|
|
|
![]() |
|
Strumenti |
![]() |
#201 |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
Se ne era già parlato in questo thread... son curioso di vedere dei test
http://www.theinquirer.net/inquirer/...d-hydra-on-x58 IT LOOKS LIKE Intel is not content with their x58 'Smackover' board, and is going to add a Lucid Hydra chip to the next revision. Yup, improved in every way, plus mix-and-match GPUs, what's not to love? Early word from people who claim to have seen boards in the orient confirms what the moles say, a Lucid Hydra chip, plus a few other goodies like a LED status readout. Moles inside second biggest blue say that the board is pretty much changed everywhere in detail, and that can't be a bad thing. This confirms what people told me a while ago, that Intel is taking gaming seriously and is listening to feedback. The last few Intel gaming boards have been fast, solid and feature laden, but lack that special pop the low-IQ set seems to crave. Luckily Intel isn't adding LEDs galore and shiny bits, just functionality and things that matter. Good for them. In the end, you will get a fast board that won't have heatpipes that double as a house for your hamster, but you will get mix and match GPUs. This board is going to be one to watch.µ
__________________
Sample is selezionated !
|
![]() |
![]() |
![]() |
#202 |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
Qualche altra slide ( non mi pare di averle già postate in passato ):
![]() ![]() ![]() ![]()
__________________
Sample is selezionated !
|
![]() |
![]() |
![]() |
#203 |
Senior Member
Iscritto dal: Mar 2001
Messaggi: 5390
|
@ Foglia Morta
hai trovato una risposta alla domanda: "come fa intel ad ovviare all'inconveniente del nuimero massimo di core che possono essere serviti da un ring bus?" Di fatto mette più ring di cui una serie (quelli più sfruttati) per trasferire dati tra un numero limitato di core (traffico locale) e uno per mettere in comunicazione tutti i ring bus locali qualora si dovessero trasferire dati tra gruppi di core. Ogni stazione di "interscambio" tra i ring è anche un gate di accesso verso l'esterno
Buon anno |
![]() |
![]() |
![]() |
#204 |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
Si infatti sembra una soluzione efficiente ed allo stesso tempo piuttosto ovvia. Penso che non andranno perduti gli studi fatti da parte di ATi sul ring bus
![]() Buon anno anche a te
__________________
Sample is selezionated !
|
![]() |
![]() |
![]() |
#205 | |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
Quote:
![]()
__________________
Sample is selezionated !
|
|
![]() |
![]() |
![]() |
#206 | |
Senior Member
Iscritto dal: Mar 2001
Messaggi: 5390
|
Quote:
|
|
![]() |
![]() |
![]() |
#207 | |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
Quote:
Se fossero 32 core e 16 ALU ciascuno fanno in tutto 512 ALU. Ad essere molto ottimisti direi che andrà a 2,5Ghz e fanno 2,5 - 2,6 Teraflop in fp32. IMHO un po pochini a fine 2009 , soprattutto se si considera che dovrebbe avere un efficienza più simile all'architettura ATi piuttosto che a quella nVidia che fa scheduling via hardware. Con Larrabee praticamente cercano minimizzare le fixed functions ma poi questo risparmio non si traduce in un reale vantaggio se riescono a metterci così poche ALU . Insomma gli svantaggi dell'architettura ATi + gli svantaggi dell'architettura nVidia , tutto per massimizzare la flessibilità. Non è un po pochino ?
__________________
Sample is selezionated !
|
|
![]() |
![]() |
![]() |
#208 | |
Senior Member
Iscritto dal: Mar 2001
Messaggi: 5390
|
Quote:
Inoltre il reale vantaggio della minimizzazione delle fixed function è che, in teoria (ovviamente), così facendo si dovrebbe riuscire a raggiungere un livello di occupazione delle unità di calcolo ancora maggiore rispetto a quanto avviene con architetture miste con alu programmabili e fixed pipeline. Infine, con l'adozione di algoritmi fdi filtraggio di tipo custom, l'importanza delle fixed function nelle rbe dovrebbe andare gradulmente scemando |
|
![]() |
![]() |
![]() |
#209 | |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
Quote:
![]()
__________________
Sample is selezionated !
|
|
![]() |
![]() |
![]() |
#210 | |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
Quote:
http://www.fudzilla.com/index.php?op...11408&Itemid=1 We had an interesting meeting with the president and CIO of Lucid, a company that gained a lot of attention at IDF 2008. Lucid promises a small chip that can make your two GPU scale almost linear. After a long discussion we learned that Linear is almost linear, and that linear is best case scenario and that real world performance should be close to linear. This would mean that two GPUs, let's say in SLI or Crossfire, powered by Lucid chip and driver, could score close to 100 percent faster than a single GPU. Let's say that if a single GPU would score 100 FPS in one game, two would score close to 200FPS. The chips that will find their place in both motherboards and graphics cards will appear on the market in 2H 2009. The prototypes are already finished, but the company didn't want to talk about any specific partners. Anyone who has a PLX chip that makes two cards run together could be Lucid’s potential customer, and obviously some guys are interested. This leaves a lot of unclear questions to be answered, as we suspect that both ATI and Nvidia might have something against a chip that makes SLI and Crossfire look like child's play, and that this could lead to a driver that will block Lucid’s ability to shine. We don’t want to talk about moral dilemmas, but we suspect that this might be the case. Lucid promised to show us a demo in the near future and we believe that Intel is keeping an eye on Lucid for its Larrabee school science project.
__________________
Sample is selezionated !
|
|
![]() |
![]() |
![]() |
#211 |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
http://www.fudzilla.com/index.php?op...11450&Itemid=1
Sources close to Intel have told us that Intel is working together with Lucid on some secret project. If Lucid can pull of its promised close to linear performance out of two chips, more GPUs could mean even more performance virtually linear so to say. Last week when we meet companies President and VP of Business development Mr. Offir Remez he did confirmed us that more chip means more linear performance. They don’t have to stop at three or four GPUs like Nvidia and ATI do, they can surpass this number. With this in mind imagine how much good this chip could do to Larrabee and we’ve already reported that Larrabee has 50 to 80 cores. With the help of Lucid Hydra technology there might be place for some sort of collaboration, but if Lucid is so good, someone will buy it out sooner rather than later. In case that Intel doesn’t need its internal Lucid, which we do believe will be the case, Intel might use the Lucid chip on its motherboards simply to stab both ATI and Nvidia and its Crossfire and SLI in the back. Once Larrabee comes out, there will always be possibility to put two Larrabee cards together and get more performance.
__________________
Sample is selezionated !
|
![]() |
![]() |
![]() |
#212 |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
non sarà larrabee ma pare che Intel ci tenga a dire la sua nel gaming anche con i processori... magari incoraggerà lo sfruttamento dei quad core: http://news.cnet.com/8301-13924_3-10153373-64.html
__________________
Sample is selezionated !
|
![]() |
![]() |
![]() |
#213 |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
http://channel.hexus.net/content/ite...m=17124&page=5
HC: Is there still an opportunity for AMD's Fusion concept? JP: Absolutely - and it will revolutionize the low end and even the midrange. Both AMD and Intel have pushed their GPU/CPU integrated parts back to 2011. That's not something to raise your eyebrows about or use as a predictive clue about the demise of the companies or the concept, but rather a clear indication of the benefit those integrated CPU/GPUs will get from 32nm. HC: What are your expectations for Intel's Larrabee this year? JP: Absolutely nothing. But next, OMG - watch out. This year will be a continuation of well placed and timed leaks with more tidbits of data, we may see a real spec sheet by Q4, and if we do it will be with first production units. But real production, and products you and I can buy won't show up till Q1'10 at the earliest as far as my crystal ball can tell. Then when Larrabee (I like to call it Larry's Bee cause one of the prime developers is a guy named Lawrence) shows up it will give ATI and Nvidia a real run for the money - regardless of what you hear those two say about it. However, ATI and Nvidia have had over a year so far and one to go to get ready for it, and they have probably 4x the number of world class graphics engineers that Intel has so they won't be easy to knock over or out perform them. That means it comes down to a marketing war and Intel's argument about programming environment. And that will not be proven out till we get to run some benchmarks which probably won't be till CeBIT 2010.
__________________
Sample is selezionated !
|
![]() |
![]() |
![]() |
#214 |
Senior Member
Iscritto dal: Oct 2005
Messaggi: 38297
|
__________________
Corsair 5000D - Ryzen 7 7700 - Asrock B650E PG - 2x16gb G.Skill Trident Z5 ddr5 6000 mhz - GeForce Rtx 4070Ti S. - Samsung 980 pro 1tb + Crucial mx500 1tb + WD 1tb - Corsair rm850w - LG oled C4 48 le vga che ho avuto |
![]() |
![]() |
![]() |
#215 |
Senior Member
Iscritto dal: Feb 2006
Città: Looking for a place to call home
Messaggi: 5325
|
La presentazione da cui sono tratte alcune delle slides viste finora.
http://s08.idav.ucdavis.edu/forsyth-...chitecture.pdf Ci sono tutte quelle del SIGGRAPH, per chi fosse interessato. ![]() http://s08.idav.ucdavis.edu/ Direi che è meglio se ci spostiamo tutti di qua, una volta fusi i 2 thread... ![]() Aggiungo solo un link ai miei passati interventi nell'altro thread, dato che credo che qualcosa non fosse stato postato: http://www.hwupgrade.it/forum/showpo...30&postcount=5 http://www.hwupgrade.it/forum/showpo...38&postcount=6 E rinnovo il mio invito a yossarian, ma anche a nAo e gli altri utenti superesperti di HWU, contenuto nei post di cui sopra... ![]()
__________________
A.L.M. @ HWBOT | Personal PC: Asus N56VZ | Work PC: Lenovo Thinkpad T420 (Core i5 2520M, 4GB ram, 320GB 7200rpm) | Mobile device: iPhone 4S Work It Harder, Make It Better, Do It Faster, Makes Us Stronger, More Than Ever Hour After Hour Work Is Never Over Ultima modifica di A.L.M. : 01-03-2009 alle 15:08. |
![]() |
![]() |
![]() |
#216 |
Senior Member
Iscritto dal: Jul 2005
Messaggi: 7819
|
Un po di slide dal GDC: http://pc.watch.impress.co.jp/docs/2.../kaigai498.htm
__________________
Sample is selezionated !
|
![]() |
![]() |
![]() |
#217 |
Member
Iscritto dal: Nov 2000
Messaggi: 91
|
Dal mio blog:
C++ implementation of Larrabee new instructions |
![]() |
![]() |
![]() |
#218 |
Senior Member
Iscritto dal: May 2006
Messaggi: 19401
|
Larrabee, analisi tecnica: un mostro, almeno su carta
![]() http://www.tomshw.it/graphic.php?guide=20090323 Mah, io rimango ancora molto perplesso... |
![]() |
![]() |
![]() |
#219 |
Member
Iscritto dal: Nov 2000
Messaggi: 91
|
Presentazioni di Micheal Abrash e Tom Forsyth al GDC (in fondo alla pagina):
http://software.intel.com/en-us/articles/intel-at-gdc/ Articolo di Abrash sul nuovo instruction set di Larrabee (LRBni): http://www.ddj.com/hpc-high-performa...ting/216402188 Marco |
![]() |
![]() |
![]() |
#220 |
Senior Member
Iscritto dal: Nov 1999
Città: Ceranova (PV)
Messaggi: 10382
|
Mi iscrivo, molto interessante.
__________________
Sometimes it is the people no one can imagine anything of who do the things no one can imagine. (Alan Turing) Pkappa Pc: R7 2700x, 16 Gb G.skill TridentZ RGB 2993 mhz 14-14-14-34, Rx Vega 64 8 Gb HBM2, Nzxt 340 elite, Asus MG279Q. Lord Fx: FX 8350, 16 Gb ram Hyperx 1866 10-11-10-30, Rx 580 8 Gb Nitro+ Sapphire, Corsair 400r, Samsung C24FG73. |
![]() |
![]() |
![]() |
Strumenti | |
|
|
Tutti gli orari sono GMT +1. Ora sono le: 04:44.