Colin 🇪🇺🇺🇦

@nnikolasCS @3kliksphilip Have you seen/tested the tech yourself? No, your whole opinion is based on one YouTube video showing a technology that's still in early beta. And then you accuse me of fallacious thinking. Maybe you should experience something yourself first before grabbing a pitchfork.

@yoColinZ @3kliksphilip fallacious thinking, just because one technology succeeds after being doubted by the wider community doesn't mean that another is guaranteed to succeed as long as it gets doubted how much shitty tech was shelved quietly?

Such a shame to see the tech space become so radically polarised. I've been receiving hate mail all evening simply for covering DLSS 5. It's blatantly obvious it has incredible potential and is one of the most exciting developments in recent years (and there's been fierce competition for that spot). Incredibly short-sighted for people to dismiss it for minor and misinformed points, and to turn any discourse about it into a witchhunt. The tech community deserves better.

@NOTimothyLottes Maybe they should stop chasing Nvidia, and try to innovate on their own. Because worse quality upscaling/framegen for ~$100 less isn't enough of a differentiation.

Most seem to miss that the true point of DLSS5 is to keep AMD/Intel/Sony/Microsoft busy chasing an impossible target to keep otherwise productive humans from actually developing something useful instead.

@GPUsAreMagic @IanCutress Got it. Thanks

@yoColinZ @IanCutress SOCAMM also in general has a more suitable size/outline for servers to stack more of them in a smaller footprint, contrary to the silly PMIC notch on LPCAMM.

Just held a 256 GB SOCAMM2 LP5X-7500 module. Tasty. There'll be 64 of them in an 8-way Vera blade.

@9550pro Switch revenue is reported under OEM not gaming I believe. So the difference should be even larger.

Q4 2025 Revenue AMD: PS+Xbox+5% GPU market=0.8B NVIDIA: NS+94% GPU market=3.7B

@HotAisle Why would anyone deal with AMD without demanding the same 10% deal OAI en Meta got?

Two years ago, I raised money and bet my entire business on this day. Nobody listened. Couldn’t even give away GPUs for free. Why AMD, they asked. Software sucks. Hardware can’t compete. AMD has no clue. I heard every excuse. Now, AMD just did what I knew they would do and proved my entire thesis correct. I literally predicted the future. Give me funding for the next great thing that builds upon what I already have. It should be a no brainer.

Murderers

@IanCutress @mweinbach @nvidia @groq NV 'acquired' Enfabrica some time ago, there might be some synergy on a system-level.

@mweinbach @nvidia @groq NVLink's benefit is bandwidth. The groq design isn't meant for that, and isn't the limiting factor. Unless you NVLink to memory, but then you still have to deal with VLIW deterministic processing, not variable latency. Then who would buy it/use it?

Looks like @NVIDIA just extracted @Groq's employees and a non-exclusive hardware license to bring it into NVLink for $20B. @Groq remains running GroqCloud, but everyone of note joins NVIDIA, including CTO and architects. groq.com/newsroom/groq-… What is going on? (a) If they were acquiring for the middle-east business, I'd kinda get it. Or just an acquihire - expensive though. Groq's lead architect joined NV a couple years ago. (b) If they wanted to remove a competitor, expensive for a minor player comparitively (c) There's that much cash burning a hole at NV? (d) Everyone I've spoken to so far can't work out why, given architecture mix and NVIDIA's market leadership and where it'd fit in (e) Looks like I might cut Christmas short and we get a TechPoutine tomorrow.

Today, we fined X for non-compliance with transparency obligations under the DSA. We're holding X accountable for: 🔹Deceptive design of its ‘blue checkmark’ 🔹Lack of transparency of its advertising repository 🔹Failure to provide access to public data for researchers ↓

@Borderlands @NVIDIAGeForce GeForce Season

Attention Vault Hunters! 🚨 Want to level-up your loadout with this Claptrap custom-wrapped GeForce RTX 5090? Comment "GeForce Season" below for a chance to win!

@aschilling @AMDServer Why are these so much wider than nvl72? Purely due to the use of ddrd5 dimms?

At 2025 OCP Global Summit @AMDServer shows their Helios rack-scale solution. These are liquid-cooled, 72-GPU racks powering AMD Instinct MI450 Series GPUs, AMD EPYC CPUs (Venice) and do feature integrated UALoE scale-up connectivity.

@dragonbornriver @TESRSkyblivion No enb/cs, just reworked Skyrim lighting templates.

@TESRSkyblivion @TESRSkyblivion does this have any kind of enb or community shaders?

"Excellent. A fine example of a Daedric shrine." - Elante of Alinor We've completed all of the dungeons for the Fighters Guild! Enjoy this early look at our entirely remade Brittlerock Cave.

@NvidiaANZ GeForce RTX 5090

Gamescom 2025 is a wrap, but we have one more epic prize announcement from our GeForce On Community Update: Jensen Huang signed GeForce RTX 5090 Want it?! Comment "GeForce RTX 5090" below for your chance to WIN!

@NVIDIAGeForce GeForce RTX 5090

Gamescom 2025 is a wrap, but we have one more epic prize announcement from our GeForce On Community Update: Jensen Huang signed GeForce RTX 5090 👀 Want it?! Comment "GeForce RTX 5090" below for your chance to WIN!

Stand with Ukraine

@IanCutress Do you expect HBM4 to ever be standardized?

Lol not a rumor. Literally a requirement for first wave of HBM4. We've known this for 2yrs+.

@nothing @grok Nothing phone (3)

Hey @grok, pick a winner out of every 10,000th comment, they get the Nothing product they mention. Must be following @nothing. Ends in 48 hours. Let’s go.

@getpeid @grok Oh hi

.@grok pick one like/reply in 24 hours and i'll buy them any nothing product

@TESRSkyblivion @6Glitz Holy... It looks so good

Welcome to Castle Chorrol, the seat of Countess Arriana Valga. Credit: @6Glitz & Jero Vote for which #Skyblivion castle you'd like to see next in our Discord Community Server: discord.gg/skyblivion