Project Denver - nVIDIA RISCs with ARM

Imam drugara ljubitelja igara. Ni on nema vremena da broji trouglove. Međutim, ako mu igra koja ga zanima ne radi u fulu sa svim efektima bez seckanja, odmah radi nadgradnju.

Potpis

Nije bitno koji su zaključci izvučeni, već kako se do njih došlo.

Jes' i ja se zabavio malo sa igrama ovih dana i stvarno se vrlo brzo zasitio. Ono cega ne mogu da se zasitim je sirove cpu snage koju ima ovaj sandy. Nisam ja neko ko ne moze da saceka malo duze da se nesto odradi ali 7x dobitak u brzini poredeci sa mojim C2D Mobile je zastrasujuca. Pre dva meseca sam jurio bug u webkitgtk-u i trebao da backportujem neki patch i stvarno je smor cekati sat i po vremena da odradi compile samo da bi na kraju video da mi nedostaje po neka funkcija a zatim poceo opet (ne od pocetka naravno ali sam se u nekim trenutcima toliko pogubio da sam morao da krenem iznova, mnogo promena u API-ju a kod gabaritan damn it). No fun at all...

Sad, hah, vezite se slecemo a mozda i necemo :D

I tako, mogu oni da prave sta god hoce sa ARM-om na tim tabletima i cudima - dok god ne bude imao ovakve performance po ovoj ceni - aj' zdravo... I naravno da ne bih razmisljao o ARM-u na desktop-u u narednih pola decenije - deceniju minimum.

A ovaj GTX460 cu da promenim al' kad bi se zezali, tu je kao razbibriga u trenutcima dokolice, ionako 90% vremena teram HD2000 na SB-u koji je zivotinja za matori X3100. A da ja kupujem nove grafulje zbog budjavo optimizovanih igara koje su se prerano pojavile na trzistu - a-a. Ovo cudo ce da mi zavrsava posao u naredne 4 godine minimum, taman da vidim dokle ce da stigne ARM ;-)

Potpis

make love - !war

pa ni Apple nije tuzio ShapeShifter na Amigi - u jednom trenutku je najbrzi Apple bila Amiga 4000 sa 68060 i ShapeShifterom
kao sto je i Atari ST sa Aladinom bio brzi od originalnog Mac-a



onim aplikacijama koje nisu samo za konzumiranje sadrzaja (kako dimkovic voli da pojednostavi pojam tableta)

npr. http://www.youtube.com/watch?v=RiR-ml91Dhw (augmented reality) (evo sad vidim da i dimkovic pominje isto to na 4. strani... btw necu ni da te pitam sta bi sa onim intelovim mobilnim CPUu koji si spominjao 2007. da ce da sahrani ARM jer vidim da sad pises: "Hajde da kazemo da tu 20% odlazi na ne-efikasnost Intel arhitekture u odnosu na ARM" )


na Mac OS Xu je bio dovoljan i GeForce II sa 16MB iz 2002. za kompozitni desktop na GPU...

elem, kad smo vec kod Applea... cela prica oko AppStorea i forsiranja da se koristi samo dokumentovana Cocoa mi lici na pripremu terena za maksimalno jednostavnu tranziciju sa jedne ISA na drugu.
Jeste da je Apple radio to vec dva puta za 30 godina ali je uvek imao neki translacioni layer, ovako bi bilo mnogo lakse prebaciti celu biblioteku Mac softwarea na novu arhitekturu...

NVIDIA Project Denver is a 64-bit ARM Processor Architecture
3/8/2011 by: Theo Valich - Get more from this author


nVidia's Co-founder and CEO recently hosted a dinner for selected press where he discussed the company past, present and future, revealing interesting bits about the future architectures as well. The dinner was covered on several publications, but we took note of one sentence published on VentureBeat:


"To take the ARM processor, partner with them to develop a next-generation 64-bit processor to extend it so that all of computing can have the benefits of that instruction set architecture. It is backward-compatible with today’s ARM processors."

This was a clear confirmation what our sources were saying about Project Denver: that nVidia is bringing a 64-bit instruction set to the ARM CPU architecture. According to our sources, in order to fully develop a new high-performing CPU architecture, you need to dedicate roughly between 1.1-1.8 billion dollars. The figure comes from companies that actually develop CPU architectures, thus they have previous experience. The cost of Project Denver could be much higher, as nVidia doesn't have previous experience in building a CPU architecture. The company however, has a lot of experience in building a complex GPU architecture, which are now (arguably) more complex than any other computer chip on the planet. GPUs in fact, cover first (nVidia GF100/GF110), second (AMD 6900 "Cayman") and third place (AMD 5800 "Cypress"). Intel's ultra high-end CPU, Itanium "Tukwila" CPU comes in fourth place, with only 50 million transistors more than fifth-placed nVidia's GF110/114, used in mainstream cards such as GeForce GTX 460 and 560 Titanium.

According to Jen-Hsun, nVidia is "working on a CPU internally for about three and half years or so. It takes about five years to build any full custom CPU. And Project Denver has a few hundred engineers working on it for this period of time and our strategy with Project Denver was to extend the reach of ARM beyond the mobile, the handheld computing space."

Thus, we don't expect Project Denver to appear before late 2012 or early 2013 - in line with Maxwell GPU architecture, which is expected to integrate Project Denver architecture and become the first shipping GPU which could boot an operating system. It would not be the first GPU to boot an operating system, though. According to several PR representatives, the company already managed to boot an special build of Linux using Fermi GPU, but resources for that were abandoned as it proved too much of a hassle. Kepler of Maxwell for a public demonstration? Our money is on Maxwell…

Two major reasons why nVidia wanted to develop the processor - with or without Microsoft's dedication to base Windows 8 on ARM architecture - were Tesla and console business. Tesla is a growing business unit still gives a lot of revenue to either Intel or AMD, whomever ends up being selected for CPUs that feed the Tesla GPGPU array (1 CPU feeds 4 GPGPUs). Pairing Tegra 3/4 with Kepler-based Tesla is a start, and Maxwell/Project Denver should be a viable business solution, abandoning the need for an x86 CPU in order to feed the parallel GPGPU arrays.

As far as console business is concerned, Jen-Hsun confirmed that the company is working on an next-generation console. Chances are, the only chips inside the unnamed console should come from Santa Clara. Sony or Microsoft? We'll leave that guessing game to others.
http://www.brightsideofnews.co...rm-processor-architecture.aspx

Gde nestade desktop? Ispada da ciljaju samo na tržišta servera i konzola.

Verovatno ce u pocetku sve to biti skupo poprilicno - ocekujem da sidju na desktop trziste ubrzo posle inicijalnog release-a (ne vidim razlog zasto ne bi).

Potpis

make love - !war

Ne znam, meni je i dalje teško da zamislim neku masovnu prodaju pc računara sa matičnim pločama na kojima je denver. Sa druge strane, Apple mi deluje kao neko ko je dovoljno lud da sve svoje hardverske proizvode bazira na derivatima arm-a.


Elem, više detalja
(note: na linku su linkovi ka člancima o pojedinim delovima procesora kao i navodna sličica jezgra)

nVidia Reveals 64-bit "Project Denver" CPU Silicon Die!?
3/9/2011 by: Theo Valich - Get more from this author


During his introduction speech at Financial Analyst Day 2011, Jen-Hsun Huang revealed an interesting slide to the semiconductor analysts - silicon die of nVidia's first CPU. Without further a due, picture below shows what nVidia presented as Project Denver 64-bit CPU core:

As expected, no die size or manufacturing process were disclosed - given that there is are a lot of features subject to change - thus take this slide with a grain of salt. In order to help you understand the information shown on diagram, we're including links to sites that explain the functionality of particular units:
Fetch
Decode
Scheduler
ALU (Arithmetic Logic Unit)
FPU (Floating Point Unit)
Load/Store Logic
MMU (Memory Management Unit)
L1 Instruction Cache
L1 Data Cache
L2 Cache Logic Instruction/Fetch

The largest block on the die belongs to L1 Data Cache, while there is also L1 Instruction cache. While it is too early to discuss the actual features of this CPU core, we're really intrigued by some of the statements about the Project Denver core - mostly ranging in the field of attaching four Project Denver cores on a high-bandwidth interface such as existing GPU memory controller.

In theory, Project Denver cores inside the Maxwell GPU die should enjoy access to 2+TB/s of internal bandwidth and potentially beyond currently possible 320GB/s of external memory bandwidth (using 512-bit interface and high-speed GDDR5 memory). If nVidia delivers this architecture as planned, we might see quite a change in the market - given that neither CPUs from AMD or Intel don't have as high system bandwidth as contemporary graphics cards.

Given that Project Denver exists in silicon form, proper execution and launch should bury any doubts whether or not nVidia can make a CPU- which are still present with some semiconductor analysts who also claimed that AMD is going to die, AMD is not going to buy ATI, and that Intel will blow everyone out of the water with Larrabee and of course - that nVidia cannot make a CPU. These theories often forget that Intel is a largest company of companies mentioned and sometimes projects end in blind alley (WiMAX, Larrabee), AMD did the logical thing and acquired ATI and that was a good move - judging by initial Fusion APUs and of lastly - nVidia showing that they can execute as well.

Another big questions is whether or not they can effectively support an CPU ecosystem? That's another ballgame entirely, since we heard that even with 3D Vision, GeForce and Tegra, as there are many partners that are not satisfied with the level of commitment given by the company. The true measure of success for any product lies in what support can be given to the ecosystem. In that way, Project Denver has to shine not just on the hardware side.

All in all, things are heating up and nothing is better for consumer than the holy word of consumerism: competition.
http://www.brightsideofnews.co...ct-denver-cpu-silicon-die.aspx

Nvidia Project Denver May Feature x86 Hardware Emulation
Tuesday, May 24th, 2011 | Hardware


In the 2013-2014 timeframe, Nvidia is set to launch its highly anticipated Nvidia Project Denver CPU and a recent report comes to claim that this processor will feature specialized x86 emulation hardware, to better compete with its Intel and AMD rivals.

The emulation will be done via a hardware layer that will include some special instructions, just as China has done with Loongson processors.According to Donanim Haber, this hardware layer should be able to emulate x86 instructions with just a 15 to 20 percent speed penalty, which is far better that the 30% overhead achieved by the Chinese MIPS processor.

This technology was presumably developed by the Transmeta engineers who joined the company back in 2009.

Nvidia has designed Nvidia Project Denver so that the CPU can be easily embedded inside a much larger pool of GPU vector hardware and it will feature two of four ARM processing cores.

These are based on a 64-bit architecture and an early die shot of the chip revealed that it would feature a massive amount of L1 cache, which will most probably be much larger than the L2 cache size.

The chip also features a high-bandwidth interface which allows Nvidia project denver to link it to a GPU memory controller.

While nothing is certain at this point, since Nvidia Project Denver is expected to launch at about the same time as Nvidia’s Maxwell GPU architecture, the two could be inter-linked to form a high-performance computing solution (with 2TB/s of internal bandwidth and 320GB/s of external memory bandwidth).

Nvidia Project Denver is designed to support future products ranging from personal computers and servers to workstations and supercomputers.

As mentioned previously, the chip is expected to launch sometime in 2013 or 2014, but Nvidia didn’t commit to any release dates.

Nvidia Project Denver May Feature x86 Hardware Emulation.

http://ontechnow.com/hardware/...re-x86-hardware-emulation.html

Izgleda da više nisu gadljivi na x86 softver. Ova boldovana rečenica me jako podseća na power pc scenario.



I ispostavilo se da nisam bio u pravu, odličan članak na temu zašto će se Apple držati x86 sveta.
http://realworldtech.com/page....icleID=RWT050911220752&p=1

Sam procesor ne čini kompletan sistem, a interesantna mi je ova Intelova Thunderbolt tehnologija, jel' neko imao prilike da to ošacuje uživo? Da li će se nvidia naći u situaciji da prodaje čip koji neće imati podršku ostatka industrije kojom dominiraju konkurenti?

iako volim da citam realworldtech.com ovaj tekst o ARMu je pun rupa.

1. "The first is a more unified software ecosystem, where iOS and OS X can eventually be merged together." ovo se moze desiti i sada. nema potrebe da u svim Apple uredjajima bude isti CPU da bi se spojili iOS i Mac OS X ali ovo se nece desiti jer ce Mac OS X ostati za high end aplikacije (tipa: PhotoShop, xCode, FinalCut, Logic, Cinema 4D... da bi sve ove aplikacije funkcionisale bar podjednako efikasno uz pomoc toucha kao sto danas funkcionisu pomocu tastature i misa bice potrebno vise godina tako da one zasigurno ostaju na Mac OS Xu i samim tim nema govori o potpunom spajanju iOSa i Mac OS Xa)

2. " The second is tighter control over their hardware and greater integration." - ovo je po meni najvalidniji razlog zasto bi Apple forsirao non-PC CPU - da sto je moguce vise oteza kopiranje svoje platforme.

Ovo je u direktnoj vezi sa onim sto prodaje njihove uredjaje: razlicitost! (u totalna pozitivnom svetlu) na ovaj nacin bi se samo jos vise odvojili od konkurencije po inovativnosti i funkcionalnosti i podgli na jos visi nivo svoju razlicitost ako ukljuce i hardware u celu pricu - vise nece biti dovoljno da neko iskopira samo softwerski funkcionalnost njihovih proizvoda (citaj Android kopirao iOS filozofiju rada) vec ce u celu pricu morati da ukljuci i custom hardware.

3. usteda u ceni CPUa je najverovatnije zanemarljiva u poredjenju sa prethodnim razlogom.

4. ostatak teksta je ok ali se postavlja pitanje:

vec danas Apple prodaje vise iOS uredjaja nego Mac OS X. Ovaj trend ce se u buducnosti samo nastaviti i jednostavno potraznja za Mac OS X uredjajima (x86) ce opadati u korist iOS uredjaja (ARM).

5. nema nikakvg smisla terati korisnike da koriste Logic ili Cinema 4D na 10x slabije procesoru tako da se prvenstveno iz ovog razloga nece preci na ARM u Mac OS X uredjajima.

Logic-u i Cinema 4D ce trebati godine da na tabletima (touch) uredjajima dostignu istu funkcinalnost kao verzije za desktop OS gde se koristi mis, tastatura... i Mac OS C

Mislim da jako lepo oslikava koje bi eventualne razloge Apple imao da pređe na arm kao i zašto ti razlozi nisu potreban i dovoljan uslov da se to i desi. Ukratko, prelazak sa x86 na arm je prelazak sa konja na magarca.


Ovo se neće desiti. Kucanje po ravnoj površini duže od par minuta je ubistvo kakva god da je podloga. Upravljanje glasom (koje najavljuju već više decenija) ne može svuda i uvek da se koristi.


Različitost po svaku cenu i bez obzira na žrtve. Ne hvala.


To insistiranje na različitom CPU-u od ostatka industrije ih je već jednom umalo koštalo glave. Da ne pričam da CPU ne čini ceo sistem.


Za njih, ne i za korisnike.Eventualni arm laptop ne bi bio jeftiniji od x86 laptopa, bez obzira na performanse. Samo ta situacija kada jedna firma određuje šta njenim korisnicima treba se nikada ne završi lepo. Čak ni IBM u vremenima kada je držao preko 70% desktop proizvodnje nije mogao sam sve da razvija a pogotovo ne brže od konkurencije.


Ta tablet budućnost (evo sada su najavili i iCloud) lepo deluje na papiru ali ne funkcioniše u praksi. Ta opsesija kompanija da imaju potpunu kontrolu nad korisnicima i koju u najnovijem pokušaju zovu tablet i cloud umesto glupi terminal i centralni server nikako ne mirišu korisnici kojima kompjuter služi kao sredstvo za rad. Zašto bi tražnja za OS X opala? Postoji kvalitetnije rešenje od desktopa?



Zašto bi bilo ko sa dva grama mozga koristio Autocad ili nešto slično na tacni za gledanje youtube spotova umesto na 24" monitoru? Lakše se programira i piše nego na laptopu? Ne daj bože da srećni korisnik poželi da unapredi to sokoćalo.

za content kreatore ostaje Mac OS X. za sve ostale korisnike: iOS (ili neki drugi tablet OS cim ga naprave).

Mac OS X sigurno ostaje na x86 (ili nekom drugom highend procesoru) - iOS ostaje na ARMu.

a komentari tipa: "Ne daj bože da srećni korisnik poželi da unapredi to sokoćalo." su takooo... Bil Gejtovski ti kao da nikada nisi imao Atari


Apple worth more than Microsoft and Intel combined

http://www.theregister.co.uk/2...osoft_and_intel_in_market_cap/

U prevodu, svi kojima je računar alatka za obavljanje posla.


Pošto su tacne neupotrebljive za bilo šta ozbiljno, hajde da vidimo kako se porede sa postojećim uređajima u zabavnom segmentu:

gledanje video materijala

tacna vs televizor 0:1
tacna vs monitor 0:1
tacna vs laptop/notebook/netbook 0:1
tacna vs mobilni telefon 1:0
tacna vs džepna konzola 1:0

čitanje

tacna vs monitor 0:1
tacna vs laptop/notebook/netbook 0:1
tacna vs ebook reader 0:1

fotografisanje:

tacna vs fotoaparat 0:1
tacna vs mobilni telefon 0:1

slušanje muzike

tacna vs mp3 player 0:1
tacna vs radio 0:1
tacna vs stub 0:1
tacna vs walkman 0:1

igranje:

tacna vs konzola 0:1
tacna vs desktop 0:1
tacna vs džepna konzola 0:1
tacna vs mobilni telefon 0:1
tacna vs riziko 0:1

Tableti neće pomoći većoj rasprostranjenosti arm-a. Klasa uređaja koja je potpuno nepotrebna a gura se jer proizvođači misle da svakih par godina moraju da predstave novu klasu uređaja. Od mobilnih mašina smo imali prenosnike, pa laptopove, notebookove, subnotebookove i sad netbookove. O raznim varijantama personal organizera neću ni da govorim.


Nisam. Moj poslednji kompjuter koji nije imao boju je bio zx 81. A i on je mogao da se proširuje. Posle iskustva sa njim (a i spektrumom) nije mi padalo na pamet da nabavljam nešto gde mi proširenja zauzmu sto i po a kablovi još pola stola. Ili crno beli računar od par hiljada dolara gde ti treba macola i let lampa da bi mu dodao pola mega memorije. Na to koji je koncept bolji jasno ukazuje činjenica da je današnji mek (ostali su pokojni) ispolirana kanta (za razliku od ove moje trenutne neispolirane) koji tera drugi os.


Ovaj, kakve ovo veze ima sa temom? Najveći deo tog profita je nastao prodajom muzičkih plejera i mobilnih telefona. Kakve to veze ima sa desktop os-evima i procesorima?

@Ivan



Nisu konzole bitno lakše za upotrebu, kad uz tu konzolu masa igrača ima i PC da tera WoW ili neku drugu MMO. Radi se o drugoj stvari, developeri forsiraju konzole iz drugih razloga.

1. Ako je konzola iz 2004. za nju je potrebna grafika iz 2004. ne iz 2011. što umnogome smanjuje troškove oko dizajna tekstura i modela. Manji posao!
2. Na konzolama je piraterija teža, mada i za PC postoje rešenja kao Steam.
3. Igre na konzolama su daleko jednostavnije od igara na PC-ju, i daleko ih je lakše napraviti. Hoćeš FPS? Kupiš engine, skriptuješ ga malo, napraviš modele , testure i nivoe, i kad je gotovo ide na test i QA. Već je sav posao završio onaj klo je napisao engine. MMORPG? Ajoj. Treba ti tim za klijent, tim za server i nekoliko godina rada, plus content i QA, naravno, i opet može da se desi da igru ne napraviš, jer odoše troškovi u nebesa, a deadline je tako blizu... Ne možeš da uporediš nijednu igru za konzolu sa MMO za PC...
4. Igrica uvek radi perfektno, da, ali je to igrica iz 2004.

Dalje, ugodjaj na konzolama je ugodjaj iz 2004. Nije uopšte ono što je 2011. Što ljudi ne mare, to je zato što ne znaju i kupuju te nikakve konzole. Konzole će da urnišu PC znaš kad? Nikad. nemaš jednu valjanu MMO ne konzolama, i tek će sledeća generacija konzola to možda da ispravi, probao je nešto Sony Online Entertainment, i naravno, ništa nije napravio. Ajd da vidim da konzola izrenda običan WoW kad je sto-dvesta ljudi na jednom mestu, kako je obično u WoW na prometnim mestima, naravno u 1080p. Problem broj dva za konzolnu MMO, kako koristiti chat ako nemaš tastaturu, a ekonomija, i gathering/crafting su važne stavke u većini MMO? Dalje MMO je kompleksnija igra od obične FPS. Nekad ni desetak tastera nije dovoljno za igranje. Kako da se to uopšte igra na joypadu?

Ljudi kupuju konzole zbog igara pisanih specijalno za njih, kojih nema za PC, ne zato što je bolje, ili znatno lakše za upotrebu.

Medjutim, radi se o tome da se ljudima prodaje bofl, ne samo kad su igre u pitanju, već sve. Kvalitet više nije na ceni. Pogledaj video u 720p odnosno 1080p. To je 2011. i konzola to ne može.



I pogledaj Carbon u 1080p (Eve Online).



@Calimero

Odgledaj video i reci kako da na tom iPadu igraš MMORPG kao Eve Online? A na OSX može BTW.

Još da dodam nešto, MMORPG plaćaš 15$ ili 15e mesečno za neograničeno igranje, koje teško može da dosadi. Na konzoli plaćaš jednu plitku sportsku igru ili FPS 50-60 $, da je igraš nedelju dana, nego ajde, nemamo mi PC igrači pojma.

Sad nešto vezano za temu, ARM je veliki korak unazad. Ima svoje specifično tržište embedded uredjaja koje godinama drži zbog svoje male potrošnje, ali kako bi se snašao sa nekom ovakvom MMORPG? I mene ovo podseća na stari PowerPC.

_{[Ovu poruku je menjao bojan_bozovic dana 09.06.2011. u 14:36 GMT+1]}

Ja imam Atari STe. Ima ACSI port i ROM port + SIMM slotove za upgrejd memorije. Cigla nema nista od toga.

Potpis

<sub>Q: HSP56 Micromodem nece da radi kompjuter ga prepozna a kad treba da se konektujem nece ne daje ni znaka zivota. u cemu je problem.

A: Crko mozda od grmljavine mozda od spanaca. Uglavnom baci ga u WC solju jako povuci vodu. Skupi 5e i uzmi drugi i ne postuj temu na pogresno mesto.</sub>

1st Silicon with NVIDIA Project Denver is an 8-Core ARM, 256 CUDA Core APU?
7/18/2011 by: Theo Valich - Get more from this author


According to information we heard from multiple sources, NVIDIA recently started to open out and disclose design targets for their first "CPU" i.e. SoC codenamed "Project Denver". As always, take this rumors with a grain of salt.

The information we have at hand is that Project Denver CPU core is looking to be very much aligned with T40, i.e. "Tegra 4" i.e. Wayne. According to internal schedule, Wayne silicon is going to be taped out in the next couple of weeks, with developers getting their hands on prototype silicon in December 2011.Wayne silicon is consisted out of four ARM cores (NVIDIA is not disclosing the core type, teasing that it might be either A15 or PD) and up to 64 GPU cores.

During the same month (December 2011), NVIDIA plans to tape out the first silicon based on Project Denver, which combines up to 8-core custom NVIDIA-ARM 64-bit CPU with a GeForce 600-class GPU. The company had a lot of issues in development of a CPU and the general consensus is that NVIDIA is take a conservative approach with a single 28nm PD CPU design and the 28nm Fermi-based design, i.e. the rumored Fermi-refresh in the form of notebook and lower-end desktop GeForce 600 Series cards (remember "GeForce 300"?). The interesting bit that we heard is that Project Denver is geared towards "full PhysX support", whatever that might be.

According to another source close to the subject, target for GPU part of the silicon is "at least 256 CUDA cores" which would put the product on pair with AMD's Trinity APU which will pair a Bulldozer-Enhanced CPU core with "Northern Islands" VLIW4 architecture and will be the key APU for AMD in 2012. Compute power-wise, NVIDIA doesn't want to clock it to heavens' high, but rather to squeeze each IPC (Instruction Per Clock) as possible. Still, it is realistic to expect 2.0-2.5GHz for CPU and similar clock for the GPU part, with memory controller and the rest of the silicon working at a lower rate to keep everything well fed.

Unlike AMD's APU design, where CPU and GPU parts connect to the memory controller at the speed of DDR3 memory, Project Denver is looking for a more direct communication between CPU and GPU cores, i.e. relying on the best GPU design can offer: high-bandwidth connection. NVIDIA is not taking the conventional route with L1, L2 and L3 cache design, since the GPUs have 1TB/s+ connections to its cache memory, similar approach is rumored for Denver core design as well. Just like in the GPUs, memory controller takes the larger portion of the die and connects CUDA cores with CPU cores and CPU will have priority access to the bandwidth needed.

However, NVIDIA needs only 10-20% of the total bandwidth available for the CPU portion, keeping the GPU well fed with system memory.

The design for the motherboards for notebooks, desktops and servers is also under way, with PCIe 3.0 being touted as the main connectivity feature, alongside USB 3.0 and SATA 6Gbps. According to the information given, nV is going to attack all remaining fronts - light notebooks will be covered by lower-scaled Tegra 3 and 4 (smartphone/tablet/netbook/t'n'l notebooks), more powerful silicon will be used for lighter-use notebooks and desktops, while the server play is probably the most interesting strategy.

From one side, within one generation of PD CPU designs, NV wants to enter the blade market and offer a powerful GPU, but from another, the company wants to get rid of the demands for Intel or AMD x86 cores in their Tesla business (one Tesla GPGPU requires one x86 CPU - Xeon or Opteron for "feeding"). Removing X86 from that equation will significantly boost the profit outlook, just like AMD and Intel are enjoying charging an arm and a leg for the same silicon used in their desktop/notebook computers.

http://www.brightsideofnews.co...e-arm2c-256-cuda-core-apu.aspx

hmm, ovi ipak misle da napadnu kompletan spektar x86 mašina. Da ne ponavljam šta mislim o njihovim izgledima na uspeh, ostaje samo da vidimo šta će predstaviti za dve godine.

Told you so

Potpis

Sloba je za 12 godina promenio antropološki kod srpskog naroda. On je od jednog
naroda koji je bio veseo, pomalo površan, od jednog naroda koji je bio znatiželjan, koji
je voleo da vidi, da putuje, da upozna,
od naroda koji je bio kosmopolitski napravio narod koji je namršten, mrzovoljan,
sumnjicav, zaplašen, narod koji se stalno nešto žali, kome je stalno neko kriv - Z.Đinđić

Nvidia: Project Denver is ARMv8-based, is ''going great''
by Cyril Kowaliski — 9:35 AM on November 9, 2012


Late yesterday, Nvidia CEO Jen-Hsun Huang got on the phone to talk to stockholders and journalists about the company's latest quarterly results. During the Q&A session, Huang revealed that Project Denver, the company's upcoming desktop processor, will be a custom chip based on the 64-bit ARMv8 architecture—and work on it is "going great." Here's the transcript, courtesy of Seeking Alpha:
Well, Project Denver is a 64-bit processor. And it's a full-custom v8 64-bit processor with some exciting secret sauce. And we haven't revealed much about it. There are no 64 bits processors shipping today, as you know. And v8 was just announced. The next 64-bit was just announced. And so, I guess, the only thing I could tell you is Denver is going great.

ARM announced its own ARMv8-based cores, the Cortex-A50 series, late last month, and it expects the first chips based on them to be out in 2014. Licensees include AMD, which plans to use them for upcoming Opteron server chips, and Samsung, which could very well have ARMv8-based server silicon in the pipeline, too.

We've known about Project Denver since Nvidia announced it at CES 2011, but the company has kept surprisingly quiet about it. In an interview with VentureBeat in March 2011, however, Nvidia's Huang said the chip was about three-and-a-half years through its five-year design cycle. If everything is going according to schedule, then Project Denver might now be complete—and perhaps we could see it released before Cortex-A50-based offerings.

Meanwhile, Microsoft is working with ARM to add support for the ARMv8 architecture to Windows. The existing ARM-compatible version of Windows, Windows RT, supports only current 32-bit ARM processors. There's no indication of exactly when 64-bit support will be implemented, though. Given Project Denver's desktop aspirations, Windows support seems like it's going to be paramount. (Thanks to X-bit labs for the heads up.)

http://techreport.com/news/238...-is-armv8-based-is-going-great

Čekaju Win RT-64.

Prvo tromesečje 2013 dođe i skoro prođe, a od Denvera ni traga. Sada ispada da treba da ga očekujemo 2015.

Nvidia CEO unveils next two gens of Tegra, dubbed Logan, Parker

Jen-Hsun Huang says the company's mobile chip will show 100-times improvement from the first generation to the fifth.
by Shara Tibken | March 19, 2013 10:50 AM PDT


Nvidia discussed its Tegra roadmap during its developers' conference.
(Credit: Screenshot by Shara Tibken/CNET)

Nvidia CEO Jen-Hsun Huang today talked up the chipmaker's next two generations of mobile chip, saying the Tegra chip line will show a 100-times improvement in strength from the first chip to the fifth. he current Tegra chip, Tegra 4, hit the market earlier this year. The next generation, coming later this year, is code-named Logan, while the fifth generation of Tegra is code-named Parker, Huang said at the company's GPU developers' conference. Logan will incorporate Nvidia's Kepler GPU, which currently is dominating the discrete graphics market. It also will include CUDA, Nvidia's programming model for GPUs that allows users to do many tasks at once. Parker, meanwhile, will be Nvidia's first processor from its Project Denver push. That includes full 64-bit capabilities, making it more attractive for use as the brains in one's computer or in a server or supercomputer. Nvidia first started talking about Project Denver during the Consumer Electronics Show in 2011, but it hadn't provided many updates since that time.

http://news.cnet.com/8301-1035...-of-tegra-dubbed-logan-parker/

Kad se je pojavio Duke Nukem Forever, valjda će i to. Živi bili pa videli!

Jedino x86 kompatibilnost može pomoći. Naime, glupo je govoriti o x86 (tj. x64) procesorima, kada AMD i Intel jedino zajedničko što imaju je instrukcijski set (inače se potpuno razlikuju). Neka nVidia može biti treća, ma šta koristili kao osnovu, ali jesu li sposobni za tako nešto?

Nvidia’s 64-bit Tegra 6 might power the first 64-bit Android devices in 2014
By Ryan Whitwam on October 11, 2013 at 10:11

Apple was accused of employing a marketing gimmick recently when it announced a new 64-bit processing core for the A7 ARM chip, which powers the new iPhone 5S. The benefits of moving to 64-bit processing on mobile devices aren’t yet clear, but that hasn’t stopped other ARM chip makers from lining up to mention their own 64-bit plans. Nvidia announced several years ago that it had licensed ARMv8 technology to build its own custom 64-bit core, and the time draws near when that endeavor will come to fruition. Tegra could finally be ahead of the pack.

All of Nvidia’s current and past ARM chips have used CPU designs licensed directly from ARM Holdings, which are then paired with Nvidia’s custom ultra low-power GeForce GPU. Together these components make up the Tegra system-on-a-chip (SoC). Tegra 4 uses the fastest ARM-designed cores currently available — the Cortex-A15 — but even with these top-of-the-line cores Nvidia is at a disadvantage thanks to custom mobile CPUs.

Snapdragon SoC

Apple and Qualcomm both design their own CPU cores based on the ARM instruction set, which they pay a pretty penny to license. Apple stepped it up to the ARMv8 set with the Cyclone CPU core in the A7, while Qualcomm is still focusing on ARMv7 with the Krait 400 core (but it is expected to announce an ARMv8-based core soon). It’s a lot of work to design a CPU core rather than just license one directly from ARM, but 64-bit is only the most recent example of how a custom core can help a company iterate and add new features. For example, Qualcomm’s custom chips can dynamically change clock speed on each core based on system load. Nvidia had to jury rig an extra companion core in Tegra to handle low-power tasks.

It’s believed that Nvidia started work on ARMv8 back in late 2010 or early 2011. If true, the company could possibly have samples of the so-called “Project Denver” 64-bit processing core ready in early 2014 as part of the Parker SoC. This is actually two generations past current ARM chips. Tegra 5 (code named Logan) is expected to hit devices in early 2014, based again on the Cortex-A15. Parker is supposed to be Tegra 6, but Nvidia could be in a position to move quickly toward this SoC and cut Tegra 5′s run a bit short. Although, Parker was slated for a 2015 release the last time Nvidia spoke about it.


No one can deny that Tegra 4 is a much more competitive chip than the Tegra 3 was last year, but Qualcomm’s custom Krait cores and quick design iteration have allowed it to nearly take over the Android device ecosystem. There are only a handful of devices announced with Tegra 4 chips, and one of them is Nvidia’s own Shield console. Meanwhile Snapdragon 600 and 800 power almost every high-end smartphone and tablet from 2013.

Over the last few years, Nvidia has no doubt poured many millions of dollars into Project Denver and the Parker SoC. Qualcomm is surely working toward a 64-bit chip down the road, but the company has been tight-lipped about future plans. The last thing Nvidia wants is for the market leader to beat it to ARMv8 with 64-bit. Whether or not 64-bit has real utility on mobile devices running 32-bit software, getting this right could set up Tegra for undeniable success, a distinction that has so far eluded it.

http://www.extremetech.com/com...64-bit-android-devices-in-2014

Intel i AMD grizu nokte, Denver stiže sledeće godine.

Hm, NVIDIA ce imati zagarantovan "niche" market (koji za njih nije "niche") u vidu CUDA kutija.

Dodavanje ARM-a i unificirana memorija ce drasticno pojednostaviti CUDA kod, tako da samo sa tog stanovista za NVIDIA-u ima smisla da prckaju sa 64-bitnim ARM-om.

Kao neka zamena za Intel servere - za to ce prvo morati malo da skoce, pa da kazu hop. Postoje neke druge stvari koje su tu u igri, kao sto je recimo gustina procesora sa malom latencijom izmedju njih. Sa Ivy Bridge EX-om koji izlazi za mesec-dva, nije nikakav problem strpati i do 140 jezgara (280 thread-ova) glueless.

Naravno, ti prvi ARM-ovi sigurno nece da konkurisu Xeon EX liniji, ali cak i najmanji E3 Xeon-i (ili ovi novi Atom-bazirani) su i dalje prilicno jaka platforma sa odlicnim performansama po vatu i odlicnim I/O mogucnostima - ARM tu jos nije ni usao u igru, tu su otprilike kao Intel i mobilno trziste 2006-te pre izbacivanja Atoma... drugim recima, upadaju azdaji na teren, to nece biti walk in the park :)

Potpis

DigiCortex (ex. SpikeFun) - Cortical Neural Network Simulator:
http://www.digicortex.net/node/1 Videos: http://www.digicortex.net/node/17 Gallery: http://www.digicortex.net/node/25
PowerMonkey - Redyce CPU Power Waste and gain performance! - https://github.com/psyq321/PowerMonkey