From memory, in Australia, companies are not allowed to set a minimum price level. They are allowed to set a recommended retail price, which is only a recommendation and I believe that they can also set a maximum price that is also considered a recommendation. Resellers are allowed to price it at cheaper than recommended retail price, as long as it is not below the cost price.
Of course in reality we know this is difficult to enforce (looking at you, Apple!)
Wholesalers set a suggested retail price. Retail can charge whatever they want and do based on supply/demand and their market strategy.
If 5090 is in short supply expect it to hit 5k
Serious question:
Is anyone else reverting back to native 1440P?
or is the 4K crowd content in getting reamed?
What's the sweet spot for gaming on a 1440p UWQHD display?
For UWQHD I think you should be pretty safe with a 5080. You are unlikely to need more VRAM at that rez and performance wise you should be getting right rule of thumb 4090 ish levels of performance before MFG.
In terms of raw pixels to push is is about 60 percent of that required by 4k.
Ref future proof and potential 4k gaming. If price is no object then get a 5090.
Retail can charge whatever they want and do based on supply/demand and their market strategy.
In practical terms they usually can, but it is permitted under Australian law for suppliers to impose a maximum retail price, above which the goods are not allowed to be sold. This doesn’t seem to happen much in the PC parts industry, but is allowed. Suppliers are not permitted to impose a minimum retail price, although if the retail price is set below cost suppliers are allowed to withhold supply on that basis.
Is anyone else reverting back to native 1440P?
or is the 4K crowd content in getting reamed?
Yep just bought a new 240 hz UWQHD OLED and it is beautiful.
Should hold me until the 7Xxx generation gets released. I will moveto 4k when we are told that 8k is the centre if the universe!!
Is anyone else reverting back to native 1440P?
or is the 4K crowd content in getting reamed
The beauty of having a 1440p (ultrawide) and 4K display :) For games that use mouse and keyboard or are particularly demanding I use the ultrawide and for games that are less taxing and play well with a controller I use the 42” 4K, which sits next to my desk (making Mouse/keyboard usage impractical.)
Is anyone else reverting back to native 1440P?
or is the 4K crowd content in getting reamed?
No need to downgrade if you already own a 4090, more than capable at 4K. If you want to enable Path Tracing in games you need to revert to DLSS Quality, which is worth it for the visuals that Path Tracing provides in my opinion.
How much you pay for your hobbies is relative to your own personal circumstances and the enjoyment/value it provides, there is no right or wrong..... what works for you is all that matters.
Is anyone else reverting back to native 1440P?
Definitely thinking about it but finding it hard to make up my mind between that and just going 4K and sticking it out. It's really that it comes as a package, if I go 4K i'm going to want a 5090 but if I go 1440p I can settle back with a 5080.
I do yearn for a higher pixel density, need to get myself to a shop that has decent models and check out the 27/32" 4K screens and see how they look.
Serious question:
Is anyone else reverting back to native 1440P?
or is the 4K crowd content in getting reamed?
Yeah. Staying 1440p but upgrading to an OLED HDR. I want to enjoy RT so I'll be pairing it with a 5080. We're just not quite there yet for quality 4K gaming without upscaling and other software compromises. I'll consider 4K when the 6080 comes out.
s anyone else reverting back to native 1440P?
What's native 1440p? Is that DLSS quality settings on a 4k screen?
Seriously I haven't used a PC monitor since the GTX 10xxx. hahaha
Still with my old trusty OLED TV for gaming.
Serious question:
Is anyone else reverting back to native 1440P?
or is the 4K crowd content in getting reamed?
I'm already running DLSS with my 3840x1440 monitor. I will have to keep doing this for a while unless I can get my hands on an FE card from overseas.
My 3840x1600 is a nice balance between the two
Me too, and my 4070ti seems to do the job on it pretty well.
I'm already running DLSS with my 3840x1440 monitor. I will have to keep doing this for a while unless I can get my hands on an FE card from overseas
Will be waiting for a 5080TI or 6 series me thinks. Will probably go down the path of the frame Gen mod to hopefully drag the 3080TI till that happens. I did read somewhere that nvidia are considering offering frame Gen to the 3000 series
Is anyone else reverting back to native 1440P?
If the input lag for DLSS 4 isn't too bad, I'll try with 4k and see how it goes.
https://www.reddit.com/r/har
I just came across this interesting Reddit post, analysing Blackwell in a way I haven't really seen discussed much yet: efficiency. It shines a light on how even though a GPU like the 5090 now has a 575W TDP vs the outgoing 4090s 450W, efficiency remains at least the same generationally when looking at generational performance uplift of the x90 class which is important given how much more powerful (in some aspects) Blackwell is compared to Ada, and is impressive considering they've only gone from N5 to N4 (essentially the same node, just with minor tweaks).
It's a bit of a long read, well worth it IMO especially considering it's sourced, but if you want something of a shorter TL;DR:
- Blackwell adds a lot of new functionality that enhances efficiency. This will be most evident in laptop and power-constrained systems, but will still net results for us regular folk as well, especially in idle and very light workloads
- It includes (now also on desktop) previous improvements that Ada Lovelace's Max-Q mobile designs brang, including a tri-speed memory controller and improved SRAM clock gating (allowing SRAM to go into standby mode when the GPU is idle)
- It adds improved clock gating (the entire clock tree can now be disabled, essentially disabling the clock signal for memory controllers and cache if they're idle), improved power gating (Blackwell can now shut off entire parts of the GPU core entirely when they're inactive), power rail changes, improved low-latency sleep (allowing it to enter/exit power states 10x faster than before), improved frequency switching (by a factor of 1000x apparently, also leading to improved boost frequencies under load) and voltage-optimised GDDR7 with ultra-low-voltage states (reducing standby power used by the VRAM a further 20% over Ada)
- When FPS capped or CPU-limited, power draw should be much lower. In lighter games it'll be lower as well, and on the opposite end in compute-heavy and RT tasks a lot of the power savings will boost performance instead.
To borrow from the post:
How the above Impacts Blackwell's Power Draw
1. When the ALUs (GPU core logic) don't need more data from L2 and memory and execute the threads parts of the L2 cache and memory controllers are clock gated saving power. When individual SMs have completed workloads and idle, they are clock gated saving power.
2. With power gating SMs can be turned off completely when workloads don't scale across many SMs and/or saturate SMs very poorly leaving many of them idle for many milliseconds in a row. This helps lower leakage power.
3. Secondary voltage rail will allow for a dynamic and adaptive decoupled voltage frequency curve on a per workload basis which maximizes performance. If some of the GPU logic is idle it helps lower leakage by turning it off 15x faster.
4. Low latency sleep ensures idling SMs can rapidly switch to a lower power states (Low Power 1+2) or deep sleep which saves a lot of power.
5. Accelerated frequency switching makes #4 possible.
6. GDDR7 being more efficient increases the GPU cores power budget and the improved ultra low voltage states allow the memory to use less power when idle and it's also possible that they optimize power draw at lower memory speeds.