AMD FX-Series (Bulldozer) & 990FX Chipset (AM3+) Discussion

[Cephas]

From that xbit article...

Early this week it turned out that AMD had advised its partners that its next-generation high-performance desktop chips would only be launched in September. Apparently, the currently available B0 and B1 stepping Zambezi/Bulldozer processors can function at around 2.50GHz/3.50GHz (nominal/turbo) clock-speeds and at such frequencies they cannot deliver performance AMD considers competitive

Assuming that these figures have some truth to them, should we be worried about the OC potential of the B2 chips?

If the current stepping is having issues being SB competitive at 3.5, how high is B2 realistically going to get? And will it have any headroom when it gets there?

[DiGiTaL MoNkEY]

Originally Posted by Cephas

From that xbit article...



Assuming that these figures have some truth to them, should we be worried about the OC potential of the B2 chips?

If the current stepping is having issues being SB competitive at 3.5, how high is B2 realistically going to get? And will it have any headroom when it gets there?

It's hard to say really.

I'm sure AMD have a specific voltage/power envelope they are trying to hit with a certain performance level. So it's tricky for them.

[Cephas]

That's true.

What it'll do under 125w on a stock cooler vs what we're prepared to do could be very different.

[Ethan W]

Originally Posted by Cephas

That's true.

What it'll do under 125w on a stock cooler vs what we're prepared to do could be very different.

But if it's anything like the Phenoms, heat will drastically affect the max OC. That needs to stay down. If they were to keep the same stepping (theoretically) and they're hitting 125W TDP at 2.5GHz, I don't want to imagine the same chip at 3.5GHz. Well over 200W, that's for sure.

[Cephas]

32nm might make heat less of an issue though.

[Ethan W]

Originally Posted by Cephas

32nm might make heat less of an issue though.

Clearly not, if they're having issues running past 2.5GHz and staying under 125W.

[Gillyske]

wow this does look very bad indeed.
Because currently what amd considers competitive is like 30% weaker clock for clock.

[James_B]

Semiaccurate has a piece saying the launch dates for the server and desktop chips got flipped around, if true not so bad but pretty horrific communication.

[.Radiant]

Originally Posted by James_B

Semiaccurate has a piece saying the launch dates for the server and desktop chips got flipped around, if true not so bad but pretty horrific communication.

well given that most of the news is sourced from "Our guy in the industry" I'm not surprised.

[Cephas]

It might not be a heat issue.
Don't most SB chips hit some electrical limit well before heat becomes an issue?
Heat can be cooled, other issues aren't so easily addressed.

[Linkin]

Creator and Author of the highly successful CPU and Hardware Monitoring software CPU-z “Franck Delattre” has released the first official Bulldozer FX-Processor CPU-z Screenshot. The CPU-z screenshot details a FX-based processors (Engineering Sample) featuring 8 Cores and 8 Threads.

The new CPU-z 1.56.4 will add support for the FX-Series processors. Details in the screenshot show that the CPU is running at a clock speed of 1.4Ghz consuming 0.950V (TDP isn’t listed). Either the chip used was part of the older revision which had performance issues or due to power gating feature the cores were down clocked due to idle usage. Socket AM3+(942) has been mentioned and the chip has a total of 8mb L3 cache. The author also detailed some on the new features of Bulldozer CPU’s and added CPU-z support as follows:

- Support 2 Module 4-core (FX-4110), 3 Module 6-core (FX-6110) and 4 Module 8-core (FX-8110 8130P) Zambezi processor
– Support for Socket AM3 +
– 32-nm SOI HKMG Technology
– 213 million transistors per module
– Support HTT 3.1 bus (3.2GHz, 25.6GB / s, 16bit bi-directional transmission)
– Supports up to 8M L3
– Fundamental frequency of 3.2GHz (may be higher?)
– Turbo Core (up 1GHz, load up all the core 500MHz)
– Supports a maximum TDP 95W and 125W
– Support voltage VID 0.8V-1.3V
– Support for Advanced Power Management (Power Gating / Clock Gating / Sectored L3)

[Deja Vu]

Originally Posted by Ethan W

Clearly not, if they're having issues running past 2.5GHz and staying under 125W.

That might have less to do with the 32nm process as more to do with the current CPU layout, such as high-speed high-leakage transistors used unnecessarily causing higher power use. Alternatively it might not be a heat issue at all, maybe some area of the chip just cannot get past that speed currently because of its design.

[Ethan W]

Originally Posted by Deja Vu

That might have less to do with the 32nm process as more to do with the current CPU layout, such as high-speed high-leakage transistors used unnecessarily causing higher power use. Alternatively it might not be a heat issue at all, maybe some area of the chip just cannot get past that speed currently because of its design.

I wasn't saying that the 32nm had anything to do with causing the heat issue, but more that it isn't proving enough to keep things cool enough.

From that CPU-Z shot, do the FXs have newer SSE instructions? That could help things.

[goatman]

Originally Posted by Ethan W

But if it's anything like the Phenoms, heat will drastically affect the max OC. That needs to stay down. If they were to keep the same stepping (theoretically) and they're hitting 125W TDP at 2.5GHz, I don't want to imagine the same chip at 3.5GHz. Well over 200W, that's for sure.

you can have the same TDP if you lower the voltage

[goatman]

Originally Posted by Cephas

It might not be a heat issue.
Don't most SB chips hit some electrical limit well before heat becomes an issue?
Heat can be cooled, other issues aren't so easily addressed.

if by SB you mean southbridge then yes, the clocks for sata, pci etc come from the SB clock so if you increase that you increase them.
HOWEVER yes if the cpu has an unusually slow "path" then you're going to need to raise the voltage of the entire chip to make it any faster thus increasing the TDP.

Remember P=(V^2)/R, any increase in voltage will have an exponential increase on the TDP