AMD FX-8150 Breaks New Clock Speed Record at 8.8 GHz

[saint19]

That's a bug program confirmed on XS.

 

[-Fran-]

[citation][nom]redyellowblueblast[/nom]I use a AMD-FX machine with Linux on it and I don't see any of the performance problems that I did on Windows 7. Which leads me top believe that the problem is more with Windows 7 not understanding the BD architecture, and not the architecture itself. The Linux kernel seems to be better suited at the moment at understanding the chip. Hopefully Windows 8 will fix it. (that is if Windows 8 will even be worth it.) And I'm still waiting on my Piledriver upgrade, AMD![/citation]

It's not Windows 7 kernel per se, but the programs themselves not supporting the new instruction set in the FX series.

As usual, AMD trusted Intel and went along with FMA4, but Intel did a turncoat and went with FMA3 for it's new Archs. I don't know about the AVX instructions though, but I'm sure 99% of Windows programs don't use it on AMD CPUs.

Cheers!

 

[nebun]

all this great potential does not make up for the mediocre coding....so sad

 

[kohvipaus]

Looking at all these comments makes me wonder, does anyone besides Intel fanbois even read this site anymore?

And Perry, he hit 9GHz with the same chip aswell

 

[kohvipaus]

[citation][nom]kohvipaus[/nom]Looking at all these comments makes me wonder, does anyone besides Intel fanbois even read this site anymore?And Perry, he hit 9GHz with the same chip aswell[/citation]


http://valid.canardpc.com/show_oc.php?id=2385966

 

[dgingeri]

...and it still lagged behind the Core i3 processor in gaming.

 

[ashinms]

[citation][nom]Razor512[/nom]an 8 core AMD CPU when overclocked to around 4.8GHz, can perform similarly to a core i7 2600k (and even beat it in many cases when the task is highly threaded (mainly workstation stuff that can efficiently use 8 cores)the core i7 still has the better per core performance, but the AMD makes up for it by the 4 extra half cores.[/citation]

When the original reviews came out, it actually beat the 2600 at stock clocks on multithreaded workloads.

 

[ashinms]

[citation][nom]redyellowblueblast[/nom]I use a AMD-FX machine with Linux on it and I don't see any of the performance problems that I did on Windows 7. Which leads me top believe that the problem is more with Windows 7 not understanding the BD architecture, and not the architecture itself. The Linux kernel seems to be better suited at the moment at understanding the chip. Hopefully Windows 8 will fix it. (that is if Windows 8 will even be worth it.) And I'm still waiting on my Piledriver upgrade, AMD![/citation]

Read this: http://www.youtube.com/watch?v=92_LIlBIO84

Yeah, a 20 percent gain on several benchmarks when manual thread scheduling happens, but only about 1 percent with the windows fixes? Something smells fishy here...

 

[ashinms]

...I mean this... http://techreport.com/articles.x/21865

 

[may1]

It's interesting to know that a BMW mini can run so fast with liquid nitrogen rocket booster - but can it beat a Ferrari? That's the issue.

 

[longshotthe1st]

[citation][nom]tinnitus[/nom]being fanatic of something helps your self-esteem[/citation]

Or it could hurt it.

 

[eddieroolz]

Just take it for what it is - a record. That's all it really is. No one questions what utility the highest skateboard jump holds, or what the point of the largest hamburger prepared really is.

 

[Outlander_04]

[citation][nom]rohitbaran[/nom]Meh. Beat a SandyBridge first and then we will talk.[/citation]
Depends which benchmark you use but that happens regularly

[citation][nom]subasteve5800[/nom]We get it. If you disable most of the chip and submerge the rest in LN2, you can achieve a ridiculously high overclock. It's completely unrealistic and useless, but you can do it.[/citation]
A bit like taking a sandybridge processor and a huge graphics card , connecting it to a low resolution monitor and claiming its a gaming champion because it makes 200fps when the Bulldozer will only run to 120 fps?
You only have to completely ignore that no one with a brain builds that computer , and that the 60 Hz monitor its connected too will only show 60 FPS with either set up so the user experience would be identical ...and then you can claim sb is the better chip . LOL !

 

[iam2thecrowe]

[citation][nom]kartu[/nom]Bullshit.[/citation]
no BS, 1x 8.8ghz BD module that is not stable can not beat a 3.3ghz i5 with 4 cores.

 

[blazorthon]

[citation][nom]ashinms[/nom]Read this: http://www.youtube.com/watch?v=92_LIlBIO84 Yeah, a 20 percent gain on several benchmarks when manual thread scheduling happens, but only about 1 percent with the windows fixes? Something smells fishy here...[/citation]

[citation][nom]ashinms[/nom]...I mean this... http://techreport.com/articles.x/21865[/citation]

Read your link again. Chances are that the Windows 7 patches invoke the 0F mask rather than the 55 mask or regular auto-scheduling. That would be why they usually don't help very much and why in some cases, they are slightly detrimental to performance.

 

[leeashton]

[citation][nom]iam2thecrowe[/nom]no BS, 1x 8.8ghz BD module that is not stable can not beat a 3.3ghz i5 with 4 cores.[/citation]
you ae right its cant beat a core i5 at 3.3 GHz, its f**ken slaughters it, the BD at STOCK can beat the Core i5 at 3.3 GHz EASY!

 

[dftsa]

What benchmarks DONT show... I have a PC shop and personally own two pricey i7's heavily OC'd. Work everyday with 3960's and rampage boards etc BUT where the AMD's have impressed is in stuff NOT benchmarked such as Windows responsiveness. Eg, on the i7 there are still lags even though CPU is around 3% and has 4x Constellation drives in raid with 32GB OC'd gaming ram. Yet a simple APU doing the same job at a higher % shows no such lag both on fresh installs!? Heck we did an install yesterday of Win 7 via flash on 3400APU. Windows install took 7 mins from start to end. Impressive - and yet I'm an Intel boy. But credit where credit is due sometimes I'm embarrassed by the Intel Zealots. Ignorance is bliss I suppose.

As for only two cores, nothing uncommon MANY overclockers advertise 5Ghz on their 2011 sockets yet thats for two cores and for 4 / 6 normally around 4.6Ghz. But it will be advertised as 5Ghz reached.

Benchmarks can really be misleading.

 

[blazorthon]

[citation][nom]iam2thecrowe[/nom]no BS, 1x 8.8ghz BD module that is not stable can not beat a 3.3ghz i5 with 4 cores.[/citation]

If it's not stable, then they couldn't have shown us a CPUz screen shot, could they? Also, a single 8.8GHz BD module would beat the i7 in single/dual threaded performance, meaning that it would win in gaming performance in almost all games because even most quad threaded games can't make good use of more than one or two threads. The other threads are inefficient and don't do as much. That's why an i3 can't really beat a Pentium G860 and why a stock i5 doesn't really beat an i3 in most gaming situations.

 

[Guest]

gaming is not the only thing you do with CPUs. in 2d/3d creation/editing applications amd destroys intel. intel only gets this much attention because of dumb kiddies who only play crysis and mw on their computers and use (bad) games as benchmarks

 

[dftsa]

why a stock i5 doesn't really beat an i3 in most situations.

Agreed, Prime example , yesterday we had a client with an i3 2100 and Gigabyte Z68. Most of the benches and frame rates showed it being around 20% faster than i5 2500 on an H67 board. It is SO silly people actually think CPU makes much of difference in gaming. Motherboard and GPU is for gamers. My i7 (waste for gaming) uses around 15% on the most demanding games and forget silly SSD's for gaming.

 

[ern88]

It's still slow!!!!

 

[aviral]

Adding few more Hz ?
Are you not satisfied with the result to which it has been overclocked.

 

[holyprof]

[citation][nom]redyellowblueblast[/nom]I use a AMD-FX machine with Linux on it and I don't see any of the performance problems that I did on Windows 7. Which leads me top believe that the problem is more with Windows 7 not understanding the BD architecture, and not the architecture itself. The Linux kernel seems to be better suited at the moment at understanding the chip. Hopefully Windows 8 will fix it. (that is if Windows 8 will even be worth it.) And I'm still waiting on my Piledriver upgrade, AMD![/citation]

That proves once more what most of us already know - Bulldozer is a quick adaptation of server (highly parallel low IPC) processor to desktop.
How I miss the old Athlon XP and Athlon-FX days. Now Intel is king and hardcore AMD fans like me buy Intel Core processors. Still remember my first PC - it had a 20MHz AMD 80286 and was much faster than similarly priced Intel machines. At least my last graphics card is AMD (my last 3 cards were nVidia).

 

[whysobluepandabear]

I hit 597.64 GHz on my FX-over-9000. Where's my news article?

 

[Shin-san]

I saw someone heavily OCing Bulldozer and benchmarking it (Tomshardware I think it was), and the signs are there. If AMD can actually get the clocks to get to 4.2 GHz and higher, then Piledriver has a clear chance. Of course, most chip makers are having a hard time making chips consistently above 3.2 GHz, so that remains to be seen.