Maybe with liquid helium, he'll achieve ludicrous speed.
So, about that Guinness World Record-beating clock speed of 8.46 GHz we talked about last week – that's been beaten again.
The very same Andre Yang that achieved that remarkable speed has upped his efforts – and his AMD FX-8150's limits – to an astounding 8.58 GHz.
Yang kept the same Asus Crosshair V Formula motherboard, but this time he cranked the voltage up from 1.992V to 2.076V. This was also done with liquid nitrogen, so it's possible that there's still room for more with the even-more-effective liquid helium.
Sponsored
Even though I mostly use AMD hardware, I don't think I can argue much with what you say there. I'm still scratching my head as to what actually happened with Bulldozer, and how it will take AMD can flatten the competition rather than massage it. I think part of the answer to that is AMD not keeping all it's eggs in one basket - their overall profits are still looking good despite the BD debacle. If they can significantly improve the design then I'd be a lot more positive about stories like this.
I'm not saying it isn't impressive to see such high clocks, but as anyone in the game knows 'speed' and 'performance' are two different things. For example look at the way the 2.0GHz Athlon 64 could devour a 3.0GHz Pentium 4 in most tests. especially games. As many will testify, the situation with Bulldozer vs. Sandy Bridge is very different.. and it doesn't exactly go in AMD's favor I think.
I agree. I think AMD is trying its best to be all things to everyone, but doesn't have the requisite resource or funding. Their overall profits are up thanks to the APU and mobile sectors so they're doing something right. They also had to setup a 32nm plant in what are dire and uncertain economic times.
But when I see a next-gen architecture - one optimised for threaded apps - struggling to out-run a Phenom II x6, something is not right. Neither BD nor Phenom II are particuly bad parts and their pricing seems quite competitive, but there doesn't seem to be a huge selling point in favour of BD.
On the positive side, I think AMD can can really hit even harder in other, increasingly markets. I think they know BD is not everything people expected it to be and that it has a long way to go, but they can't risk pooling resources into one market. Intel is just too big. In the meantime, even if they could run this chip at 10GHz I'd have the same reservations; I could run a Slot-1 Celeron 266 @ 400MHz.. but it was still a Celeron, and the performance gain wasn't linear. Get the per-core IPC performance right first, as well as the per-watt performance. Then crank the core to oblivion. In honor of the Athlon64, go for it
It's the opposite really. AMD's engineers deliberately designed a CPU that would excel in threaded applications. They had hoped to offset the lower single core performance by raising the clocks, but GlobalFoundries' poor yields forced them to accept lower clocked parts.
AMD did a great job with Bulldozer. The design is decent and on OS'es that fully support Bulldozer's semi-cores (Linux and Windows 8) it shines even more. However, low yields and poor manufacturing got us stuck with a poorly performing part.