Soldato
I'll be undervolting that bad boy day 1. Hopefully can lose at least 100w with little performance loss.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
NVIDIA 5000 SERIES
- Thread starter scubes
- Start date
I'll be undervolting that bad boy day 1. Hopefully can lose at least 100w with little performance loss.
Associate
Retailers, AIBs, and System Integrators warn of GeForce RTX 50 series shortages and delays - VideoCardz.com
GeForce RTX 50 in short supply NVIDIA has a problem with GPU shipments. NVIDIA made several mistakes during the announcement of the RTX 50 series. Board partners were unable to prepare stock before the end of the year because the company failed to provide the final BIOS on time. Now, it seems...
videocardz.com
Some additional sources on the stock situation. Effectively zilch.
Just wait for the 5090Ti/Titan, it's rumoured to be 800W!
Soldato
Yeah, that just seems mental.
I think 250w - 350w GPU, with at least 16GB, and around £800, is the sweet spot for me. But honestly thinking of waiting it out, to see what the landscape is like in a few months, after all the launches. Not got much faith in AMD to make anything desirable either. Guess have to wait and see how it all plays out.
All RTX5090 3D marks scores leak out from reviewer
All scores in link below, but as a summary
* In pure Rasterization tests like Timespy and Firestrike, the 5090 is 35% faster than 4090
* In pure Ray Tracing tests, the 5090 is 39% faster than 4090
* In mixed raster+RT tests, the 5090 is 40% faster than 4090
videocardz.com
Keep in mind 3d marks scores in general tend to be a bit better than games due to superior optimisation, so the difference between the cards in actual games may be smaller
All scores in link below, but as a summary
* In pure Rasterization tests like Timespy and Firestrike, the 5090 is 35% faster than 4090
* In pure Ray Tracing tests, the 5090 is 39% faster than 4090
* In mixed raster+RT tests, the 5090 is 40% faster than 4090
NVIDIA GeForce RTX 5090 3DMark performance leaks out - VideoCardz.com
NVIDIA RTX 5090 in 3DMark Straight from friendly reviewers. In just two days, NVIDIA will lift the embargo on the RTX 5090 Founders Edition and the so-called MSRP cards. This marks the first review coverage of the Blackwell architecture and offers an initial, although somewhat expensive, look...
videocardz.com
Keep in mind 3d marks scores in general tend to be a bit better than games due to superior optimisation, so the difference between the cards in actual games may be smaller
Last edited:
but then another poster is saying the 4090 goes up to 600w as a good thing.
It doesn’t work like that with synthetic benchmarks. A 4090 in those graphs shows about 40% to 50% faster than a 4080 Super. Yet in real gaming tests it’s more like 20% - 30%.
So a 5090 will not be literally twice as fast as a 4080 Super.
Last edited:
The timespy extreme score is pretty impressive by itself…. over 25k!! Phwoar.
It does a cpu test as well so you can’t scale it directly to your own hardware, but for reference I get around 10k with a 3090 and got about 19k with a 4090 when I had one.
It does a cpu test as well so you can’t scale it directly to your own hardware, but for reference I get around 10k with a 3090 and got about 19k with a 4090 when I had one.
At least the rest of your kids will be safe for a couple of years!
That’s right online with the 4090 results in those graphs. So a 5090 is 30ish percent faster than a 4090.
Given the 25% increase on MSRP, not exactly amazing. Still you want the absolute fastest and don’t care about value, the 5090 is the card to get.
If money so no object and epeen is the aim, you may as well splurge the extra £500 for the few percent increase on the extreme OC models… right?
Last edited:
Seems to be 24-25% in timespy according to those ‘leaks’, assuming they used the same hardware.
3Dmark has the core clocks listed for each result.
600W can do 3-3.1+ ghz whereas 450W is about 2.4-2.8 depending on the bench.
Yeah, but 5% difference is just a mild overclock
I’m not sure of the point you’re making - we don’t know anything about whether these scores were using overclocked HW or not. A different cpu will also alter the score because of the multi core element of the test.
So it’s all just indicative. Not long until we have a better idea thankfully!
No issue / worries!
Soldato
Looks like paper launch if those stock rumors are true, a few FE for the msrp of £1939 then AIB going near £3k as they are barely in stock could be on the cards.
I wonder if Nvidia stipulated to partners that they can only have 1 x 12V-2x6 as don't think anything shown had more than that, which will really limit OC.
I wonder if Nvidia stipulated to partners that they can only have 1 x 12V-2x6 as don't think anything shown had more than that, which will really limit OC.
so for the results videocardz showed in their article what were they using, I can't tell.