AMD unleashes first ever commercial “5GHz” CPU, the FX-9590

[Makaveli]

Not sure if serious or just trolling :hmm:

I was not being serious

How about this instead?
http://www.anandtech.com/bench/Product/697?vs=157

:wub: Gulftown still holding its own

not bad for 6 vs 8 cores .....

 

[Zucker2k]

These are not 5GHZ chips! These are 4.4GHZ and 4.7GHZ. The 5GHZ figure is surely going to be accompanied by a disclaimer, or I smell lawsuits.

 

[Centauri]

^Hmmm... I'm gonna go out on a limb and guess that you didn't bother to read a single post of this thread before replying to it.

 

[Zucker2k]

^Hmmm... I'm gonna go out on a limb and guess that you didn't bother to read a single post of this thread before replying to it.

Educate me, please.

Edit: I read most of it. Did I miss something?

 

[Magic Carpet]

 

[SiliconWars]

These are not 5GHZ chips! These are 4.4GHZ and 4.7GHZ. The 5GHZ figure is surely going to be accompanied by a disclaimer, or I smell lawsuits.

Had this been an issue both companies (and Nvidia) would have been sued many times over for their creative accounting of TDP, cores etc.

The chip probably spends a lot of its time at 5 GHz sitting on Facebook, like most chips.

 

[SlowSpyder]

These are not 5GHZ chips! These are 4.4GHZ and 4.7GHZ. The 5GHZ figure is surely going to be accompanied by a disclaimer, or I smell lawsuits.

I don't know about other countries, but here in America it seems like companies can get away with advertising just about anything as long as there is small print or a fast talking disclaimer at the end.

 

[galego]

These are not 5GHZ chips! These are 4.4GHZ and 4.7GHZ. The 5GHZ figure is surely going to be accompanied by a disclaimer, or I smell lawsuits.

AMD presented the new FX line as "up to 5GHz" and gave us a table reporting 5.0GHZ as the turbo frequency and 4.7 as base clock.

That is the same than Intel does when advertise their chips with the turbo frequency and latter gives the base clock. E.g. the i5-3570k is advertised as a "up to 3.8 GHz" processor.

http://ark.intel.com/products/65520

The only difference is that Intel advertising is based in a higher gap between turbo and base (about 1 GHz in some models).

 

[ShintaiDK]

AMD presented the new FX line as "up to 5GHz" and gave us a table reporting 5.0GHZ as the turbo frequency and 4.7 as base clock.

That is the same than Intel does when advertise their chips with the turbo frequency and latter gives the base clock. E.g. the i5-3570k is advertised as a "up to 3.8 GHz" processor.

http://ark.intel.com/products/65520

The only difference is that Intel advertising is based in a higher gap between turbo and base (about 1 GHz in some models).

My i5 4670 box says 3.4Ghz and turbo frequency is not even mentioned on it.

But again, only a fool would buy these 800-900$ 220W slow space heaters.

 

[inf64]

AMD Boxed CPUs have both base and Turbo listed clearly on the box itself.

 

[galego]

My i5 4670 box says 3.4Ghz and turbo frequency is not even mentioned on it.

<snip insult>

Boxes? Who mentioned Boxes? We are discussing the official presentation at E3 and slides.

Google your processor and you obtain

ARK | Intel® Core™ i5-4670 Processor (6M Cache, up to 3.80 GHz)

ark.intel.com/products/75047/ - Cached
ARK | Intel® Core™ i5-4670 Processor (6M Cache, up to 3.80 GHz) quick
reference guide including specifications, features, pricing, compatibility, design ...

It is only when you go to the page that you find the based clock of 3.4 GHz mentioned in small font

http://ark.intel.com/products/75047/

Thus Intel advertising turbo is fine but AMD doing it and you are going to call police? Stop. I am ignoring further nonsense about this.

 

[guskline]

Folks, think about it! AMD needs something to divert the attention away from Haswell on the desktop front. The SteamRoller must not be ready. The stategy is to divert attention away from Intel and the Haswell.. Simple. Anounce a new chip called the 9590 that can Turbo to 5Ghz. Up the vcore and the TDP and whala! Seriously, they might have a few more tweaks but that's it. I don't know what kind of silicon I really have in my FX 8350 and FX 8320 chips. However, I was able to crank the 8350 up to 5 Ghz for all 8 cores and boot into Win 8 AND I run it all the time at 4.6Ghz (21 x 219). My 8320 is running at stock voltage at 4.3Ghz. I use water coolers on both. Why all of the outcry about a 4.7Ghz base and 5Ghz Turbo chip? The only real amazing thing is the possible price of $800. Owning what I do, (3770k/8350/8320) the $800 price would make me upgrade to a LG2011 mb and a 3930k.

 

[mrmt]

Folks, think about it! AMD needs something to divert the attention away from Haswell on the desktop front.

And find buyers for those single die Piledriver SKUs they have in stock, because AMD just killed single die "unmitigated failure" SKUs on the server market. Only MCM chips from now on.

 

[ShintaiDK]

So AMD is now scaling down on the desktop from "8" to "4" cores. That company sure knows how to make fun of people.

 

[guskline]

BTW, to put some perspective to a 4.7Ghz base clock for the FX 9570 chip (4.7base/5 turbo) I ran Aida64, Cinebench 11.5 and Passmark 8 cpu tests on my FX 8350 OC'd to 4.7ghz (23.5 x 200) with the vcore at 1.48 for stability. I ran the 8350 stock, at 4.7Ghz, my 3770k at stock and OC'd to 4.4Ghz. My results:

I would love if a Haswell 4770k owner could run the same tests at stock and at @4.2 Ghz for some comparison. I realize that these tests are a small snippet and perhaps not a fair comparison. However I wanted to focus on cpu tests. When you introduce tests on games the type of gpu seems to play a much more important role than the cpu.

Finally, because the FX 9570 will be 4.7 base and turbo 5, I expect its scores to be higher than what I posted for a FX 8350 at 4.7Ghz but not significantly higher.

 

[Enigmoid]

Weird that in the passmark FP math test the 8350 comes out ahead considering it does so bad in the other FPU tests.

 

[inf64]

So AMD is now scaling down on the desktop from "8" to "4" cores. That company sure knows how to make fun of people.

So you think they won't be making 8T PD based FX any more? Please come back to reality.

Weird that in the passmark FP math test the 8350 comes out ahead considering it does so bad in the other FPU tests.

It depends how software is written. If FP benchmarks were fully optimized for AMD 15h the difference could be much higher.

 

[ShintaiDK]

So you think they won't be making 8T PD based FX any more? Please come back to reality.

Thats not what I said. But AMDs new desktop line for a foreseeable future will be 4T.

 

[sushiwarrior]

Thats not what I said. But AMDs new desktop line for a foreseeable future will be 4T.

AMD's new desktop line for the forseeable future is all Kaveri, which is indeed a 4C design. So what's your point? "Durr durr AMD sux they forgot how to make 8C CPUs lawl"? FX Steamroller will certainly be 8C.

 

[inf64]

AMD's performance desktop will continue to be 8T PD based. Plus we have no official 2014 desktop roadmap yet. Kaveri is ~204mm^2 on 28nm node with SR cores and 512SP iGPU. Do you think AMD is unable to stitch one more module if needs be? SR module @ 28nm with full L2 cache is between 25 and 30mm^2 (roughly). It's peanuts. 3M Kaveri would probably be equal to 4M PD 83xx in MT apps and crush it in ST ones.

 

[Sweepr]

AMD's performance desktop will continue to be 8T PD based. Plus we have no official 2014 desktop roadmap yet. Kaveri is ~204mm^2 on 28nm node with SR cores and 512SP iGPU. Do you think AMD is unable to stitch one more module if needs be? SR module @ 28nm with full L2 cache is between 25 and 30mm^2 (roughly). It's peanuts. 3M Kaveri would probably be equal to 4M PD 83xx in MT apps and crush it in ST ones.

There is no 3M APU & 4M/6M SR in AMD server roadmaps. They wont release server versions of these imaginary chips (in 2014), so chances of an enthusiast focused desktop release are pretty much non-existent . They might be able to do it, but they probably dont want to (same as the LGA1150 desktop Broadwell drama). Please accept that before it gets embarassing, your Steamroller FX 6M/12T @ 4GHz beating Haswell-EP 8C/16T in 2014 predictions were already fun enough.

 

[Tsavo]

AMD should have called this the Emmett Brown Special.

It doesn't post unless it has a 1.21 jiggawatt power source.

Free lightning rod with every purchase!

 

[ShintaiDK]

AMD's new desktop line for the forseeable future is all Kaveri, which is indeed a 4C design. So what's your point? "Durr durr AMD sux they forgot how to make 8C CPUs lawl"? FX Steamroller will certainly be 8C.

And what chip and socket should this 8T steamroller FX come from? Servers will be 4T in the 1P segment. So why would the desktop get 8T? Seems counterintuitive.

 

[mrmt]

AMD's performance desktop will continue to be 8T PD based. Plus we have no official 2014 desktop roadmap yet. Kaveri is ~204mm^2 on 28nm node with SR cores and 512SP iGPU. Do you think AMD is unable to stitch one more module if needs be? SR module @ 28nm with full L2 cache is between 25 and 30mm^2 (roughly). It's peanuts. 3M Kaveri would probably be equal to 4M PD 83xx in MT apps and crush it in ST ones.

What's your point in telling the size of the core here? Do you think that AMD engineers are a bunch of screw ups wasting silicon space with the uncore? You don't know how much infrastructure you need to support 3M, so in fact 2M might be the limit of AMD small designs.

 

[NTMBK]

AMD should have called this the Emmett Brown Special.

It doesn't post unless it has a 1.21 jiggawatt power source.

Free lightning rod with every purchase!

Where we're going, we don't need IPC.