Slashdot Mirror
AMD Launches Fastest Phenom Yet, Phenom II X4 980
MojoKid writes "Although much of the buzz lately has revolved around AMD's upcoming Llano and Bulldozer-based APUs, AMD isn't done pushing the envelope with their existing processor designs. Over the last few months AMD has continued to ramp up frequencies on their current bread-and-butter Phenom II processor line-up to the point where they're now flirting with the 4GHz mark. The Phenom II X4 980 Black Edition marks the release of AMD's highest clocked processor yet. The new quad-core Phenom II X4 980 Black Edition's default clock on all four of its cores is 3.7GHz. Like previous Deneb-based Phenom II processors, the X4 980 BE sports a total of 512K of L1 cache with 2MB of L2 cache, and 6MB of shared L3 cache. Performance-wise, for under $200, the processor holds up pretty well versus others in its class and it's an easy upgrade for AM2+ and AM3 socket systems."
I'll be waiting for the dust to clear with Bulldozer before I make a commitment for my next build. No reason to buy a $200 Phenom II X4 980 now when there is no application that needs that much power. If you buy a Sandy Bridge or a higher-end AM3 board/processor now, your average gamer or office worker won't be able to max it out for years -- unless he does video editing or extensive photo shop or if he has to get his DVD rips down to a 10 minute rip vs a 15 minute rip per feature film...
Might as well wait for the dust to clear or for prices to fall.
This [arstechnica.com] thread has some interesting information on possible BD performance.
.....
This is 301 posts with back and forth that looks basically to be speculation. Prove me wrong by quoting specific statements of those that have benched the [unreleased] bulldozer. Because otherwise, this link is basically a bunch of AMD fanboys fighting against Intel fanboys. But prove me wrong...
So clock speed means everything when comparing different CPUs and not their raw performance. Got it.
Furthermore, there is no 10 year old CPU that runs at 3ghz unless you did some absurd overclocking.
Bah, the speeders are wasting their money. Cores all the way. Half a GHz doesn't make up for two cores. When you've seen a 6-core kernel building it's hard to go back to start-stop. Even if you are a dumb Windows user, you will enjoy the shovelware being unable to slow you down.
Read this excerpt from an AMD management blog:
"Thanks to Damon at AMD for this link to a blog from AMD's Godfrey Cheng.
We are no longer chasing the Phantom x86 Bottleneck. Our goal is to provide good headroom for video and graphics workloads, and to this effect “Llano” is designed to be successful. To be clear, AMD continues to invest in x86 performance. With our “Bulldozer” core and in future Bulldozer-based products, we are designing for faster and more efficient x86 performance; however, AMD is seeking to deliver a balance of graphics, video, compute and x86 capabilities and we are confident our APUs provide the best recipe for the great majority of consumers. "
People, read between the lines.
What he is saying is that they can no longer compete with Intel on speed and have decided to concentrate on a balance at the low end priced points.
The days of the cpu wars are in fact over and Intel has won with Sandy Bridge.
Yes, I have been an AMD only fan for years but you have to face the reality that times have changed permanently in Intels favor and AMD's days are numbered.
Why else do you think Bulldozer is over a year late!
Oh, and AMD is prime for a buyout right now and there are rumors.
When AMD fails Intel will have a monopoly and the consumer will loose in the end.
Sad but true.
Prime95, in this context, is for convincing 0v3rcl0ckz0r kiddiez that their massive overclock is stable even though it's a terrible stability test. A prime number search program is not exactly the world's best method of achieving full test coverage of a CPU, no matter what a billion leetboy forums may tell you.
Just for example, according to its webpage, prime95 only uses 32MB of memory, which means it basically runs from cache on any modern CPU. Which in turn means you're not really exercising memory access much at all. Guess what's really, really important to test if you want to know how stable your system is, especially given that modern CPUs have integrated memory controllers? (Some overclockers are more sane and only do multiplier overclocking, but the focus of most is speed at any cost and the memory gets it too, and if they rely on prime95, well... not good.)
And then there's the issue that as a program which does nothing but manipulate large integer numbers, prime95 probably isn't touching anything other than the integer ALUs. Maybe MMX/SSE if you're lucky. So huge chunks of the CPU's datapath go untested.
Another consequence of that limited memory use is that it probably doesn't thrash the TLBs much, which means that OS pagefault handlers are rarely called, which means you're not testing the stability of all the VM machinery.
I could go on. Next to no I/O or interaction with peripherals. So on and so forth. Prime95 has a reputation vastly in excess of its true usefulness.
But the real problem is this:
Say you're doing something like the GP: using your computer to do scientific calculations which have to be right. You want to overclock, but because the results matter you want to find software which can help you validate that your computer is so stable that there's no chance of a crash. Or worse: silent data corruption. (Which I've personally observed when overclocking. Not fun when you don't discover it until after it's trashed a lot of data.)
Problem is, there is literally no end-user software which is an adequate stress test for this purpose. The only known way to get that kind of reassurance is to use factory automated test equipment (ATE). ATEs don't typically run software on the CPU under test. Instead, they make use of special test mode circuitry to quickly perform direct pass/fail tests on most circuits in the chip at any desired voltage/frequency/temperature operating point.
Well designed ATE tests can cover essentially every circuit. The factory uses ATE testers both to identify rejects and bin good chips into speed grades, but they're also the only way to be truly sure that an overclock will be 100% stable.
But you can't buy factory ATEs, and you can't get them to tell you the ATE test data for your chip, beyond a guarantee that it passed at the frequency they sold it to you as.
Which is why, if you're doing work which is important, such as scientific research, you damn well shouldn't overclock no matter how safe you think it is.
Don't forget that IPC isn't the be all and end all. If you're stalled due to cache misses, then IPC goes out the Window. Modern CPUs have much more cache and much faster buses to main memory than we had in 2004. That is a large reason as to why they're faster. They also have additional instructions that can do more work per instruction - so comparing IPC from CPUs released today to CPUs released last decade is even more meaningless.
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.