Printable View
Six-core Phenom II to feature Turbo Core tech, high frequencies
by Scott Wasson — 11:00 PM on April 7, 2010
Looks like the rumor mill was right to be buzzing about AMD incorporating a dynamic clock speed boosting mechanism similar to Intel's Turbo Boost into its upcoming six-core processors. AMD has confirmed to us that the chip code-named Thuban does indeed include such a feature, dubbed Turbo Core technology.
http://www.techreport.com/r.x/2010_4_7_/slide.jpg
Hope I don't need to use DThermomtr to bench it:up:
noice! can't wait for these beasts to come out :)
Release the Kraken!Quote:
Originally Posted by ridney
would be nice to have one to test these new boards from AMD i have here hehehe
I am getting some good feeback from the channel and media about 6core CPUs and 890 boards. It looks as though it might be a popular bundle in our market. Let's see how it works out with pricing and performance. :up:
Simon, how it is with turbo and AM2+? Its not only for AM3 board active? Thx. O.C. AM2+ and Thuban is nothing good for me ;), waiting for 890FX Giga:), but most users maybe upgrade with AM2+ board at Thuban 1035/1055
Fine with proper bios support which we should have in all of the enthusiast boards.Quote:
Originally Posted by FlanK3r
is that means that the 3 cores itself are capable to jump up 150+Mhz each (totalling 500Mhz)?Quote:
The processor becomes eligible to enter the "boost" state when three or fewer cores are heavily occupied with work. This boost state is a simple, binary condition. When the CPU enters this state, the peak clock speed for the three affected cores will jump by as much as 500MHz above the baseline, improving performance
without the even raising the multiplier?
so if someone somehow can unlock the 6 cores to jump up to 1000Mhz? without increasing the multiplier
i'm pretty sure the cores fabricated are identical ...
or have i gotten it wrong?
Turbo CORE provides a higher top frequency each of the three cores of up to 500MHz (depending on which CPU model only). Turbo CORE works automatically, maintaining TDP while dynamically selecting which cores are in Turbo, or if all six cores are online at nominal frequencies.Quote:
Originally Posted by Serpentarius
CNQ works as always, so each core has independent frequency. Turbo CORE simply raises the top frequency limit for three cores, again by up to 500MHz.
Which of the three cores are in Turbo is dynamic and shifts rapidly. Moving in and out of Turbo happens extremely quickly.
and launch day its 23.4.(friday) or 26.4.(monday)? I heard this two termines:shrug:
Very goooood article answered all question i had about TC....
These are the important quotes:
"Thuban does not include the ability to gate off power entirely to an idle core like Intel's Nehalem-derived CPUs"
and
"500MHz from going into a boost state, which means effective clock speeds with lightly multithreaded applications could approach or exceed 4GHz."
Why the the other one imp well because when a PhII is OCed around 4Ghz the performance in most games equal that of a i7 @ 4Ghz so in effect a 3.2Ghz 1095T which has a 500mhz boost will kick some serious arse in games performance in which not more than 3 cores are used.
This is a very calculated move by AMD because its not that intensive as Turbo nor as unpredictable and it basically gives you 3 cores @ stock+500mhz or gives you 6 cores @ stock. AMD most likely will use a similar approach with the BD also but add in other things like power gate's, etc.
http://images.anandtech.com/reviews/...whatiscore.jpg
http://images.anandtech.com/reviews/...coreatwork.jpg
"Last month Intel introduced its first desktop 6-core CPU, the 32nm Gulftown Core i7 980X. Running at 3.33GHz we loved the fact that it’s quite possibly the first Extreme Edition part that is able to justify its price. For $999 you get six cores and better performance all in the same power envelope as the current high end quad-core i7s."
http://anandtech.com/show/3641/amd-d...o-core-enabled
Ooooooo
Because you run into the gpu limit... :p:Quote:
Originally Posted by ajaidev
Anyway, i wana see banchmarks. :cool:
Would be cool to be able to configure these states by ourselves.
Pretty smart implementation either way! :up:
http://img176.imageshack.us/img176/5...cc6a6624a8.jpg
This is how i see it the simplest way to represent AMD's turbo core :D
I like Intel's turbo also but i thing the all core at higher freq as a unfair and unrepresentable determinant. Imagine you go and buy a i7 980x and it says it runs at 3.33GHz on the box the only thing is it never really does.
Hell everyone can OC a i7 980x and get the same effect of all cores at a higher freq plane.
AMD Phenom II X6 processor surfaces at UK retail
http://www.hexus.net/content/item.php?item=24143
it's annoying when these sites put up the info knowing they have not bloody stock :(
What is unfair exactly? It will run at higher freq if a power/temperature conditions will allove it. Otherwise 3.33GHz always guaranteed (throttling excluded).Quote:
Originally Posted by ajaidev
Everyone can overclock every x86 cpu. What is the point?Quote:
Hell everyone can OC a i7 980x and get the same effect of all cores at a higher freq plane.
These 6 core AMD CPU's sounds very exciting. I really want one, though I'll wait for reviews and benchmarks before deciding.
Not quite sure what you are saying here, but 980x is advertised as 3.33Ghz when it runs 3.46Ghz stock with Turbo enabled.Quote:
Originally Posted by ajaidev
what about Cold bug 64NOMIS? :D
If someone is interested, I wrote about these news in my latest blog entry to combine all important information available.
It seems, some article authors think, that the idle and loaded cores will have different voltages, but according to AT this is not the case and it still works.
Is the relation between the least significant digit "0" or "5" in the model number and the frequency boost already known here?
Last time i checked stock frequenzy of the 980x was 3.33ghz and with turbo 3.46 on all cores and 3,60 for singel (dual?) core load. And intel didn't sold it as 3,46ghz.Quote:
Originally Posted by ajaidev
Most interesting info in AT article(also confirmed by Simon in this thread ):
Quote:
The switch to turbo up cores apparently happens fast enough to deal with Windows moving threads around from core to core.
Its marked as 3.33Ghz stock speed, yes but with turbo this speed is never kept is it its always around 3.46Ghz. In essence stock speed is never really used for anything is it?Quote:
Originally Posted by kl0012
The point is that when turbo overclock all cores its not fair, turbo should overclock when the core count is less than total cores like the 3.6ghz for 2 cores is acceptable.
Yes thats exactly what i am trying to say it never really suns at 3.33Ghz and is advertised as a 3.33Ghz cpu.Quote:
Originally Posted by tool_462
Among all the ES's i had time to spent i9 aka i7 980x was one of the best but at the same time the i had a horrible time with Lynnfield cpu's and it was all because of Intel's turbo implementation. I dont want to go in detail but the fact is Turbo soon will not hold as much importance that it does today.
See the above responses and to add to it how do you predict Turbo i mean how do you know when it should start/stop/etc. Intel has taken a very grainy road on this and i dont like it one bit.Quote:
Originally Posted by Hornet331
The Turbo speed can flex "specially in all cores mode" when ever the cpu wants it to based on the environment. This gives one benefit that is when the environment has not warmed enough or is still in the process the speed of the cpu will work at 3.46Ghz but once the environment is hot enough the speed drops to stock.
Also the power consumed and TDP change with stock freq. on all cores and the turbo freq. on all cores take out a wall meter and check for yourself.
Some good news for overclockers in here:
The big achievement is how they manage to get six cores running at
3.2 GHz at the same TDP as a quad core 3.2 GHz Phenom II X4 955.
The 140W 965 is only just now going EOL being entirely replaced by
a 125W version.
It's almost comparable with a full process node improvement like Intel
did by going from a 130W quad core 3.33 GHz 45nm Nehalem to the
130W six core 3.33 GHz 32nm Westmere.
http://www.tomshardware.com/reviews/...i-x6,2604.html
http://www.techreport.com/discussions.x/18731
http://anandtech.com/show/3641/amd-d...o-core-enabled
http://hothardware.com/News/AMD-Turb...logy-Revealed/
The first two sides mention low-k and THG specifically mentions it as
going between the metal lines so most likely the speed improvements
are indeed due to to lower RC-wire delays.
This looks to be very good news for overclockers since the RC delays
are basically temperature independent. You can't improve them by cooling,
so LN and LHe results are for a large part determined by RC delays.
Regards, Hans.
how would turbo be affected if i were to up the bus speed from say 200mhz to 220mhz? for my current PII with CnQ when it drops cores to 4x the speed is 880mhz instead of 800. so can i safely assume that if i buy a 1035T, i can OC it just a hair, and have it perform about the same as 1095T? (the only thing i would worry about is what voltage it would be at, being a 95W part, i assume its about .5v less at max speed compared to the 125W parts)
I think that the more important question is, will it enable to use higher multipliers on NON-BE chips when manually overclocking.
so basically your saying its bad to you get more then you pay for...Quote:
Originally Posted by ajaidev
I don't see a "grainy road" at all, you pay for a 3.0ghz cpu and you get that. If the conditions are right (temp/consumption) you get more performance. How can this be bad thing... for most reviews it was disabled anyway or tested with and without turbo so even compareabillity is given...
Imho ever since intel introduced turbo certain people where in knee-jerk mode -> turbo = bad...
At least there "arguments" have run out of full now... :p:
Sounds good but what is "RC-wire"?:DQuote:
Originally Posted by Hans de Vries
960t will oc nicely !!!!
Cant wait to get one of these and see how they do.
it's on legit reviews, it was using a 5970 at 2560 x 1600.Quote:
Originally Posted by Hornet331
6% difference up at that resolution.
is there such thing as a balanced system anymore ?....
http://paddyk.files.wordpress.com/20...o_no_spoon.jpgQuote:
Originally Posted by FatAlbert
ok :D
thanks :D
Ohnoes! Ohnoes! I ordered a 3.33GHZ cpu and de evil Intel gave me a cpu that runs at 3.46GHZ! Ohnoes! :rofl:Quote:
Originally Posted by ajaidev
Seriously, dude!
If turbo was such a bad idea, why did AMD implement it? Surely there must be a conspiracy somewhere! Wait...... ah I see..... AMD's Turbo Core does not boost power on all 6 cores. That could be the only reason.
Fact is turbo boost is a genius feature and AMD would have been stupid not to implement it seeing that they're staking their future on many core (just as Intel) in a world dominated by lazy programmers fixated on single-thread.
I like this answer!Quote:
Originally Posted by 64NOMIS
Was laughing hard! :rofl:
Quote:
Originally Posted by Hornet331
Oh No i did not say i dont like turbo i just said that the all core boost is not fair that's it. It alters the speed of all cores if a correct environment is presented and that is grainy in it self.Quote:
Originally Posted by OhNoes!
The box should have been marked 3.46Ghz instead then should it not, different users of same cpu with same multi thread app will get different performance depending on the turbo mode.
As for AMD not supporting all core turbo, look at my posts when i had a intel Bloomfield ES and i reported odd bench results in cinebench, i hate all core turbo since then. I supported less than 4 core turbo from that time because single threaded app's worked better as a result.
Quote:
Originally Posted by ajaidev
Oh Nooooes....it's the "It's not fair!" complaint. Or the "Intel cheats!" complaint.
And, no, the cpu in question is marked 3.33GHz, and it runs at 3.33GHz base. Who cares how much boost it gives different people in differing situations? It's guaranteed to provide a base speed of 3.33GHz....and you get it. The performance boost is gravy.
Or do you prefer the AMD solution......downclocking cores but not turning them off and shoving whatever overvoltage the turbo'd cores need into the downclocked cores, despite them not needing anywhere near that voltage? Seems to me that AMD's solution is almost half-baked and a knee-jerk response to Intel's rather elegant solution.
i think they are just different. AMD is staying in its power envelope, inlets is about getting out as much perf as possible. how about the i7 laptops where they had to underclock due to heat, thats not very elegantQuote:
Originally Posted by Humminn55
edit: somehow i put in a thumbs down, didnt mean too, lol
Intel's turbo also considers the power envelop, but the i7 can get a reasonable measurement of the actual power usage through an onboard chip.Quote:
Originally Posted by Manicdan
That's not really anything to do with turbo; more of the laptop designers not designing the laptop to handle the maximum possible load situations.Quote:
how about the i7 laptops where they had to underclock due to heat, thats not very elegant
AMD's method also uses dedicated HW which AMD is not publicly specifying .The core is a new Shanghai/Istanbul revision code named Pharaoh Hound btw.Quote:
Originally Posted by accord99
AMD guarantees each chip won't cross over its maximum electrical specification in any circumstance. Also,we now know the changes in clocks are dynamically distributed over various cores and are very fast and thus should be "immune" to poor thread scheduling of Vista/Windows 7 OS.
Quote:
Originally Posted by Oliverda
http://en.wikipedia.org/wiki/Repeater_insertion
The RC wire delay is the delay through a wire due to it's (R)esistance
and parasitic (C)apacitance.
A low-k dielectric reduces the parasitic capacitance.
Regards, Hans
Thanks. :)Quote:
Originally Posted by Hans de Vries
If I'm not mistaken GF's 32nm SOI process tech will use high-k dielectric instead of low-k dielectric.
The key words here are: 'up to' and 'depending on CPU model'
It will use both :)Quote:
Originally Posted by Oliverda
High-k in the HKMG transistors and low-k to embed the wires.
Regards, Hans
Good to know.:DQuote:
Originally Posted by Hans de Vries
Hasn't the AMD/GF utilized the low-k dielectric in any SOI process tech so far?
Yes, for many years actually. This one has a just a lower k as the previous one.Quote:
Originally Posted by Oliverda
I think it should be at around 2.4 this time.
Regards, Hans
yes, some models are good for 400mhz, some 500, but sofar i dont know if that means we will always see the max turbo, or no turbo, or if there are stages, but since its a P state, i assume it means its max or no turbo, just how many cores is the only variableQuote:
Originally Posted by TRICHEESE
Thanks again.:up:Quote:
Originally Posted by Hans de Vries
I think the last digit determines if its 400mhz or 500mhz, because the 960T has 400mhz boost and so does the 1090T.... :shrug:Quote:
Originally Posted by Manicdan
yeah we went over all that before, it kinda sucks that at launch we wont have the 1095, or 1075, and the 1055 will be the 125W version instead of the 95W version. so all the ones i would want will be out in Q3Quote:
Originally Posted by ajaidev
I specially miss the 1075T, now my choice will be between a i7 930 or a 1075T. I wanted the E5620 instead of the i7 930 as a option but it will be more expensive than i want to go.Quote:
Originally Posted by Manicdan
Yeah it's very complicated :rolleyes: : 400Mhz over stock clock across 3 cores if model number ends with "0" or 500Mhz over stock clock across 3 cores if model number ends with "5" . So with 1090T model you have 125W 3.2Ghz hex core OR 3.6Ghz three core,depending on the load scenario.The TDP is never crossed ,it's a guarantee(dedicated HW inside the chip,not temperature based).It can only be the rated value or below the rated value(95W or 125W),no matter if Turbo kicks in or not.Quote:
Originally Posted by TRICHEESE
So even if you have just a single compute-intensive software thread its going to turbo 3 cores? That doesn't seem very efficient to me... just sayin.Quote:
Originally Posted by informal
that's orange to apples... What you said can be reversed when looking at games today, most of them use 2 cores so they will get a nice boost in CPU power.
id bet u will see 5 cores at .8 ghz, and 1 core at 3.x ghzQuote:
Originally Posted by Firestrider
I bet you will never have a case where there is a single compute-intensive thread, unless the intention is to intentionally isolate one core, for some strange reason. Or if you run out of date and obsolete software. Both seem pointless, unless you are bias for or against one architecture.Quote:
Originally Posted by Firestrider
One word: virtualization.Quote:
Originally Posted by flippin_waffles
so? You run 6 instances and your done.
Turbo on Intel or AMD is not for well threaded apps, it's for old/low-threaded apps, to give better perf.
In virtualization you need more cores mainly, so you can run more instances.
Quote:
Originally Posted by zalbard
probably thinking like a top of the line cool n quiet for the oc'er ???? 4ghz at a certain volt ... and when you put more load certain core will overvolt and oc to 5ghz if the heat can be taken out of equation??? if so we are thinking about the same thing
Quote:
Originally Posted by FatAlbert
same 45nm fab process so i doubt that they have one .... queue the LHE joke
I think this will be my next upgrade :)
Im going back to AMD!
Belgian stores already have stock ;)Quote:
Originally Posted by dinos22
Can I just say it's a pleasure to read your posts? :)Quote:
Originally Posted by Hans de Vries
really daaamn okQuote:
Originally Posted by jmke
nothing here
@jmke
And reviewers have their samples?
2 more weeks to go ...
BTW it will be interesting to see how 6-core is coping with DDR-2 memory. Will it show bigger difference going to DDR-3 than Phenoms X4 were showing?
If AMD will use faster NB clocks then I bet it will!
No. Review samples have yet to be send out.Quote:
Originally Posted by Lightman
No cold bug again :D
hm, il like for overclocking, better than PII C3? Or simillary (air/watter)?
come on! show us the results already! :D
how will the 960T price compare to the 965?
that will really be awesome if these unlock
960T + asus 890gx for the win
will these be available April 26th or is that just a paper launch?
960T with unlocked to 6cores will be my next cpu
i think, it will not possible unlock 960T up x6 (different solutions than core Denebs)
960T is Thuban based,so unlock is a possibility.
surely if an application only uses three cores and three throttle up there is only a 1 in 2 chance of the program using the overclocked cores, or does the CPU throttle up the 3 cores with the highest usage
:rofl:Quote:
Originally Posted by TRICHEESE
Sometimes it amazes me how ppl think they are so clever they know something that those who as a work desing cpu:s wont.
It would be fine to have it like this:
"How does the cpu know to boost most utilized cores all the time"
yes, but im meaning diferent system lock cores. We will see
Read the thread again,it's been explained before. There is dedicated HW in Thubans that dynamically clocks used cores to 3.6Ghz (for 1090T model) and these clock changes are extremely fast so the thread bouncing that happens in Vista/7 OS won't be an issue since the boosted cores can be random(the ones that are used in any moment of time if the condition for the Turbo Core is met: CPU utilization is less or equal to 50%).Quote:
Originally Posted by TRICHEESE
Turbo CORE rapidly adjust which 3 cores are in Turbo. Realy fast.Quote:
Originally Posted by TRICHEESE
If an application has multiple, demanding simultaneous threads then it is likely to benefit from all six true cores. If these threads are intermittent or if there are few threads, Turbo CORE is ready.
Turbo CORE is designed not to harm performance on highly threaded loads. Do not think of three cores being put to idle; the OS places the cores in a low P-state and Turbo CORE smartly accelerates the remaining cores. Turbo CORE is opportunistic.
AMD Phenom II X6 1055T Processor Hits Retail in China
http://en.expreview.com/2010/04/10/a...hina/7035.html
According to OPN it will be a new stepping. Probably "E".Quote:
Originally Posted by onethreehill
Thuban is revE
This is definitely going to be in my next ground up build for my Eyefinity build! Awesome.
BTW Informal, do you have any idea what CORE stands for? I guess its an acronym for something?
As per Simons post, maybe the O stands for opportunistic and the C for core. E for execution?
you forgot R which stands for Rapid :DQuote:
Originally Posted by flippin_waffles
Turbo Core Opportunistic Rapid Execution...lol
There is no such thing as an 'RC-wire', Hans means RC wire delay. Chips are wired up by etching or depositing metal lines within a dielectric material that is electrically isolating. Wires themselves have an intrinsic resistance and, also run parallel or one above another, due to the proximity two separate conductors separated in space setup a capacitance.Quote:
Originally Posted by Oliverda
The result is a complicated RC circuit, where the capacitor/resistor have an intrinsic charge time. The effect is to delay propogation of signals through the matrix but some fundamental frequency or time delay. The time constant (tc) is proportional to the capacitance and resistance of that arrangement. tc=R*C (R=resistance and C = capacitance). The concept of RC delay in an RC circuit is taught in most freshman level physics courses when they study electricity and magnitism.
The industry must work to minimize this effect because a) it causes latency in signal propogation, b) works to limit clock speed, and c) increases power dissipation. This is why there was a move to copper lines 8-10 years ago, and this is why you hear all the hubbub about low-k and ultra-low k material (not to be confused with high-K for gates). To decrease the time constant you want low R and low C, copper has lower R per unit length than does aluminum (which was used prior to copper). Capacitance gets lower with lower k (dielectric constant) material.
Jack
Heh yeah I couldn't come up with something good for R, but Rapid is good, I like! :)Quote:
Originally Posted by ridney
Really well explained in simple language :up:Quote:
Originally Posted by JumpingJack
affordable hexacore .. interestingg :p:
any tips as to what motherboard (chipset) and memory i should combine it with ? For overclocking of course !
I'm coming over from long time intel .. and those 2 worlds are very different nowadays i think .
You'll need AM2+/AM3 board that supports Thuban hexa core chips (BIOS level) and DDR2(in case of AM2+ board) or DDR3(in the case of AM3 board).
Thuban goes both ways,AM2+ and AM3,since the integrated memory contr. inside the chip supports both DDR standards :).
http://prohardver.hu/dl/s/rl.gifQuote:
Originally Posted by JumpingJack
Btw I've found a strange identity.
Opteron 170 vs X6 1055T
hmmmm when will AMD Desktop CPU has a multithread like Intel HT
A. neverQuote:
Originally Posted by PatRaceTin
B. when Heck freezes over.
C. see bulldozer for better options.
lol jk
programing still playing catch up.
so hyper threading can't replace real cores, yet.
And it shouldn't. HyperThreading is only good for servers, in which more servers could be hosted with dedicated threads or crunching/folding, in which more threads could run more efficiently.Quote:
Originally Posted by demonkevy666
If multi-threaded apps benifit from more threads without more cores there is something wrong with the CPU architecture or the programming.
Like has been said before, it's a band-aid for pipelines which are too long. Core i7 is like the Pentium 4 of it's day. :rolleyes:
And also, real man use real cores xD.
These are being sold in china still no review??
I mean no forum info, etc?
I'm curious that the 960T will be Black Edition or not...
Sorry I missed your question before :) . It should be acronym but I have no info on what it could mean. I can ask around and see if I get any info on it :). The launch is soon so we are going to know anything there is to know about Tubans soon enough :DQuote:
Originally Posted by flippin_waffles
Quote:
Originally Posted by Oliverda
hmm will these OC like opterons? xmas coming early? :up::up: haha
im sure memory will OC and be stable at stock speeds for alot more budget kits
This chip sounds pretty exciting to me with the Turbo features. I am going to have to check this CPU out and see what heat loads it has. Looks like a very nice chip ;)
It appears from those retail box shots that is has a stock heat sink like the ones that come with 965/955's, thats a good sign its not a really hot running chip. Compared to the new Gulftown Heat Sink.