AMD Shanghai/Deneb Review Thread

[Caveman787]

Zucker....

As always the AMD downer pill....

I would think most people wouldn't want to run their processor's 70*c+. Why because theey're likely to run into heat build up problem's and possible overheating of other component's.

Or as the case of some out there frying alive in thier houses

Sound like fun to you? then enjoy!

The fact is still that keeping either within sensible temps deneb will clock a few mhz higher and that's just on air.

[qurious63ss]

Zucker....

As always the AMD downer pill....

I would think most people wouldn't want to run their processor's 70*c+. Why because theey're likely to run into heat build up problem's and possible overheating of other component's.

Or as the case of some out there frying alive in thier houses

Sound like fun to you? then enjoy!

The fact is still that keeping either within sensible temps deneb will clock a few mhz higher and that's just on air.

Ahh... that's the beauty of C2Q, it will overclock crazy and stay cool while on average giving higher ICP then PII.

[Caveman787]

I still think deneb fully overclocked with overclocked nb/hypertransport will have a chance of being equal or a little faster then c2q.

[imsochobo]

Got phenom 1.

Crash is @ 84 C.

Orthos can run 24H at temperatures of 76, so no issue there, the limit is clearly not 60's.

Temperature decreases the max clock, there is no shock there, if you can run 4.2 ghz on air the first 10 minutes with orthos @ 60's but then goes over 70 and fail, you can probaly run 4-4,1 somewhere, easy as that.

Athlon x2's stop at 71 for my windsor 5600+ and the 5000+ brisbane stopped at 69.

according to their temp sensors.

My pentium 4 70 C idle and load, just no diffrence, and it died after a year, stock cooler, all stock settings.

[iocedmyself]

Yes zucker good summation, i7's kinda said "power efficiency....with performance....forget that, that's just greedy talk!"

No matter what the benchmarks may show on i7, truth is that it's still very far from perfected, the thermal envelope should be enough to show that fact. However, deneb can take alot more volts then I7 from what i've seen.

http://hothardware.com/Articles/Over...cessor/?page=5

While i've already shown that you can do 3.2 at 1.0875 on deneb (where as i7 takes 1.13 for 2.6) though you can actually go lower. keep in mind that that 3.8 i7 runs at around 1.37v.....and still gets that hot...and still has that high of power consumption. The link above shows 100-114w increase with cpu under load over idle. 316-368 between 3 and 4ghz with load. Phenom II load power consumption is below to I7's idle.


Denebs have a kill temp of 60 or 65c this is true, but as caveman pointed out, who the hell wants to run their chip at 70c if it can be avoided.

[Astennu]

Yes zucker good summation, i7's kinda said "power efficiency....with performance....forget that, that's just greedy talk!"

No matter what the benchmarks may show on i7, truth is that it's still very far from perfected, the thermal envelope should be enough to show that fact. However, deneb can take alot more volts then I7 from what i've seen.

http://hothardware.com/Articles/Over...cessor/?page=5

While i've already shown that you can do 3.2 at 1.0875 on deneb (where as i7 takes 1.13 for 2.6) though you can actually go lower. keep in mind that that 3.8 i7 runs at around 1.37v.....and still gets that hot...and still has that high of power consumption. The link above shows 100-114w increase with cpu under load over idle. 316-368 between 3 and 4ghz with load. Phenom II load power consumption is below to I7's idle.


Denebs have a kill temp of 60 or 65c this is true, but as caveman pointed out, who the hell wants to run their chip at 70c if it can be avoided.

You cant compare voltages from a AMD chip with those of a Intel chip.
Intel uses High K and AMD low k SOI. And there are many more differences. Chips can be designed for low or high voltages. Same goes for temperatures. So you cant compare them 1:1.

[qurious63ss]

Yes zucker good summation, i7's kinda said "power efficiency....with performance....forget that, that's just greedy talk!"

No matter what the benchmarks may show on i7, truth is that it's still very far from perfected, the thermal envelope should be enough to show that fact. However, deneb can take alot more volts then I7 from what i've seen.

http://hothardware.com/Articles/Over...cessor/?page=5

While i've already shown that you can do 3.2 at 1.0875 on deneb (where as i7 takes 1.13 for 2.6) though you can actually go lower. keep in mind that that 3.8 i7 runs at around 1.37v.....and still gets that hot...and still has that high of power consumption. The link above shows 100-114w increase with cpu under load over idle. 316-368 between 3 and 4ghz with load. Phenom II load power consumption is below to I7's idle.


Denebs have a kill temp of 60 or 65c this is true, but as caveman pointed out, who the hell wants to run their chip at 70c if it can be avoided.

I wonder how C2Q would do?? Only reason I'm asking is because it appears that C2Q performs slightly faster then PII so it would make it a fair comparison.

[TL1000S]

Seriously guys!
Are you really believing "temps" (*numbers*) reported by software?
Supposedly "measured" by some hardware within the core(s)?

I have several K8's that shows 5-10 degrees idle, and 30-40 full load (depending on clocks/volts).
I have had numerous Intel CPU's (P4, C2D, C2Q...) that shows completely different "numbers" between similar models/revisions.
Even seen that upgrading Bios results in 20+ degrees differences.

The *ONLY* way to compare *true* temps are by hardware mods to the CPU (installing probes).
ONLY compare temps for CPU's from a given model/installation!

Please also remember that many are not that clever when it comes to mounting heatsinks (ecpecially on S775).

As long as the system shows no signs of instability absolute numbers ("temps") are guidelines for you system/OC/volts.
This said most people (including myself) wants to see low numbers.