Jump to content

Trixanity

Member
  • Posts

    3,333
  • Joined

  • Last visited

Everything posted by Trixanity

  1. I wonder what the thought process behind USB naming conventions are. You'd think they could find a common ground between engineering and marketing because if you give the job to either you get stupid names like this. Really makes you think how a guy could come up with this and some other guys patted him on the shoulder and said good job.
  2. That's true but it never stopped AMD going forward with HBM. A limited launch, while painful and would draw criticism, would cement Nvidia's lead unanimously. By the time supply stabilizes (if we assume it isn't all gobbled up by miners) people would gave forgotten about it already.
  3. Yeah, I'm not holding my breath. That way I can only be pleasantly surprised if they knock it out of the park for once (on the GPU side that is). I don't expect anything until they ditch GCN in 2020 and even then I'm not entirely confident until we know any details at least - if not outright benchmarks.
  4. The rumored date was July 30th but making the announcement the week leading up to HotChips (starts on a Sunday) would make sense with Gamescom starting just as HotChips ends the following Tuesday. I would assume they have FE cards available then (or soon after) and can showcase them there.
  5. Nvidia doesn't (as far as I know) present technical details on unreleased products so I imagine that they'll at least pull a paper launch beforehand; perhaps with a FE card in limited quantities.
  6. Well, we're probably seeing a product launch within two months considering that they're presenting technical details on said product at HotChips in August. Unless Nvidia scrapped those plans at the last minute of course but that would still mean they'd have spent tons of money sampling a product only to not release it.
  7. Probably less than 2 if we don't count unboxing and plugging in. It's getting ridiculously easy to setup consumer products. Depends on the product though of course. Just the idea of asking the average joe to read/watch a couple of tutorials and spend a couple of hours fiddling with a product that's overkill for their needs sounds absolutely ludicrous. That's when you realize how enthusiasts have lost perspective.
  8. I would assume if your ISP uses CGN it would be pretty difficult to get infected? It seems to me that while some routers are vulnerable the primary target was Ukraine and the US. So all things considered and the amount of layers they'd have to go through: it's pretty unlikely to be infected if you check the right boxes. That's my understanding of the situation. It seems like no one has officially explained the prerequisites for infection, how to detect infection and how to get rid of infection. At least not in any certain terms.
  9. Yeah we've known that for like a year now. What would have been interesting if something had changed eg. delays or the time table moving up. Pretty much guaranteed to be Navi and pretty much guaranteed to be 2H which is way too late considering the state of AMD graphics and with Nvidia apparently going in for the kill next month with a next gen launch.
  10. The logo is already RGB on this so the next step would be to make the entire thing a giant RGB stick with power connectors attached.
  11. This is eerily similar to the headphone jack debate which Apple is also in the center of. Removing features as if they haven't already coexisted and worked perfectly. If the 'new' feature/technology was superior in every way then it would make the predecessor disappear without forced deprecation. Apple is just stubbornly trying to make their investment in their own API bear fruit (pun not intended but I suspect a subconscious effort) despite it being a stupid idea to begin with considering they could have contributed to Vulkan instead. However Apple wants things to be unnecessarily exclusive and proprietary for maximum obnoxiousness.
  12. Oh, that's cool. I thought it was still very 'proof-of-concept'-like in terms of adoption. I stand corrected. I mean I hope it takes off - don't get me wrong - but Apple wants to destroy everything they can't control in their ecosystem. Their treatment of OpenGL which this very thread is about pretty much spells that out. Another example is that I have a Macbook in which I can't replace the SSD because uses a proprietary connector. It's stupid.
  13. Who's holding what? I'm certainly not. Just telling you how it actually is. It isn't the same as a native implementation. While it may run decently enough and, with work put into it, might perform 99% the same in some scenarios: that still doesn't adequately describe it as being Vulkan considering it's actually translating Vulkan to Metal. I haven't heard of wider adoption of MoltenVK either and if Apple gets their way they may actively try to block any application that uses it from their app store. Apple uses every opportunity to be absolute douche nozzles. Ultimately it may as well not exist if developers either opt for Metal or avoid the platform altogether. Perhaps the porting studios will spearhead adoption but for now any pseudo Vulkan support remains to be seen. Tl;dr it needs mainstream adoption to be considered a valid part of the ecosystem and even if that happens it's still using Metal to get there. It's kinda like claiming you can speak French because you have an app that can do live translation.
  14. I knew that was the idea behind the statement but it's not really the same. Ask Apple if their platform supports Vulkan or any graphics partner for their platform if the driver supports Vulkan. The answer is no. By your logic AMD supports CUDA (as an example). Meanwhile it doesn't really but there are tools to make it work one way or another. That goes for a lot of things with varying degrees of adoption and success. However, the real answer is no with an asterisk.
  15. Historically no. They're usually opting for 28W processors with the bigger iGPUs and eDRAM. And it just so happens that Intel launched a little something two months ago. The non-Touch bar version is a bit uncertain though. They'd probably want to put a 15W with Iris graphics into it but no such SKU exists right now as far as I know. So it's either 28W with Iris or 15W with standard UHD 620. I think Intel intends to move Iris to 28W exclusively this time around to avoid overlap and to allow Iris to stretch its legs as 15W is pretty constrained. They always throttle during games for example (or any other workload utilizing both CPU and GPU).
  16. They're probably still unveiling cards within a few months. The Hot Chips story pretty much confirmed that and no CEO is gonna go out and say "stop buying graphics cards - we'll reveal something next month" nor can he legally talk about any unreleased product outside of any official press release.
  17. I think mine was 2003 as well or early 2004. I remember quitting CS when the transition from 1.5 to 1.6 required Steam and I thought "I'm not downloading this weird Steam thing". Yeah I was that bothered by needing this client just to launch a game that I stopped playing. I think it took like 6 months before I bit the bullet and created my account anyway. Good times. I guess that's similar to how people think about all the other game clients we have today.
  18. Very true. Probably one of the most brilliant marketing pushes I've seen. Power was never a serious talking point until it suddenly was. A smart way to differentiate when performance was still similar.
  19. A unique username would be better than a deprecated email address. I've had nothing but trouble with this. Everytime people want to add me to my friends list I have to give out my email. That's not the only problem. Many people can't find my account to add either so I've always had to ask for theirs instead. For some reason the friends systems doesn't like email addresses yet any old account will have an email address as the username. However that's beside the point. Tying a username to a primary key in a database is poor design flaw. It's a broken system and not even a manual petition to get it changed will work. They're afraid if they change the primary key it'll straight up break your entire account. That's how bad it is to do that.
  20. Not only CCleaner. I believe various Linux distros on multiple occasions have been hijacked as well.
  21. Well, there's a ton of legacy code and other design choices in Steam that are incredibly stupid today. Like how usernames were the primary keys in accounts and that your email address used to be your username too. So my username is still my old email address. It's stupid as hell and Valve has said they'll fix it but only after they've released HL3.
  22. Z370 is a generation older in many aspects and launched much earlier. Remember CFL and Z370 was a limited initial launch. That explains why newer chipsets are on 14nm. 14nm while of course better isn't necessary for chipsets. I don't know the internals of Intel's foundry business but in theory it should be more expensive to use. Combine that with the obviously limited capacity that actually required them to pull a product. That begs the question I asked. I don't know what Intel does with their old equipment but that shit is expensive so putting it to use churning out chipsets is a great idea. As far as I know the chipset doesn't need the power or area gains from 14nm especially not in a desktop. I could understand it for laptops to squeeze every ounce of battery life you can. That's an area where AMD is struggling for example. They're posting less battery life despite a power efficient core design but something somewhere is using power that gives average to poor battery life (should be noted that the uncore stuff on die use a good deal of power during load but doesn't explain the numbers overall). Could AMD benefit from using something a bit more modern than 55nm? Possibly but then again there's a big chunk of the chipset on die which does benefit from the 14nm process.
  23. That laptop is obsolete. This is the AMD flagship now It only requires this to power it.
×