Procesory na architekture K10 - informace, spekulace

[GaGy]

Jenže když už CPU tak aspon 2,8ghz

[Aros]

Když Phenom ,tak fakt jen 45nm ...

Jenže když už CPU tak aspon 2,8ghz

Jenže když 2,8 GHz, tak jedině 45nm

[DEVastor]

jsem rad ze AMD zase zaclo s reklamni kampani, podle me je to dost dulezite http://www.amd-phenom.cz/x3/index.htm

[flanker]

Krasne nataktovaný Phenom X4 9950, chlazeno vzduchem (ovsem docela vysokych 1.45V ) !!!!



http://valid.x86-secret.com/show_oc.php?id=384836

edit//ten clovek to momentalne provozuje na cca 3480MHz (asi kvuli stabilite a nebo aby ho prilis netrapil s napetim?)

[socket]

myslim si ze to velmi stabilne nebolo

[DEVastor]

na 65nm je to zajimave

[yuri.cs]

doufam, ze to muze byt v sekci o K10

AMD SB750 Preview
http://www.anandtech.com/weblog/showpost.aspx?i=471

[flanker]

brzy by tam mela byt cela recenze, ale vypada to zajimave, co ctu ve forech, par vlastniku techto MB chipsetu SB750 nabehne s default napetim na 3.2GHz stabilne s x4 9950, coz je supr .

[flanker]

AMD to Bring 45nm Products in Early Q4 2008 dle noveho sefa Dirka Meyera.
http://www.xbitlabs.com/news/cpu/dis...5nm_Chips.html

Takze jeste jedno ctvrtleti a je tu Deneb oficialne :ú)

[flanker]

AMD Deneb a spotreba



Phenom 9600 BE 2.3Ghz


Deneb ES 2.3Ghz


spotreba sestavy 9600BE


A spotreba sestavy Deneb ES


A teploty? Phenom: 45 stupnu Deneb: 35 stupnu

http://www.itocp.com/thread-11603-1-1.html

[socket]

Najblbsia otazka: preco je na CPU Deneb napisane AMD (c) 2008 ked to je len die-shrink?? to potom 90nm procaky do 939 mali mat 2004, ale oni maju 2001
Dokonca aj cela K7 mala (c) 1999 aj ked bol barton urobeny az v 2001

[flanker]

AMD Deneb na velmi solidni frekvenci...
http://www.overclock.net/attachments/am ... 4.0-2-.jpg

konecne 4GHz, i kdyz pri 1.475V, chlazeni BOX !!!!

[axloth]

Kdy se maji zacit 45nm prodavat?

[Wolruf]

huste , cekam na deneb jak na smilovani bozi. To by mohla byt nahrada za meho optase

[flanker]

asi tak v rijnu az v listopadu. Ja furt cekam, az se objevi treba Cinebench taky

[flanker]

A mame tu dalsi benchmarky procesoru Deneb, nyni na zakladni frekvenci 2.3GHz a nasledne pretaktovane na 3.25GHz. CPU-Z nejspis zase ukazuje vyssi napeti, nez je skutecne.

http://www.itocp.com/html/amd45nm/cpuz.jpg

testování bylo na sestave nize a s take nize uvedenymi testy:

http://www.itocp.com/html/amd45nm/pingtai.gif

http://www.itocp.com/html/amd45nm/autodaiji.jpg

http://www.itocp.com/html/amd45nm/IMG_2755.JPG
http://www.itocp.com/html/amd45nm/IMG_2766.JPG
http://www.itocp.com/html/amd45nm/automanzais.jpg

po 10minutach Orthosu stale nizka teplota (teplota okoli byla 26 stupnu)
http://www.itocp.com/html/amd45nm/IMG_2769.JPG

Pretaktovany stav:
http://www.itocp.com/html/amd45nm/ocdaiji.jpg
a teplota po Orthosu po 10minutach
http://www.itocp.com/html/amd45nm/IMG_2772.JPG

tentokrat superpi "zadny zazrak"
http://www.itocp.com/html/amd45nm/superpis.jpg (i kdyz je to stale o 1.5s lepsi nez napr x3 8650 na stejne frekvenci, ovsem s nize taktovanymi pametmi!)

Everest pameti:
http://www.itocp.com/html/amd45nm/everest.jpg

Sisoft Sandra, tam je solidni vysledek:
http://www.itocp.com/html/amd45nm/siss.jpg

winrar uvadet nebudu, nebot nepouzil k testu MT...pro info vysledek 518 a 563KB/s

Fritz chess benchmark-znovu spis nevalny vysledek vzhledem k soucasnym Phenomum (prakticky stejny, presneji uvedeno dokonce nizsi! )
http://www.itocp.com/html/amd45nm/frz.jpg

Cinebench R10, na ktery jsem byl osobne natesen mi prinasi zklamani...Znovu horsi nez soucasny X4 9650..


A jako posledni 3D mark Vantage
http://www.itocp.com/html/amd45nm/3dmarks.jpg

Zaverem jsem se docetl, ze soucasne bezne pouzivane desky, maji s temito CPU problemy (nejspis BIOS a jeho stare verze) a udajne to ma vliv na vykon (proto takove rozporuplne vysledky?). Problem byl i s otevrenym nasobicem, pry nesel nastavit na vyssi hodnotu a tak zbytek byl limit FSB...Kladny poznatek uzivatele jsou prave nizke teploty,ktere se pry pri testovani neprehoupli v OC stavu nad 50 stupnu a udajne AMD na to v prezentacich bude poukazovat. Pry i Cinebench neprobihal korektne jak by mel, takze vykonove testy nam moc dnes nerekly

[yuri.cs]

vykonove testy nam moc dnes nerekly

bude to zrejme chtit pockat na nove verze biosu.

[flanker]

Tak údajně se dočkáme i vylepšení HT u příchozích procesorů K10.5 (Deneb)
A to HT3.1
http://www.techpowerup.com/68933/New_Hy ... s_get.html

New HyperTransport specification reveals updates for AMD's 45 nm CPUs, Fusion's secret

Chicago (IL) - As the release of AMD's 45 nm processors Shanghai and, more importantly, Deneb is drawing closer, we are getting a better idea of what to expect from AMD’s next generation of CPUs. The HyperTransport Consortium just released details of the HyperTransport 3.1 specification, which increases the clock speed from 2.6 to 3.2 GHz (6.4 GTransfers/s).

Since the HyperTransport (HT) bus is 32-bit wide, the total aggregate bandwidth grows to a massive 51.6 GB/s (25.8 GB/s in both directions), which is a 10 GB/s improvement over the original 3.0 standard and a powerful tool to improve the performance of HT components such as chipsets and CPUs – including the upcoming Fusion processor.

It is widely expected that motherboard vendors will have no trouble adopting 3.1 support in 3.0-enabled chipsets. AMD’s existing 790FX and GX chipsets should support native bandwidth of HT 3.1 as soon as you pair them with a 45 nm CPU.

The HyperTransport Consortium also revealed the HTX3 specification, lifting the potential aggregate bandwidth of HTX cards to the level of HT3.1: 51.6 GB/s is an impressive number for an extension connector and we are hearing that the standard is already getting some attention in the field of enterprise NAS-boxes with SSD technology.

Possibly most interesting, HT3.1 and HTX3 clear up some of the questions surrounding key features of AMD’s Fusion processor. Without doubt, HyperTransport 3.1 will be used as a communication interface between CPU and GPU and a bandwidth of 51.6 GB/s may open a whole new world of possibilities and an opportunity to be more competitive with Intel in terms of overall performance.

HyperTransport 3.1 will also be included with every Shanghai processor. According to our sources however, the technology will disabled and only HyperTransport 1.1 and 2.0 (up to 22.4 GB/s) will be supported.

[flanker]

Dnes se to novinkama jen hemží .Sotva jsem nekolik jich postnul, je tu dalsi a to ohledne AMD serverove oblasti.
Obrazky mluvi za vse:


nyní by se mělo jednat o aktuální a platnou roadmapu!


A teď úplná novinka, tzv chipset Firrano, založená na AMD890 (jak vidíte, již existuje ) Mimo jiné bude podporovat 4-channel DDR3. Jedná se asi o platformu, kam osadíme procesory 2xSao Paulo resp Magny Cours


//edit:je to tak, ověřeno, jedná se o socket F (1207), bude tu cca za rok

[Tom1122]

(Snad) k 1.9. uvede AMD nové BE procesory. Phenom X4 9950 BE 125W za $186, X3 8750BE 95W za $139,
Athlon X2 6500 BE 95W - $105.
Dále zlevňuje tyto zhruba o 20%

Product............................ Old price ($)...... New price
Phenom X4 9950 BE 140W..... 235.................. 186
Phenom X4 9850 BE................. 194.................. 180
Phenom X4 9750..................... 215.................. 170
Phenom X4 9650..................... 195.................. 156
Phenom X3 8750 .....................175.................. 134
Phenom X3 8650..................... 145.................. 123
Athlon X2 6000+..................... 112 .................. 95
Athlon X2 5600+..................... 102................... 88
Athlon X2 5400+...................... 87.................... 78
Athlon X2 5200+...................... 76.................... 68

Zdroj:http://channel.hexus.net/content/item.php?item=15174
http://www.ddworld.cz/procesory-cpu/amd ... -20-2.html