What I be most interested in Lakefield, I believe this is some ways will be Intel's little sleeping giant. ARM has be at Intel where it hurts the most, and this is suppose to be Intel's answer to it. Only thing I seen so far is this is Intel next class of mobile machines - and last time Intel did has led us to 2in1 revolution in mobile computers.
I also interesting no discussion about Intel 10nm - but Lakefield is 10nm and 10nm Icelake stuff is suppose before in June supposedly.
smilingcrow who knows.. but ONLY HStewart would be more interested in what intel is presenting at hot chips, when most others.. are more interested in what AMD is presenting.... im almost expecting intel to just feed everyone more lies and promises they will break cause they cant deliver.
If AMD can give me a $320 Ryzen 7-3700X 12c/24t that beats Intel $485 i9-9900K 8c/16t... Until then, all that stupid hype will just make AMD look worse if they can’t deliver.
AMD hasn't promised anything. It's all rumors and speculation at this point. Setting arbitrary price and core requirements (fueled by baseless rumors) as your "win condition" is just asinine.
For that matter even if they only undercut Intel by $50 and matched performance, that's still a win, not a loss. For their sake I hope they secure a more definite victory, but the point is that reviewers aren't going to base their results on whackjob rumor-fueled conditions.
It would be very interesting if Intel addressed security considerations for CPU design, what the trade offs between speed and security are and how to chose which one....
"You choose speed until you get in the news because someone notices all the security holes you created. Then you tell everyone that it was unforeseeable and affects all processors instead of just badly-designed ones."
Which security issues ? Did not see anything on Anandtech or in the staff tweets about this, so must not be anything important and definitely far less serious than "Ryzenfall"
And yeah, I've been wondering why that was, especially since it concerns all but the latest CPUs...
I like to think that AnandTech is waiting for a test to be done so they can show impact benchmarks. There's probably only one or two staff who can do that; they may be elsewhere, plus it takes time.
I was being sarcastic :) Phoronix have run test and performance is down - on top of the already peformance degrading previous patches. And this with HT still enabled....
As for not concerning the latest Intel CPU - reseachers do not seem to agree with Intel and Intel was using rather vague language. Perhaps the fact that most of their 9th gen core CPU come with HT disabled is the fix ?
I've understood that Jim Keller didn't lead the developement of the FSD chip, but led some other project. He is neither mentioned in any patents nor was he mentioned at the autonomy day presentations.
I am more interested in "Upmem True Processing In Memory with DRAM Accelerator" part. Fanboy f(igh)(ar)ts between outdated CPU architectures is boring.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
17 Comments
Back to Article
HStewart - Thursday, May 16, 2019 - link
What I be most interested in Lakefield, I believe this is some ways will be Intel's little sleeping giant. ARM has be at Intel where it hurts the most, and this is suppose to be Intel's answer to it. Only thing I seen so far is this is Intel next class of mobile machines - and last time Intel did has led us to 2in1 revolution in mobile computers.I also interesting no discussion about Intel 10nm - but Lakefield is 10nm and 10nm Icelake stuff is suppose before in June supposedly.
Aephe - Thursday, May 16, 2019 - link
Didn't see that coming!smilingcrow - Thursday, May 16, 2019 - link
I wonder if HStewart is paid in dollars, bitcoin, failed Intel 10nm dies or just broken dreams?Korguz - Thursday, May 16, 2019 - link
smilingcrow who knows.. but ONLY HStewart would be more interested in what intel is presenting at hot chips, when most others.. are more interested in what AMD is presenting.... im almost expecting intel to just feed everyone more lies and promises they will break cause they cant deliver.sonny73n - Friday, May 17, 2019 - link
If AMD can give me a $320 Ryzen 7-3700X 12c/24t that beats Intel $485 i9-9900K 8c/16t...Until then, all that stupid hype will just make AMD look worse if they can’t deliver.
Alexvrb - Friday, May 17, 2019 - link
AMD hasn't promised anything. It's all rumors and speculation at this point. Setting arbitrary price and core requirements (fueled by baseless rumors) as your "win condition" is just asinine.For that matter even if they only undercut Intel by $50 and matched performance, that's still a win, not a loss. For their sake I hope they secure a more definite victory, but the point is that reviewers aren't going to base their results on whackjob rumor-fueled conditions.
Korguz - Friday, May 17, 2019 - link
sonny73n even then.. cant look worse then intel has the last few years.. :)Irata - Friday, May 17, 2019 - link
It would be very interesting if Intel addressed security considerations for CPU design, what the trade offs between speed and security are and how to chose which one....Lord of the Bored - Saturday, May 18, 2019 - link
"You choose speed until you get in the news because someone notices all the security holes you created. Then you tell everyone that it was unforeseeable and affects all processors instead of just badly-designed ones."GreenReaper - Friday, May 17, 2019 - link
Let's hope Jintide's hardware security isn't compromised by the recent issues exposed in Intel cores.Irata - Friday, May 17, 2019 - link
Which security issues ? Did not see anything on Anandtech or in the staff tweets about this, so must not be anything important and definitely far less serious than "Ryzenfall"GreenReaper - Friday, May 17, 2019 - link
Ars covered it a few days ago:https://arstechnica.com/gadgets/2019/05/new-specul...
AMD claims they are, as far as they can see, unaffected:
https://www.amd.com/en/corporate/product-security
And yeah, I've been wondering why that was, especially since it concerns all but the latest CPUs...
I like to think that AnandTech is waiting for a test to be done so they can show impact benchmarks. There's probably only one or two staff who can do that; they may be elsewhere, plus it takes time.
Irata - Friday, May 17, 2019 - link
I was being sarcastic :) Phoronix have run test and performance is down - on top of the already peformance degrading previous patches. And this with HT still enabled....As for not concerning the latest Intel CPU - reseachers do not seem to agree with Intel and Intel was using rather vague language. Perhaps the fact that most of their 9th gen core CPU come with HT disabled is the fix ?
Lord of the Bored - Saturday, May 18, 2019 - link
I thought you were HStewart at first.Irata - Saturday, May 18, 2019 - link
Will make sure to add the [/sarcasm] tag next time :-)Rudde - Friday, May 17, 2019 - link
I've understood that Jim Keller didn't lead the developement of the FSD chip, but led some other project. He is neither mentioned in any patents nor was he mentioned at the autonomy day presentations.peevee - Monday, May 20, 2019 - link
I am more interested in "Upmem True Processing In Memory with DRAM Accelerator" part. Fanboy f(igh)(ar)ts between outdated CPU architectures is boring.