• 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 4000 Series

There would also be no speedy advances in gaming technology if nobody bought the expensive stuff too. Where do you think the money comes from for these companies to spend R&D time! Budget model sales aren't going to cover that sadly. This applies to most things. If nobody buys the flagship model then the trickle down effect of near-future advances is less obvious.

It's like cars, if nobody buys the high specced models, then all that's left for the used market are crap spec used options down the line.
fair :)
 
Tbh it took me a month of RMA's to get my inaudible whine card and it can hit 3000mhz so I'm really in no rush to replace it.
Next upgrade will (probably) be the next 3D CPU.

Was lucky enough to hit the jackpot on my first purchase with the 4090 FE. Even more so when i waterblocked it, what little coil whine i had completely disappeared! I'm able to hit max oc on the memory plus 3045mhz on the core, i call that a win even though i spend most of the time undervolted lol.
 
Was lucky enough to hit the jackpot on my first purchase with the 4090 FE. Even more so when i waterblocked it, what little coil whine i had completely disappeared! I'm able to hit max oc on the memory plus 3045mhz on the core, i call that a win even though i spend most of the time undervolted lol.

Yeah I hear you. Forgot I run mine at like 80%. TBh, could probs put down to 10% for some of the games I play lol.
 
Yeah I hear you. Forgot I run mine at like 80%. TBh, could probs put down to 10% for some of the games I play lol.

Yeah easily lol, i enjoy old games a lot more than the newer triple A stuff tbh. I mean playing something like Stronghold uses about 10% gpu power if that :cry:.
 
The 5080 would then stay at $1200 offering 4090 performance so those already on a 4080 and wanting to upgrade would end up having paid 2400 instead of just going for a 4090 in the first place.

You don't really spend $2400 as you'll sell 4080 and recoup some of the money.
Going on the path of two cards, you'll also get probable improvements and tech (ie. 3xxx vs 4000 series was SER, DLSS (FG), vRAM), perhaps even less power consumption (which equals less heat!).

But it's all just a bubble!!!!!!!!

It is very scary just how quick AI has improved in the space of a year and the developments seem to be speeding up even more now, scary times ahead for the real world tbh.

Anything worthwhile? I haven't seen anything really "wow" besides was already shown and ChatGPT v4 and now v5, while giving somewhat better responses than v3.5, is still in its infancy.

There would also be no speedy advances in gaming technology if nobody bought the expensive stuff too. Where do you think the money comes from for these companies to spend R&D time! Budget model sales aren't going to cover that sadly. This applies to most things. If nobody buys the flagship model then the trickle down effect of near-future advances is less obvious.

It's like cars, if nobody buys the high specced models, then all that's left for the used market are crap spec used options down the line.

Advances in gaming tech are usually on the pace of a console's generation. Leaving outside nVIDIA's push for RT/AI (which is only truly seen in CB77 within the new games), what else is there? Ah, the SQ42/Star Citizen engine. The rest goes as consoles go. No high end stuff? Well, just better optimize your "trash".

The high end being overpriced isn't such an issue with things like cars,because a Fiat Punto is not priced based on a Ferrari(even though it is the same company).

The issue is how the entry level and mainstream dGPUs are being priced relative to the high end.

A Fiat Punto will also run about the same as day one, no matter the road. The use scenario remains the same. These examples don't really translate properly into tech world.
 
Anything worthwhile? I haven't seen anything really "wow" besides was already shown and ChatGPT v4 and now v5, while giving somewhat better responses than v3.5, is still in its infancy.

Well in work, we're all using copilot (github and for microsoft azure), duet AI (GCPs ai) and so far, it's very impressive for technical assistance, it has come along way in terms of being more precise and in some of our products, we're using open ai and Googles vertex ai for chat like functionality, basically things like being able to summarise documents and picking out key information rather than trolling through thousands of pages. Seeing various companies wanting to adapt it in their products especially where first line/entry level kind of roles are concerned. A lot of the drive around wanting to adapt the tech is being centered around making existing processes "more efficient and making people more productive" thus "cutting costs".

In terms of gaming? So far, we still only really have nvidias ray reconstruction, DLSS and the npc narrative/facial expressions. I imagine we'll see some more gaming use cases come nvidias announcement though.
 
The vast majority of games with ray tracing for starters, some even come with RT enabled by default with no way to turn it off. The consoles have a cut down version of said tech whilst the PC gets the full whack as documented many times by the usual big name tech review outlets.

Yeah path tracing is limited to just two games really right now, but as a showcase of what is possible when combining all of these technologies to deliver graphics that no console can dream of matching, but ultimately you need to have an RTX PC to be able to get decent fps out of, yeah that I can see.

Edit*
Yes 2024 is going to be the year of AI infused gaming. Just wait and see what comes out of the generative AI stuff in relation to NPC interactions. Nexusmods will probably need a new section just for AI mods for existing games. Maybe Starfield will actually get some decent dialogue finally :p
 
Last edited:
You don't really spend $2400 as you'll sell 4080 and recoup some of the money.
Going on the path of two cards, you'll also get probable improvements and tech (ie. 3xxx vs 4000 series was SER, DLSS (FG), vRAM), perhaps even less power consumption (which equals less heat!).
That’s what Nvidia will get though so it works for them.

You’ll probably be able to get 400 quid back if the 5070 is £600 and matches the 4080 with better efficiency.
Anything worthwhile? I haven't seen anything really "wow" besides was already shown and ChatGPT v4 and now v5, while giving somewhat better responses than v3.5, is still in its infancy.
Apart from Nvidia none else has made any money out of AI yet, open AI are losing $700000 a day and they are supposed to be one of the successes, they would have probably gone bust already without Microsoft’s backing.
 
Apart from Nvidia none else has made any money out of AI yet, open AI are losing $700000 a day and they are supposed to be one of the successes, they would have probably gone bust already without Microsoft’s backing.
This is complete nonsense, there are plenty of companies who are making money from providing products with AI functionality or/and more importantly being able to reduce operating costs thus save money.
 
Last edited:
And they are?

Google is your friend, it's more about understanding what AI is seeking to resolve and more importantly why companies are wanting to use it in their workflows and/or products.

Why do you think so many companies are cutting jobs especially in the tech sector? I'll give you a clue, it begins with A.... By directly implementing a more efficient process of carrying out tasks, said companies are able to cut staff. There is even rumour that google are going to lay of 30k people soon because of AI:


You forget that a lot of companies have partnerships with GCP, Microsoft, Open AI and so on which provides discounts to such companies and so operating costs are not anywhere near what Open AI costs are especially since the traffic for a lot of products will not be anywhere near what chatgpt is.

One of our internal products sold to other areas of the business is making money because of the AI functionality provided.

The biggest money maker is actually from people doing blogs/articles on niche topics (because AI is able to use all the right keywords for SEO and getting the hits)

There is a lot more to Open AI than just chatgpt too i.e. getting the api used in products.... They are in the phase of growth right now and probably won't get any profit for a long time. Just look at uber, first time it made actual profit was just last year in summer there.
 
I do wonder whether companies worldwide are thinking things through. They could be using AI to expand their businesses and improving the working conditions of their workers,so people have less need to slave away at their job. Instead they are more interested in short term overhead reductions,by cutting their workforces even more. The issue is if you end making large amounts of the workforce redundant,where is the magic money tree going to come from? There is only so much debt and magic money governments can create before the system turns on itself.
 
Last edited:
Common advice used to be to spend ~ 30-40% of system cost on a GPU so I wondered what that looked like today.
Taking this the GPU costs 40% of the total tower cost as this is a gaming PC i.e. no peripherals or monitor.

Aimed for generic pricing, not the absolute cheapest of everything, but somewhat representative of that Tier.

Conclusion, everything is expensive now.... just that it's really obvious on GPU's as it's a single item.

For a new build basic
7600X £230
B650 £180
32GB DDR5 £ 100
Case £80
2TB M2 £120
650W branded PSU £80
AIO £80
Fans £30

Total £900 + GPU £600 @ 40% ~ RTX4070, RX7800XT
=£1500

7800X3D £380
B650 £250
64GB DDR5 £ 200
Case £100
4TB M2 £220
850W branded modular PSU £140
AIO £100
Fans £30

Total £1420 + GPU £950 @ 40% ~ 7900XT/XTX , 4070 TI/4080


7950X3D £600
X670E £400
64GB DDR5 £300
Case £150
4TB M2 £300
1000W branded modular PSU £160
AIO £150
Fans £30

Total £2090 + GPU £1400 @ 40% ~ 4090
 
I do wonder whether companies worldwide are thinking things through. They could be using AI to expand their businesses and improving the working conditions of their workers,so people have need to slave away at their job. Instead they are more interested in short term overhead reductions,by cutting their workforces even more. The issue is if you end making large amounts of the workforce redundant,where is the magic money tree going to come from? There is only so much debt and magic money governments can create before the system turns on itself.

Well ultimately the biggest problem we're going to have is unemployment which in return means no one buying things so economy is going to go down the toilet even more, a lot of people are saying about introducing a universal basic income as it is a very real threat ai to the economy long term but then handing out everyone an income is going to introduce its own problems too. Sadly a lot of companies/leadership are stating things like "AI won't take jobs, just means people can work on more important less boring things" which is a load of **** as evidenced with all the layoffs. The other big problem we're going to face is that AI is enabling lesser skilled/knowledgeable folk to do higher skilled work that they would have never been able to do before, which means salaries could very well stagnate or possibly even see certain job salaries go down particularly in the software development industry since why pay more for someone who can just ask a tool to write the code for them.
 
Back
Top Bottom