Papusan

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

Hmmm. More cuda cores will undoubtedly choke the boost clocks further down due the measly Total Graphics Power (W). Not so sure those extra cores will do much without dual power connector, increased board power and better cooling. And nvidia want prefer cute small boards, so...

The best way is to avoid the AI and fake frame upscaling if this come as benchmark software... 5000 series GPU If you use the fake frame features all time.... 5080 should be good enough for most gamers. But the 16GB vram limit isn't nice. Should have been minimum 20GB. The 5080 reminds me all too much about the 3080@10GB stupidity.

I wonder how hwbot shall handle fake frame benchmarks. They will come. Because newer is always better. Intel fails to raise its head with its Core Ultra 200S CPUs: FPS losses after the wave of patches and firmware, where is the promised performance? This is yet another demonstration that what was seen at CES 2025 is just an attempt to dress up a product, the Core Ultra 200S, which present FPS losses compared to the Core 14 that are impossible for the blues to overcome, not even with overclocking and increased consumption. Gone is the old days with 1080Ti with 50-100$ extra for the better custom cards. Yup.... Newer has to be better. NVIDIA takes custom RTX 5080 binning to the next level: they will have a price difference never seen before with up to 400 euros between models And Zotac is so kind give you 25W more for the more expensive custom cards. Isn't it fantastic? The only thing we need now from Nvidia is a proper signing key to stop cross-flashing. ZOTAC RTX 5090 AMP Extreme INFINITY: this is the first custom graphics card to reach 600W as standard

Yep, more cores cost money. Same for vram. Also this gen the flagship card got slight better love than the lower end SKUs. But only because of the $400 price increase. The perfect fix for vram castrated cards from 4000 series. Nvidia know how to make profits. Oh'well. The bus castrated 8GB cards will still have some life. But the 3000 series cards are ready for the trash bin. NVIDIA Will Update DLSS 3 for RTX 40 Series with Improved Frame Generation Performance GPUs with more than 12GB of video memory may not notice a big change, but machines with less memory will benefit from the reduction in memory use and higher frame rates. The update allows for maintaining high graphic settings without overloading the system’s memory, preventing severe slowdowns

*rumors suggest Intel might consider backporting Lion Cove architecture to the platform*. Does this mean trash P-cores as for Arrow lake? Aka follow same trend...... Nvidia's RTX 5070 laptops give me little hope for desktop RTX 5060s The disappointing specs for RTX 5070 laptops may spell trouble for desktop RTX 5060 GPUs. Will Trump open up for helping China with AI computers. Making Nvidia happy? And Trump already leffling with allow TikTok in USA. So everything is possible when big money is involved. We all know China will pick up everything and everywhere of western tech if it's benefits them. And they will use all the power they have to get it. A harsh ban is mandatory. And cannot be negotiated. Full ban. Nvidia blasts Biden’s new AI restrictions as ‘misguided’ – GPU giant praises Trump’s hands-off approach to tech industry

Better scores from the old. And no need for Arrow Lake🙂 3DMark 11 - GTX 1080Ti https://hwbot.org/submission/5752378_papusan_3dmark11___performance_geforce_gtx_1080_ti_44367_marks?recalculate=true https://hwbot.org/submission/5752391_papusan_3dmark11___entry_geforce_gtx_1080_ti_60920_marks?recalculate=true What with this world fastest China-box? Maybe a laptop for bro @Mr. Fox😁 World’s fastest gaming laptops Jokebook with AMD Ryzen 9 9955HX3D and GeForce RTX 5090 announced, up to 280W power They said you will get 125W for the CPU. Fantastic, isn't it?

All the talk about new and modern. Old is gold. https://hwbot.org/submission/5752354_papusan_3dmark11___performance_geforce_gtx_1080_ti_43831_marks?recalculate=true

Only a true good company would do that. None cater to performance enthusiasts. And they'll all get rid of the cards anyway without too much work. And regardless of price. Mining, scalpers, monetary inflation and AI have destroyed the GPU market https://web.archive.org/web/20220521031800/https://www.evga.com/articles/archive/00944/EVGA-GeForce-GTX-980-Ti-KINGPIN/default.asp AMD didn't know what to do. Maybe fight fire with fire?😁

The only advantage for the AIB partners is.... They sell their cards all over the world. Many places you can't get an FE card. Either empty in the store or you can't buy them. And with the silicon lottery... All over the place. The more expensive cards offer better bling bling, better power delivery and cooler but don't be sure you get a good binned chips. And then you have the vram lottery. Not so sure the better cards (the more expensive one) will get the better binned IC's. But prepare for scalpers. 2000$ cards is cheaper than expected. They want their bite of the cake. Gigabyte’s custom GeForce RTX 5080 cards listed in EU, only one model at MSRP so far

The HOF is probably dead because of the U:S AI war with China. We all will be stuck with what the different AIB partners offer of silicon lottery and firmware.... Is this the end of the GALAX HOF series?

This is in fact great news. People buying less jokebooks is a good thing. And less e-waste. And I hope Microsoft see the paint on the wall. No one need more modern trash (to be able to use win 11). The problem? The U.S. doesn’t have a computer-manufacturing industry. Since it does not, retailers and consumers would be forced to pay the tariffs, regardless. The CTA estimates that U.S. consumers would instead curtail their spending dramatically, reducing laptop purchases by about 44 percent. https://www.pcworld.com/article/2573327/tariffs-could-cause-tech-prices-to-skyrocket-cta-warns-again.html

You never know when you will be next on the melting list. But some of the AIB partners have done the same what Galax did 2 years ago for the HOF. Asus followed up with the Matrix later. A chip to stop the melting. This should be the minimum with the stupid tiny power connector for <nvidia graphics. This should be a minimum for a card that cost +1000$ Maybe stupid. But it is what it is. Bling bling, and not cheap but choices here home is low for faster 48GB stics. I never opted for Intel's failure (Arrow Lake) so I invest in old Raptor Lake. Higher speed memory means people will buy it in the used market when I move on (when, I don't know). I can just turn the RGB to all white. I wanted 2 x 24 GB Teamgroup T-Force Xtreem 8200 but as usual these have never been a choice here home. I have the nickel plated EK-Quantum Velocity² so it will fit well with the sticks https://www.proshop.no/RAM/GSkill-Trident-Z5-Royal-DDR5-8400-48GB-CL40-Dual-Channel-2-pcs-Intel-XMP-Soelv-med-RGB/3275216 Edit. The forum speed sucks @Reciever Btw. I hope Trump have the balls to stop Nvidia.... Nvidia and SIA fire back at US gov's new export restrictions on AI GPUs to China https://www.tomshardware.com/tech-industry/artificial-intelligence/nvidia-and-sia-fire-back-at-u-s-govs-new-export-restrictions-on-ai-gpus-to-china