CPU aging a problem?
"Such degradation might involve variations in supply voltage, temperature changes, or simply aging transistors."
Yeah, we all have a bunch of PCs where CPU has died due the old age in our corners, don't we?
Intel has developed a research microprocessor that it claims can improve throughput of degraded chips or chip environments by over 40 per cent. Such degradation might involve variations in supply voltage, temperature changes, or simply aging transistors. As described by Intel staff research scientist Keith Bowman on Tuesday at …
Border line chips can be sold at relatively low prices. Of course as the error rate climbs the clock slows down untill it ramps up again (can you say limite cycle oscillation)
But most importantly it retains the ability to charge customers for faster chips. Dumping the clock and going asynchrounous would be such *bad* marketing.
Presumably 1/2 ofthe transistors will continue to be for clock distribution as they are now.
Does anyone else think this sounds a bit desperate? Clever but definitely running full tilt into a brick wall.
Me and a 2 people I worked with bought TUV4x's when they came out. 2 months after warranty ended my caps blew, 3 weeks after that my one buddies blew (he got his 2 weeks after I got mine), and around a month later the other guys blew. 3rd guy was lucky he ordered a new motherboard after he saw mine and my one buddies blow withing weeks of each other.
I have yet to buy a Asus product since...
I've never had any capacitor issue on my motherboards, ASUS or otherwise. And I've bought plenty of motherboards in the past 15 years. 17, to be precise, and at least 5 were ASUS. Yeah, I was on the upgrade treadmill.
Most of my hardware problems were essentially due to bad power units - until I learned my lesson and started buying proper power units, quality and over -dimensioned for my needs. And a UPS to control the quality of the electrical current.
Since that time, about six years ago, I have never had any hardware problem of any kind.
Or even better - Intel's Friends.
Isn't this technique the University of Michigan research in conjunction with Arm Holdings and the 'research CPU' is in fact an ARM ISA (which I guess is why the Reg article doesn't mention the architecture in more detail)?
Since Intel no longer muddies its paws with power-efficient processors, I would be surprised if it would build a research processor based on a competitor architecture it isn't licenced for. So I figure the article needs a retitle - "Intel takes credit for rival's geriatric cpu cure" ( All Your Research R belong to us ) ;-)
Biting the hand that feeds IT © 1998–2020