Prednosti grafičnih kartic AMD vs NVIDIA

Gigybyte 7950 Boost iz škatle dela na 1.25V
Non boost pa na 1.05V.

---

Jaz sem pred leti kupil Asus DirectCU II 670 za dobrih 400EUR.

Kolega je približno v istem času kupil Gigabyte 7950 WF3.

Ker mu je 7950 WF3 tako tulila (3 venti), sem dal idejo, če bi za nekaj časa zamenjala grafe. In sva.

Njemu je bila 670 zelo všeč, ker je bila tako tiha. Jaz sem takrat minal *coine. In mi je bila Gigabyte 7950 všeč, ker ima dual bios in če jo flashah z non-boost BIOSom, postane vsaj za polovico tišja. (OC se pa sploh ne spremeni kaj preveč.) Original je imela 1.250V, non boost pa 1.050V. Kar je skoraj dva volta manj.

In ker še do sedaj nisem slišal od kolega, da bi rad svojo Gigabyte 7950 nazaj, imam 7950 že dolgo časa. Od takrat smo dobili 3 driverje, ki so konkretno povišali performanse. Prvi takšen driver update je bil še preden so javno izjavili, da imajo probleme z framepacingom. Drugi je bil takrat, ko so izdali driverje, kjer naj bi popravili (omilili) težave z fracepacingom. Tretji driver update, ki je prinesel ene 5% več performansa, je bil pa nekaj izdaj pred Crimson izdajo. (Crimson za 7950 ni prinesel omembe vredno pohitritev, mogoče kakšen odstotek, statistična motnja.)

Če malo računam, so od izdaje "non boost" 7950, do sedaj pohitrili kartico za ene 25%. OC iz 900/1250 na 1100/1300 prinese 10%. Vseskupaj se mi zdi, da je od izdaje 7950, performans z driver update-ti narasel okoli 15%. Tako, da driver update-ti + OC = ~20%.

Ko sem kupil 670, je bila le-ta hitrejša od 7950. Sedaj, z najnovejšimi driverji tudi za Nvidio, očitno ni.

(Upam le, da kolega misli še naprej obdržati 670, da jaz lahko uporabljam 7950, dokler ne kupim novejše arhitekture.)

moja GB wf3 GTX 780 OC ne parira undervoltani in OC powercolor 7950 v dirt rally po banchmarku...razlika v ceni mislim pa, da je vsem jasna

za gcn rabis vbe7

http://gamegpu.com/action-/-fps-/-tps/h...
http://pclab.pl/art68473-7.html
...

uf, R9 380X se dobi za 230EUR. Pa še 4Gb ima.

D3m je 20. feb 2016 ob 21:19 izjavil:

???

Joke je bil na tele benchmarke @ GTX 970 vs R9 380X.
R9 380X je za 10-15% počasnejša od GTX 970 pa je 100€ manj.

Jst je 20. feb 2016 ob 22:18 izjavil:

Trenutno 7950 JE hitrejša od 670.

Sicer se ne spomnim dobro, ampak ko sem 7950 spustil skozi 3DMark, je dobila okrog 1500 točk več, kot na začetku, ob menjavi.

Jap, konzole so pripomorle k temu "ZELO".

Jst je 20. feb 2016 ob 22:18 izjavil:

Trenutno 7950 JE hitrejša od 670.

Sicer se ne spomnim dobro, ampak ko sem 7950 spustil skozi 3DMark, je dobila okrog 1500 točk več, kot na začetku, ob menjavi.

Based on se ne spomnim dobro?

D3m je 21. feb 2016 ob 08:39 izjavil:

Ajd Predator za 225€ jo pa res dobiš.

http://www.mindfactory.de/product_info....

kdo bi pa zdej tole kupi? Pozabil si da ni semipasive.
http://www.mindfactory.de/product_info....
229€

Tole je pa že kr cena.

Senitel je 21. feb 2016 ob 07:24 izjavil:

Jst je 20. feb 2016 ob 22:18 izjavil:

Trenutno 7950 JE hitrejša od 670.

Sicer se ne spomnim dobro, ampak ko sem 7950 spustil skozi 3DMark, je dobila okrog 1500 točk več, kot na začetku, ob menjavi.

Based on se ne spomnim dobro?

benchmarkih.

A je 380X že eno leto zunaj?

Pričakujem, da bo ceneje konec avgusta tam cca. 190 eur.

Zanimivo branje: http://www.computerbase.de/2016-02/graf...

Milisekunden = milisekunda
4gb gddr5 radeon != 4gb gddr5 nvidia

malo pomoci

Kakor za koga, sam je treba znat raz_brat slike, ce ne ves _brat tekst

@Predator





Neverjetno a?

https://m.youtube.com/watch?v=fq6GyUzyu...

1. igra je beta

2. igra je gaming evolved, tako da, če ne bi delala dobro na amd karticah bi bilo pa že žalostno...

3. ne vemo na kakšni frekvenci delujeta kartici v videu, ker stock 970 vs. stock 390 ni niti pravilno primerjat, ker je 390 že overclockana kartica.

Alien123 je 23. feb 2016 ob 16:20 izjavil:

brez driverjev 390ka poje 970ko.. resda so framedropi, sam to je kriv špil.. Beta? čez 2 tedna prije špil ven..
ne ker je gaming evolved, ker ni gameworksa, zato..

Bwahaha sanja svinja o koruzi...za frame drope pa se tako ve kako je pri AMD...ni nič špil kriv ampak njihovi "fini" driveri

So pa kr poštimal CPU optimizacijo pr hitmanu ...

https://www.amd4u.com/amdgamepromo/

Eto prvi ASYNC benchmarki
etola več benchmarkov
http://www.extremetech.com/gaming/22356...
http://www.pcgameshardware.de/Ashes-of-...
http://www.computerbase.de/2016-02/ashe...
http://www.nordichardware.se/Grafikkort...

Jap NVIDIA fura tud async shaderje... si imel prov Senitel.

D3m je 24. feb 2016 ob 22:11 izjavil:

Ne bit no nesramen.

Sej po eni strani ma prov. NVIDIA lahko fura Async shaderje, ampak nima od tega koristi in zgodba je precej enostavna... ni AMD dal kr tko notr ACE.

Je požrešen špil, ampak vse ima svojo ceno.

No, da nebom samo govoril v prazno.
Res, da so tako rečeni "hacki", ki bi omogčil furat NVIDIAji async shaderje.
http://ext3h.makegames.de/DX12_Compute....

The following chapter is based only on vague assumptions about the hardware platforms. Don't take these for given, and read carefully on which assumptions these are based on. These features might not be implemented at all, or may even be plain impossible to provide. There might also be other technical difficulties which prevent this from happening.

Some of the weaknesses regarding Nvidias platform can be avoided using additional resources in the driver. This is especially true for the second Maxwell generation which has some yet unused additional features.

Using the grid management unit to boost the 3D engine.
Since GK110, Nvidias card are equipped with something called a "Grid Management Unit". This unit is comparable to the "Asynchronous Compute Engines" found on GCN cards, with a few differences. Unlike the ACE, the GMU has no own kernel execution slots but shares them with the 3D engine. The GMU has 32 truly async compute queues, but it is incompatible with DX12 for unknown reasons12.
Prior to the second Maxwell generation, the GMU could only access all execution slots if the 3D engine was inactive. Otherwise it was limited to a single execution slot. This limitation was supposed to be lifted on GM2xx, so that the GMU could access all slots not in active use by the graphics command processor.
In order to avoid the low concurrency between compute commands in the 3D engine, it would be possible to extract compute commands from the command lists and offload them to the GMU. This only works for sets of compute commands not separated by any split barriers, regular barriers or fences. It's only worth the effort if these sets contain more than a single compute command.
Doing so does not come free of cost. The analysis and extraction of compute commands from a command list causes additional work upfront. Additional effort is required to ensure synchronisation between the GMU and the 3D engine. This requires additional signals and fences13 to be inserted into the 3D queue, which are then resolved on the CPU side as the synchronisation capabilties of the GMU are limited.
This has nothing to do with async compute, despite the GMU being used. It is merely a method for offloading compute commands from the main queue to avoid the hardware limitations. The GMU is explicitly synchronized with the 3D engine by the newly introduced signals and fences.
For game engines making heavy use of compute commands in the 3D queue, this can help getting the runtime behaviour of the 3D engine closer to AMDs equivalent. It's not a generic optimization though, as the increased CPU load can backfire. There is also still no guarantee that compute commands and draw calls will run on the same SMM.
Use the grid management unit for barrier free command lists.
This is based on the same assumptions as the previous hack.
Even though command lists submitted to a compute queue may include resource barriers, they don't necessarily do. If they don't include any barriers, it should be safe to schedule to the GMU instead. This avoids deactivating the 3D engine and allows for true async and concurrent execution.
Similar to the previous approach, this one also comes at an slightly increased CPU load for synchronisation. It is also not applicable to every application, as barriers can't necessarily be avoided.

Že spet pa je vprašanje ali se bo to res obrestovalo?

Sej od malce navite oziroma R9 280X 800/1000core boost je pa GTX 780TI/GTX 970 samo 10% hitrejša.

Čakamo na odgovor nvidie.

AapocalypseE je 23. feb 2016 ob 16:38 izjavil:

Pol pa kar zbriši driverje in igraj igre

Kaj ti bo dober HW, če je SW za ku**c...

Lahko maš 1000HP+ avto, ko brez pravega driverja ne bo dosegal peak performance

Popravljeno :)

Najpomembnejše je pa to kolk hitro teče en DX12 tech demo na raznem hardware-u ane.