Intel Broadwell Thread

[Nothingness]

Why is SHA grayed out?

Shouldn't RDSEED be available too?

 

[Dufus]

Possibly the Author of HWInfo32 knows something most others don't or it's a bug. Going by the latter and SDE default CPUID's for relevant RDSEED, SHA and AVX512 gives...

Broadwell
RDSEED

Skylake Client
RDSEED

Skylake Server
RDSEED, AVX512VL, AVX512BW, AVX512CD, AVX512DQ, AVX512F

Knights Landing
RDSEED, AVX512CD, AVX512ER, AVX512PF, AVX512F

Cannonlake
RDSEED, SHA, AVX512VL, AVX512BW, AVX512CD, AVX512IFMA, AVX512DQ, AVX512F

Goldmont
RDSEED, SHA

 

[Kallogan]

http://www.fanlesstech.com/2014/09/entire-broadwell-u-lineup.html

most interesting cpu for me :
Core i3-5157U (2.5 GHz / 2C4T / Iris Graphics 6100 / 28W TDP)

and

Core i5-5250U (1.6-2.7 GHz / 2C4T / HD Graphics 6000 / 15W TDP)

15W with full 48 EUs

🙂

 

[Arachnotronic]

Possibly the Author of HWInfo32 knows something most others don't or it's a bug. Going by the latter and SDE default CPUID's for relevant RDSEED, SHA and AVX512 gives...

Broadwell
RDSEED

Skylake Client
RDSEED

Skylake Server
RDSEED, AVX512VL, AVX512BW, AVX512CD, AVX512DQ, AVX512F

Knights Landing
RDSEED, AVX512CD, AVX512ER, AVX512PF, AVX512F

Cannonlake
RDSEED, SHA, AVX512VL, AVX512BW, AVX512CD, AVX512IFMA, AVX512DQ, AVX512F

Goldmont
RDSEED, SHA

So no SHA for Skylake, but at least they're putting it in Goldmont to fend off the ARMy in Geekbench 😀

 

[witeken]

http://www.fanlesstech.com/2014/09/entire-broadwell-u-lineup.html

most interesting cpu for me :
Core i3-5157U (2.5 GHz / 2C4T / Iris Graphics 6100 / 28W TDP)
🙂

Not bad. That is up 25% from

Core i3-4158U Iris 5100 (GT3) 28 W / 2.0 GHz

So no SHA for Skylake, but at least they're putting it in Goldmont to fend off the ARMy in Geekbench 😀

There are no architectural details known about Goldmont.

 

[Arachnotronic]

There are no architectural details known about Goldmont.

One of the big weaknesses for Intel chips on Geekbench 3 is lack of hardware SHA support for the SHA subtests, so I'd expect a big improvement in that section of the benchmark.

 

[Nothingness]

http://www.fanlesstech.com/2014/09/entire-broadwell-u-lineup.html

Funny, the column are mislabeled on the first slide.

Not bad. That is up 25% from

Core i3-4158U Iris 5100 (GT3) 28 W / 2.0 GHz

It's better than that: it's 25% up with almost half TDP (the text says 28W, but the slide says 15W).

 

[Nothingness]

One of the big weaknesses for Intel chips on Geekbench 3 is lack of hardware SHA support for the SHA subtests, so I'd expect a big improvement in that section of the benchmark.

The impact won't be that big on total score since Geekbench uses geometric mean.

 

[ShintaiDK]

No HDMI 2.0 tho, so turn off for me.

My old i5 661 HTPC would like to be replaced with a NUC type one. But seems it needs to wait 1-1½ year still.

 

[witeken]

One of the big weaknesses for Intel chips on Geekbench 3 is lack of hardware SHA support for the SHA subtests, so I'd expect a big improvement in that section of the benchmark.

So we're having benchmark driven innovation?

It's better than that: it's 25% up with almost half TDP (the text says 28W, but the slide says 15W).

I don't think it will be 15W since the name suggests that this is the successor in Intel's 28W U series.

 

[Nothingness]

I don't think it will be 15W since the name suggests that this is the successor in Intel's 28W U series.

So the slides are fake 😉

 

[Arachnotronic]

So we're having benchmark driven innovation?

I was being slightly facetious 🙂

The impact won't be that big on total score since Geekbench uses geometric mean.

Ah, interesting! Thanks, Nothingness!

 

[kimmel]

So we're having benchmark driven innovation?

Widely used terrible benchmarks drive the expected and equally terrible results.

 

[Dufus]

http://www.fanlesstech.com/2014/09/entire-broadwell-u-lineup.html

most interesting cpu for me :
Core i3-5157U (2.5 GHz / 2C4T / Iris Graphics 6100 / 28W TDP)

and

Core i5-5250U (1.6-2.7 GHz / 2C4T / HD Graphics 6000 / 15W TDP)

15W with full 48 EUs

🙂

Remember that TDP is just a specification and not the maximum power that can be used (providing the hardware is up to it).

There are no architectural details known about Goldmont.

The SDE provides a means for emulating newer instructions without the hardware. While having those instructions doesn't mean they are written in stone (TSX for example), it does add a lot of weight to them being there.

From the SDE release notes 2013

Emulation support for the Intel® Secure Hash Algorithm (Intel® SHA) extensions present on the Intel Goldmont microarchtiecture.

 

[AtenRa]

Really, all those high-end iGPUs at 15W TDP are meaningless. Those SKUs cannot use 100% of that iGPU performance capabilities.

 

[witeken]

Acer Will Launch World’s First 15.6 Inch Chromebook In March

Processor choices are listed as either a 5th generation Intel Broadwell-U Celeron or Core i3 processor.

RAM & onboard will span the usual 2GB/16GB, 4GB/32GB offerings, with 802.11ac Wi-Fi, Bluetooth 4.0 and a smattering of USB 3.0 and HDMI out ports to cater to expansion needs coming as standard on all variants.

Finally, the only other information our friendly neighbourhood Chrome Bandit has passed to us is the release date: March 2015.

‘Rugged’ Acer C740 Chromebook With Broadwell CPU Coming March

The big sell here is that at least one of these devices will come bearing Intel’s fifth-generation Broadwell-based processors, providing users with a huge boost in performance, battery life and graphics performance.

That will be a very early spring launch indeed! (but nonetheless still a spring later than promised)

 

[witeken]

Really, all those high-end iGPUs at 15W TDP are meaningless. Those SKUs cannot use 100% of that iGPU performance capabilities.

They might be (fairly) meaningless at 22nm, but likely not a 14nm, certainly not with Duty Cycle Control and the improved efficiency of Gen8. We'll see.

 

[Nothingness]

Ah, interesting! Thanks, Nothingness!

To give some perspective if I put the best SHA numbers from Cortex-A57 into the best Moorefield score, the Moorefield integer score will go from 1026 up to 1223, and the total score will go from 944 up to 1001. Certainly not insignificant, but far from allowing to reach A57 (or even A17 which doesn't even have AES and SHA) scores.

I know you were kidding, but I think that might be nonetheless interesting 🙂

 

[AtenRa]

They might be (fairly) meaningless at 22nm, but likely not a 14nm, certainly not with Duty Cycle Control and the improved efficiency of Gen8. We'll see.

They will have the same problem. They are using 14nm with 30% less power usage than 22nm but at the same time they increased performance capabilities by more than 20-30% over Haswell.

 

[NTMBK]

They might be (fairly) meaningless at 22nm, but likely not a 14nm, certainly not with Duty Cycle Control and the improved efficiency of Gen8. We'll see.

Duty Cycle Control only helps with reducing power use in low utilization situations- it won't help when the GPU is used in a performance intensive situation.

 

[NTMBK]

No HDMI 2.0 tho, so turn off for me.

My old i5 661 HTPC would like to be replaced with a NUC type one. But seems it needs to wait 1-1½ year still.

There's always AMD 😉 But tbh if you want to drive a 4k display, I'd wait for something with full h.265 hardware decode.

 

[Arachnotronic]

There's always AMD 😉 But tbh if you want to drive a 4k display, I'd wait for something with full h.265 hardware decode.

i.e. Skylake.

 

[Nothingness]

i.e. Skylake.

Nah just pick this now: http://www.aliexpress.com/item/Blue...tex-A17-1-8GHz-TV-BOX-4K-x-2K/2021695712.html
😀

 

[ShintaiDK]

There's always AMD 😉 But tbh if you want to drive a 4k display, I'd wait for something with full h.265 hardware decode.

AMD doesnt have 4K decode ability in any APU nor do any AMD product have HDMI 2.0.

 

[Arachnotronic]

Nah just pick this now: http://www.aliexpress.com/item/Blue...tex-A17-1-8GHz-TV-BOX-4K-x-2K/2021695712.html
😀

Aha! Nice!