AMD - Piledriver chitchat

[DanceswithUnix]

I did wonder about that - either small die area increase, or reduce the number of shaders in the IGP (6 cores + 384 shaders? sounds reasonable to me....)

128 shaders, 8 cores, and some L3 cache please

[scaryjim]

And quad-channel DDR4, presumably? Sounds good to me - way better than the nasty 960G northbridges we're currently having to put up with if we want 8 cores and onboard graphics....

EDIT:
Wonder if we should start talking to AMD's semi-custom business? Create our own line of CPUs. Think they'll work on a commission only basis?

[DanceswithUnix]

And quad-channel DDR4, presumably? Sounds good to me - way better than the nasty 960G northbridges we're currently having to put up with if we want 8 cores and onboard graphics....

EDIT:
Wonder if we should start talking to AMD's semi-custom business? Create our own line of CPUs. Think they'll work on a commission only basis?

Apparently it costs $1M to make a set of masks, and I will happily stick with DDR3 for now and take FM2+ socket compatibility. So AMD, do a kickstarter, I will be there.

[watercooled]

I find it hard to believe something like an 8C, 128SP processor wouldn't sell well enough to at least offset the costs of producing it, and I would've thought it would be good for brand image.

However it seems AMD is just being persistent in trying to bootstrap the APU/HSA market with their current strategy?

[Bonebreaker777]

128 shaders, 8 cores, and some L3 cache please

Fat chance I would say. Or you could try to write to Santa Claus, Christmas is just around the corner.

[Bonebreaker777]

Apparently it costs $1M to make a set of masks, and I will happily stick with DDR3 for now and take FM2+ socket compatibility. So AMD, do a kickstarter, I will be there.

I would jump in too (only if the minimal contribution for an actual product would be round £100). FM2+ ITX and wooo...

[scaryjim]

Sounds to me like we need to put together a HEXUS consortium...

[Bonebreaker777]

I would jump in too (only if the minimal contribution for an actual product would be round £100). FM2+ ITX and wooo...

I meant maximal.

[abaxas]

Hmmm,

Desktop PC needs and upgrade... possible understatement as it's at atom 330 with 1 gig of ram.

So the question is..

Althlon 5350 + itx board + 4gig
or.
A8-7600 + itx board + 8 gig.

Anyone know the cheapest itx board that supports the A8 in 45w mode?

[Bonebreaker777]

Hmmm,

Desktop PC needs and upgrade... possible understatement as it's at atom 330 with 1 gig of ram.

So the question is..

Althlon 5350 + itx board + 4gig
or.
A8-7600 + itx board + 8 gig.

Anyone know the cheapest itx board that supports the A8 in 45w mode?

Depends on the final utilisation but I would say the A8 7600 plus AsRock FM2A88X-ITX plus the 8GB of RAM (the MoBo should't have problems with the 45W mode as it was used in the following review: http://www.legitreviews.com/amd-kave...-review_147879).

[abaxas]

It runs a pc based oscilloscope and logic analyzer + basic office/browsing.

The current machine is fine up to the point that browsers have got slower and slower over the years and I quite fancy playing a few games too. Not really modern stuff but quite fancy playing DS1/DS2 and neverwinter nights again.

[scaryjim]

I used to play NWN on a 1.3GHz dual core Core 2 and Intel G45 graphics (can't remember what the actual IGP was called it but it was pretty poor ) so either would cope with that fine. The A8 will obviously give you more longevity and overall better performance, so really it's a question of whether you're happy spending the extra money.

[badass]

128 shaders, 8 cores, and some L3 cache please

Naaaah. More IPC with the 4 cores (maybe the L3 cache can help there) and 128-256MB Stacked huge bandwidth, ultra low latency RAM

Suspect what will actually happen is better performance at low power, barely changed performance at higher power, no stacked RAM and DDR3 (possibly DDR3/DDR4)
Why oh why do you strangle your APU's AMD!
It's like asking Mo Farah to run when breathing through a straw!

[abaxas]

I used to play NWN on a 1.3GHz dual core Core 2 and Intel G45 graphics (can't remember what the actual IGP was called it but it was pretty poor ) so either would cope with that fine. The A8 will obviously give you more longevity and overall better performance, so really it's a question of whether you're happy spending the extra money.

I guess the question should be...

Will elite dangerous run at 1280x1024 on an a8/7600

If so, decision made. If not, I'll just skip it.

[CAT-THE-FIFTH]

I guess the question should be...

Will elite dangerous run at 1280x1024 on an a8/7600

If so, decision made. If not, I'll just skip it.

http://www.hotukdeals.com/deals/amd-...ebuyer-2052858

[CAT-THE-FIFTH]

It appears a supercomputer powerered by FirePro S9150 compute cards is now top of the Green500 list:

http://www.sys-con.com/node/3237476
http://www.marketwatch.com/story/amd...k=MW_news_stmp

A10-7700K discontinued:

http://www.kitguru.net/components/cp...nue-a10-7700k/

Carrizo and Carrizo-L announced:

http://www.anandtech.com/show/8742/a...us-for-h1-2015

http://techreport.com/news/27392/amd...eema-successor

In an accompanying press release, AMD indicates that the first systems won't be available until "mid-year 2015."

So quite a while for actual systems though.

. A video by AMD’s VP/GM for Computing and Graphics, John Byrne, states that Carrizo and Carrizo-L are currently being tested internally ready for a 1H 2015 release, along with support for DirectX 12, OpenCL 2.0, Mantle and Freesync.

It supports DX12.

CN 1.2's lossless delta color compression algorithms help improve the performance in memory bandwidth limited scenarios, such as in APUs.

It seems that the next APUs will suppport memory bandwidth compression.

However people doubted me about what GPU changes AMD would make in one thread:

One example of the efficiency improvement was provided by AMD’s Voltage Adaptive Operation. Rather than compensate for voltage variations which wastes energy, this technology takes the average operating voltage and detects when the voltage increases beyond a smaller margin. To compensate for this increase, the CPU speed is reduced until the voltage drops below the threshold and then the CPU speed is moved back up.

The changes in speed are designed to be so minute that it does not affect overall performance, however it might only take an errant voltage delivery component to consistently make the voltage go above that threshold, causing erratic slowdown that might be statistically significant. It will be interesting to see how AMD implements the latest version of this feature.