Overclockers Vie for World Record with AMD's "Vishera" FX-8350 CPU

24

Comments

+ Add a Comment
avatar

Hilarity

Fail. Can barely match (if that) a 3570K but at higher power consumption. Overclock it and it sucks an even bigger pile of power. Utterly useless compared to Intel. Have thrown out and replaced all my rubbish AMD systems with Intel and I've never been more satisfied.

avatar

Andrew.Hodge

You people just can't get enough attention, can you?

avatar

praack

wow - so with all the overclocking is the visha- thingie now finally equal to a i5? oh wait - it is now only dual core.,,, sigh

I am a sad sad AMD guy, looking at a continued wasteland of amd parts- living off my old gen, and seeing nothing from the AMD stable i can recommend to anyone.

Now we know why AMD has had such a bleed of execs as of late, and unless they get a new architecture in for the desktop there will not be anything of note for AMD x86 brand for another 5 years

sigh

avatar

Andrew.Hodge

http://www.cpubenchmark.net/cpu.php?cpu=AMD+FX-8320+Eight-Core&id=1782

It's almost the exact same as the 3770K in well threaded software and costs less money. I wouldn't exactly call that a failure. I really can't think of anything today that would need a lot of performance that is single threaded anymore, at least that I use. It seems that anything that needs a lot of power is multithreaded these days, and if it is single threaded, you won't notice the difference. I sure as hell don't. In that respect, Piledriver is an exelent choice for anyone who wants to save a huge amount of money.

If you can think of an explicitly single threaded program that is of any count today, I would love to hear about it, though.

avatar

illusionslayer

Have a look at some benchmarks that actually mean something: http://www.anandtech.com/bench/Product/697?vs=551&i=499.501.502.503.504.505.506.327.25.26.29.38.39.40.41.42.43.45.46.344.345.53.54.55.60.61.62.129

Not only does the Intel top of the line offer more performance at a lower wattage in most threaded tests, it kicks the crap out of single threaded tests.

But it's really not fair to compare these two based on price since intel likes to add a premium to their flagships.

In fact, if you compare the 3570k, the chip most closely related in price, you see that not only does it perform about as well on threaded tests with half the cores, it beats AMD at it's own game by offering better gaming numbers and still winning single-threaded tests.

avatar

Andrew.Hodge

I see a lot of really high single threaded winds for the 3570, but not much of a reason for me to buy it based on multithreaded scores. My point still remains: the 8350 isn't crap. If you want a good performing computer with hardware that doesn't suck your bank account dry, the 8350 is the way to go. So far as power draw I can only see that mattering in two situations: One, if I were a folder and I was running my computer 24/7, the other being if I had the same power supply in an upgrade and was teetering on the edge of what it could supply for my computer.

avatar

dgrmouse

Andrew.Hodge said, "If you can think of an explicitly single threaded program that is of any count today, I would love to hear about it, though."

You've gotten it backwards. Nothing is explicitly single-threaded - software only becomes usefully (discounting threading for UI responsiveness, etc) multi-threaded with explicit design. Aside from the complications required to manage data sharing (even with shared memory models) and a massive jump in the difficulty of troubleshooting errors, a program must be structured in one of several patterns designed to maximize the effectiveness of concurrency.

That said, there might possibly be one or two or a few million folks using iTunes to rip their CDs.

Andrew.Hodge also said, "It seems that anything that needs a lot of power is multithreaded these days, and if it is single threaded, you won't notice the difference."

I can't understand your reasoning here, but your conclusions are very much wrong. If single-threaded task X takes ten hours on single-core processor A and ten minutes on single-core processor B, you're going to notice. All other variables held equal, a 1GHz single-core system will smoke a 500MHz dual-processor system on every task, every day of the week and twice on Sunday. Bearing this in mind, of course single-threaded performance is important.

avatar

Andrew.Hodge

"You've gotten it backwards. Nothing is explicitly single-threaded - software only becomes usefully (discounting threading for UI responsiveness, etc) multi-threaded with explicit design. Aside from the complications required to manage data sharing (even with shared memory models) and a massive jump in the difficulty of troubleshooting errors, a program must be structured in one of several patterns designed to maximize the effectiveness of concurrency."

What I meant by that was that the majority of software written to be sold was written to perform as needed. That being said, multicore processors have been selling for years now (since '05, I believe), and multiprocessor computers have been around even longer. Almost every program out there has some form of performance-related threading going on somewhere in the mix. With that being said, there are very few modern programs that don't have some form of multithreading going on. As for iTunes, it's Apple. What did you expect? XD
Either way, it's bottlenecked by how quickly the DVD drive can get data to the CPU, so I don't really see how multithreading can make any major difference, although I don't use iTunes, so I'm not really qualified to comment on that piece of software.

Andrew.Hodge also said, "It seems that anything that needs a lot of power is multithreaded these days, and if it is single threaded, you won't notice the difference."

"I can't understand your reasoning here, but your conclusions are very much wrong. If single-threaded task X takes ten hours on single-core processor A and ten minutes on single-core processor B, you're going to notice. All other variables held equal, a 1GHz single-core system will smoke a 500MHz dual-processor system on every task, every day of the week and twice on Sunday. Bearing this in mind, of course single-threaded performance is important."

The differences in transcoding times and FPS scores between the 8350 and 3770k on single threaded benchmarks are within a few seconds and FPS on almost every benchmark. I could understand one hour and ten, but that isn't the case. I's more like three and a half minutes VS three and three quarters.

avatar

limitbreaker

I'd personally take a 8350 over a i5 and a i7 over an 8350 based poorly on performance and not even price.

avatar

lordfirefox

Disabling cores to overclock makes the overclocking completely pointless. And only 2GB of RAM? Still running XP?

avatar

streetking

regardless, an oc record is an oc record. and the record breakers usually use small amounts of ram. im not sure why, but they do

avatar

harz3000gt

I am pretty sure that they use only a single DIMM as it is easier to keep cool and just one less piece to potentially fail under such extreme conditions. They aren't using these games to play Modern Warfare, just enough to reach a record. Like land-speed records, they don't continue to drive 700+ MPH, they do what they need to to obtain the record then back off.

avatar

joshnorem

Thanks for everyone's feedback. We've ammended the article :)

avatar

Andrew.Hodge

Wow. Very polite an controlled after such agression. I'm inpressed!

avatar

limitbreaker

They behave professionally enough not to be sucked in by trolls, that would make me lose respect if it happened.

avatar

limitbreaker

Take it easy guys... Mistakes happen and we're all merely just humans.

avatar

Nastyman

...so the guy is dyslexic...give him a break....

Nasty

avatar

markstrelecki

How the mighty magazines have fallen!

These kinda proofreading gaffes are more typical of the "semi-pro" review websites, not the hallowed and holy MaximumPC.

To get the headline one way, and every reference in the article wrong is inexcusable, lackadaisical, unprofessional and just plain LAZY.

How many times do I have to offer my proofreading services to you guys before you realize that this is the kinda stuff the amateurs do, not the pros.

Josh, are you aware of this crap, and can you please do something to fix it? And Paul - C'MON, MAN!

MARK (EDITOR IN TRAINING) STRELECKI
ATLANTA, GA. USA

avatar

harz3000gt

Not sure what the article originally said since it has since been amended, but what you said was a whole lot of douche-baggery. And yes, I know that "baggery" is not a word, but I'm sure the sight of it is driving you nuts right now.

avatar

Paul_Lilly

I transposed numbers and referenced the FX-8350 as an FX-8530. In my defense, I typed the post while standing on my head in front of a mirror, so that may have played a role in the gaffe.

avatar

Shadow Death

You could just politely respond and keep your self-righteous self-promoting comments to yourself. The little gimmick you threw in the middle of your name is something a child would do and is laughable. That's a unprofessional way of saying INTERN.

avatar

illusionslayer

Edit: Wrong post to reply to.

avatar

T3RR0RH4WK

Did you mean FX-8350 where it says "8530?"

avatar

limitbreaker

Nice, a 8530fx and it has exactly the same specs as the 8350!! Okay sorry

Log in to MaximumPC directly or log in using Facebook

Forgot your username or password?
Click here for help.

Login with Facebook
Log in using Facebook to share comments and articles easily with your Facebook feed.