i have a feeling trinity will be great performer
but TDP i hear will be increased to 125w at least for the desktop side of things
i have a feeling trinity will be great performer
but TDP i hear will be increased to 125w at least for the desktop side of things
125W will be another option afaik, so we'll still have 100W and 65W (possible 45W?).Quote:
Originally Posted by tbone8ty
125W is fine to me, I personally won't even use Turbo, as I will clock it manually and shut it off.
The difference in operating cost isn't worth discussing for me on the desktop...the cost to run an OC'd BD may be another thing altogether, as it is a really large jump in power.
I'm hopeful Trinity will be a good chip.....she looked bangin' in those leather pants!
Seriously, I started looking at Inhell offerings, as they can be had rather cheaply, and they perform....come on AMD, please don't make me do it!
If you take a very close look at the FX Next slide that was shown some time ago(at FX launch),you can see them mentioning 8MB of L2 cache(4 modules x 2MB per module) . No mention of L3 cache at all :). So they seem to canned the L3 in Vishera and still get the good core improvement without it. That's pretty good IMO. Server variant will definitely have L3 and feature 10C/20C models,so it can only be faster than Vishera is. Vishera on the other hand,if no L3 onboard tidbit is true, may end up considerably smaller than Orochi. In Orochi,those 8MBs of L3 not only consume a lot of power,they eat up a lot of die area too. So if they canned it they may gain some on the frequency side(due to TDP headroom they would get) and they can reduce the cost of the chip (smaller die).
without l3 we get lots of modules even if it doesnt make sense :D
I wanna more than 3MB L2 if without L3.:DQuote:
Originally Posted by Tomasis
This shouldn't be needed.Quote:
Originally Posted by undone
During the brief AMD mentioned that the L3 cache was overkill for desktop and very rarely showed any real benefit. The L3 was/is there for the server SKU's and should be seen as a bonus for desktop. So don't be disappointed when Trinity arrives without L3 ;)
I think it will be a good riddance on desktop :). Give us some decent latency L2 and give us some TDP headroom from cutting L3 off :D. Trinity was said to have a bit higher IPC and more clock ,good job in my opinion.
will their be a new stepping in this generation of bulldozers? i.e. FX-8170?
maybe yes....In documentation is info about B3 rev., but nothing more.
Charlie posted a die shot for Trinity
http://semiaccurate.com/2012/01/05/e...about-trinity/
looks like the gpu portion got even bigger, and cpu even smaller. even though amd is behind intel with their mobile cpu line, their gpus just crush them, and this will do an even better job of it.
whats the chances that its GCN?
Manicdan about this much <1%:D
edit: or rather 0%, it's pretty similar to Llano's IGP at least that's what I think.
booooo!
One more interesting part is the die size which Charlie lists as ~240mm^2. Roughly the half of this is GPU portion. If FX Next is based on PD with no L3 (like leaked slide suggests),then we can expect similar die size for 4CU/8T part( 240-250mm^2). If clock and IPC actually do go up by 15-20% (combined #) and die area goes down by ~30% this would considerably up the performance/mm^2 ratio number versus today's FX (1.15*1.3=~1.5x or 50% better perf./mm^2).
Manicdan That's not necessarily bad. GCN is a really nice architecture even more so for general computing but I think GCN IGP would end up bigger than this one. Charlie said the size is ~240mm2 so just a bit bigger than Llano(~228mm2), yet IGP at least 30% better + you get VCE, eyefinity maybe UVD3 is also better who knows. Even the cpu should be faster but it depends on clocks.
Someone mentioned the shader count is only 384. It's a bit low to me, maybe the clock is higher than expected.
http://semiaccurate.com/forums/showp...57&postcount=2
undone They said ~30% above or a bit more so 384 or 6Simd VliW4 is +25%, then the rest will be clocks, if they can clock it at 660-720Mhz and still stay in a comparable TDP that's another 10-20%.
well theres a few things to think about that i realized after my post.Quote:
Originally Posted by TESKATLIPOKA
if its based on GCN, then how would they do xfire if the lower end 7000 gpus are based on the previous arch. which makes sense why they lower end 7000s are the previous arch, so its all compatible. and what informal said, the size of the 2 PD cores is really small, so hopefully this means we shall soon see 10 core non L3 chips with great turbo clocks and maybe some extra single threaded perf.
Vishera will be PD core with L3 :)Quote:
Originally Posted by informal
Yeah it could be but I meant this slide:Quote:
Originally Posted by FlanK3r
http://img651.imageshack.us/img651/6...tachmentdz.jpg
Only L2 is listed in the slide. Maybe they didn't mention L3 on purpose ,who knows. Anyway,15-20% over 8150 is good enough to put AMD back into the mid-high range.
http://img707.imageshack.us/img707/9711/63332322.jpg
edit: note that this slides says 10% better x86 performance. Other slides state Piledriver will have 10-15% better core performance than Bulldozer. So It's not really clear if the FX Next slide is referring just to IPC or clock+IPC (which should be very low improvemet). Another possible explanation can be that AMD knows they will launch 8170 at higher clock(around 3.8-3.9Ghz base clock) so they figured this in the PD performance estimate: PD @ 4Ghz should be ~10% faster than 8170 @ 3.8Ghz.
That slide really says 10% better x86 DIGITAL MEDIA WORKLOAD. Not 10% better x86 GENERAL WORKLOAD. I do not trust or like seeing caveats like that.
AMD disclosed the IPC improvement to THG awhile back. They stated a third of 15% improvement PD will bring over BD will be IPC(so 5%), while two thirds will be power improvement(read more clock within same TDP). So the 10% in the slide must be clock+IPC over Bulldozer "8C" at undisclosed clock (maybe 8150 or 8170,where the later one is more logical choice as PD will come right after this part).
We still may see an additional boost of 3-7% in less multithreaded apps( thread count lower than 8) , as per Microsoft. The patch will come with 8170 or a bit sooner.
It looks a little smaller, with marginally lesser performance (Hopefully PD is much improved!)
If those two pictures ARE really to scale, look at how far down the die the Llano CPU part goes vs PD.
Oh, he used the I/O pads on the left (which are not the same) for reference...
Do you really think performance in gaming scales linearly to expected performance from an early marketing slide, even if it was real?Quote:
Originally Posted by informal
There is a 9% clock speed difference with Turbo on and 6.5% without turbo clock between i5 2400 and 2500. The real recorded difference is 4%...
Think 20-25% for that gain you are looking for.
Of top of this, BD loses to i5 2300 by 3%, which is clocked at 2.8/3.1 vs FX-8150's 3.6/3.9/4.2.
The FX-8150's base clock is 28.5% higher than the i5 2300, and 26% higher with mid turbo vs i5 turbo. Make that 35% for max BD turbo vs i5 turbo. The i5 also will have an intermediary Turbo state(s) however that is not published by intel and the 3.1 only applies to 1 thread used.
That is, of course with about twice the real world power consumption as well compared to the lower end intel quad i5 2300.
Gaming is only one part of the workloads. I think for most games ,users will NOT play at low-res . And @ 16x10 the difference,even with 7970 will be not that big between i7/i5/BD. Somewhere you may have a bit lower fps but while all the CPUs are OCed to their maximum I think the gaming experience at resolutions 7970/78xx users will play most games, difference would be negligible or not that noticeable.
In real workloads that are not gaming, difference will indeed go by the clock and IPC uplift. So 10-15% as AMD stated will be true in some of the workloads and less true in others , depending by what part of the CPU the application is bound.
Look at the results of Multi GPU tests....
Results aren't good, THOSE are the folks looking for maximum performance, and are willing to spend the cash on top tier cpus.
Just sayin'
Good point.Quote:
Originally Posted by muzz
FX-8150 being $270, it's not exactly a cheap CPU and SB obliterates any hope for it and PD when overclocked in that area.
Here you go,end user results with pair of OCed 6990. If BD can't bottleneck a pair of these monsters it won't be a bad CPU.Quote:
Originally Posted by muzz
http://forums.anandtech.com/showthread.php?t=2217355
Lol, you're so funny informal ... :DQuote:
Originally Posted by informal
We all know that BD is an awful gaming CPU, they're plenty reviewz on the net from trusties respected sites.
But you, not ...
You prefer using a chart with 1 game on it, done by an 8's post guy on AT forum who is doing promo for his own amdfx.blogspot.com ...
You really think we're stupid ? :rofl:
Got a link to this? Doesn't sound anything near from what I've heard.Quote:
Originally Posted by informal
Thank You :clap:
Quote:
Originally Posted by Olivon
i wonder about AMD A10 APU with Piledriver Cores. A friend tell me it park with 2 Modules and 384 Stream Processor VLIW4. Good to go?
Smartidiot89The info is from Anand and AMD slides if I am not mistaken they can be found in this thread at the beginning.Quote:
Got a link to this? Doesn't sound anything near from what I've heard.
What did you hear? If I remember correctly you said you heard something about 15% will be just from IPC or am I wrong?
About BD and gaming. Techpowerup did a comparision between SB,BD and Nehalem on Radeon 7970. Here is the link:
http://www.techpowerup.com/reviews/A...Scaling/1.html
It's on this page from THG article:Quote:
Originally Posted by Smartidiot89
http://www.tomshardware.com/reviews/...fx,3043-9.html
@informal
In most situations FX loses to Phenom II X6 and even Phenom II X4 980 in gaming as well as overclocked. No need to make BD out to be a gaming monster, we all know it's a failure.
It's not a monster at all. It's just good enough. There is no question that i7/i5 is better. As for Phenoms,they are just not THAT better than FX,that's all.
It's not like in Cyrix time when it was totally useless for gaming while K6 and Celerons were great in this tasks(while Cyrix was again faster in pure integer workloads,it only had a poor FPU).
different with the test informal posted is actually the ddr-2200 cl7 ram. That makes a difference over these crappy cl8 1333 that are used elsewhere for BD benching :shakes:
dunno if it changes anything tho, but i guess there are games that are not cpu nor gpu limited but ram. I have a good experience with the BD over my x6 in my unoptimized, oldfashioned but alltime favorite game, that improved a lot.
realistically the difference in gaming is 2-3 FPS or none in some gamesQuote:
Originally Posted by Oese
thats about what is seen as difference on some gaming tests anyway at resolution > 1024x860 except some singlethreaded games.
No sugarcoat, but thats what it is, good enough...
Anyway, this is PD thread, looking forward to it :)
it'd be nice!Quote:
Originally Posted by Manicdan
If next version of trinity (not PD) bases on GCN, it would be good htpc that does anything you throw at. Instead of cpu they uses raw Gpu power for heavier loads when we get new softwares for that (it takes long time i guess). I saw Boxx machine calculating Vray 3d in almost real time using Nvidia Cuda on Quadro/Tesla cards. Alone for that workload, cpu is not close to that tflop calculation power (not even a hundred high end as Intel 39xx can compete to triple/quad sli cards).
I believe more on GPU in future. It is taking more care of heavy load, at the same time cpu is getting less imporant. What we need is high clocked ,leaner, smaller cpu that does required work between ram,nb and card.
the future looks interesting :up:
Well yeah,but that smaller and fater cpu is SB no way bulldozer.SB is fast as hell,super optimized IMC.Not that thing called bulldozer,which costs the same,consumes more power,and is damn ineffective..
Actually a lot of games separate BD and SB by 10FPS+, and BD is often 1-5 FPS behind Phenom II stock/overclocked, despite Phenom II's IMC being 60% slower.Quote:
Originally Posted by Oese
mhm k
http://tpucdn.com/reviews/AMD/HD_797...rfrel_1920.gifQuote:
Originally Posted by Oese
:shrug:
In the games where FPS difference is 2-4 FPS, memory isn't going to do anything. (Besides, memory was equal on all systems except Nehalem which was in triple channel but 1066 instead of 1333)
Ye i saw it. looks horrible :)
That is again from the same brief I heard, so I heard the same information as TomsHardware but it looks like you can interpret it both ways... ;)Quote:
Originally Posted by informal
I don't get you guys. This thread is about Piledriver, and you're ragging on Bulldozer. Bulldozer may not be the greatest thing since sliced bread, but it's far from the worst also. I've had a TON of fun tweaking it and just playing around. And it actually runs great.
I built a "Budget" BD system and it has been rock solid for me. I love it. Yes I've read all the "reviews" out there and what not. Frankly I don't much care. I go for how it feels. And BD is feeling mighty fun atm.
And I cant wait to see what Piledriver does, should be better than the Phenom - Phenom II advances.
Friggen relax guys, have a shot of vodka and chill and enjoy tweaking and benching and having fun.
Yeah I'm bout halfway through a bottle right now and feeling mighty fine. Till I started reading all the emo crap in this thread.
Lol, took like 30 mins to type this post.
Peace
you know, the might of excel bars.... enjoying bd much also but you are right too in the aspect that this is offtopic... just, waiting for pd :)
Well i sure love vodka!
What is the current projected release date of PD? I've seen so many dates I'm :confused2
You might be a little too drunk...nobody is raging or "emo" :rolleyes: or even worked up :ROTF:Quote:
Originally Posted by Plywood99
:DQuote:
Originally Posted by Oese
Talking about BD is at least 50% relevant, PD is like "BD+" :yepp:
Vishera is to Zambezi as Deneb is to Agena (minus process shrink and a whole lot of cache :p:)
I think, it will be in Q3Quote:
Originally Posted by Sparky
FAB8 is the new one (third AMD fab?)
GloFo has 8 fabs now, 3-300mm and 5-200mm fabs that used to be Charter, and there was talk of a 9th fab.Quote:
Originally Posted by FlanK3r
I wouldn't call them AMD fabs, I think AMD only owns around 10-15% of GloFo now.
GLOBALFOUNDRIES Fabs , An aerial view of Fab 8
A future built on silicon
Trinity with PD cores
http://hothardware.com/News/AMD-Fusi...olt-In-Action/
25% increase in TDP range on desktop products, 15% CPU gain + 30% GPU gain?
omg where did you read the 25% increase in tdp......Quote:
Originally Posted by BeepBeep2
lower tdp. they mean increase in power efficiency.Quote:
Originally Posted by undone
FMA is supported in BDver1 CPUs(FMA4),so CPU world made a mistake. FMA3 is a new ISA extension supported by Piledriver which is now compatible with intel's haswell FMA3.
edit: I see CPU world corrected the FMA mistake,nice.
I think this statement is from here:Quote:
— Mobile (35W): +25% for the CPU and +50% for the GPU vs a mobile Llano (35-45W);
http://www.computerbase.de/bildstrecke/39091/5/
25% increase in PCmark Vantage(Productivity), 50% increase with GPU in 3dMark Vantage(Performance)
Hardware.fr got this directly from AMD, they got a reporter at CES.
For Desktop vs Desktop parts(PD vs Llano), AMD states 15% faster CPU side and 30% faster GPU side.
Before it was reported that trinity will have 65w, 100w, and 125w parts.Quote:
Originally Posted by undone
One would have to imagine the 15% CPU gain and 30% GPU gain are seen only with the extra 25w headroom given. I'm sure there will be a significant increase with GPU at 100w too, but we will see.
EDIT:
I see that from CES, news states 100w TDP on desktop...if true then I welcome x86 improvements PD brings over STARS with open arms ;D
"The GPU portion will be based on AMD's Southern Islands architecture..."
http://arstechnica.com/gadgets/news/...ultrathins.ars
could be?
No. Fab 8 was a joint venture between Global Foundries and IBM along with other JDA (joint development alliance) companies like Toshiba, Samsung, and Renesas.Quote:
Originally Posted by FlanK3r
And contrary to what the article says, Fab 8 is nowhere near producing chips that will see the consumer market. It's going to be another 6 months before everything becomes fully operational, and a full year before they begin actual revenue production. Only some time after that will they reach high volume production in 2013.
To be honest, I have absolutely no idea where Charlie gets these so called "facts" that he put in his articles. I have a feeling he was wearing his pink bunny suit while typing it up and forgot that Journalists must maintain some sort of professionalism and accuracy.
And yes... this is an actual photo of Charlie in the previously mentioned suit. It was punishment for publicly challenging executives @ Intel while having his facts wrong. :) He lost the bet.
http://www.theinquirer.net/img/8492/chax.jpg
Here are my facts btw...
http://www.globalfoundries.com/manuf..._overview.aspx
What do you think about these?:
http://apps.corel.com/lp/amd/index.html
http://www.imaging-resource.com/NEWS/1294863829.html
Corel support AMD and his APU/FX processor.
Or, Red Hat in this document: http://www.redhat.com/f/pdf/RHEL_6_2...s_benefits.pdf
Red Hat says that RHEL 6.2 supports Bulldozer and Piledriver. Is it true?
"Gcc, glibc, gdb, oprofile — include optimizations to support new chip set capabilities — Intel Westmere & Ivy Bridge, AMD Bulldozer & Piledriver"
What I'm wondering about is the Pcmark benchmark, I have seeking many results about this bench but I can't reach a reasonable conclusion that what this bench is testing, single core or multithread? What I can sure is the result has no relationship with GPU.Quote:
Originally Posted by undone
EDIT: could these two pics describe something? I thought the A8 comparison in second pic is the source of '15% increase' statement.
http://images.hardwarecanucks.com/im...ge/chart15.jpg
http://www.techpowerup.com/img/11-11-21/110c.jpg
looks like pcmark doesn't utilize more than 4 threads, and very bias to single core performance.
Nope, VLIW4.Quote:
Originally Posted by cesariuth
AMD Describes Piledriver Architecture Peculiarities to Software Developers
http://www.xbitlabs.com/news/cpu/dis...evelopers.html
http://www.xbitlabs.com/images/news/...y_features.jpg
Vishera(FX) is indeed a ten core cpu.Quote:
Interestingly, but the maximum amount of cores inside Vishera (next-gen FX-class desktop-chip), Sepang (1-die Opteron chip for 1-2 socket servers) and Terramar (2-die Opteron MCM for 2-4 socket servers) microprocessors is considered to be ten, not eight in case of Vishera as noted earlier. Moreover, it looks like the silicon behind all three of the products seems to be very same.
Based on new, old and very old data compiled by Xbit-Labs.Quote:
Originally Posted by undone
Think, we will be smarter after AMD Analyst day next month :)
I think Vishera will stay 8 core...I thought the newer slides all said 8 core, only the very first around bulldozer release said anything about 10.Quote:
Originally Posted by Smartidiot89
10 will be maybe only for servers (so Teramar/Sepang)
Correct, I should've made my sarcasm more obvious :rofl:Quote:
Originally Posted by BeepBeep2
Komodo was a 10-core part including the Northbridge (PCI-E hub if you'd like) integrated, with Tri-channel memory. Sepang would be the server variant, while Terramer would be a 20-core one. Now without Komodo, I am not too sure Sepang and Terramer are still on the list to come to servers so 16-core might still be the highest AMD has to offer. According to the new plans, or shall we say leaks... Vishera will be an AM3+ compatible 8-core Piledriver CPU. Of course I am basing this all on leaks, all I know for sure is that Vishera is very real... ;)
Like Flank3r said, all of this will probably be clear after Analyst Day which will be held the 2nd of february :D Until then it's hardly worth speculating in anything AMD-related as all their official roadmaps released last Analyst Day are useless today due to all changes.
I'll take this quad-channel rumor as a grain of salt. If amd wanna do this, what I worried about is the platform cost. Unless the second generation FX is an intel-killer.Quote:
Originally Posted by erek
still cheaper than tri channel tho?Quote:
Originally Posted by undone
can buy 4's instead of 6's ?
I dont believe this. G34 will be server socket, Vishera in desktop at AM3+ (990FX or maybe 1090FX).
1090FX will come with USB3.0 integrated,several tweaks maybe,but no pci-ex 3.0...
piledriver cores using the same cores as trinity?
difference is having L3 or notQuote:
Originally Posted by demonkevy666
There are some results in 3dMark11 of a supposed Trinity eng sample(1D32246W4K44_38/32/22_2/8):
http://3dmark.com/3dm11/2417633
http://3dmark.com/3dm11/2584681
but this GPU is what???? This can not be 7950 radeon...
Finally! i was thinking about the G34 option too!!!
it has what AMD needs against Intel 2011 socket
Quad channel Memory with 8 Dims!
and you can get up too 10~16 cores like in servers!
and G34 platform not so expensive! you can get single socket MB for less then 300$
My Photoshop Images!
http://img689.imageshack.us/img689/4554/amdfxx2.jpg
http://img835.imageshack.us/img835/7...2bulldozer.jpg
Lol
not what I asked but that would really only be about 5% less cpu performance. with out L3 cache.Quote:
Originally Posted by undone
without L3 cache, the core logic is the same between Trinity and second generation FX(Komodo/Vishera).Quote:
Originally Posted by demonkevy666
I find another one.Quote:
Originally Posted by sergiojr
http://wccftech.com/amd-trinity-a8-e...-11-benchmark/
http://wccftech.com/wp-content/uploa...1/trinity2.png
the graphic score is kidding me right?
Update: someones find out the CPUID is for 8 physical cores, which represent an ES FX with 4 cores disabled.
http://www.planet3dnow.de/vbulletin/...&postcount=751
http://www.planet3dnow.de/vbulletin/...&postcount=754
(pls translate by yourself)
seem to be fake.
result is from Mister O..B.......:rolleyes:
thx for link :)
You don't need to look at the Physics test It's misleading, you won't know how It really performs.
Here are two link with the same CPU(APU) and look at the physics score
http://3dmark.com/3dm11/2584681
http://3dmark.com/3dm11/2417633
The difference is 11.5% for the same cpu(apu) :shrug:
I found an interesting comparison for this test on the same cpu but different gpus
http://www.lostcircuits.com/mambo//i...1&limitstart=2
P.S. It's not even sure this is a Trinity
Physics test is just not good metric for comparison since it varies wildly with different graphics cards(as the LS link above demonstrates). Not to mention that in 2 instances of Trinity 3dmark11 runs we have 2 scores that are more than 10% apart... This tells us that either the user OCed or down-clocked the chip or something else was responsible for the discrepancy.
PS Never mind that dumb OBR didn't see the freaking title of the 3dmark11 which has Radeon 7950 listed as a GPU. He thinks this is the iGPU in Trinity... Yeah right,3x faster than Llano while AMD officially said 30% better on desktop. Not even a magic pixie could make Trinity's iGPU 3x faster. But hey,he is "good old" OBR :D
Attachment 123808Quote:
Originally Posted by informal
EDIT: image y u show up as attachment, y u no?!
I doubt any IPC increase and only a decrease in power.
Piledriver will have a moderate IPC increase (e.g. ~5-8% is predicted) mostly coupled to module optimization...then again, I only care about server side stuff so....ymmv on the desktop. ;)Quote:
Originally Posted by demonkevy666
d
what applications? game? compare to what? llano or bdver1 or deneb? and btw i'm fed up with hearing any comment or complain about IPC, let's say 'performance per clock' rather than IPC, ok? cuz these two are different.Quote:
Originally Posted by demonkevy666
Which means higher clock is possible. Certainly higher clock with lower power is much impressive than just lower power with same clock. need to wait and see.Quote:
and only a decrease in power.
-----unrelated message----
anyone see FX-8170(or second wave of bdver1) news or rumor?
There's an ETA for piledriver yet? 6 months from now? 1 year?
Thanks.
I have no idea how fast the Piledriver core might be over Bulldozer, but I expect a decent rise in clock speed for Trinity over Llano- the A10 should end up at 3.6G atleast.
Current info talks of Mid-2012 launch - whether that means April-May or September I have no clue.Quote:
Originally Posted by Piotrsama
I am thinking about getting a unlocked Trinity and OCing the hell out of it, if its not as power hungry as bulldozer.
so Llano is awesome, and trinity will wipe the floor with it <3
10% higher In single thread would be my guess that's over "Phenom II/LLANO" not bulldozer.Quote:
Originally Posted by undone
higher clock bulldozer would need cool down with lower power too.
not always possible with lower VIDs on chips.