{"video_id":"fp_j9QMANnGpB","title":"RTX 5080 reviews, Deepseek, Qwen2.5 Max + more!","channel":"TechLinked","show":"TechLinked","published_at":"2025-01-30T03:44:00.018Z","duration_s":346,"segments":[{"start_s":0.0,"end_s":8.88,"text":"Sorry I haven't been hosting for a while guys, I was busy growing this mustache.","speaker":null,"is_sponsor":0},{"start_s":8.88,"end_s":15.84,"text":"Reviews are up for the GeForce RTX 5080, NVIDIA's latest AI product formerly known as the GPU,","speaker":null,"is_sponsor":0},{"start_s":15.84,"end_s":20.4,"text":"and for many, including Linus Tech Tip Sebastian, it's falling a bit flat.","speaker":null,"is_sponsor":0},{"start_s":20.4,"end_s":26.12,"text":"In native rendering performance at 1440p, which is a lot of P, the RTX 5080 is roughly","speaker":null,"is_sponsor":0},{"start_s":26.12,"end_s":32.92,"text":"10% better than the 4080 Super and 7% better than AMD's RX 7900 XTX.","speaker":null,"is_sponsor":0},{"start_s":32.92,"end_s":37.28,"text":"Those leads increased to 20% and 10% respectively at 4K.","speaker":null,"is_sponsor":0},{"start_s":37.28,"end_s":42.2,"text":"But the 5080 is still well behind the 4090, thanks in no small part to NVIDIA keeping the","speaker":null,"is_sponsor":0},{"start_s":42.2,"end_s":50.0,"text":"VRAM limited to 16GB, while the 4090 has 24GB, and it does not like to share.","speaker":null,"is_sponsor":0},{"start_s":50.0,"end_s":54.28,"text":"This release is disappointing to enthusiasts who remember that the RTX 4080 outperformed","speaker":null,"is_sponsor":0},{"start_s":54.28,"end_s":60.64,"text":"the 3090 and 3090 Ti at launch, although NVIDIA gave it a $1,200 price tag, the 5080","speaker":null,"is_sponsor":0},{"start_s":60.64,"end_s":65.36,"text":"only costs $1,000, so it's missing a whole $200 worth of GPU juice apparently, but no","speaker":null,"is_sponsor":0},{"start_s":65.36,"end_s":69.8,"text":"matter how mid reviewers say the RTX 50 series cards are so far, they're still going to","speaker":null,"is_sponsor":0},{"start_s":69.8,"end_s":76.0,"text":"be hard to find. NVIDIA has warned of availability issues to significant demand, probably because people","speaker":null,"is_sponsor":0},{"start_s":76.0,"end_s":79.84,"text":"just love AI so much and the bubble is definitely not popping, can't do that.","speaker":null,"is_sponsor":0},{"start_s":79.84,"end_s":83.56,"text":"I didn't hear anything, it's solid. I'll be over here AI-ing guys.","speaker":null,"is_sponsor":0},{"start_s":83.56,"end_s":88.72,"text":"Speaking of which, there's a fresh new Chinese AI model to give US investors more panic attacks.","speaker":null,"is_sponsor":0},{"start_s":88.72,"end_s":94.8,"text":"E-commerce giant Alibaba has released Quen 2.5 Max, which the company says outperforms","speaker":null,"is_sponsor":0},{"start_s":94.8,"end_s":98.32,"text":"DeepSeq V3, although it can't be run locally.","speaker":null,"is_sponsor":0},{"start_s":98.32,"end_s":104.8,"text":"How are they doing this? Well after DeepSeq's own model indicated to redditors that it sometimes confuses itself","speaker":null,"is_sponsor":0},{"start_s":104.8,"end_s":109.96,"text":"with the chat GPT, OpenAI told the Financial Times, it has evidence that DeepSeq trained","speaker":null,"is_sponsor":0},{"start_s":109.96,"end_s":115.24,"text":"models on data generated by chat GPT, which was famously trained only on original text","speaker":null,"is_sponsor":0},{"start_s":115.24,"end_s":118.28,"text":"handwritten by Sam Altman, the manifesto we call it.","speaker":null,"is_sponsor":0},{"start_s":118.28,"end_s":122.72,"text":"It's artisanal. Microsoft says they're investigating these claims, in case you forgot, they're best","speaker":null,"is_sponsor":0},{"start_s":122.72,"end_s":127.0,"text":"friends with OpenAI, which we know because thanks to social media we can follow the friendships","speaker":null,"is_sponsor":0},{"start_s":127.0,"end_s":130.6,"text":"and rivalries of tech CEOs like their Minecraft YouTubers.","speaker":null,"is_sponsor":0},{"start_s":130.6,"end_s":135.28,"text":"Now as we said on Monday, the US stock market panicked in response to the release of DeepSeq","speaker":null,"is_sponsor":0},{"start_s":135.28,"end_s":143.16,"text":"AI's models. But does that make sense? I mean, there are some reports that a while DeepSeq's chatbots were trained on NVIDIA GPUs,","speaker":null,"is_sponsor":0},{"start_s":143.16,"end_s":148.48,"text":"when you use one on the web, now it's running on AI chips made by Huawei, which would give","speaker":null,"is_sponsor":0},{"start_s":148.48,"end_s":152.08,"text":"US investors some reason to be worried about NVIDIA's monopoly.","speaker":null,"is_sponsor":0},{"start_s":152.08,"end_s":156.08,"text":"However, whether these Chinese models are actually as cheap to train and use as DeepSeq","speaker":null,"is_sponsor":0},{"start_s":156.08,"end_s":163.56,"text":"claims is being debated by analysts. An anthropologist, Dario Amode, argues the fact Chinese companies have to turn to less","speaker":null,"is_sponsor":0},{"start_s":163.56,"end_s":167.6,"text":"powerful hardware is proof that American restrictions on the export of AI chips are","speaker":null,"is_sponsor":0},{"start_s":167.6,"end_s":173.52,"text":"working. But let's say, sure, DeepSeq is way more efficient, as explained by Sam Maltman, that doesn't","speaker":null,"is_sponsor":0},{"start_s":173.52,"end_s":178.32,"text":"mean AI companies are going to buy less hardware. Uh, that $500 billion server?","speaker":null,"is_sponsor":0},{"start_s":178.32,"end_s":181.32,"text":"Yeah. We could do 250. Keep it coming.","speaker":null,"is_sponsor":0},{"start_s":181.32,"end_s":185.12,"text":"I thought that was all the stories, but it turns out I saved a quick bit in my mustache.","speaker":null,"is_sponsor":0},{"start_s":185.12,"end_s":191.96,"text":"Mustache, mustache! Enough! Google has open-source PebbleOS, the operating system that powered Pebble smartwatches, which","speaker":null,"is_sponsor":0},{"start_s":192.2,"end_s":197.08,"text":"you may not be old enough to remember. They were bought by Fitbit, which was bought by Google in 2021.","speaker":null,"is_sponsor":0},{"start_s":197.08,"end_s":202.24,"text":"Well, the original founder of Pebble, and super gigachat, Eric Megakofsky, says the","speaker":null,"is_sponsor":0},{"start_s":202.24,"end_s":205.24,"text":"open-sourcening means he's bringing Pebble back.","speaker":null,"is_sponsor":0},{"start_s":205.24,"end_s":209.16,"text":"His new team is working on an open-source smartwatch with a focused core set of features that","speaker":null,"is_sponsor":0},{"start_s":209.16,"end_s":213.36,"text":"users can tinker with so they don't have to depend on companies to fix stuff like the","speaker":null,"is_sponsor":0},{"start_s":213.36,"end_s":216.96,"text":"Blue Triangle of Death that briefly afflicted Garmin wearables this week.","speaker":null,"is_sponsor":0},{"start_s":216.96,"end_s":219.96,"text":"I mean, Blue Triangle? That's not even a thing.","speaker":null,"is_sponsor":0},{"start_s":219.96,"end_s":227.28,"text":"They made it up. Next, Pink Tetrahedrons. U.S. President Donald Trump said in a speech on Monday that his government will place","speaker":null,"is_sponsor":0},{"start_s":227.28,"end_s":233.44,"text":"tariffs on the import of computer chips and semiconductors to return production of these","speaker":null,"is_sponsor":0},{"start_s":233.44,"end_s":240.8,"text":"essential goods to the United States. The Biden Administration's CHIPS Act tried to do this by planning to invest 52 billion","speaker":null,"is_sponsor":0},{"start_s":240.8,"end_s":245.7,"text":"dollars in domestic chip foundries. But Trump says they had it the wrong way around.","speaker":null,"is_sponsor":0},{"start_s":245.7,"end_s":254.68,"text":"Chip companies don't need money. They need an incentive to not pay what Trump says could be a 25, 50, even 100 percent tax.","speaker":null,"is_sponsor":0},{"start_s":254.68,"end_s":260.76,"text":"Wow, it's like he's here. And that's why the U.S. is collaborating with every country in the world to make sure","speaker":null,"is_sponsor":0},{"start_s":260.76,"end_s":265.2,"text":"they all place those tariffs too. That way, TSMC will have no choice.","speaker":null,"is_sponsor":0},{"start_s":265.2,"end_s":272.72,"text":"It's foolproof. Comcast is rolling out new tech in select cities that it says could reduce latency by 78 percent","speaker":null,"is_sponsor":0},{"start_s":272.72,"end_s":277.08,"text":"in what the telecom giant calls an ultra-low-leg connectivity experience.","speaker":null,"is_sponsor":0},{"start_s":277.08,"end_s":283.16,"text":"They put experience on the end and it makes it funny. The experience uses a tech standard that's been in the works for a while called L4S,","speaker":null,"is_sponsor":0},{"start_s":283.16,"end_s":287.56,"text":"which stands for low-latency, low-loss, scalable throughput, girl boss.","speaker":null,"is_sponsor":0},{"start_s":287.56,"end_s":292.72,"text":"L4S sounds like a Craigslist list. You tell me in the comments what it means.","speaker":null,"is_sponsor":0},{"start_s":292.72,"end_s":298.72,"text":"I'm an L, I'm looking for an S. If Comcast claims are true, this might make you happier with your existing internet, which","speaker":null,"is_sponsor":0},{"start_s":298.72,"end_s":303.4,"text":"is good because the new chair of the FCC is killing his predecessor's proposal, making","speaker":null,"is_sponsor":0},{"start_s":303.4,"end_s":306.72,"text":"it easier for renters to switch internet service providers.","speaker":null,"is_sponsor":0},{"start_s":306.72,"end_s":309.76,"text":"Of the one you're with? Is that so hard?","speaker":null,"is_sponsor":0},{"start_s":309.76,"end_s":316.24,"text":"No one commits anymore. Anacoder has published a new open-source tar pit tool that tackles the problem of AI training","speaker":null,"is_sponsor":0},{"start_s":316.24,"end_s":321.44,"text":"webcrawlers hogging websites resources by trapping those webcrawlers in an infinite, randomly","speaker":null,"is_sponsor":0},{"start_s":321.44,"end_s":329.72,"text":"generated maze of linked pages. The tool is called Nepentheeth, as in the biological name for a genus of carnivorous plants.","speaker":null,"is_sponsor":0},{"start_s":329.72,"end_s":333.64,"text":"The tool is just like those, except while tropical insects are scary and huge, they're","speaker":null,"is_sponsor":0},{"start_s":333.64,"end_s":336.92,"text":"not known for hitting the same servers a million times in 24 hours.","speaker":null,"is_sponsor":0},{"start_s":336.92,"end_s":340.84,"text":"Although whoever wrote this has clearly never been to Botswana.","speaker":null,"is_sponsor":0},{"start_s":340.84,"end_s":344.48,"text":"No botflies though, just keep your carapaces out of my skin.","speaker":null,"is_sponsor":0}],"full_text":"Sorry I haven't been hosting for a while guys, I was busy growing this mustache. Reviews are up for the GeForce RTX 5080, NVIDIA's latest AI product formerly known as the GPU, and for many, including Linus Tech Tip Sebastian, it's falling a bit flat. In native rendering performance at 1440p, which is a lot of P, the RTX 5080 is roughly 10% better than the 4080 Super and 7% better than AMD's RX 7900 XTX. Those leads increased to 20% and 10% respectively at 4K. But the 5080 is still well behind the 4090, thanks in no small part to NVIDIA keeping the VRAM limited to 16GB, while the 4090 has 24GB, and it does not like to share. This release is disappointing to enthusiasts who remember that the RTX 4080 outperformed the 3090 and 3090 Ti at launch, although NVIDIA gave it a $1,200 price tag, the 5080 only costs $1,000, so it's missing a whole $200 worth of GPU juice apparently, but no matter how mid reviewers say the RTX 50 series cards are so far, they're still going to be hard to find. NVIDIA has warned of availability issues to significant demand, probably because people just love AI so much and the bubble is definitely not popping, can't do that. I didn't hear anything, it's solid. I'll be over here AI-ing guys. Speaking of which, there's a fresh new Chinese AI model to give US investors more panic attacks. E-commerce giant Alibaba has released Quen 2.5 Max, which the company says outperforms DeepSeq V3, although it can't be run locally. How are they doing this? Well after DeepSeq's own model indicated to redditors that it sometimes confuses itself with the chat GPT, OpenAI told the Financial Times, it has evidence that DeepSeq trained models on data generated by chat GPT, which was famously trained only on original text handwritten by Sam Altman, the manifesto we call it. It's artisanal. Microsoft says they're investigating these claims, in case you forgot, they're best friends with OpenAI, which we know because thanks to social media we can follow the friendships and rivalries of tech CEOs like their Minecraft YouTubers. Now as we said on Monday, the US stock market panicked in response to the release of DeepSeq AI's models. But does that make sense? I mean, there are some reports that a while DeepSeq's chatbots were trained on NVIDIA GPUs, when you use one on the web, now it's running on AI chips made by Huawei, which would give US investors some reason to be worried about NVIDIA's monopoly. However, whether these Chinese models are actually as cheap to train and use as DeepSeq claims is being debated by analysts. An anthropologist, Dario Amode, argues the fact Chinese companies have to turn to less powerful hardware is proof that American restrictions on the export of AI chips are working. But let's say, sure, DeepSeq is way more efficient, as explained by Sam Maltman, that doesn't mean AI companies are going to buy less hardware. Uh, that $500 billion server? Yeah. We could do 250. Keep it coming. I thought that was all the stories, but it turns out I saved a quick bit in my mustache. Mustache, mustache! Enough! Google has open-source PebbleOS, the operating system that powered Pebble smartwatches, which you may not be old enough to remember. They were bought by Fitbit, which was bought by Google in 2021. Well, the original founder of Pebble, and super gigachat, Eric Megakofsky, says the open-sourcening means he's bringing Pebble back. His new team is working on an open-source smartwatch with a focused core set of features that users can tinker with so they don't have to depend on companies to fix stuff like the Blue Triangle of Death that briefly afflicted Garmin wearables this week. I mean, Blue Triangle? That's not even a thing. They made it up. Next, Pink Tetrahedrons. U.S. President Donald Trump said in a speech on Monday that his government will place tariffs on the import of computer chips and semiconductors to return production of these essential goods to the United States. The Biden Administration's CHIPS Act tried to do this by planning to invest 52 billion dollars in domestic chip foundries. But Trump says they had it the wrong way around. Chip companies don't need money. They need an incentive to not pay what Trump says could be a 25, 50, even 100 percent tax. Wow, it's like he's here. And that's why the U.S. is collaborating with every country in the world to make sure they all place those tariffs too. That way, TSMC will have no choice. It's foolproof. Comcast is rolling out new tech in select cities that it says could reduce latency by 78 percent in what the telecom giant calls an ultra-low-leg connectivity experience. They put experience on the end and it makes it funny. The experience uses a tech standard that's been in the works for a while called L4S, which stands for low-latency, low-loss, scalable throughput, girl boss. L4S sounds like a Craigslist list. You tell me in the comments what it means. I'm an L, I'm looking for an S. If Comcast claims are true, this might make you happier with your existing internet, which is good because the new chair of the FCC is killing his predecessor's proposal, making it easier for renters to switch internet service providers. Of the one you're with? Is that so hard? No one commits anymore. Anacoder has published a new open-source tar pit tool that tackles the problem of AI training webcrawlers hogging websites resources by trapping those webcrawlers in an infinite, randomly generated maze of linked pages. The tool is called Nepentheeth, as in the biological name for a genus of carnivorous plants. The tool is just like those, except while tropical insects are scary and huge, they're not known for hitting the same servers a million times in 24 hours. Although whoever wrote this has clearly never been to Botswana. No botflies though, just keep your carapaces out of my skin."}