the same time. Consider this:
Agena 4C 2.6GHz 2Mb L3 65nm
Thuban 6C 3.3GHz 6Mb L3 45nm
Thats 50% more cores, 27% higher frequency and 200% more cache, and still cooler than Agena. Of course you can't just extrapolate this gains into 32nm, but I think it's reasonable to say that we would be able to see 33% more cores, higher frequencies (even more with turbo) and an extra 2Mb cache and still have a smaller chip than Thuban. Throw in some core enhancements and you're set.
Would it be as competitive as BD? Maybe, maybe not, but I think it's a bad sign when you first really new architecture in 12 years isn't clearly a better choice than the old one. Intel did this once, they crippled their old P6 in favor for Pentium 4 a century ago, it turned out that then the improved P6 came back 5 years later it was twice as fast as their current Pentium 4, and that's dual vs. dual, their P6 derived architecture had more thermal room for extra cores and came as C2Q delivering even more performance.
Well I think even a beefed up high frequency 4C/6C Phenom III with doubled FMAC capable FPU and better prefetchers.would actually perform better against Intels quads than BD will considering the leaks and price. But AMD made the choice to focus on cores to make it an Apples to Oranges comparison they could benefit from. I think it's their server focus that makes them do bad desktop processors, in server world multithreading is everything.
has anyone taken a look at this?
http://www.donanimhaber.com/islemci/...onuclari_1.htm
It seems "official" Amd slides...
some people post ONLY the cinebench slide, because fx seems to perform worse than i7 2600k and similar to i5 2500k, but take a look at the other slides.
Attachment 120446
http://img.donanimhaber.com/images/h...2a_dh_fx57.jpg
Attachment 120447
Fair comparision, but with a brand new architecture the important point is the 2nd or even 3rd generation. The first one is very often buggy and not well optimized. Intels very first S423 Pentium4 was crap, the Socket 478's Northwoods were quite good, but then Preshott was the end, because they could not control the power consumption, and the pipeline got ridiculously long.Quote:
Originally Posted by -Boris-
I do not see these problems here, the only problem is low-IPC, and a still not well running process at GF. Both could be fixed in the next revision. Then you still have superiour clocks. One should not forget, that Llano's CPU part @32nm is neither running well. Only ~3GHz clock, with OC you can reach maybe 3.6, Bulldozer seems to reach easily 4.8Ghz, with 8cores... and it has AES, AVX, SSE4.1/4.2... IMO really the better option to have build Zambezi, especially if the next revisions will gain some IPC.
Looks fake in design, it's nothing professional about it. But the numbers look real, game benches are clearly GPU limited and the rest is hand picked FMA4 with cut Y-axis. Looks like official charts since they try to hide performance disadvantages.
These leaks are useless, most are surely crippled and they mean nothing. However, the pricing does seem official and points to BD performing very close to the level of a 2600K in MT apps. If it overclocks well, its all good. I also believe the release date being very near, most likely two to three weeks at max from now.
I dont know if these are true or fake.. I assume that they re fake, unitl we have a confirmation.Quote:
Originally Posted by -Boris-
Ok game benches are gpu limited but full hd is the resolution that gamers use most.. so I dont care about benches in 1280x1024 only to see the best cpu, useless IMHO.
what about multithread slide?
Would recommend taking a look at this one, as well...
Attachment 120448
Isn't affected at all!? Then why it is in short supply? While they're concentrating on these as being mainstream parts they face the biggest demand here.Quote:
Originally Posted by -Boris-
Then, most Orochi dies that produced are going to server OEM's and Cray, as Interlagoses and perhaps Valencias, at much higher profit margins than desktop CPU's.
No cut in that diagram, base is at 0. And why hand-picked is 15 different kind of tests?Quote:
hand picked FMA4 with cut Y-axis.
Meh, I was hoping for an advantage in multithread over the 2600K at least, this is mediocre performance.
But brand new architectures isn't usually worse or equal to the old, compare K7 to K6, K8 to K7, Core i7 to Core 2, Core 2 to Core, the only example of this is the failed P4 architecture. See the pattern? And your comparison with Llano is unfair. It has low stock clocks because it has a full 400 shaders eating power, and at the same time it isn't meant to be high performance. And you can't compare overclocking either, Llanos bad overclocking abilities is due to lack of dividers and frequency locks. An unlocked Llano without graphics part would overclock much much better.Quote:
Originally Posted by Opteron146
I think you have to consider that FX-8150 is 245 $ and i7 2600k is over 300$.Quote:
Originally Posted by AKM
I dont think it is mediocre, I think it is a balanced cpu, with that price it should be a cpu between i2500k and i7 2600k, and from these slides it seems to be compared like that.
Based on these slides it is a good step in the right direction! Not a leap forward but a solid step up.
With that performance I won't be disappointed. Plenty of knobs to turn and keep me entertained with new BD CPU coming to my system ;].
Can't wait really because 3 days ago I've sold my 1090T and temporary putted X2 555 in it's place. I was shocked how much slower this CPU is in games even @4GHz compared to 1090T. In Crysis2 FPS cut in HALF! Civilization 5 visibly slower, even EVE Online is jerky now! Game engines are definitely going in the right direction with MT support :yepp:
Game benches which are not GPU limited are relevant, because you can see which processor which will be bottlenecking first in future games. That both a P4B 2.4GHz and Athlon X2 at 2.4GHz is able to get the same result in a GPU limited Q3 bench doesn't mean they perform the same in Crysis.Quote:
Originally Posted by liberato87
GF isn't in full 32nm production yet, a large part of their capacity is still 45nm. And Llano is mostly laptop, and I don't have a hard time finding those. Nothing unusual for an AMD node transition. AMD can't have a large 32nm production so short time after the transition. At least the last 4 node transitions has been low volume the first months. Well 45nm went unusually fast. But 65nm, 90nm and 130nm was low volume for several month before they had transitioned to full capacity on the new node.Quote:
Originally Posted by dess
You're right, my mistake. And it's 15 different tests, but not different kinds of test. It all measures FMA4 performance, which is unfair against SB. So still handpicked.Quote:
Originally Posted by dess
But I feel AMD with it's better multi core performance inflate the total score across the board compared to i7 and i5. So even if it perform worse than i5 in say games they're still more expensive since benches show them faster in other areas.Quote:
Originally Posted by liberato87
the world+dog hope that AMD could make something faster than Intel and sell dirt cheap.Quote:
Originally Posted by liberato87
Ok big step - no discussion.Quote:
Originally Posted by -Boris-
Basically only an IMC and SSE2 was added.Quote:
, K8 to K7
Same as above, just SSE4.2 instead of SSE2 ;-)Quote:
Core i7 to Core 2
Hmm that was only a Shrink and SSE4.1. Nothing big. Or do you mean Yonah?Quote:
, Core 2 to Core
It was quite good, after intel pured lots into software companies to use SSE2. However, in the end with Prescott, power went out of control. Imo mostly because a bad, leaky manufacturing process.Quote:
the only example of this is the failed P4 architecture.
Of course I meant with disabled graphics. Well it isnt meant to have high-perf. yes, but you wanted to have a high-perf. K10@32nm. What would be the difference to Llano's CPU part? If you exchange the GPU for four more cores it wont have a much better TDP headroom for high-clocks, either.Quote:
And your comparison with Llano is unfair. It has low stock clocks because it has a full 400 shaders eating power, and at the same time it isn't meant to be high performance.
Hmm, ok let's see and wait for the first BE or K models. If they also lower vcore considerably (current ones are as high as 1.4V which is too much, imo), then it could be better. I also heard about a Rev. B1 on the way.Quote:
And you can't compare overclocking either, Llanos bad overclocking abilities is due to lack of dividers and frequency locks. An unlocked Llano without graphics part would overclock much much better.
I just wonder, the launch date is in October 12 or not. And I just know these slides are in August.Quote:
Originally Posted by JF-AMD
http://tof.canardpc.com/preview/4bd5...da62539221.jpg http://tof.canardpc.com/preview/a53a...1068e8af28.jpg http://tof.canardpc.com/preview/0670...6f23783311.jpg http://tof.canardpc.com/preview/7576...b8563c4893.jpg http://tof.canardpc.com/preview/7c2b...65cd855cc1.jpg http://tof.canardpc.com/preview/d8f4...3bd2477145.jpg http://tof.canardpc.com/preview/4f17...8852809258.jpg http://tof.canardpc.com/preview/149f...2e81618d87.jpg http://tof.canardpc.com/preview/9c78...cd82874148.jpg http://tof.canardpc.com/preview/83ce...c0682be6a2.jpg http://tof.canardpc.com/preview/e3a4...ad7216fa0f.jpg http://tof.canardpc.com/preview/0844...348613fec4.jpg http://tof.canardpc.com/preview/bd57...8dcd7f5186.jpg http://tof.canardpc.com/preview/659f...cc09d313a0.jpg http://tof.canardpc.com/preview/47d3...5115469673.jpg http://tof.canardpc.com/preview/74db...9c856105ce.jpg http://tof.canardpc.com/preview/be00...dfc69a70c4.jpg http://tof.canardpc.com/preview/82f1...2b3a98aa5b.jpg http://tof.canardpc.com/preview/ccd6...c581dd4bd0.jpg http://tof.canardpc.com/preview/2f5d...cdc9194a89.jpg http://tof.canardpc.com/preview/73f4...3c5b7eff36.jpg http://tof.canardpc.com/preview/4860...046774b32c.jpg
Chew has already stated elsewhere that unlocked ES Llano doesn't overclock any better than current retail locked parts. I wouldn't expect miracles from the A8-3870.Quote:
Originally Posted by -Boris-
980X - 999.99 - 800 = 199.99Quote:
Originally Posted by liberato87
Umm the 8 core is only going to be 200 dollars? ;)
Also i want reviews from members on this site. Not Amd Propaganda slides.
That's not the most encouraging slides I've seen. We'll see how it goes. Comparing game cost/perf ratios to a 980x instead of a 2600k isn't encouraging either.
Well, are you suggesting 980X isn't good enough at gaming? It's probably better marketing to compare FX to 980X than 2600K since the gap in price is much bigger... If you compare Civ V which has been poor to say the least for AMD and 980X is best at this according to Anandtech I think it's quite safe to say it will be at least good in gaming, much better than PhII.Quote:
Originally Posted by spicypixel
Oh I know that is has a bigger price to performance ratio difference so its gold for PR, but it's competing in price with different products to the 980x.
While i agree comparing it to a 980X in gaming is just juvenile. Intel fans will say that you can't compare it because of the price but on November 15th when the 3930K gets released at 582 dollars Amd fans will be the ones who will be declaring that it's not in the same price bracket.Quote:
Originally Posted by marten_larsson
Amd needs to man up.
8150 - 240 Dollars
2500K - 219 Dollars
Thats what you compare it against to see what has the best price/performance
I cannot stress enough on how I HATE such kind of marketing/slides.Quote:
Originally Posted by Olivon
Oh, 980x costs 4 times as our cpu, lets compare it in price and in gaming
Oh, FX beats a 4 core Intel CPU(2500) in MT, lets use it for those benches
Oh, there is one game which uses some of those 8 cores, lets use it, and also use a competing product which only has 4 cores. Why not use 8 thread 2600K there?
IF those slides are true and not faked and they will be used to present FX line up during the launch I feel sorry for the person(s) who created those as this is a disgrace.
At least have the balls to come out and say, that yeah, our new CPU will be great in certain scenarios and worse in others against competition, but don't treat us like a complete idiots.
I mean am I or anyone else running 2500K, 2600K and 980X in one system and just flipping the switch and using those CPUs scenarios where they do worst?
We know that 980X market is in hundreds and whoever bought them are either folding or doing some workstationy stuff, NOT gaming.
We know that 2600K in most cases is better for gaming and is also 3-4 times less expensive than 980x, thus people buy this for games or some MT stuff. Also we know that 2500 is cheapest of them all and a lot of people buy those NOT for some MT, but mainly for games.
So IF those slides are correct, they are a pathetic way to introduce your product. Otherwise, please continue :D
980X has 6 cores plus hyperthreading (12 threads). It is faster in multithreading than a 2600KQuote:
Originally Posted by muziqaz
Yes it is faster in MT, do we see AMD comparing it in MT? Nooo, we see them compare it in gaming :)Quote:
Originally Posted by Apokalipse
Double post, laggy server? perhaps you guys need to migrate to FX CPUs instead of current i5 2500 you are running this forum on :D
If those slides are fake..someone have a hell of free time. :rofl:
I think the performance is like in the slides. And it just show to us that it's a 4core/8thread.
Does anyone have gaming benchmarks comparing 980X, Deneb and Thuban at 1920x1080?
Yep, agreed, if the slides are TRUE, which i still doubt it quite strongly. If they are, then BD is a flop overall, a 315 mm^2 chip without graphic doing so poorly vs a 220 mm^2 chip with graphic, fact that both are the most up to date design you can buy in the next few weeks.Quote:
Originally Posted by AKM
Yeah it's cheaper, because it HAS to be, i think AMD would certainly want it priced higher as a 140% bigger chip on the same process supposed to be, should the performance warrant it in the first place.
AMD Delay launch because they are busy making those Powerpoint slide? haha
some guys are talking about GPU limited gaming at 1920x1080. But look at this
http://media.bestofmicro.com/5/Q/274...010%201920.png
and this:
http://images.anandtech.com/graphs/graph4310/37380.png
(this one is 1680x1050, but high settings too)
On those games the nehalem smashes the K10s, the difference is huge even on high resolutions and high details. On Civ5 the Core i7 980x is almost two time faster than the 1100T. And those slides says that the FX8150 is able to match 980X on those games. it's clearly a HUGE step forward, and a doubt a Phenom II X8 @ 32nm could do that.
I don't think you would need much time to do that kinda slides :)Quote:
Originally Posted by PatRaceTin
But if we forget about die sizes and look at them as a consumer, they should be really great an exciting chips for us. They won't be dull to clock for sure.
Let's consider the possibility, that those slides are true and someone from AMD done them.
By the looks of it, they were done in a hurry with all the weird stuff used in them, so we might assume that the benches done for those quick slides were done quickly and in a hurry as well.
I think we can all agree that we as the enthusiasts can get more out of those chips than any of the PR people of given company. You know, most of us have basements in our moms' houses were we tweak and test every bit of the system to get out the best perf for our taste. So I again I would believe that we would get more balanced perf number across the benchmarks/apps after our tweaking.
Actually the results in the slides are not bad at all. Take a look at this one:
http://img.donanimhaber.com/images/h...2a_dh_fx57.jpg
You can see that SMT in case of 2600K vs 2500K brings a lot (8 threads vs 4) in terms of performance and not many of those listed applications scale perfectly with more cores/threads. So SMT kind of already extracted a lot of parallelism from the code and I doubt that bringing more cores in the case of intel (say SB-E) would increase the results by 50% . Well maybe in few cases it would,like wprime and Pov ray.In the rest I really doubt it would.
If you look at the 8150 vs 2500K numbers,the average is around 23.5% in favor of FX,give or take a few. That's pretty good advantage on desktop and it's more than what SMT (2600K) gets over non-SMT (2500K). We do need more applications that scale well with many cores to see this more clearly, but the point is that AMD's "cores" will scale better and will also use SMT in SIMD workloads to extract the ILP from applications that are not coded so well for many threads (just like i7 does).
I have a question though. Why is there no model number for FX listed in the slide above?? :confused: All other slides have 8150 listed but they are not compute intensive test but gaming ones. In this case FX6xxx and even 4xxx is just as good as FX8xxx. Anyone seen the "back up" slides that have system info listed?
I don't know about F1, but I do know about CIV5, that it has something to do with the drivers, and or AMD cpus, as company who created the Civ5(firaxis?) said it can use all the cores you can give them, and game engine does just that with nvidia hardware, but somehow only uses 1 core of AMD cpu when amd gpu is used.Quote:
Originally Posted by Lokinhow
You can see from every AMD Catalyst release that CIV5 is mentioned, and we all know that none of the IT websites come back to retest their systems when new drivers come out, unless something major is announced. I haven't played that game with 11.8 drivers, but even with few previous drivers CIV5 was still somehow limited on AMD hardware. Maybe they have a fix with newest drivers, who knows :)
That's a nice information. But I think it's irrelevant on this case. Because on the test on anadtech all CPUs were tested with a GTX580 GPU. So no drivers problems here. ;)Quote:
Originally Posted by muziqaz
So it might be some limitations on current AMD cpus, who knows, as using only one thuban core on my system is not cool while engine can use most of the core from the competition.:)
These are slides meant for the press/media briefing. The "back up" is presentation slides which AMD don't have time/won't talk show during the phone brief which is usually 40-50 minutes + Q&A - but will be available in the same PDF :)Quote:
Originally Posted by informal
It's generally more detailed in terms of the technical stuff and some extra slides and information. It's also there you can read about all disclaimers, exact system information used for each test etc.
Thanks. I asked because the most important slide ,the one with application performance comparison, is somehow missing the model number for FX :). It just says "FX",nothing else. WHich FX? 8150,8100? 6100 (I'm joking :D) ?Quote:
Originally Posted by Smartidiot89
Until we see the reviews all of this is still not official . But gives us a ballpark though. And by the looks of it it's not THAT awesome or anything special,but it's not that bad either. It's competitive enough with 2600K and it allows great OCing on air and even better on sub zero. For 245$ it is pretty good I have to say.
the 8120 looks like its going to be the more popular option, unless we see a massive difference in OC results.
the stock speed is going to be only 5% slower than the 8150 for lower thread counts, and in multi threaded the difference is still just 15% slower, but the price is a good 20% cheaper
From past experience with these slides and briefings, I do believe that "FX" is infact FX-8150. I'd be more then happy if I was wrong though.Quote:
Originally Posted by informal
I think it looks good for a $245 CPU like you said. But it would've looked even better (in my mind atleast) if it was marketed as a Quad, not Octo-core :D
But stil....when is the release......
I think nobody can sure even amd itself.Quote:
Originally Posted by FaYt
There's a date ;)Quote:
Originally Posted by undone
Quote:
Originally Posted by Smartidiot89
Shame it's the date when I'm flying on holiday for a week! You guys will be playing with new hardware and I will be just reading about it instead of participating!!!! BAD TIMING AMD BAAAD!!
Move the date one more time by a week :)
:sofa:
Attachment 120468Quote:
Originally Posted by Lightman
Lightman, haha. Oh its on tuesday, hm, tuesday not good for me. i am off for holiday, can you move it to, say next wednesday? yeah :D
Don't think so. Not familiar with Ilano, but if it has the usual dividers/multipliers (as of K10), and one knows how to play with them, it shouldn't OC worse than with an unlocked core multiplier. The latter is just making OC easier. IMHO.Quote:
Originally Posted by -Boris-
Not because of lack of demand, that's for sure, but that the transition needs time.Quote:
GF isn't in full 32nm production yet, a large part of their capacity is still 45nm.
There aren't as many as it could be.Quote:
And Llano is mostly laptop, and I don't have a hard time finding those.
Hence the title of the diagram.Quote:
And it's 15 different tests, but not different kinds of test. It all measures FMA4 performance, which is unfair against SB. So still handpicked.
Does it do?Quote:
So even if it perform worse than i5 in say games
And x64? :)Quote:
Originally Posted by Opteron146
What about the rest of the system...?Quote:
Originally Posted by Pestilence
Don't worry, I'll OC some for you too... :p:Quote:
Originally Posted by Lightman
OC on Llano is a bit harder it seems than K10 / AM platformQuote:
Originally Posted by dess
Ah no, this is getting difficult to read, I have been reading since page one and only commented a few times but I really want to make this point.....
I need to start by asking "how many years ago did AMD have a better chip than Intel?"
1999??
There is so much negative talk and speculation pointing towards how rubbish bulldozer will be if it cant beat sandy bridge. OK, it would be awesome to see an amazing SB killer but so many people are missing the point that AMD for so long have been miles away from Intel, it has taken AMD a year or two to catch up in terms of performance (as in AMD's 'current chips' always perform as well as Intels 2 year old chips...... I hope everyone will agree that a 2 year lag in terms of technology (running at the rate it has in recent years) is a MASSIVE gap.
The point I am trying to make is that if AMD equal or even get close to equaling SB then they have done an amazing job, jumping from 2 years behind to 6 months.
Wow, what an achievement :) especially for a company that has something like 10% of the corporate power of Intel.
Please stop getting into arguments about whether it scores 100 more points in a benchmark and appreciate the fact that AMD are back in it, we all love AMD on this thread and it is obvious so lets put our hands together for a company who will deliver at least on par performance with intels current stock at 50 - 25% less price (thats including the extreme pices of intel MBs)
Top job AMD, even if it is late, its a massive achievement to catch up 2 years and respect to you :)
Edit: I was aiming the relevance of this post to past posts, not the most recent that arent so pessimistic!
So, these are real then, OK. From the buyer perspective, it offers decent value, like any other AMD's CPUs before, but from technical standpoint, while the concept is certainly interesting, promising stars and the moon, the end result (or execution ?) seems quite bad IMHO. If the diesize is around 250 mm^2, that might still be acceptable, but not as the current incarnation. Quite disappointed TBH.
How useful are the xop and fma instruction sets for desktop apps, and if they are, what are the chances that developers will bother optimizing only for AMD cpus?
Actually 2004-2005...and to be honest K7 did quite well vs Pentium III...Quote:
Originally Posted by james333
let's hope software devs will be approached by AMD representatives showing the potential of those new instructions. It is up to AMD themselves to kick some ass and push devs to try and optimise for those new instructions ;)Quote:
Originally Posted by AKM
me every time im here now. >.>Quote:
Originally Posted by Smartidiot89
I don't know about you, but I see my self as an enthusiast. As an enthusiast I like "kick-ass-hardware", basically, I don't mind buying the very best, no matter the price - even if we are talking about $999.Quote:
Originally Posted by james333
And as an enthusiast AMD will disappoint me if Bulldozer not will delivery performance that is worth an upgrade over a Phenom II x6. I would rather see a chip at the price level of $500-$999 that out perform intels i7 processors then this comprimise in price/performance.
We want performance, but gets price/performance. IMO that sucks.
*Still I'm quite sure that I will pick up a 8150 at the release day, if it will be available :(*
(Based on what these slides and other leaks have shown so far)
your not the market they were aiming for. their market was mainstream and highend. the chips that go for 1000$ are only like 2% of the income because of how rare they are sold.Quote:
Originally Posted by hirsch
And I respect that, but performance sells. Intel is a very good example of that.Quote:
Originally Posted by Manicdan
thats just marketing, i have yet to see a single commercial about their 1000$ products, everything i see is about "second generation core i series being visually better"
They sell caue that is all anyone knows, they sell cause every windows based software company has recomendations of Intel bla bla, or better in the hard ware requirements sheets (excpet some games now days). That's the only reason my company sells and install Intel, is cause "it said we had to"Quote:
Originally Posted by hirsch
I totally see your point :) I was just thinking about market share and relative company strength. Think underdog fianally being able to keep up with the top shooter :)
^^ Well said... They sell because of their huge marketing budget. They have managed to brain wash and condition the masses of people into thinking that Intel is the only way.Quote:
Originally Posted by charged3800z24
Oh, and lets not forget their anti-competitive / illegal / bullying practices not so very long ago which gave them a huge market advantage.
I didn't mean the performance of an i7 990x - just in general, they have the fastest low-, mid- and high-end CPU and as all of you understand that's why they rule the market and can have the highest price, so in the end, real performance controls the market.Quote:
Originally Posted by Manicdan
(Btw, you should start looking at ESLs commercials, it's all bout Intels extreme stuff, but maybe that's because of the names of their tournaments :))
--
I've a question, am I the only one that think it's embarrassing/funny/weird that they compare there new top model against a 1,5 year old processor that probably performs worse then the 2500k/2600k in the comparing games. Imo that make this (http://tof.canardpc.com/view/a53a68e...1068e8af28.jpg) slide worthless because if they would have compared it with the better performing 2500/2600 the slide would have looked very very different...
EDIT:
And why you to think the situation looks like that? :)Quote:
Originally Posted by charged3800z24
The market (Intel) needs competition in stuff that matters to customers!
ok i dont plan on derailing a thread thats in the AMD section, so my last point is that the markets for low medium and high is set by prices, not the size of the chip. to say intel is better in all of them is no where near true, because AMD is always trying to offer better value (meaning perf/price).
Don't you think it's embarrassing/funny/weird that their $1K top model loses more than it wins in gaming to their $220 model?Quote:
Originally Posted by hirsch
It's marketing. People are extremely stupid and actually believe that price indicates performance. They will see slides like this and think "WOW thats AMAZING". Enthusiasts know the score so why bother with it unless you are just looking to nit-pick irrelevent stuff? Believe it or not, it's a marketing teams job to make their chip look AS BEST AS POSSIBLE. That's what they are doing. Remember the 36% faster in gaming physics benchmark intel did for the new SB-E's? Or did you just deliberately forget that?
http://www.xbitlabs.com/images/cpu/l...df/slide-5.png
Guys, Wprime is integer or floating point intensive?
I think you are right so this will be my last point/post about this too.Quote:
Originally Posted by Manicdan
First of all I've to say that I only looked at the game benchmarks because imo that's real performance and that what matters to most people, not some synthetic tests (even if they matters to me :)).
Lets compare i3 2100 vs. AMD's flagship, 1090T
Attachment 120471
First round goes to AMD (Full HD)
Attachment 120472
Second round goes to Intel
Attachment 120473
Third round goes to Intel
Maybe you are right, AMD TRIES to offer performance/price, but if their flagship can't even win against Intel's low/mid-end segment, that's bad and that's not what I call for competition. I can tell you one other thing, here in sweden the Intel-2100 costs about 1 000SEK vs. AMD 1090T costs about 1 500SEK, so basically it's 50% more expensive and performs worse in real applications, real performance. Sure the 1090T performs better in almost all the synthetic tests, but what do you think the market cares about? Let's say all of the 465 000 000 gamers out there, FPS in games or score in Cinebench 11.5R?
--
Please, don't missunderstand me or take me for a hater. I really like AMD and I have lots of chips from them myself, but what I don't like is when people are trying to convince me that AMD is better then Intel in any perspective of what market want (which includes low-/mid- and high-end as I said). I buy AMD for different reasons then performance, so no I don't have "two faces"... :)
(Charts from here http://www.legitreviews.com/article/1501/1/)
Ehm, why should that be embarrassing/funny/weird? As you probably know, most of the games out there (today) can't use more than 4 cores, in best case. And games dosen't benefit that much from CPU-power if it isn't the bottleneck - right? And what do you mean with this "People are extremely stupid and actually believe that price indicates performance"? In my point of view, price indicates performance VERY well. For example compare the i3-2100 to the i7-2600k - which have the best performance and cost the most... <--- PRICE indicate PERFORMANCE, end of that discussion!Quote:
Originally Posted by jimbo75
Except it doesn't. See my previous example where the Gulftowns and SB-E's massively higher price clearly does not lead to higher gaming performance. See the higher clocked, cheaper Deneb quads that outperform the more expensive Thuban's.Quote:
Originally Posted by hirsch
Price does not equate to performance in gaming, but the general public has zero clue about that. All they will see is a much cheaper chip performing the same as an incredibly expensive chip. What did you say in your last few posts? Only gaming performance matters, and you are an "enthusiast" who would pay $1k for the best cpu? Did you buy Gulftown 18 months ago? Cuz until SB it WAS the fastest gaming cpu and still is in the higher threaded ones.
Gulftown/SB-e maybe isn't worth their price if we are talking gaming performance but still the price indicates their performance because SB-e (6 core) will probably be the best performing chip in the world when it releases (just like 980x/990x are or were) in any applications (games/benchmarks/what ever). "See the higher clocked, cheaper Deneb quads that outperform the more expensive Thuban's." as I said, most of todays games can't benefit from more then 4 cores, so ofcourse Denebs will benefit from higher freq. when both deneb and thuban are based on the same articheture. But in that scenario we are talking about 1 or maybe 5 FPS at most, right?Quote:
Originally Posted by jimbo75
And the only reason that I used gaming performance for my comparison was that the review I founded only included gaming benchmarks that should show real performance. Earlier this day I read that it's more than 465 000 000 gamers in the world today, so just ask yourself if they would've read that review, would they care about FPS in games or score in Cinebench?
Yes I did, and I've chips from both Intel and AMD (ex. 980x/930/2600k/1090T/965/940).
Not sure about the importance of FMA on desktop. Also, some say it will be neglected because of Intel's FMA3, that AFAIK also Piledriver will support (a year before Intel itself!). Altough, I would think if one is going to use FMA in his codes, it's not as hard to support both, indeed in case of OpenCL.Quote:
Originally Posted by AKM
In the other hand, XOP being an additional integer SIMD extension has probably more place on the desktop, think f.ex. videoconverters, games... Althogh, some also bring up AVX2, but that's to come with Haswell only. (Didn't heard if Piledriver were going to support it, as well.)
Hirsch the point is, this is marketing. You are failing to understand what the point of marketing is.
If you take 100 random people who were only interested in gaming, and show them 2 pc's, each with a price sticker on them, which one would they choose? The Gulftown or the SB?
99 of them would take the Gulftown because they would believe it MUST be the faster gaming cpu. 99/100 people have zero concept of multi-threading, or that games only use a few cores. They would simply see the price difference and assume that the more expensive pc MUST be the faster gaming pc.
As an enthusiast you know it's not the case but it's not "embarrassing" that AMD chose to compare to Gulftown, it's just pure marketing sense.
and how many gamers need 300fps? which is where half those benchmarks ended, only 2 of them in one you mentioned were under 60fps. you just pointed out the fact that gamers can be happy with any modern cpu.
price reflects a cpus ability to perform in low and high threaded tasks,so thuban is cheap because in single threaded its not nearly as competitive as it is in multithreaded.
can we pls stop with the "whats better" argument, thanks
You are right and that's what I've been trying to say.Quote:
Originally Posted by jimbo75
And that is exactly what is embarrassing, their marketing team shouldn't need to do that if bulldozer would have been that good I was hopping for! A real comparsation would've been against the 2600K, that was my point and ofcourse I understand why they compared it with the 980x (that's why i talked about the price/performance-slide, I said it would've been worthless if they compared it against the 2600K). This is why I'm started to get so disapointed with bulldozer if all the facts are true....
Better performance today, means better performance tomorrow. Let's say that, setup #1 gives you 120fps in games 2011 and setup #2 gives you 70fps in games 2011. Which setup should perform best in future games?Quote:
Originally Posted by Manicdan
What I say now is that a regular user can never get too "future-proof". It isn't everyone that change hardware every week... :)
Why did someone just put out a chart with a 1090T vs i3 2100, then talk about how bad the 1090T is compared to the 2100?
Why are people arguing? Why are we having these stupid discussions? Launch is 3 weeks away...while I still stand on my hypothesis you all can talk about how bad Bulldozer will be without me.
If you have a 5.2 Ghz BD on water...which runs like a 4 core with HT (by the way, I was told not to utter HT anymore earlier in this thread because apparently comparing AMD's CMT to Intel's SMT better known as HT is not valid)
and you put it up against a Core i7 2600K @ 5 Ghz, which will win? Probably the 2600K by a factor of 5-10% in single thread and ~5% in multithread.
If you go LN2, 7.3 Ghz BD vs 2600K @ 5.8, which will win?
Bulldozer will most likely claim the sub-zero performance crown from intel for a while, at least...until 3930K is released, which will still be locked to 63x or less.
if your looking for perf in future games, you would never recommend just 2 cores
and none of them had a gap that big while being in the same price range
ive already recommended multiple people to get a 2500k as a gaming cpu since its not very expensive and should last 4-5 years (basically i expect it to handle almost every game of the next gen consoles, excluding games with crap coding). once BD comes out i might recommend something different, maybe a 6100 depending on scaling of CMT and overclockability
I've so many arguements for what you are saying right now, but please let us go back to topic because this is meaningless, we have different thoughts and clearly, no one of us is ready to change them.Quote:
Originally Posted by Manicdan
Ehh you are giving 3 slides of Game benches and talk about "real applications" ? And then you call Cinebench synthetic? Way off, games are still badly optimized for more than 4 threads and Cinebench is not only a bench, the Cinema4D Engine is used in real-world movie productions. If you only game then yes, a i3 is a nice choice, however I personally like to overclock, hence I am more interested in a FX4000.Quote:
Originally Posted by hirsch
next thing is that again AMD will deliver latest and most exiting tech, which will really be fun to test out and tweak. Intel will eventually again copy this tech like they did with x64, HyperTransport, internal IMC, Fusion... bring it on a smaller node, better process... And win, eventually.
Its pure economical brute force that lets Intel win at the moment, they are not anyway near as innovative as AMD. Some prefetch/HT frontend optimizations bein the exeption from that rule.
As a tech enthusiast, i see AMD brings more exiting stuff on the table. Really, i thought about if this time a BD would really pay off. Only dealbreaker could be power consumption if it is much worse then intel, but then i'd rather stay with my 1055T overclocked then buying a new system, until AMD comes with a better version of this new tech...
I am exited to say the least and urging for news like with every release before...
curious to see what kind of difference Core Affinity will make on applications. For instance, say a game uses 4 cores max. I would think that assigning cores 1, 3, 5, 7 would work better than Auto or 1-4 due to shared resources in the new architecture
It was told it's probably not the case. The overhead is only some 10%, but you could gain double of that with a 4-core only turbo.Quote:
Originally Posted by Miwo
These latest slides don't make me feel any different about my earlier conclusion. I feel the most likely explanation is that GF missed AMD's clock target by ~25% or so, while keeping it within existing TDP envelopes. Obviously these chips can clock WAY higher (with enough cooling). And this is an architecture that is designed for clocks, yet the stock clocks aren't even any higher than their current arch.
The upshot is that we may get some really high clocking chips for cheap. AMD has to stay within TDP limits to satisfy partners, but we don't. So the real battle will be which chips perform the best at the average max overclocks. If SB tops out around 4.5-5GHz and BD tops out around 5.5-6Ghz on high end water I'll have to see how they both bench in my apps (3d deconvolution and volume rendering, microscopic digital holography, VMs, massive multitasking, etc) because I couldn't likely predict that outcome. The downside is that for the stock speeds to improve and thus AMD's market position and margins they may have to wait for the next process from GF - and that's quite the downer.
This is going to be the last time I post about BD until after release. I've said all I need to say.
BD is not going to do 6 Ghz on water.
5.5 benching would be a stretch.
chew* already posted frequency numbers, what he's done so far is low-mid 5 Ghz on air/water bench and 6+ phase, 6.5-6.9 dice, LN2 ~ 7.5-8 Ghz, LHe gave them 8.4 valid.
I'd expect 5.0 for high end air, high end water maybe 5.2-5.3?
BTW, no X6 or X4 turbos to 4.2 Ghz, and we only have X4's at 3.7... actually some of us cant get 4.2 Ghz within a 200w TDP overclocking...
Can BD bench 3D @ 7Ghz 2C/2T??...Thats the (simple) question.
We all know that 2600K that max out/valid @ 5.8Ghz will (99% sure) bench 2C/2T @ 5.75Ghz all day
Has this already been posted? For the game benches with the 980X they used only a HD6870:
http://www.abload.de/thumb/fakers2q7oa.jpg
For the other benchmarks, they mostly used 6970 CF or a 6950. AMD must really think people are stupid...
I agree completely, if those slides are true, person who done those tests really does not respect their customers.Quote:
Originally Posted by boxleitnerb
If it was me who was shown those slides, I would consider it an insult to me :/
Maybe it was Rick, who made those slides? :D That's why he is out of AMD :D
the specs of the cinebench slide (20) are the only one missing.. why?
WHAT THE xxxx amd why are you lying to us?:wierd::banana3::cord:Quote:
Originally Posted by boxleitnerb