Discussion in 'Article Discussion' started by bit-tech, 17 Aug 2021.
"Intel has announced that it will enter the high-performance consumer graphics card fray"
How many times have they declared this now?
Up against the 1030 and rx550 mobile knowing them...
I've always been rooting for an Intel discrete card as I believed they would do a better job on Linux drivers than the others. Although AMD has done some good stuff with their open source drivers more recently. I don't expect 3090 tier performance straight out of the gate, but it will be interesting to see what they come up with and how they progress and compete in the future.
I assume their main business focus is to get their cards into servers for AI and machine learning applications.
I would love a 3rd enthusiast GPU maker… but for the GPUs to be worthwhile to them they will firstly focus on high volume.
We can hope though!
Just a thought - is it ever possible for a manufacturer to ‘fix’ a range of cards to achieve higher FPS during testing, to make their card ‘appear’ better? And not get caught out?
Yup! In fact, it kinda happens already: parts are often released with high-quality Samsung RAM, which is what gets reviewed, then quietly switched to cheaper Micron RAM. Shouldn't affect stock performance, but can have an impact on overclocking.
They do it with motherboards too, building a Rev.1 and then reusing the same SKU for a Rev.2 which drops some of the power circuitry or what have you.
Then there's the fact the drivers reviewers are using aren't necessarily the drivers buyers will be using...
What Gareth said plus
Only offering top of the line products in the first place (can't have someone talking about what corners where cut on that £100 mobo now can we?)
Testing the hardware to make sure reviewers don't receive DOA parts
Sending cherry picked samples that clock higher
Putting it in a fancy box to "flavour" the opinion of the reviewer
Suggesting certain tests be done (no points for guessing they'll be ones favouring a particular piece of hw)
Telling reviewers a fake price that won't exist in retail (remember Vega 56 & 64 launch?)
And of course if the review ain't sucking enough d**k there is the good old blacklist (Hardware Unboxed not spewing enough Raytraing propaganda for the taste of Nvidia comes to mind)
I mean, you'd think. I reviewed the RasPad 3 tablet conversion kit a while back, before the Kickstarter launched, and it mostly worked - except the Ethernet port was completely deaded. Pining for the fjords. Turns out I wasn't alone, either.
Testing for DOA before sending out to reviewers is less about skewing the reviews (unless you have serious quality control problems) and more about making sure the reviewer can... y'know, review it.
But yeah, the rest are all fair points.
I always wanted more competition again in this market to bring prices down but as we have seen with AMD that when they finally got a good product they just basically copied Nvidia pricing. PC gaming is on life support and Intel could help things but they likely won't. As long as shareholders ridiculously expect gains each and every year on their share prices these companies will nickel and dime us and Intel will do just that once they have something worth buying. I hope I am wrong on this but I doubt it. As for Intel really being in the battle in Q1 2022 I doubt it. Like others have said they will likely have something very low end.
I hope Intel is genuinely going to compete - I did wonder if they are aiming to cash in on mining though …
I doubt they give a flying banana about mining.
The long term goal will be more along the lines of giving the Nvidia DGX stuff a run for its money.
That being said, Intel will no doubt also raise prices of consumer GPUs during every mining craze, if they didn't then the shareholders would have the CEOs head on a plate.
Separate names with a comma.