HW News - Intel CPU Recall, 7nm GPU, AMD Ray Tracing Bench, & HDD Failure Stats
video description
Date: 2020-05-06
Related videos
Comments and reviews: 10
bobhumplick
the crytek raytracing thing is alright and everything but its not really helpfu. everybody has just equated raytracing and rtx even though there is a big difference. you hear comments all the time raytracing is a gimmick, like hairworks. no. raytracing is a rendering technique that the gpu business has been slowly getting close to for the last 10 years or. rtx is an nvidia feature that may or may not be a gimmick depending on your outlook. but if EITHER of these are a gimmick then it would be rtx, not raytracing. rtx is a set of hardware that is used to accelerate raytracing effects in windows DXR. any card with driver support turned on can do this. amd could do it easily on any card but it just doesnt make sense. you could say that the 20 series rtx\raytracing features dont make sense either because performance is too low (at least for gaming. and this is true, it is too slow for gaming right now except if you are doing 60hz or so at lower than expected res for the tier of card youre running. ive seen articles and videos saying why is raytracing considered an nvidia thing when really its rtx. and the reason they give is that nvidia has done really good at marketing raytracing as an nvidia thing when its not, rtx is. i have to really disaggree there and how. raytracing being called bad or a fad or whatever is mostly because people just assume stuff and repeat what they have heard. and of course people that know what they mean are just making typos. but its mostly customer\commenter ignorance about the product. THE REAL REASON RTX CAME OUT NOW, before the tech was really done or before the cards were at the level they needed to be is probably mostly because the professional\quadro cards are the same die as the gaming versions. using 1 die for pro and gaming both saves money and means you will have higher binned turbos etc and any product with defective cores hav a place to go. if 7nm is actually a huge step up in raytracing performance then maybe it will be practical. also maybe these cards were meant to be on 7nm from the start but things go held up. nvidia make a lot more dies than amd and their current arch is very power efficient so 7nm is not needed as badly.
reply
the crytek raytracing thing is alright and everything but its not really helpfu. everybody has just equated raytracing and rtx even though there is a big difference. you hear comments all the time raytracing is a gimmick, like hairworks. no. raytracing is a rendering technique that the gpu business has been slowly getting close to for the last 10 years or. rtx is an nvidia feature that may or may not be a gimmick depending on your outlook. but if EITHER of these are a gimmick then it would be rtx, not raytracing. rtx is a set of hardware that is used to accelerate raytracing effects in windows DXR. any card with driver support turned on can do this. amd could do it easily on any card but it just doesnt make sense. you could say that the 20 series rtx\raytracing features dont make sense either because performance is too low (at least for gaming. and this is true, it is too slow for gaming right now except if you are doing 60hz or so at lower than expected res for the tier of card youre running. ive seen articles and videos saying why is raytracing considered an nvidia thing when really its rtx. and the reason they give is that nvidia has done really good at marketing raytracing as an nvidia thing when its not, rtx is. i have to really disaggree there and how. raytracing being called bad or a fad or whatever is mostly because people just assume stuff and repeat what they have heard. and of course people that know what they mean are just making typos. but its mostly customer\commenter ignorance about the product. THE REAL REASON RTX CAME OUT NOW, before the tech was really done or before the cards were at the level they needed to be is probably mostly because the professional\quadro cards are the same die as the gaming versions. using 1 die for pro and gaming both saves money and means you will have higher binned turbos etc and any product with defective cores hav a place to go. if 7nm is actually a huge step up in raytracing performance then maybe it will be practical. also maybe these cards were meant to be on 7nm from the start but things go held up. nvidia make a lot more dies than amd and their current arch is very power efficient so 7nm is not needed as badly.
reply
Mel
Bizzarre? I am highly suspicious of INTEL since building their largest most modern plant in IS rael. I think the whole wannacry BS is fishy and maybe related to this recall. I think the intel fix for wannacry actually opens the security between the cpu and ram and sends a copy of that data to a server farm in IS rael. This would be undetectable but slightly slow system performance just like the intel patch does. This technology can also snatch that cpu data and remotlely manipulate it before sending it to ram with low latancy. I also dont know of individuals having their computers hacked, just government offices and large buisnesses which usually have tighter security. Yet this performance decreasing patch is forced on everyone on windows and linux distros like to sneak it in. My suspicion is maybe the security onboard these chips cant be programmed to become listening devices with the patch so Intels replacing them with ones that w ill. another thought is why wouldnt intel just throttle the speed 10% to keep them cool. They thought it was ok to do this with all other chips over a security flaw? One other thing is these chips slow down when hot, i cant find complaints on the web of customers reporting this. not bizarre in the least but very suspicious.
reply
Bizzarre? I am highly suspicious of INTEL since building their largest most modern plant in IS rael. I think the whole wannacry BS is fishy and maybe related to this recall. I think the intel fix for wannacry actually opens the security between the cpu and ram and sends a copy of that data to a server farm in IS rael. This would be undetectable but slightly slow system performance just like the intel patch does. This technology can also snatch that cpu data and remotlely manipulate it before sending it to ram with low latancy. I also dont know of individuals having their computers hacked, just government offices and large buisnesses which usually have tighter security. Yet this performance decreasing patch is forced on everyone on windows and linux distros like to sneak it in. My suspicion is maybe the security onboard these chips cant be programmed to become listening devices with the patch so Intels replacing them with ones that w ill. another thought is why wouldnt intel just throttle the speed 10% to keep them cool. They thought it was ok to do this with all other chips over a security flaw? One other thing is these chips slow down when hot, i cant find complaints on the web of customers reporting this. not bizarre in the least but very suspicious.
reply
S.A.
I dont know why a Third Party (test house, or the companies themselves, doesn't just measure Waste Heat (WH. A WH rating would be worth a million times more than TDP. TDP should either be clarified to mean power draw minus power disipation as heat waste, or not used. Bare die temperature measurements at the stock frequency in lab conditions (all of which, AMD, Nvidia, Intel and other ASIC manufacturers are capable of, with a controlled atmosphere would be best for this. Clean room, climate control, a known load, tested on a test jig, with all voltage inputs measured, current draw (and thus power) measured and calculated, heat disipation measured. Humidity would vary a few decimal places and should be accounted for. It's possible, given how much money these companies put into testing already. And MAYBE, people will actually get a proper idea of HOW to cool their hardware properly and we'd also have an even more competitive cooler market. Not to mention, measuring WH is a good way for them to identify problems in an ASIC, also helping their R&D departments resolve problems to deliver us better products. I hope you've all read this AMD/Nvidia/Intel/Samsung/SK Hynix/Micron etc.
reply
I dont know why a Third Party (test house, or the companies themselves, doesn't just measure Waste Heat (WH. A WH rating would be worth a million times more than TDP. TDP should either be clarified to mean power draw minus power disipation as heat waste, or not used. Bare die temperature measurements at the stock frequency in lab conditions (all of which, AMD, Nvidia, Intel and other ASIC manufacturers are capable of, with a controlled atmosphere would be best for this. Clean room, climate control, a known load, tested on a test jig, with all voltage inputs measured, current draw (and thus power) measured and calculated, heat disipation measured. Humidity would vary a few decimal places and should be accounted for. It's possible, given how much money these companies put into testing already. And MAYBE, people will actually get a proper idea of HOW to cool their hardware properly and we'd also have an even more competitive cooler market. Not to mention, measuring WH is a good way for them to identify problems in an ASIC, also helping their R&D departments resolve problems to deliver us better products. I hope you've all read this AMD/Nvidia/Intel/Samsung/SK Hynix/Micron etc.
reply
scythelord
The whole star engine thing wasn't crytek's central complaint. They claimed CIG was required to use Cryengine for SC and also claimed that CIG is developing two games when they only agreed to one. The only issue is that there is no language in the agreement that restricts CIG to ONLY use Cryengine for SC, nor is the other claim vaild because the original game license agreed to defines the game as Space Citizen (name was not yet finalized as Star Citizen) and Squadron 42. Crytek's claims were broadly dismissed but some are progressing. Most who pay attention to it don't expect Crytek to win as it largely seems as though Crytek just wants to extort cash to remain solvent.
reply
The whole star engine thing wasn't crytek's central complaint. They claimed CIG was required to use Cryengine for SC and also claimed that CIG is developing two games when they only agreed to one. The only issue is that there is no language in the agreement that restricts CIG to ONLY use Cryengine for SC, nor is the other claim vaild because the original game license agreed to defines the game as Space Citizen (name was not yet finalized as Star Citizen) and Squadron 42. Crytek's claims were broadly dismissed but some are progressing. Most who pay attention to it don't expect Crytek to win as it largely seems as though Crytek just wants to extort cash to remain solvent.
reply
MI
That HDD Report makes me wonder. no matter which brand, 3 of 2 of my HDDs over 1TB gave up, in the third year. I kept older Drives, a 10, 2GB Samsung from 2002 for example, it still runs. The Brand wit the most defects of what i bought, is Western Digital. 4 of 5 Drives broke, Seagate 2 of 5 Toshiba 1 of 3, Samsung 1 of 5, Maxtor 2 of 4. When i think about my SSDs, none of them died yet. my first SSD was an OCZ Agility 2, a little later i got a Samsung SSD830 256(or250)GB those ar my Veterans, they still work fine, of course i havve newer models running now. This twisted my Picture of reliablity for SSDs and HDDs.
reply
That HDD Report makes me wonder. no matter which brand, 3 of 2 of my HDDs over 1TB gave up, in the third year. I kept older Drives, a 10, 2GB Samsung from 2002 for example, it still runs. The Brand wit the most defects of what i bought, is Western Digital. 4 of 5 Drives broke, Seagate 2 of 5 Toshiba 1 of 3, Samsung 1 of 5, Maxtor 2 of 4. When i think about my SSDs, none of them died yet. my first SSD was an OCZ Agility 2, a little later i got a Samsung SSD830 256(or250)GB those ar my Veterans, they still work fine, of course i havve newer models running now. This twisted my Picture of reliablity for SSDs and HDDs.
reply
SkillTweak
Hey, I m looking for help. I ve purchased a 5700XT and the card still has massive issues. While playing almost any game the card goes to 0mHz about once a minute for 1-2s. I first thought my card was broken but then I went to the amd subreddit and found lots and lots of people have this issue. It s been a known issue in the amd driver notes for almost 3 months now and AMD still has not fixed it. I feel like this is completely gutting the card and this topic needs way more traction.
reply
Hey, I m looking for help. I ve purchased a 5700XT and the card still has massive issues. While playing almost any game the card goes to 0mHz about once a minute for 1-2s. I first thought my card was broken but then I went to the amd subreddit and found lots and lots of people have this issue. It s been a known issue in the amd driver notes for almost 3 months now and AMD still has not fixed it. I feel like this is completely gutting the card and this topic needs way more traction.
reply
Z80
4: 55 This wont work as expected, all programmers hats works with new API/frameworks it's very annoying ( I'm talking about hobbies and freelancers ) for companies they could force their stuff to use the new API but what benefits they will got ( even on the future ) are these API will become the industry slandered on somehow! if not they will just ignore it unless Intel give them some motivations ( money ).
reply
4: 55 This wont work as expected, all programmers hats works with new API/frameworks it's very annoying ( I'm talking about hobbies and freelancers ) for companies they could force their stuff to use the new API but what benefits they will got ( even on the future ) are these API will become the industry slandered on somehow! if not they will just ignore it unless Intel give them some motivations ( money ).
reply
Michel
I know Intel's the fastest fast for ludicrous epic fortnite performance and that these security breaches don't really actually particularly seem to matter very much whatsoever for casual epic gaming consumers and everything, but this is getting ridiculous baby. One can't help but wonder if this dominance in the e-peen contest that is FPS ratings is purely caused by these security flaws.
reply
I know Intel's the fastest fast for ludicrous epic fortnite performance and that these security breaches don't really actually particularly seem to matter very much whatsoever for casual epic gaming consumers and everything, but this is getting ridiculous baby. One can't help but wonder if this dominance in the e-peen contest that is FPS ratings is purely caused by these security flaws.
reply
Jake
I made a comment on backblaze and got banned from their site. I guess they didn't like my opinion regarding garbage Seagate drives. As for Corsair, I absolutely love their products, but honestly have no need for LED lighting. Why bother making your computer glittery? Only thing that matters to me is functionality and quality, as I never look at anything except my monitor.
reply
I made a comment on backblaze and got banned from their site. I guess they didn't like my opinion regarding garbage Seagate drives. As for Corsair, I absolutely love their products, but honestly have no need for LED lighting. Why bother making your computer glittery? Only thing that matters to me is functionality and quality, as I never look at anything except my monitor.
reply
Stephen
New vulnerability. CPU makers respond (A)no big deal - (B) Don't tell anyone till we have a fix. (Later) We have a fix and you REALLY need to install it this stuff is dangerous. (Still Later) Oh yeah the vulnerability is no big deal but you really should stop using all your threads! I think the industry suffers from some form of mental disease and should seek help!
reply
New vulnerability. CPU makers respond (A)no big deal - (B) Don't tell anyone till we have a fix. (Later) We have a fix and you REALLY need to install it this stuff is dangerous. (Still Later) Oh yeah the vulnerability is no big deal but you really should stop using all your threads! I think the industry suffers from some form of mental disease and should seek help!
reply
Add a review, comment
Other channel videos
