Conroe Vs FX-62 Benchmark - HKEPC

[TheSlamma]

The power consumption has me sold.

Isn't there supposed to be LV models out also? like in the 35-40W range?

 

[DidlySquat]

Originally posted by: Viditor

Originally posted by: Furen

Originally posted by: DidlySquat
ROFL !!!! what speed bump ? Are you talking about a 3 GHz FX ? because if you do then you don't understand that with the current Athlon64 core revision and manufacturing process it is quite hard to reach 3 GHz. Even hardcore overclockers that try out many cherry picked CPUs have a hard time to reach 3 GHz, especially on air cooling. So I don't think AMD will be able to release a 3 GHz FX CPU in any significant quantities, until they move to 65 nm or significant;y optimize their manufacturing process.

My post was actually in reply to kidcool321's, who said that he'd bet AMD has something better than Conroe lined up for the end of July. I think a speed bump is doable though, of course, the chips would be insanely cherry-picked. Now, I don't think it LIKELY, but it is possible. Notice that I suggested a 5400+ (which would be a 2.8GHz/512kBx2 part) because it is simply much more likely than an FX-64.

While I can't link to it, I'm fairly sure that FX-64 is on the roadmap for before Xmas...

ROFLMAO @ morons ! B4 xmas is not the same as what he said that AMD will release a speed bump within a month. And x2 5400+ (2.8 GHz 2x512KB) is irrelevant to counter conroe because it's not even as fast as a FX-62. The question is if they can quickly release a faster top of the line CPU and the answer is NO.

 

[Furen]

Originally posted by: DidlySquat
ROFLMAO @ morons ! B4 xmas is not the same as what he said that AMD will release a speed bump within a month. And x2 5400+ (2.8 GHz 2x512KB) is irrelevant to counter conroe because it's not even as fast as a FX-62. The question is if they can quickly release a faster top of the line CPU and the answer is NO.

Wow, you're a little asshole, huh? A 5400+ speed bump anywhere will "help" things because AMD is not only competing in the high end but throughout all of the product line. I never mentioned an FX-64 as a viable product for august (which happens to be Q3) and I dont think it likely unless AMD introduces embedded SiGe into its 90nm process (kind of like it did with DSL stressing on the FX-55).

 

[coldpower27]

Originally posted by: Furen

Originally posted by: DidlySquat
ROFLMAO @ morons ! B4 xmas is not the same as what he said that AMD will release a speed bump within a month. And x2 5400+ (2.8 GHz 2x512KB) is irrelevant to counter conroe because it's not even as fast as a FX-62. The question is if they can quickly release a faster top of the line CPU and the answer is NO.

Wow, you're a little asshole, huh? A 5400+ speed bump anywhere will "help" things because AMD is not only competing in the high end but throughout all of the product line. I never mentioned an FX-64 as a viable product for august (which happens to be Q3) and I dont think it likely unless AMD introduces embedded SiGe into its 90nm process (kind of like it did with DSL stressing on the FX-55).

Yeah, if it were possible a bump to 2.8GHZ for a mainstream part would be nice. Though from the data we have now the jump from 2.6GHZ to 2.8GHZ for a Dual Core seems to take a huge toll on power consumption, I wonder if a 2x512KB part has as much issue as typically cache doesn't require that much energy to operate and sholudn't be the the largest factor contirubuting to the power consumption increase..

A 5400+ if introduced, would be just about matching Intel's lowest Conroe processor which isn't too bad to say the least.

Regarding the FX64 I think that should probably be the last Core on the 90nm process, for Dual Core, which would be a nice place to end the 90nm process on.

 

[dexvx]

Originally posted by: PlatinumGold
ya, and who's "fault" was that?

no one told intel they had to abandon the p3 architecture for the netburst.

shoot, if intel had put all their resources into developing the p3 architecture instead of develping and marketing netburst, they would have BURIED AMD by now.

LoL?

Conroe is nothing like the P3 Architecture. Pentium-M optimized binaries perform worse than Pentium-4 optimized binaries. If anything, its still closer to Pentium-4 than to Pentium-M, which is a far far off-shoot of what you claim to be based off of Pentium-3.

 

[DrMrLordX]

Originally posted by: TheSlamma
The power consumption has me sold.

Isn't there supposed to be LV models out also? like in the 35-40W range?

You mean on Conroe? There's a ULV Merom variant supposedly slated for release that will have a 5W TDP. I thought there were going to be LV and ULV Woodcrests as well, but all I see out there is talk of Sossaman which is just a Yonah variant for servers that should already be on the market.

 

[DrMrLordX]

Originally posted by: dexvx

LoL?

Conroe is nothing like the P3 Architecture. Pentium-M optimized binaries perform worse than Pentium-4 optimized binaries. If anything, its still closer to Pentium-4 than to Pentium-M, which is a far far off-shoot of what you claim to be based off of Pentium-3.

This is true, but I think PlatinumGold's sentiment may have a grain of truth to it. Had Intel supported development of the old P6 core(instead of diverting resources to Netburst and Itanium), they could have made the jump to Tualatin much faster than they did (instead of getting owned by the Thunderbird for so long). What they would have released after Tualatin is anyone's guess, but considering how long AMD sat on the 1.4 ghz Thunderbird, Intel would have a lot of wiggle room.

 

[dmens]

all this foresight on "if there were no P4" is self-evident and annoying. perhaps intel should have kept timna alive, but ditched rdram support, and at the same time have the backend widened to 4 uop (and take a severe clock speed hit), but time the release with 90nm to regain the cycle time. oh hell, might as well toss in all the good stuff from P4, like in-flight RRF, and the trace cache, since it would've been cool to predict the explosion of front latency. Don't forget SMT too, but make sure it was more configurable the first time over...

damn thats some sharp 20/10.

 

[archcommus]

Originally posted by: DrMrLordX

Originally posted by: dexvx

LoL?

Conroe is nothing like the P3 Architecture. Pentium-M optimized binaries perform worse than Pentium-4 optimized binaries. If anything, its still closer to Pentium-4 than to Pentium-M, which is a far far off-shoot of what you claim to be based off of Pentium-3.

This is true, but I think PlatinumGold's sentiment may have a grain of truth to it. Had Intel supported development of the old P6 core(instead of diverting resources to Netburst and Itanium), they could have made the jump to Tualatin much faster than they did (instead of getting owned by the Thunderbird for so long). What they would have released after Tualatin is anyone's guess, but considering how long AMD sat on the 1.4 ghz Thunderbird, Intel would have a lot of wiggle room.

Oh man...I miss my 1.4 Tbird, second AMD processor I ever owned. Ran hot but performed great.

 

[Howard]

I want to see more minimum FPS benches at 1680x1050+ resolutions...