I already replied to you, Ian, on Twitter but I'll comment here again.
I've always liked AT's articles etc. in text form but I don't think it would hurt if you guys branched out a little bit, poking your head into "the Youtube space" and having these kinds of conversations on video with people for example like Wendell, David Kanter, people from WikiChip, Bryan Lunduke, Linus Sebastian, Steve Burke and so on about tech and the industry, hardware, software, semiconductors, computer engineering in general and obviously about microarchitectures and so on and so forth.
But this in particular was a really interesting conversation as I haven't really seen something like this before. I've known you, Ian, are a really sharp guy but now I've heard it and I and probably many others would like to hear more about what you have to say. So answering your question, definitely do more of this. Anandtech obviously comes first but if you have the time then definitely do it. Either in real life or on Skype or whatever.
And the same goes for other members of the Anandtech team, for example Ryan and/or Nate chatting about GPU's and stuff with other people. Would be interesting and I'd definitely watch that.
I second this, though I'm sure it's an investment that would take a long time to see any meaningful return.
Anandtech and Level1 (all iterations) are my favorite places to go for tech because of the detail and depth, like the threadripper scheduling saga which I loved digging into.
Kudos to you both for all the great content and the time and effort it took to generate and maintain such expertise.
Agreed, really enjoyed this would love to see more. It can be hard to invest 45 minutes, that is probably the upper end of my attention span. But really fascinating topics and I think with the appropriate level of depth, speaking for myself at least, but will allow me to dig in further on some of the items you mentioned.
Personally i found a little bit more structure on the topic flow would have been nice for a novice like me. For example , discuss the entire AMD side first , now the Intel one felt sometimes like side notes interwoven between AMD stuff
Also Ian has the proper posh voice when speaking. When you hear him talk without knowing his background, you think this guy know his stuff. Add in the wit and you get someone who can the best Youtube vids about tecnology.
Also i got a small hint that Ian will be joining AMD very soon.
I opened this up to read on the bus and saw quite disappointed by having nothing to read. I'd love to see a transcript alongside this sort of video interview content.
Great video with you and Wendell. I watched it before I found this article (I subscribe to Level1Techs on YT). Really hoping you can convince Wendell to make it to Hot Chips. I'd love if you two could sit down and do a chat like this after that conference. Isn't there also an interconnect conference around that time also?
In any case, great work and thank you for your reporting. I'm definitely a fan of your work and look forward to the upcoming coverage.
Don't know if you remember me Ian, but we used to bench together when you were on Benchtec UK, in fact we have benched many times since then lol, the last time was at Cambridge i think ? Anyway, just wanted to say that that the Video i have just watched is easily one of the most interesting and entertaining video's i have watched in years. Two guys just talking about stuff they love, there simply has to be more video's like this.
Thanks for the open discussion. It's good to hear the thoughts behind all those technical articles every now and then. Question: after all the Spectre-related mitigations, the improvement in the scheduler in W10 1903, wouldn't it make sense to re-run a CPU comparison using the last could of generations from both AMD and Intel, before we start digging into Ryzen 3xxx and Core 9xxx?
I saw this on Wendell's channel when it was posted, and I really enjoyed it. Especially since I've never heard you talk on camera, Ian. Very insightful video, please do some more of these in the future (AT doesn't even necessarily need to produce their own videos, you could continue doing collaborations with Youtubers like this).
Awesome video, thank guys! This is exactly the sort of behind the scenes stuff and info that's great to hear about especially with how it ties into the hardware! You two should make this a regular thing after big events!
Wendell, speaking about Intel: "chiplets would be implemented years ago" - well, they were, remember Core 2 Quad? And many others for different reasons, including North Bridge and last level cache. This idea is old, and not all that smart as you have to have more SKUs instead of letting user pick and choose different chips for their system (of course you need small sockets to match power and latency of chiplets instead of the current monsters).
Your comment makes no sense. Even though gluing two cores together long ago happened, you say that it is "not all that smart," while Intel is now looking at doing it in a couple years and incorporating ASICs when they do chiplets this time. The industry is moving as a whole to 2.5D and 3D stacking, which is based on chiplets, in part.
Then you continue on to say they "have to have more SKUs instead of letting user pick and choose different chips for their system (of course you need small sockets to match power and latency of chiplets instead of the current monsters)." Intel with monolithic dies has more SKUs than AMD. That doesn't end with chiplets. In fact, by making the core counts smaller per chiplet, you can drastically increase yields per wafer on smaller process nodes, allowing for the harvesting and binning of those dies, which can easily fill the product stack. Also, more SKUs means more choice for the consumer, rather than less.
Yes, they have to develop ways to get around issues like latency, which AMD has done significant work on with Zen 2. So, you saying it is dumb to do chiplets is irrelevant and shows you lack an understanding of where the industry is going as a whole. Moore's law is dead. We are reaching a point where clock speeds are likely to start regressing, as seen with Intel's 10nm and that AMD said they thought would happen until TSMC figured it out on 7nm. With future nodes, especially 5nm and below at TSMC and 7nm and below at Intel (even though Intel's 10nm has crap on yields), that problem is amplified. To combat that, Intel went wide on architecture, increasing IPC, that way to attempt to make up for the frequency regression. Sunny Cove achieved this well. But, there will be a point where due to yields, likely at 7nm on Intel's process, where the only way to achieve reasonable use of the wafer is to disintegrate the cores into chiplets, bin them, and incorporate them on their specific lines.
Even Foveros uses this disintegration as I/O was moved to the base die, suggesting that those are two different chiplets, just organized in a different way. Yet you cannot even acknowledge that.
c2d, c2q and pentium D was intel's desperate approach to fight the multi-core uma athlon with a numa design.... heh, coherency and consistency was fixed with s/w... an Itanium designer's wet dream. "chiplet" is a similar concept to mcm skus from the 90s and even before that. there are many success stories from IBM, with multi-role chips on one sku, including a "chiplet" with cache only. The rest of your comment doesn't make any sense and I cannot figure what you are trying to say. Anyhow, Wendell is just a guy like the tech support you have in companies, he can tell you which hdd or cpu to buy and he will install ubuntu for you, same applies for Dr. Professor Patrick Cutress, they read whatever's on the box.
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.
33 Comments
Back to Article
lagittaja - Friday, June 14, 2019 - link
I already replied to you, Ian, on Twitter but I'll comment here again.I've always liked AT's articles etc. in text form but I don't think it would hurt if you guys branched out a little bit, poking your head into "the Youtube space" and having these kinds of conversations on video with people for example like Wendell, David Kanter, people from WikiChip, Bryan Lunduke, Linus Sebastian, Steve Burke and so on about tech and the industry, hardware, software, semiconductors, computer engineering in general and obviously about microarchitectures and so on and so forth.
But this in particular was a really interesting conversation as I haven't really seen something like this before. I've known you, Ian, are a really sharp guy but now I've heard it and I and probably many others would like to hear more about what you have to say.
So answering your question, definitely do more of this. Anandtech obviously comes first but if you have the time then definitely do it. Either in real life or on Skype or whatever.
And the same goes for other members of the Anandtech team, for example Ryan and/or Nate chatting about GPU's and stuff with other people. Would be interesting and I'd definitely watch that.
lipscomb88 - Friday, June 14, 2019 - link
I second this, though I'm sure it's an investment that would take a long time to see any meaningful return.Anandtech and Level1 (all iterations) are my favorite places to go for tech because of the detail and depth, like the threadripper scheduling saga which I loved digging into.
Kudos to you both for all the great content and the time and effort it took to generate and maintain such expertise.
vanilla_gorilla - Saturday, June 15, 2019 - link
Agreed, really enjoyed this would love to see more. It can be hard to invest 45 minutes, that is probably the upper end of my attention span. But really fascinating topics and I think with the appropriate level of depth, speaking for myself at least, but will allow me to dig in further on some of the items you mentioned.0iron - Sunday, June 16, 2019 - link
Is Nate Oh still at AT? Haven't see his article for a whilenagi603 - Monday, June 17, 2019 - link
Agreed, I'd love to see more of chats like these.FreckledTrout - Friday, June 14, 2019 - link
Really enjoyed your talk.chatman1 - Friday, June 14, 2019 - link
Cutress, where is your Cutter?Ian Cutress - Friday, June 14, 2019 - link
What?plopke - Friday, June 14, 2019 - link
That was fun :) !A lot of teasing of threadripper zen 2 !
Personally i found a little bit more structure on the topic flow would have been nice for a novice like me.
For example , discuss the entire AMD side first , now the Intel one felt sometimes like side notes interwoven between AMD stuff
bleh0 - Friday, June 14, 2019 - link
Really cool video Ian! Having someone like Wendell to bounce off for stuff like this make for some good content.AmazighQ - Friday, June 14, 2019 - link
Also Ian has the proper posh voice when speaking. When you hear him talk without knowing his background, you think this guy know his stuff. Add in the wit and you get someone who can the best Youtube vids about tecnology.Also i got a small hint that Ian will be joining AMD very soon.
phoenix_rizzen - Friday, June 14, 2019 - link
Is there a written transcript anywhere for those of us who want to skim through it?Ian Cutress - Friday, June 14, 2019 - link
There's the autogenerated subtitles. It calls me Dr. Ian Tetriskepstin - Friday, June 14, 2019 - link
I opened this up to read on the bus and saw quite disappointed by having nothing to read. I'd love to see a transcript alongside this sort of video interview content.hamiltenor - Friday, June 14, 2019 - link
I enjoyed this video, thanks for being a part of it!ajc9988 - Friday, June 14, 2019 - link
Great video with you and Wendell. I watched it before I found this article (I subscribe to Level1Techs on YT). Really hoping you can convince Wendell to make it to Hot Chips. I'd love if you two could sit down and do a chat like this after that conference. Isn't there also an interconnect conference around that time also?In any case, great work and thank you for your reporting. I'm definitely a fan of your work and look forward to the upcoming coverage.
Drazick - Saturday, June 15, 2019 - link
This is a great video and I'd love to see more!Keep them coming...
BigMamaInHouse - Saturday, June 15, 2019 - link
I saw on twitter that you are working on Z390 AORUS XTREME WATERFORCE + 5.1Ghz 9900K Bundle review - Am I right? :-).kitfit1 - Saturday, June 15, 2019 - link
Don't know if you remember me Ian, but we used to bench together when you were on Benchtec UK, in fact we have benched many times since then lol, the last time was at Cambridge i think ?Anyway, just wanted to say that that the Video i have just watched is easily one of the most interesting and entertaining video's i have watched in years. Two guys just talking about stuff they love, there simply has to be more video's like this.
Dr. Swag - Saturday, June 15, 2019 - link
Ian almost makes Wendell sound like a tech noob, and this is Wendell we're talking abouthanselltc - Sunday, June 16, 2019 - link
oh HELL YEA.yankeeDDL - Monday, June 17, 2019 - link
Thanks for the open discussion. It's good to hear the thoughts behind all those technical articles every now and then.Question: after all the Spectre-related mitigations, the improvement in the scheduler in W10 1903, wouldn't it make sense to re-run a CPU comparison using the last could of generations from both AMD and Intel, before we start digging into Ryzen 3xxx and Core 9xxx?
rarson - Tuesday, June 18, 2019 - link
I saw this on Wendell's channel when it was posted, and I really enjoyed it. Especially since I've never heard you talk on camera, Ian. Very insightful video, please do some more of these in the future (AT doesn't even necessarily need to produce their own videos, you could continue doing collaborations with Youtubers like this).Kougar - Tuesday, June 18, 2019 - link
Awesome video, thank guys! This is exactly the sort of behind the scenes stuff and info that's great to hear about especially with how it ties into the hardware! You two should make this a regular thing after big events!just4U - Tuesday, June 18, 2019 - link
It was a excellent video.. You need to do more of these Ian.Rukur - Wednesday, June 19, 2019 - link
Was an interesting chat. Hope you guys catch up again across the world.WaltC - Wednesday, June 19, 2019 - link
Excellent! Most interesting...we need...more...!peevee - Thursday, June 20, 2019 - link
Video sucks because the talk is not searchable and not easily quotable. Video makes sense when you have something to SHOW, not just to TELL.peevee - Thursday, June 20, 2019 - link
Wendell, speaking about Intel: "chiplets would be implemented years ago" - well, they were, remember Core 2 Quad? And many others for different reasons, including North Bridge and last level cache. This idea is old, and not all that smart as you have to have more SKUs instead of letting user pick and choose different chips for their system (of course you need small sockets to match power and latency of chiplets instead of the current monsters).ajc9988 - Friday, June 21, 2019 - link
Your comment makes no sense. Even though gluing two cores together long ago happened, you say that it is "not all that smart," while Intel is now looking at doing it in a couple years and incorporating ASICs when they do chiplets this time. The industry is moving as a whole to 2.5D and 3D stacking, which is based on chiplets, in part.Then you continue on to say they "have to have more SKUs instead of letting user pick and choose different chips for their system (of course you need small sockets to match power and latency of chiplets instead of the current monsters)." Intel with monolithic dies has more SKUs than AMD. That doesn't end with chiplets. In fact, by making the core counts smaller per chiplet, you can drastically increase yields per wafer on smaller process nodes, allowing for the harvesting and binning of those dies, which can easily fill the product stack. Also, more SKUs means more choice for the consumer, rather than less.
Yes, they have to develop ways to get around issues like latency, which AMD has done significant work on with Zen 2. So, you saying it is dumb to do chiplets is irrelevant and shows you lack an understanding of where the industry is going as a whole. Moore's law is dead. We are reaching a point where clock speeds are likely to start regressing, as seen with Intel's 10nm and that AMD said they thought would happen until TSMC figured it out on 7nm. With future nodes, especially 5nm and below at TSMC and 7nm and below at Intel (even though Intel's 10nm has crap on yields), that problem is amplified. To combat that, Intel went wide on architecture, increasing IPC, that way to attempt to make up for the frequency regression. Sunny Cove achieved this well. But, there will be a point where due to yields, likely at 7nm on Intel's process, where the only way to achieve reasonable use of the wafer is to disintegrate the cores into chiplets, bin them, and incorporate them on their specific lines.
Even Foveros uses this disintegration as I/O was moved to the base die, suggesting that those are two different chiplets, just organized in a different way. Yet you cannot even acknowledge that.
Yorgos - Sunday, June 23, 2019 - link
c2d, c2q and pentium D was intel's desperate approach to fight the multi-core uma athlon with a numa design.... heh, coherency and consistency was fixed with s/w... an Itanium designer's wet dream."chiplet" is a similar concept to mcm skus from the 90s and even before that. there are many success stories from IBM, with multi-role chips on one sku, including a "chiplet" with cache only.
The rest of your comment doesn't make any sense and I cannot figure what you are trying to say.
Anyhow, Wendell is just a guy like the tech support you have in companies, he can tell you which hdd or cpu to buy and he will install ubuntu for you, same applies for Dr. Professor Patrick Cutress, they read whatever's on the box.