Clockability

[mali]

We are using the 3530, other devices like those Nokia phones use 3430. The wikipedia article is misleading, as there apparently are 600MHz versions and 720MHz versions of 3530 available.

 

[Dead1nside]

We are using the 3530, other devices like those Nokia phones use 3430. The wikipedia article is misleading, as there apparently are 600MHz versions and 720MHz versions of 3530 available.

Thank you for clarifying it for me. Does this mean that there are also multiple clockspeed versions of the DSP on 3530s also and the Pandora's DSP is only certified for 430Mhz operation, as per the official website?

 

[mali]

Afaik, the DSP clock ist tied to CPU clock and depends on multipliers/dividers. So overclocking a 600MHz 3530 to 720MHz should also raise the DSP's clock to 520MHz if you don't set a different multiplier/divider. I'm no expert so this is just layman's talk an guesswork

Edit:
Page 144,145 of the datasheet shows the frequency tables
http://focus.ti.com/lit/ds/symlink/omap3530.pdf

 

[kingoddball]

Can the GPU be overclocked?
I have not seen it mentioned anywhere but alot of cards these days can have the Mhz bumped.
it's on the offical OP homepage about overclocking. Upto 50%.

But GPU - ??

 

[mali]

See post #13

 

[kingoddball]

Beautiful.
Thanks Mali!

It's awkward reading a thread on a phone screen.
200mhz - Yay! Another quick thing: If it does get clocked to 200mhz and it crashed will it harm the GPU or is it fine and will just reboot to normal 166mhz without issue like the CPU (My GP2X crashed over 240mhz and is fine once I restart it).

 

[fischju2000]

Has anybody used the SGX at those speeds? It's always been either 55mhz or 110mhz in tests that I've seen. 166 and 200mhz is a stretch.

 

[Exophase]

Has anybody used the SGX at those speeds? It's always been either 55mhz or 110mhz in tests that I've seen. 166 and 200mhz is a stretch.

110MHz is the maximum speed. If you use the other clock source the best you can get is 90MHz. MWeston was confusing CORE_CLK (333MHz max) with the CPU clock (500MHz).

 

[kingoddball]

If it can run reliably at 166-200mhz then it could really help performance on many games (including the fabled Doom3 port [source dependent]) and even some of the higher end emulators possibly!

 

[andys]

Has anybody used the SGX at those speeds? It's always been either 55mhz or 110mhz in tests that I've seen. 166 and 200mhz is a stretch.

110MHz is the maximum speed. If you use the other clock source the best you can get is 90MHz. MWeston was confusing CORE_CLK (333MHz max) with the CPU clock (500MHz).

The 3630 actually has a better clock. This might also be the case for the 720Mhz versions of the 3530 - but I don't know

http://patchwork.kernel.org/patch/61019/

I don't exactly know what the difference is between the 3530 and the 3430, other than the name - maybe it is just an SKU?

 

[Exophase]

The 3630 actually has a better clock. This might also be the case for the 720Mhz versions of the 3530 - but I don't know

http://patchwork.ker...rg/patch/61019/

720MHz versions of 3530 are the same core, just higher binning. So it won't have different clock options. Interesting find on the 3630 though.

 

[WizardStan]

The processors in the Pandora are guaranteed to be stable at 600 mhz. TI has done some further testing and found that these processors are stable at 720 mhz as well. What we're not sure about is whether this 720mhz applies to existing chips or if it's a change in their process that only applies to future chips.

I think you got it wrong.

Chances are high that the 720MHz chips are out from the exact same fab lines as the 600MHz ones. Out of the line some tests are done to qualify chips at one frequency or the other (binning).

What this means is that if one buys a 600MHz part now, there are reduced chances to OC it as high as previous parts. OTOH if TI can't sell enough 720MHz parts, they might badge them as 600MHz parts.

My understanding is that OP had already bought the chips before TI started doing that. But the next batch might very well suffer from less overclockability.

I can't be wrong, you just said almost exactly what I said.
The key point being that new chips are coming off that are 720 that may or may not be the exact same as the 600s, just labeled higher, and the point still stands: assume 600 is maximum until

 

[ldesnogu]

I was commenting on this: "What we're not sure about is whether this 720mhz applies to existing chips or if it's a change in their process that only applies to future chips."

For me a process change means some change in the characteristics of the chip implementation (that's how the term is used in the EDA world), you perhaps meant something different

I bet there was no change except binning.

 

[andys]

The 3630 actually has a better clock. This might also be the case for the 720Mhz versions of the 3530 - but I don't know

http://patchwork.ker...rg/patch/61019/

720MHz versions of 3530 are the same core, just higher binning. So it won't have different clock options. Interesting find on the 3630 though.

I think it pretty much shows that the OMAP4 will have substantially better graphics performance than the OMAP3 - if they hold the same clockspeeds (~x2) and fit a 540 rather than a 530 you'd be looking at a rough quadrupling of performance. (Although of course, nothing is ever *that* simple)

I still want to see how good the SGX is when people push it to the limits, even the one on the bog-standard OMAP3 should be pretty nippy.

 

[WizardStan]

To rephrase what I said, "either they changed something to make them consistently more stable at 720 (change in their process that only applies to future chips), or they changed nothing and just started binning the 600s that worked at 720 (this 720mhz applies to existing chips), but we don't know which"
Is that better for you?

 

[craigix]

Even if OMAP4 boards are made available soon we probably won't be able to get one for quite some time, just the way it works.

Then it will probably be 2011 before the actual chips can even be bought in quantity.

Then the software, well, the OMAP3 software isn't even complete yet.

 

[ldesnogu]

Then the software, well, the OMAP3 software isn't even complete yet.

The good news is that most of the work that has to be done for OMAP3, which was one of the first commercially available ARMv7-A SoC, will be easily portable to OMAP4 (except perhaps for the DSP). At work, we routinely use A8-compiled filesystems on A9 chips, and changes to the kernel and U-boot are really minor and well understood.