Papusan

1.100v and +700W. 600W cap and 33% more cores need really good binned chips. Using same old node as for the new and shiny doesn't make magic for the power consumption. Somewhere it has to be a trade off. And the closest I can see is lower max boost clocks. How much voltage Nvidia will allow for Blackwell chips.... I don't know. With later revisions of 4090(FE) they decreased max voltage from 1.1v down to 1.070. The 12VHPWR meltdown, cross-flashing firmware with custom nvflash software for FE cards, the China restrictions (US ban) and Intel's failure for voltage (Vmin Instability Issue)... Anyhing can pop up from a real software company as Nvidia. Even today Nvidia have the implemented voltage capp for 900 series cards intact in newer drivers (no more allowed voltage tuning with the different AIB OC software for Maxwell). Nvidia can be pretty disgusting. Hmmm. Exactly as Intel. Always right before a new model is out. Is it a coincidence, @Mr. Fox? Incredible timing... Nvidia's January GPU driver update patches multiple vulnerabilities

I expect you get lower max boost clock with more cores for the 5090. 33% faster doesn't necessarily mean 33% faster when max oc'd. We'll see how much headroom more cores will get within the max board power limits. Bin quality this gen etc. Need to see results first. And there is no more HOF cards you can grab their vbios for cross-flashing. We'll see. Not so sure Nvidia will go the same route once again with the new (4090FE with an 600W bios). Depending on how confident they are about the new thin 2 slot cooler design. Remember those FE cards is meant used also in thin and slimy pc boxes. And the best Zotac card max out at 600W, so... A measly 600W power limit can in worst case gimp your maxed (OC) boost clocks if you got an average chips.

I wonder how much performance Nvidia hold back aka rob from older gen cards to up sell the new and shiny. I don think Nvidia have squized out every drop of perfomance (be it raw performance or software based) before next gen is out. NVIDIA DLSS Boss Doesn’t Rule Out Possible Frame Generation Support for RTX 30 Series NVIDIA has used a Supercomputer 24/7 to train DLSS for 6 years straight without a break From 1:12:40

Hmmm. That will be the real price if the proposed tax takes effect for the US. But close to $800 USD extra for the Suprim over MSRP cards (FE) without tax is badly enough. Gone is the days you paid $100 extra for the better custom cards. I'm sure EVGA could do decent deals with Nvidia cards if they continued with their business as we see how the prices has becom for custom cards. Imagine +3000$ 5090 K|ngp|n cards. That would be an reality. And not forget the custom blocks from EVGA. Just aldd $500-750 on top

Take a week off and get in line outside Microcenter. No point start up the car 2.5 hours before the shope opening😁

Hmmm. Maybe Tomshardware have the correct answers? Also Paul Alcorn can't see what Intel see. Is slower an improvement bro @Mr. Fox? Is less more nowadays?😎 Our testing shows that Intel’s fix for its Arrow Lake chips isn’t effective in addressing the chips’ lackluster gaming performance, at least on the motherboards we tested with. And we found that the Core Ultra 9 285K’s updated gaming performance with one motherboard is now slightly slower than before. Intel's Arrow Lake fix doesn't 'fix' overall gaming performance or match the company's bad marketing claims - Core Ultra 200S still trails AMD and previous-gen chips Maybe Intel have tested the wrong CPU? Additionally, the required operating system update has improved gaming performance for the prior-gen Raptor Lake Refresh even more than the Arrow Lake chips, so the flagship Core Ultra 9 285K falls even further behind its predecessor

Thermal Grizzly Releases Duronaut High-Performance Thermal Paste with Exceptional Long-Term Stability Press Release Yesterday, 13:52 Thermal Grizzly's Duronaut thermal paste stands out due to its combination of exceptional long-term stability and excellent thermal conductivity. Its outstanding performance in heat transfer is further enhanced by the ability to apply Duronaut in an extremely thin layer. This is made possible by its unique composition of aluminium microparticles and zinc oxide nanoparticles, optimized for particle shapes and sizes. These features ensure excellent adhesion of the paste to the surface. The powders are mixed into a specially developed silicone oil, resulting in extraordinary long-term stability. For optimal application, the high-end thermal paste comes with the new TG Spatula Pro. The redesigned spatula allows for more pressure to be applied to the surface during application, enabling the paste to be spread more efficiently in a thinner layer. Duronaut is available in 2-gram and 6-gram packaging options.

Hmmm. I wonder what Intel can see but not the others. See above. Not the first time Intel saw totally different results vs the reviewers. Have Intel posted more improvement patches the last 4-5 days @Talon? Intel claims Core Ultra 200 patches improve gaming performance by up to 26% Synthetic benchmarks show even bigger potential gains

$2800 without tax? I will wait and see. With the Norwegian tax I see $3500 on my calculator. Add a block on top and you'll see close to $4000. That's a lot of money for more fake frames. And this is not even for the flagship models. Custom RTX 5090, RTX 5080 listed for up to $2,800 and $1,400 overseas — initial pricing exceeds MSRP Maybe Nvidia will try make a shortage to push up the ASP. And the new 32GBb cards is good for AI work. Nvidia know their drill. NVIDIA RTX 50 will arrive in Spain in dribs and drabs, with retail limitations, AI-capped, and without multi-GPU Why is NVIDIA putting so few units into circulation? Is it waiting to see what AMD does? If so, could those who buy right away see the MSRP adjust in a few weeks when the Reds show off their hardware? Yup, you'll pay a lot extra for less than expected.... The RTX 5090 and RTX 5080 have the same amount of L1 cache as their predecessors, L2 will again be a performance bottleneck

Hmmm. Edit. Equal disgusting as the weather here home now. Only a phone work https://www.nrk.no/trondelag/kraftig-vind-og-flom-stenger-skoler-og-kommuner-setter-kriseledelse-i-trondelag-1.17211331

It's for a reason they all (OEM) want all to upload their bios files on Windows servers. Even Asus fall into their trap with their desktop motherboards firmware. They even try push on you older bios version than you current have on your computer. Disgusting? Yes.

O&O ShutUp10++ 1.9.1442 released 1/7/2025 About time to scrap M$ 365. And avoid buy anything of software/apps from the Redmond based tech company. Seems they have forgot that there is Open source software outthere that can be used. @Mr. Fox+++ Microsoft is determined to screw users: the end of Windows 10 in 2025 will drag down Microsoft 365, which will only be compatible with Windows 11 If you've been using Microsoft operating systems for years, you've probably noticed that the company sometimes puts out a good product and other times creates a disaster. For years, it seemed like this was alternating with each new release. Microsoft raises Microsoft 365 Personal and Home prices as it adds Copilot and AI features Microsoft introduces Copilot Chat for Microsoft 365 Today, Microsoft introduced Microsoft 365 Copilot Chat, an offering that adds optional pay-as-you-go agents.

HaHa. If it wasn't for new gen fake frame feature it would look bad. And Nvidia use increased Total Graphics Power (W) to offer better performance if they struggle to meet their goals. Why not max out TGP for every new generation cards? Can't do this because it will destroy the performance uplift for next gen/refreshed cards. The RTX 50 has a lower transistor density than the RTX 40 despite being manufactured on the same lithographic node

ThrottleStop 9.7.2 January 16th, 2025 9.7.2 (January 16th, 2025) removed the max multiplier code from the MMIO Lock function. saving of the PROCHOT Offset Lock status is fixed. added Arrow Lake beta support.

The 4090's will be very useful in China forwards (multi-gpu setups). Don't be surprised if we see quite high prices for the 4090's in the used market forwards. This graphics card will continue beeing popular. I expect someone will vacuum the market for 5090 and plow them into China. Be prepared for a new era of shortage of real 5090's. Same with price hike. Scalpers will have a wonderful time forwards if not some find ways to unlock "nvidia's new ways" to lock the 5090D models. This could also boost sales and keep up prices of used 4090's for China market (for multi gpu AI/mining compute). Yup.... Nvidia need to be sure no-one can unlock the 5000 series flagship. Or nvidia risk a more strickt ban of Blackwell GPU's in China. This doesn't look very well for performance enthusiasts in the rest of the world. NVIDIA GeForce RTX 5090D for China has limits on AI, cryptomining and multi-GPU configs, gaming not affected Testing need to be changed... How do you properly measure the performance of fake frames? With Blackwell, Nvidia recommends everyone doing benchmarks switch to using MsBetweenDisplayChange, as it's apparently more accurate. Looking at the numbers, most of the time it's not all that different, but Nvidia says it will better capture dropped frames, frametime variation, and the new flip metering that's used by MFG. So, if you want to get DLSS 4 "framerates" — in quotes because AI-generated frames are not the same as fully rendered frames — you'll need to make the switch. Nvidia says frame generation and upscaling will require us to rethink benchmarks Yup, show benchmarks the old way just show how little the performance uplift is from 4000 series up to 5000 series. UL aka Futuremark and Ungine need to come up with new benchmarks. They will probably use AI to pick up fake vs real results for the fake frame benchmark leaderboards, LOOL