i got two trio on air and later water. both running no issue since 1st it was sold in MY.

u sure its the card??

I also wonder why Nvidia need to increase it's new card performance so much when last time it usually only increase about 20% and also maintain same price when usually they will increase price every new gen. They are already the market leader after all. Maybe they know RDNA2 will be powerful enough to easily beat RTX 2080ti hence this drastic measures. One indication of RDNA2 power will be the PS5 GPU can run up to 2.23GHz speed, this is in a console which have poor thermal dissipation compare to PC. AMD also claim RDNA2 will support ray tracing, so whether it's implementation will be same/worse/better than Nvidia still unknown.

Really curious what is AMD answer to RTX 3000 series, will it be better performance or lower performance but better pricing.

MY doesnt work like dat.

ok some more info alot ppl dont know. last 15 years before turing.. MY distro/resellers will F dat nda .. and we get it on our hands.

turing.. nvidia asia pac step in and control gpu allocation to the country. all brands had to follow. and the Fed MY big time with few giga models and some asus dual models etc..
no FE cards

i want FE card or strix. its da only two i want. dis 3080 so overpowered even throttling da card is fine for me. resellers idealtech u can forget it. their internal customer builds demand so high all aib cannot even fullfill them. wont be suprised idealtech becomes a distro for them self .. question is which brand.. so they will tahpau alot of the first few batch.

i just hope they can produce alot. fed up with this wait shit. running to da shop. etc

Hey guys, before buying a PSU for my upcoming build with an RTX 3080, would a Corsair RM 750 suffice or are there better options at that price point?

iphohps3 Send me the link to the luxxx discussion if you need to know something specific that translation doesn't explain well enough. Couldn't find what you were referring to.

To my noob understanding what he's getting at is the following:
Turing:
FP32 ALU or RT, one at a time
- OR -
FP32 + Tensor

Ampere: All 3 at the same time, bringing down the frame times massively.

So the twitter guy thinks that because L2 cache remained 6MB (2080ti vs 3090) there might be a bottleneck.

--

But you can't forget now the cores themself also do more instructions per cycle and RT cores can even to motion blur which Turing couldn't.

So at the end the datapath doubled, but with that L1/Texture bandwidth also did. So personally I think this is not as important for the gaming cards.

The only thing which will really slow down these cards is when (only/mostly) Integer instructions are being used. Then there's not much benefit over Turing, since technically a 3080 still only has 4352 Cuda Cores, just like a 2080ti. Just that the ti couldn't use the INT32 ALU's for FP32 calculations. But again, ... you're gaming.

As i said asus tuf 3080 be insane. Also dat video i dunno what to say
Asusmy
Z490 asus Fed malaysians
B550 asus Fed malaysians
2080ti during mco 7k+ fed us good

Sg/thai cheaper closer to usd mrsp.

Now dis. U turn with pr video.. they nvr did this

If its a pricing continuation on all their product in future.. by all means.. thank you. But hmmmm..

people and their prioritize

went to pick up da card
was intro to a dude building this
10900kf
strix aio 360  white
strix gaming E
zotac trinity 3080
strix psu
lianli 011 razer

asked da dude wth take strix mobo.. hero top up a bit can do.... no moolah..

i looked at the da rig was like.. better take gaygay master z490 or ace or sell ass right for rm300

fed psu?

L

That's such a weird timing.

If it's that soon, wonder why they dont just release 20GB as default model.