Intel Sandy Bridge

[Dusan Despotovic]

Intelov najnoviji nadolazeći model četvorojezgarnog procesora izrađenog u 32-nm tehnici poznatog pod kodnim imenom “Sandy Bridge” je po prvi put detaljnije predstavljen IT javnosti. Naime, procureli su prvi detalji o arhitekturi novog procesora koji bi trebao da se pojavi početkom 2011 godine. Prva verzija procesora koja će se pojaviti pod “A0” revizijom imaće 4 jezgra sa 256KB drugostepene keš memorije i radiće na frekvencijama od 2.8GHz do 3.4GHz a sa uključenim Turbo režimom jedno od angažovanih jezgara biće sposobno da radi na taktu od 3.8 – 4GHz. Intel “Sandy Bridge” će imati integrisano grafičko jezgro koje će takođe biti izrađeno u 32-nm tehnologiji dok će brzina biti postavljena na maksimalnih 1.4GHz. Grafičko jezgro će biti direktno povezano sa L3 keš memorijom i integrisanim dvokanalnim DDR3 memorijskim kontrolerom dok će arbitražu komunikacije između različitih uređaja na sebe preuzeti “Sandy Bridge System Agent” kontrolerska logika unutar samog procesora koja je ekvivalent današnjim “Northbridge” kontrolerima. Podržana je PCI Express 2.0 magistrala kao i nezaobilazni DMI. Najnoviji Intelov procesor će podržavati Hyper Threading tehnologiju, AVX set instrukcija i imaće 8MB zajedničkog trećestepene keš memorije.

[ukimafija]

Dusane valjda se daje i link sa kog sajta je vest kad se postavlja....

Elem, meni je ova njihova fora sa klokovanjem samo jednog jezgra stvarno presela vec, pa koji debil klokuje samo jedno od 4 jezgra...A i ovi taktovi nesto deluju bas mlako...Ja sam ocekivao da vidim default od 3.4 do 4ghz, a klokom da vozimo 5+ghz...Ali dobro , dosta ce se stvari promeniti do konacnog dizajna valjda...Samo sam se sa ovim dvokanalni memorijski kontroler malo zabrinuo...Necemo valjda tu korak unazad...
A i ranije sa pricalo o PCI-E 3.0 standardu jedno vreme...Bas me zanima zasto se od tog standarda odustalo...Valjda su shvatili da im ne treba za sad

[Dusan Despotovic]

Ko te pita da li ti se svidja ili ne(sto se jedno jezgro OC)..To je Intel,moze mu se..

http://www.bozzabench.com/News/tabid...S/Default.aspx

[Mickie]

Dusan Despotovic kaže:

Ko te pita da li ti se svidja ili ne(sto se jedno jezgro OC)..To je Intel,moze mu se..

http://www.bozzabench.com/News/tabid...S/Default.aspx

Taj link je odlicna vest, samo ako je tacno... AMD prakticno hoce da ponudo 6-Core Opterone masama po ceni sadasnjih Phenoma... tesko za poverovati.

Nije mi jasno sto nisti otvorio novu temu za to? Sad cu ja...

[sportista]

Intel planira da namerno ogranici OC Sandy Bridge-a

Information provided by Intel in its own presentations about its upcoming mainstream LGA1155 Sandy Bridge CPUs appears to confirm the company has designed the CPUs to deliberately limit overclocking.

A video leaked to HKEPC and posted on YouTube (see from 2mins onwards) confirms the fact that only a 2-3 per cent OC via Base Clock adjustments will be possible. This is because Intel has tied the speed of every bus (USB, SATA, PCI, PCI-E, CPU cores, Uncore, memory etc) to a single internal clock generator issuing the basic 100MHz Base Clock.

This clock gen is integrated into the P67 motherboard chipset and transmits the clock signal to the CPU via the DMI bus. This means there's no need for an external clock generator that used to allow completely separate control of all the individual hardware.

When you're overclocking, you want to be able to push certain frequencies, such as the Base Clock and memory clock, but leave others, such as SATA, completely stable as they're very sensitive to adjustment. Current motherboards allow multiple bus speeds because external clock generators are programmable via the BIOS.

According to one Taiwanese motherboard company, on a Sandy Bridge system, the fact that all the busses are linked means that turning up the Base Clock by just 5MHz caused the USB to fail and SATA bus to corrupt...

Izvor: bit-tech

[Thaedass]

^ dodatak

[ivanza]

I benchmark se oglasio.

Sa dolaskom Sandy Bridge procesora, Intel bi mogao da promeni politiku vezanu za overkloking. Nova LGA 1155 platforma će biti vezana za Cougar Point (P67) čipset koji neće imati eksterni PLL i sav menadžment klokova će biti integrisan u PCH. Svi klokovi su vezani za jedan klok generator, tako da će de facto overklokovanje biti ako ne onemogućeno, a ono znatno otežano. Novi base klok, DMICLK, će biti moguće overklokovati svega 2-3% pre nego što se pojave problemi kod pojedinih buseva, kao što su USB ili SATA. Overklokovanje će tako biti znatno otežano - makar u mainstream platformi. Prema trenutnim planovima, buduća high-end LGA2011 platforma bazirana na Sandy Bridge E i Patsburg (X68) čipovima neće imati ovih limita i menadžment klokova će biti standardan. U mesecima pre dolaska novih platformi stalno se pojavljuju informacije da Intel želi da stane na put overklokovanju. Ovaj put se radi o nečemu drugačijem, pošto se radi o drugačijoj vrsti overklokovanja, te će korisnici morati da napuste pomisao na overklokovanje pomoću busa, već će morati da kupuju skuplje procesore sa otključanim množiocima, koji će moći da se overklokuju. Na taj način će Intel povratiti kontrolu i moći će da bolje dozira napredak u performansama iz generacije u generaciju, kao i preciznije da određuje cenovne klase u skladu sa svojim željama.

http://www.benchmark.rs/benchmark-ve...ocesorima.html

Intelovcima se lose pise.

[Thaedass]

Sandy Bridge will Show up LGA 2011

We heard of JCornell over at XtremeSystems that a teaser image about Intel’s 32nm Sandy Bridge processors, which comes in the form of socket shots of the LGA 2011 Sandy Bridge chips.


It is said that these sockets are almost perfectly square, and LGA 2011 is set to supersede LGA 1366 (Socket B) in the Sandy Bridge E series processors.

In order to hold down the processor, it will be flanked by two retention latches, but it may cause some new challenges for motherboard manufacturers, and inexperienced customer will damage the motherboard under usage.
It’s imperative that LGA 2011 takes the place of LGA 1366, for the former means quad channel, which performs better than tri-channel; In addition to that, the 2011 socket will have integrated graphics and PCI-Express bus. The design will increase in pins.

Izvor: enEXPreview

[Beagle]

Ovaj socket je tiha jeza... Nisam bas upucen u mogucnosti, ali zar BGA pakovanje ne bi bilo pametnije resenje? Veca kontaktna povrsina, a "kuglice" su mnogo otpornije na mehanicka ostecenja.

[Thaedass]

Intel's Q1 Sandy Bridge processor line-up exposed

There's no doubt about it, by the end of the year Intel will start manufacturing 32nm processors based on the Sandy Bridge architecture and, just as 2011 roll in, will unleash them upon the desktop and portable PC markets to keep the money pouring in and AMD well behind on second place.

Unsurprisingly, although it has confirmed the release schedule of the first Sandy Bridge chips, Intel isn't saying anything (officially) about the specs of the upcoming models but thankfully, a leak is providing a good picture of what's coming.

The Sandy Bridge desktop and mobile offering for Q1 is detailed in the table below and it contains no less than 19 CPUs - 12 quad-cores and 7 dual-cores, going from 2.3 GHz to 3.4 GHz. The chips listed come in LGA 1155 packaging and feature an integrated (dual channel) DDR3 memory controller as well as a graphics core.

The mobile models all include Hyper-Threading technology but this will apparently be absent on certain desktop chips (the Core i5-2500s and i5-2400). To keep things complicated Intel will have several flavors of the same model available, these being set apart with an end letter (or lack of a letter). For example K means an unlocked multiplier, S - lower frequencies and a lower TDP, while T - even lower clocks and TDP ratings. Pricing info hasn't been revealed as it's probably still in the works.

Izvor: TCMagazine

[Doomche]

Svidja mi se to shto su potvrdili dvokanalni DDR3 kontroler. To potencijalno (znajuci Intel, ostaje potencijalno) znachi nizhe cene, a realne potrebe za vecim memorijskim protokom i dalje nema.

Valjda bude kakvih informacija oko ovog "ujedinjenog BCLK-a", nadam se da nece napraviti ogroman fijasko sprdajuci se na taj nachin...

[ivan90BG]

Jel moguće da je intel promeni broj pinova za slabije procesore za jedan samo da razbije kompatibilnost sa starijom generacijom komponenata.

Kako li je samo socket 775 služio 4 godine od P4 do C2Q. Čini mi se da znam šta je u pitanju. Pre nehelema Intel nije pomišljao a integriše još nešto u procesor, pa je 775 imao rezervu u pinovima za 4 jezgra procesora. Da bi se dodao memorijski kontroler i PCi-E povećava s na 1156, pošto memorijski interfejs verovatno troši dosta više pinova od PCI-e za 3. kanal i još dva jezgra broj se broj penje na 1366. Pošto se za sledeću generaciju dodaju još jedan kanal memorije, PCI-E i verovatno više od 6 jezgara (i možda još nešto što ne znam) razumljivo je novo povećanje.

Ali ostaje činjenica da je intel ovo verovatno znao sve vreme i da je kao dobar poslovni potez odlučio da ne ostavlja rezerve u broju pinova za sledeće generacije procesora. Ali pošto će specifikacije nižeg i srednjeg segmenta ostati iste i nema razloga za veliko povećanje broja pinova odlučili su samo da promene broj za jedan i razbiju kompatibilnost.

[Thaedass]

More info

Hint: GPU clocks

[Thaedass]

Intel's Patsburg chipset comes with a few unexpected surprises

TODAY WE FOUND out some details about changes to Intel's upcoming Patsburg chipset for Sandy Bridge LGA-2011 processors and we were slightly surprised at what we heard. What struck as very unusual is the fact that Intel is developing a single chipset that is intended for at least both the high-end desktop Sandy Bridge B2 processors and the Xeon destined Sandy Bridge EN processors. Traditionally Intel has developed these types of chipsets separately, but apparently this is no longer the case.

The Patsburg chipset is destined to be Intel's first single chipset solution for its high-end desktop products and mid-range server solutions, as with the rest of Intel's range of Sandy Bridge processors, we should see the PCI Express controller being moved from a separate chip into the CPU itself. Unlike the LGA-1155 parts, the LGA-2011 processors are still set to feature PCI Express 3.0, although, unless AMD launches Southern Islands with PCI Express 3.0 support, it'll be quite a while before anything will take advantage of this faster interface.

It seems like Intel has yet to make up its mind up with regards to storage connectivity. It's been suggested that Intel was set to fit as many as 10 SATA 6Gbps ports with SAS support, but from what we were told today, this isn't the case. Instead the new chipset is something of a mess, as it'll have support for two or three SATA 6Gbps ports, six SATA 3Gbps ports and six SAS ports, via three independent controllers in the chipset. This all has to do with Intel's new great idea to develop a single chipset for desktop and server products and as such both teams have added the features they want. You won't be able to make RAID arrays across the various ports either, despite the SAS ports supporting SATA 3Gbps drives. We can't quite figure out why Intel has gone down this route if this proves to be the final solution.

Luckily for Intel, we should be about 9-12 months away from the launch of the LGA-2011 platform, so the company has plenty of time to iron out whatever needs to be fixed. One thing that we had confirmed as well is that the desktop parts will get quad channel memory support, not triple channel as what was believed earlier. Intel is meant to have vastly improved the memory performance as well thanks to the new ring bus type memory controller that the new Sandy Bridge LGA-2011 processors use.

The CPU is supposedly going to use the QPI bus to interface with the chipset, unlike the LGA-1155 processors which are stuck with DMI, although a version that should offer twice the bandwidth of today's LGA-1156 platforms. The LGA-2011 processors might still use DMI, as we haven't been able to verify if QPI will be used as the chipset interconnect, but considering that the CPUs have an external QPI link, it would make sense for Intel to use this as the chipset interconnect, as otherwise it would be left unused in the desktop and single socket Xeon LGA-2011 platforms.

With eight memory slots and up to 15 SATA/SAS connectors on the board, we'd expect to see a lot more EATX boards in the market, as a standard ATX board just won't be able to fit it all. EATX boards measures 13-inches (330mm) in width, compared to 9.6-inches (244mm) for a standard ATX board, although the length remains the same. There's no other way you could fit eight memory slots – if that is what the motherboard manufacturers will go for – on a standard ATX board. It's possible that we'll see some LGA-2011 boards with just four memory slots, but this means that you'd have to scrap all your memory if you'd want to upgrade it in the future. For most people this would also mean buying a new case, as most standard PC cases can't handle EATX boards.

Although the person we talked to is a trusted source, it seems like Intel either hasn't finalised its specs for the Patsburg chipset, or it's telling different stories to its partners. In all fairness, our previous article about Patsburg was back in April and based on details from another publication and very little information seems to have trickled out since then. We can't quite make heads or tails of what's going on and as much as we trust our source in as far as what they were telling us, we're not as confident that Intel is telling the same story to all of its partners, so take this information for what it is.

Izvor: SA

[Predrag Stankovic]

Intel Readies New Xeon X5690 Processor, Fastest Six-Core

Intel is readying its fastest six-core processors ever. While the company is working to introduce the Core i7 990X Extreme Edition for launch sometime in Q4 2010, its already begun sampling the Xeon X5690. The X5690 is a dual-socket ready LGA1366 processor based on the 32 nm Westmere-EP silicon. It carries a nominal clock-speed of 3.46 GHz (26.0 x 133 MHz), with a Turbo Boost speed of 3.60 GHz (27.0 x 133 MHz). It features a QuickPath Interconnect speed of 6.4 GT/s, and supports triple-channel registered DDR3 memory with its integrated memory controller. Cache levels include 256 KB L2 cache per core, and 12 MB shared L3 cache.
The processor is compatible with most server boards based on the Intel 5500 and 5520 chipsets with BIOS updates. A quick overclocking feat by Fitseries3 on the EVGA Classified SR-2 motherboard yielded a speed of 5 GHz making use of the high BClk multiplier. The chip is likely to carry a rated TDP of 130W. Intel may release this processor in the next quarter, either displacing the X5680 from its US $1,730 price point, or occupying a higher one. The CPU-Z validation of the mentioned overclocking feat can be found here.

[Thaedass]

[starx]

http://www.itx.ba/index.php?option=c...13207&Itemid=1

Objašnjenje imena Sandy Bridge Core i7 2600K

Druga Core generacija

Ime najbržeg Sandy Bridge proizvoda je Core i7 2600K i sada vam možemo dati i detalje ovog procesa imenovanja. Kao prvo, znate da je u pitanju Sandy Bridge ako vidite ime sa četiri znamenke koje počinje brojem dva.

Podijelimo sada Core i7 2600K na segmente da objasnimo ime. Intel Core je brend CPU-a, i7 je brend modifikatora koja će vam pomoći razlikovati opcije na različitim jezgrama. Broj 2 je indikator generacije, što znali da je u pitanju druga generacija Intel Core arhitekture.

600 je broj SKU-a koji će vam reći koji CPU kupujete, i naravno, veća brojka znači bolji CPU.

Sufiks K vam govori da je CPU otključan, a druge iteracije kao što je sufiks S vam govore da je u pitanju CPU sa performansama za lifestyle computing, dok T znači Power Optimised Lifestyle, šta god da je u pitanju.

Intel ponovo ima novi čipset, novi socket, nove nazive i, naravno, ako želite Sandy Bridge morat ćete baciti ili prodati svoj Nehalem Core i sistem i kupiti novi sa socketom koji ima samo jedan pin manje od svog prethodnika.

[Thaedass]

[Rumour] Celeron to be resurrected with Sandy Bridge

With no Celeron products released for the Nehalem generation, and the introduction of Atom, rumours spread citing the demise of the Celeron branding, which Intel was quick to deny. Fudzilla now reports that the Celeron branding will indeed be back from the dead with the Sandy Bridge generation. The first Sandy Bridge Celeron is expected to release in Q3 2010, with a price tag of $49.

However, Celeron didn't really die. While no Nehalem Celerons made it to retail, there was a limited OEM version targeted at low-cost business computers in developing countries - Celeron G1101. There was also just a single Pentium - G6950. Since Nehalem was split over two generations - 45nm for quad cores; 32nm for dual/six cores - the large range and quantity of Core 2 based Pentium and Celeron have continued to be on the market, leaving little room for new Nehalem based Pentium and Celeron CPUs.

Sandy Bridge sees a return to simultaneous release for dual/quad core mainstream and performance products, and with it, we can expect a major jump for the Celeron brand.

We can expect Sandy Bridge Celeron to be dual-core, and feature much lower clock speeds, with Turbo, HyperThreading disabled, and possibly lower L3 cache. The IGP core is also expected to be downclocked.

We are still about a year away from the possible release of Sandy Bridge Celeron - so a lot could change between now and then.

Izvor: Fudzilla

[sportista]

Sandy-Bridge Preview

[Predrag Stankovic]

Tek u januaru 2011. Pfff
http://vr-zone.com/articles/intel-sa...ewed/9697.html