Prvo nekoliko komentara na urednikov post

- on je došao s X-om u paketu.
Nije istina. Jobs je napravio X, ono što je došlo s njim se više bio paravan za njegov dolazak nego temelji proizvoda koji danas zovemo X. BeOS je bio lošije utome što uz njega nije dolazio Jobs

- on je razjurio proizvođače klonova.
Jedan od mudrijih poteza. Mac brand ne podnosi klonove. Klonovi su sigurna propast Applea i Maca. Možda u budućnosti (vidi niže)

- Svaki pokušaj analiziranja intel tranzicije kroz \"Jobs je onda rekao ovo, počešljao se ovako, a sad kaže ovo i češlja se ovako\" su bezvezarije. Industrija se mijenja brzinom svjetlosti, prčkati po 20 godina starim izjavama da bi se komentirao ovaj novi potez su bezvezarije. Pogotovo uzimati u obzir reklame kao mjerilo što se govorilo nekada a što sada....

- Surova osvetoljubivost nema veze s ovim potezom. Izbaciti 10 knjiga iz dućana ne može se mjeriti s prebacivanjem na intel. Ovaj potez je odobrio nadzorni odbor, a izbacivanje knjiga niži manager koji upravlja dodatnim sadržajem u Apple Storu. Uostalom, u SVOJEM DUĆANU ti pišeš pravila. Sloboda medija? Nije Appleovo da brine o tome.



Moj komentar na traniziciju... Diskusija me interesira...



1. Apple je R&D cijele PC industrije. Niti oni sami ne znaju uvijek da li će uspjeti, ali upravo zato su danas tu gdje jesu (u pozitivnom i negativnom dostignuću). Danas nitko ne zna točni odgovor, niti Jobs, što će sve donijeti tranzicija na Intel. Ali inovativna tvrtka poput Applea sigurno će pronaći rješenja za ključne probleme. Tržište će imati zadnju, ipak.

2. Više nego rješavanje Appleovih problema (vidi 4. i 5) ovdje se radi i velikom koraku za industriju koja konačno hardver smješta na nivo koji zaslužuje - komad željeza!

3. \"Komentatori\" koji analiziraju da li će Apple želi uništiti Gatesa (ili bilo koga) ili što već - odrastite ili prestanite komentirati. Jobs nikoga ne želi uništiti već samo traži nova tržišta za Mac. Čak i ne moraju drugima oteti tržište - ima još dovoljno potpuno nesvojenih tržišta

4. Jobs nije izričito govorio da prelaze na Intel zbog sirove snage već zbog boljeg omjera watt/performance! Zar to ne govori ništa... Naravno, i snaga budućih Intela nije zanemariva

5. Novi proizvodi koje žele napraviti a ne mogu s PowerPC-ijem - tko kaže da su desktop/notebook računala? Možda se radi o PocketPC, TabletPC ili sličnim uređajima za koje žele vertikalnu kompatibilnost procesora s ostatkom ponude

6. Danas Intel ne znači da sutra neće biti AMD. Naravno da je Intel odabran za prvog dobavljača, iz hrpe mktg razloga.

7. Intel Inside brand dodati će svoju brand vrijednost na Mac brand

8. Intel Inside ne znači i Windows nesigurnost i problemi. Negativni utjecaj na Mac brand u tom pogledu ne postoji. Zar je ikada e-mail virus pod Windowsima povezan s Inte procesorom na matičnoj ploči?

9.Najveći rizik tranzicije je gubitak povjerenja/lojalnosti korisnika. Developeri će profitirati - nešto se događa, lova se vrti

10. Očuvanje vrijednosti Mac branda je prioritet za Apple. Direktiva #0. Što će se sve moći pokrenuti na x86 Macu od konkurentnih operativnih sustava i na koji način ovisiti će o Appleovoj procjeni koliko to može štetiti Apple/Mac brandu. Kada SW biznis postane X puta profitabilniji od HW biznisa - OS X na Dell PC-iju je realna opcija. A trebao bi biti ako išta od MS uspjeha Jobs preslika na Apple

11. Ako je konsolidacija PC industrije pri kraju (preuzimanja, spajanja itd) na svim područjima, zašto se ne bi dogodila i na CPU nivou. Od 2006 godine PC računala će pogoniti jedna CPU tehnologija koju će proizvoditi Intel ili AMD. Korisniku svejedno.

12. Korisnik će razlike između PC brandova percipirati primarno kroz OS koji ganjamo povrh hardvera, plus malo vanjštine na koju Apple uvijek polaže puno.

13. Ako je OS X toliko bolji od XP-a, i ostane toliko bolji od Longhorna - Intel tranzicija mu pruža prvu pravu šansu za ravnopravnu borbu. Pobijediti će bolji!


Zaključak
- ne zanima me CPU brand, ne zanima ne GPU brand, ne zanima ne mktg pljuvanje i usporedbe prije ili danas- zanimaju me samo vrhunski proizvodi za posao koji radim.

Apple, možeš li nastaviti isporučivati takve proizvode na vrijeme, uz najbolji price/performance ratio? Ako možeš, ja sam i dalje tu. Ako ne, tražim dalje...

BRAVO !!!

Čestitke i od mene. Lijepo napisano!

Hrvoje misli glavom. Samo mi je zao zasto nimalo od ovakvih objasnjenja necemo cuti s Appleove strane.

Nadodajem jos jednu misao:

* Malecka racunala kao sto su Mac Mini vape za necim boljim od G4. Intel to moze ponuditi bez ikakvih problema. Kad na to nadodamo licenciranje OS-a kojeg Apple ne mora placati nikomu, za razliu od ostalih PC proizvojaca koji XP masno placaju gladnom Microsoftu, imamo sigurnog pobjednika. Nijedna mala kutija iz PC svijeta nece biti toliko jeftina da pobijedi Mac Minija s mobilnim Intel procesorom, jer je on odmah u startu jeftiniji.

Jobsov graf s potrošnjom struje/performanse kao argument prelaska na Intel ne bi niti pas prožvakao. Naravno, nije imao petlje dati graf o snazi (te grafove imamo na apple.com pod G5, zar ne?), pa se počeo furati ne štedljivost. Meni takve stvari dižu tlak. OK, ako se govori o notebooku, ali što me boli briga da li mi stolni stroj troši ovoliko ili onoliko? Ionako danas svaki stroj ima smart energy management i dok radi lake poslove i potrošnja je minimalna. Muljao je u keynotu. Mnogima to nije baš leglo i to će imati svoje penale.

dpasaric je napisao:


Njegov keynote je supalj kao glave onih koji su mu povjerovali.
Vecu kolicinu budalastina odavna nisam cuo, a onaj graf mu je legendaran. Valjda je to jedini koji bi mu pomogao opravdati lucidno glupi potez -- zanemariti PowerPC na duge staze.
Slikovitije, ono sto je napravio vrijedja inteligenciju svake normalne osobe.

upravo procitah Jobsov komentar u Macworldu od prije tocno dvije godine(2003.)...kaze covjek; intel rumours? ma bez veze, ppc je nas izbor na duge staze i kazem vam da nas na tom putu ocekuje mnogo ugodnih iznenadjenja...

dpasaric je napisao:


...

\"As soon as I heard Steve say that the factor where Intel's road map was superior was processing power per (watt) I knew right away that it was exactly what I have been reading and saying and so have many others, that this is the real key to the future of high performance computers,\" Wozniak wrote.

As it happens, Intel's was not the only alternative chip design that Apple had explored for the Mac. An executive close to Sony said that last year Jobs met in California with both Nobuyuki Idei, then the chairman and chief executive of the Japanese consumer electronics firm, and with Kenichi Kutaragi, the creator of the Sony PlayStation.

Kutaragi tried to interest Jobs in adopting the Cell chip, which is being developed by IBM for use in the coming PlayStation 3, in exchange for access to certain Sony technologies. Jobs rejected the idea, telling Kutaragi that he was disappointed with the Cell design, which he believes will be even less effective than the PowerPC.
Now that Jobs has broken with IBM, however, Apple is free to pursue a potentially intriguing consumer electronics strategy with Intel.

Intel has been looking for ways to get its chips into devices that can compete with game consoles as living-room entertainment hubs. In fact, all three next-generation video game machines made by Microsoft, Nintendo and Sony are based on IBM chips. And analysts say that both Microsoft's Xbox 360 and the Sony PlayStation 3, scheduled to arrive next spring, will be positioned as home media hubs in addition to being video game machines--and priced far lower than the Intel-powered, Windows Media Center PC's that are also aimed at the living room.
Should the new consoles find wide acceptance as broad-based entertainment engines, Intel will need to respond--and one attractive alternative would be an inexpensive Macintosh Mini based on an Intel processor, able to run the vast library of PC games.

shlomo je napisao:



Zasto nije odabran Cell?
Evo nekoliko natuknica o Cellu ...


The reason I believe the cell was not considdered is that it multiplies the strengths and weakenesses of the PPC. You will have notcied that in benchmarks, the 970 eaither comes in at 70% perfomance of its competitors, or 200+%. This would be even more extreme in the cell, somehere at 30~50%, and 800+%. Unfortunately the number of benchmarks that show the 800% improvement would be reduced, and those that show the 50% loss would be increased.

The cell would make a good co-processor addon to get that 800% boost on code it is good for ( streaming media, vector calculations, highly paralell graphics rendering ), while having a more traditional CPU ( not the cut down one on the cell itself) handle the general computing tasks.

Given the target and volume of the cell, and the spread of low-lactency peripheral interconects ( like PCIe, which seems to be ont he new \"MacTel\" boxes ), one could get 90% of the benefits of the cell from an addon card with the same connection width as a GFX card. The local memory of the SPE's would mask the added latency of the peripheral bus nicely.

Another problem of going with a game CPU such as IBM's CELL and Xenion chips is that performance progress is very slow. You don't have better and better chips each year as with PC Chips such as Intel's Pentium. This is because a game console does not have to be updated every year - it gets updated every 6-7 years. PC's on the other hand, have to be updated each year or less in order to be competitive. THUS, the CELL and Xenion chips - despite their power today - will be STAGNANT in progress over the next 6-7 years. During this time, Intel's chips will only become more powerful and faster. IBM LOVES the idea of not having to pour development costs into making new versions of the CEL and Xenion each year.

Unfortunately, despite the current power of the CELL and Xenion chips, that is all you will get for the next 6-7 years. Note that the CELL is not even designed for PC use. It can only address 256 megabytes of memory - due to limitations of its memory bus - despite its speed. It would take another two years to be able to churn out CELLs good enough for PC use (i.e. which can address more memory). It would take another two years to be able to churn out CELLs which can have low enough power requirements to make it into laptops. To Apple, this would mean stagnation and eventual death.

Intel's processors on the other hand, have to improve each year to remain competitive. AND Intel can produce bulk numbers of their processors. AND Intel chips have low power requirements NOW for laptop use. For Intel to survive, it has to aggressively innovate and progress quickly. Look at how Intel developed a RISC core with a CISC wrapper to improve the Pentium's performance compared to RISC chips like the PowerPC. Eventually, the Pentium outstripped the PowerPC on almost all tasks. Despite the CELL's current power, I believe Intel (because it has to), will improve their chip power to the point it will also be faster than the CELL. Remember that the CELL's progress will be stagnant over the next 6-7 years because IBM will not have to innovate.

Apple absolutely made the right choice by choosing Intel. I wished they had done so years ago instead of foreseeing the dead-end that PowerPC would head into.


Graphic Processors (GPUs) are already faster than CELL


Note that the CELL is essentially a weak PowerPC core with 8 programmable Altivec units surrounding it. The 8 programmable Altivec units behave like DSPs - able to do parallel processing work at nigh speed.

The problem is that much personal computer work is serial not parallel. Outside of graphics and scientific number crunching, almost everything else is done serially in the CPU. Thus, dual or quad FULL cores are better for personal computer work.

Apple SOLVED the problem of faster parallel processing by harnessing the existing power of the programmable GPUs in the graphic cards provided by ATI and nVidia. These GPUs have much more power now than the CELL processor. Newer ones can easily outstrip the CELL processor. People don't realize how powerful their graphics cards are when used for other functions. Mac OS X no longer needs to use the Altivec processor to speed up its graphics. It instead gives the work to the GPU - which does a fantastic job of speeding up drawing the user interface.

Thus with Intel providing leading edge general purpose CPUs and ATI/nVidia providing leading edge GPUs, Apple is no longer constrained in power as it has been over the past several years.

Note that when Sony advertises the terraflops of power provided by the Playstation 3, it does not tell you that most of that power is provided by the GPU - not the CELL.

Note that when Sony advertises the terraflops of power provided by the Playstation 3, it does not tell you that most of that power is provided by the GPU - not the CELL.
Zato je napravljen CoreGraphics

Hvoje i Mak! Jako dobri članci i urednički i komentar. Obojici čestitke!