Only if your not fast enuff or greedy enuff. Remember the Rules of Acquisition...

xx80series fans rejoice! Nvidia has learnt their lesson from 4080 and now you can upgrade to 5080 at the lower MSRP of $1000 (unlike the $1200 4080).

But what you should really be aiming for is the sweet 5090. Bcoz for Cainis, 5080 is bad luck.

Its confirmed new, 12V 2x6. Nvidia & partners has transition to it since the last batches of 4090 & 4080 (4070series too?)

The CES graph charts are kinda misleading, when they compare 5090 with DLSS4 vs 4090 with DLSS3 so yeah, any FPS claims should be taken with lotsa salt until Orange to Orange comparison review are out.

With the pricing structure and hardware gap between 5090 & 5080, its clear that Nvidia is trying to discourage gamers from buying 5090 purely for gaming despite all the marketing and naming such as a gamer GPU and yes its the fastest in the planet but its really actually meant for application use (if your too poor to own a Quadro).

It harkens back the days of Titan GPU lineage down to xx90s where its not really for gaming but due to its raw processing power, games do get a boosts thus rich gamers/influencers could flex their Titans/xx90s and this caused everyone to want "the best gaming GPU existed... Titans/xx90s". And thats what this is about.
FLEX.

Sure boh. You cant even get 4080 at that low 5080 prices wei. Betcha is just a direct convert from USD to RM. Havent factor in distro street prices.

Just cant help it

Want Galax non boring design? Wait their overtop overkill HOF version.

Given fake frames trajectory, now we are doing 9x fakes to 1 true frame, by DLSS6 we might have 30 fakes to 1 true. Do you want your gameplay turn into 30FPS movie instead?

Most highly unlikely unless someone parallel import. Our market isnt big enuff for Nvidia to want to kill their AIB partners.

It depends on resolution actually. At 2K is less affected by CPU performance, at 4K more often than not its GPU limited rather than CPU.

You can refer to perf difference at various reso in comparison

Say 4080 vs 4090 (everything else same). TPU relative perf:

In 1080 - difference 5%
In 2K - diff 11%
In 4K - diff 25%

Too low CPU does bottleneck at 4K but your more likely to run out of perf with a limited GPU (or likely to get a bigger boost with a higher GPU tier upgrade). 12400 isnt something I would pair for 4K builds. At least a 12Gen i7 Ksku would not cause CPU bottleneck.

Relative to inflation costs. Apparently USD has tumbled over the years in value and some say xx80 class GPU once adjusted to inflation is cost about the same as your 2080TI back then.

I think at most will drop rm 50-100. Our resellers very resilient to price drops, it will likely maintain this level until just before swap for 5000series on the shelfs. Not sure what they do with balance 4000series stocks but for sure no fire sales. Maybe they return back to distros for full refund or something.

Then it goes back to the distros what they do with excess 4000series cards once 5000series comes, for sure there will be no fire sales just like never happen for 3000 when 4000 came etc. This unlike USA where you could still find RX6700/6750XT at really good prices. Here GPU just disappears at the high price.

Deswai Msian market sucks. Ady come here distro markup kaw2, then when got global discounts from the brands we dont get too, and then when phasing out the older stocks there is no clearance discounts. So expect to continue pay the super higher prices that distros set until next gen comes next.

Laptop "4070" is not the same as desktop 4070.

Amazon, Newegg, any Microcenter if your nearby one but need to pickup in store.

Oh but the GPU is real tho...

Its a functioning GPU so its as real as "not fake frames or AI" . And I never did say kopite was wrong, just that such a GPU did exist.

Oh yeah, LM is going on a commercial scale now. Will we see AIBS or even CPU makers switching to LM as OEM TIM moving forward?

Its oddly weird that Nvidia, that had primarily been a chip designer (TSMC makes the silicon), could come out with such innovative GPU design and coolers since the last 2-3 FE generations. It puts to shame all those traditional AIBs ie Asus, MSI, etc what is supposed to be their bread & butter.

Nvidia showed us all their 3-4 slot giganto PCIE busting cards that need hanger support doesnt have to be so huge when a higher heatload (575W) could be contained within a 2 slot cooler.

How this could happen? Innovation.

These AIBs with their excessive profits clearly did not put all that money back to R&D instead they lazily pile on more metal making the GPU unnecessarily too large. The only "innovation" you see from them is how much RGB they can cram into the tacky plasticky shroud. Would they have thought to split the traditional PCB into 3 parts? Not in a million years.

Nvidia likely had to work with these OEM AIB to produce their FE cards and yet such innovation didnt come from them. Once again, Nvidia had to push the market to create this from practically scratch as no one had done this before.

Isnt Toslink fibre optic? How can it fry your soundcard?

Get a USB external DAC

2 words: planned obsolescence. It prolly popped bcoz it reached its designed life. Which is usually after your warranty period

Tbh, this is just my suspect but audiocards have lotsa caps and these are sensitive to high temps. Certain custom WC loop system designs could lead to higher temps within the case which may have affected these caps leading to shortening their life. (Im basing this theory on caps inside PSU which is why more durable Jap caps are expensive and usually found inside higher end PSU).

This post has been edited by babylon52281: Jan 18 2025, 11:50 PM

7800x3d not much perceptible difference vs 9800x3d, why not go for it? Dotatech has combo with B650 mobo for RM 3k.