Huh, they still have full height cube walls at Intel. That won't last. Millennials like to stare at each other so now cube farms everywhere are being converted to low walls, bullpens, and small size cubes to match the times. It sucks for those of us who really want to get work done.
Working in an office that has seen partition heights and desk space steadily eroded, I can assure you that it's not "the millennials" making the decisions to do so.
I am not that old myself (35), but I hate the open space thing... When I need to focus on work done (esp. more complex tasks) I want full concentration and no distractions. Sudden interruptions or environment disruptions only increase bug rate and lower overall productivity.
LOL. There are no "millennials" making the decision to go to open space. It's cramming as many people into as small a space as possible. Nobody who gets shit done likes it.
The manager that can say "I saved $FOO in my org by packing them in like sardines" will get phat l00t.
Moore's Law is often misinterpreted as being solely about transistor count. It is actually about the $/mm^2 effect of reducing transistor size. In the next generation you can either shrink the die size by 1/2 or keep the same die size and add more transistors to add more features or performance. Either option increases the revenue you can create from the same wafer. From that perspective, Moore's law is still very much in effect today.
"The period is often quoted as 18 months because of Intel executive David House, who predicted that chip performance would double every 18 months (being a combination of the effect of more transistors and the transistors being faster)."
> the first processor capable of a single teraflop: Intel’s Pentium Pro
Ian!!! Of course, it was first 1 TFLOP ***supercomputer*** and you can find on the photo that each processor was only 300 MFLOPS. The first Teraflop processor I know was Radeon 4870 from 2008
At first glance it looked more like a 4:3 to me. I have a 5:4 1280x1024 19" Dell on my workbench and its more square. The more I look at it, the more I think you're right that its 5:4. Still, its obviously not 16:9. Not a big deal, but... its hard to not notice this when a joke about screen resolution is made on a tech site.
I was entertained but came immediately to a halt. Any more content related to this?
I have read a few articles from random sites recently that Intel's 10nm delay could get them into trouble and I'm starting doubt the company. I still believe Intel is delaying 10nm intentionally but this is giving competition the fuel for growth. AMD will be back in the server and laptop space while TSMC is already in production of 7nm parts
I believe you don't know what you're talking about.
If you had evidence of such delays, it'd be worth $Millions to shareholders who would then sue Intel for intentionally holding back. In fact, I think intentionally lying to investors could earn some Intel execs prison time.
The most likely reason they haven't mass-launched 10 nm products is because it hasn't been economically viable to do so. Moreover, the first 10 nm CPU they're rumored to have launch (in China, of all places) has an oddly low clockspeed, which could indicate issues besides yield.
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.
37 Comments
Back to Article
guidryp - Wednesday, May 30, 2018 - link
Of course the computer and monitor are an older vintage. Moore is long retired. His desk is essentially a time capsule.Jedi2155 - Thursday, May 31, 2018 - link
Despite the vintage, it was a relatively low powered machine even at time of release. I'd wouldathunk it'd be at least be a E6600 if not an E6800.bji - Wednesday, May 30, 2018 - link
Huh, they still have full height cube walls at Intel. That won't last. Millennials like to stare at each other so now cube farms everywhere are being converted to low walls, bullpens, and small size cubes to match the times. It sucks for those of us who really want to get work done.boozed - Wednesday, May 30, 2018 - link
Working in an office that has seen partition heights and desk space steadily eroded, I can assure you that it's not "the millennials" making the decisions to do so.Ian Cutress - Wednesday, May 30, 2018 - link
Most of the cubicles are neck height. Moore's is just more vintagemr_tawan - Wednesday, May 30, 2018 - link
the low wall, or no wall, are most-likely gen X's decision. They are just cheaper and space-efficient.Of course those Gen X have their own 'glass-walled' office ...
HollyDOL - Thursday, May 31, 2018 - link
I am not that old myself (35), but I hate the open space thing... When I need to focus on work done (esp. more complex tasks) I want full concentration and no distractions. Sudden interruptions or environment disruptions only increase bug rate and lower overall productivity.stadisticado - Friday, June 1, 2018 - link
His desk is basically an informal shrine for employees. All workspaces have long since been updated, save for a few very dense floors.ianken - Wednesday, June 6, 2018 - link
LOL. There are no "millennials" making the decision to go to open space. It's cramming as many people into as small a space as possible. Nobody who gets shit done likes it.The manager that can say "I saved $FOO in my org by packing them in like sardines" will get phat l00t.
boozed - Wednesday, May 30, 2018 - link
A fascinating article, thanks.ಬುಲ್ವಿಂಕಲ್ ಜೆ ಮೂಸ್ - Wednesday, May 30, 2018 - link
Moore had a reasonable theoryToo bad it never became a law
I would have loved to see the number of transistors double every 12 months
Think of where we'd be today if it were actually a Law
ಬುಲ್ವಿಂಕಲ್ ಜೆ ಮೂಸ್ - Wednesday, May 30, 2018 - link
A doubling of transistors every 12 months would give us about 48 Quadrillion transistors in single-chip processor todayquantcon - Thursday, May 31, 2018 - link
Funny how fast exponential functions grow.ಬುಲ್ವಿಂಕಲ್ ಜೆ ಮೂಸ್ - Thursday, May 31, 2018 - link
I was off by a few quadrillion but I blame the processor for the errorWill we finally get another transistor doubling after waiting 6 years for 10nm?
ಬುಲ್ವಿಂಕಲ್ ಜೆ ಮೂಸ್ - Thursday, May 31, 2018 - link
5 years (not 6)14nm began 2014
10nm goes mainstream in 2019 (hopefully)
OK, maybe 6
Kidster3001 - Wednesday, June 6, 2018 - link
Moore's Law is often misinterpreted as being solely about transistor count. It is actually about the $/mm^2 effect of reducing transistor size. In the next generation you can either shrink the die size by 1/2 or keep the same die size and add more transistors to add more features or performance. Either option increases the revenue you can create from the same wafer. From that perspective, Moore's law is still very much in effect today.UltraWide - Thursday, May 31, 2018 - link
It's every 18 months.mode_13h - Friday, June 1, 2018 - link
10x every 5 years was my favorite formulation.Ian Cutress - Saturday, June 2, 2018 - link
He initially said 'every 12 months for 10 years' in 1965 then revised it to every 24 months in 1975.mode_13h - Sunday, June 3, 2018 - link
And the next sentence in Wikipedia reads:"The period is often quoted as 18 months because of Intel executive David House, who predicted that chip performance would double every 18 months (being a combination of the effect of more transistors and the transistors being faster)."
https://en.wikipedia.org/wiki/Moore%27s_law
ianmills - Thursday, May 31, 2018 - link
Good idea Mr.Stalin let's put all those scientists in jail who disobeyed the law!Bulat Ziganshin - Wednesday, May 30, 2018 - link
> the first processor capable of a single teraflop: Intel’s Pentium ProIan!!! Of course, it was first 1 TFLOP ***supercomputer*** and you can find on the photo that each processor was only 300 MFLOPS. The first Teraflop processor I know was Radeon 4870 from 2008
mode_13h - Friday, June 1, 2018 - link
What about this, from 2007 (and Intel, no less):https://en.wikipedia.org/wiki/Teraflops_Research_C...
ozzuneoj86 - Thursday, May 31, 2018 - link
A 720P 4:3 ratio LCD huh?guidryp - Thursday, May 31, 2018 - link
Looks similar to a 5:4(1280x1024) LCD I had back then.ozzuneoj86 - Thursday, May 31, 2018 - link
At first glance it looked more like a 4:3 to me. I have a 5:4 1280x1024 19" Dell on my workbench and its more square. The more I look at it, the more I think you're right that its 5:4. Still, its obviously not 16:9. Not a big deal, but... its hard to not notice this when a joke about screen resolution is made on a tech site.Why am I wasting my time writing about this? LOL
stadisticado - Friday, June 1, 2018 - link
He's been retired for a long time and likely hasn't had cause to sit at that desk for ~10 years for any reason. I'd expect outdated equipment.Jon Tseng - Thursday, May 31, 2018 - link
Huh cool. Presumably this is in the RNB right? What floor was it maybe I'll ask to see next time I'm there!Ian Cutress - Thursday, May 31, 2018 - link
It is - second floor.wrkingclass_hero - Thursday, May 31, 2018 - link
It's like he died 10 years agogfkBill - Thursday, May 31, 2018 - link
"a life size representation of a die layout" lol. A life-size die is what's in your computer ;)zodiacfml - Thursday, May 31, 2018 - link
I was entertained but came immediately to a halt. Any more content related to this?I have read a few articles from random sites recently that Intel's 10nm delay could get them into trouble and I'm starting doubt the company.
I still believe Intel is delaying 10nm intentionally but this is giving competition the fuel for growth. AMD will be back in the server and laptop space while TSMC is already in production of 7nm parts
mode_13h - Friday, June 1, 2018 - link
I believe you don't know what you're talking about.If you had evidence of such delays, it'd be worth $Millions to shareholders who would then sue Intel for intentionally holding back. In fact, I think intentionally lying to investors could earn some Intel execs prison time.
https://en.wikipedia.org/wiki/Sarbanes%E2%80%93Oxl...
MrSpadge - Saturday, June 2, 2018 - link
They are "intentionally holding back" because the yield is still too bad for wide scale production, i.e. they'd cost them more than current chips.mode_13h - Sunday, June 3, 2018 - link
This.The most likely reason they haven't mass-launched 10 nm products is because it hasn't been economically viable to do so. Moreover, the first 10 nm CPU they're rumored to have launch (in China, of all places) has an oddly low clockspeed, which could indicate issues besides yield.
UltraWide - Thursday, May 31, 2018 - link
a VR walk-through would have been amazing!Ian Cutress - Saturday, June 2, 2018 - link
Alas I could have done a 360-video, but I do not have the equipment.