I was talking about two different benchmarks...not just superpi.
As far as it being "just superpi", a lot of overclockers/benchers would disagree :p:
That single threaded performance is still awful....Assuming Turbo is working on Single Threaded, 4.2GHz... :( . Even so , I suppose it does beat their 10% improvement vs Llano in CPU and much more on GPU. Multithreaded about 10% better as well according to the CPU score in 3DMark.
23.775s - Trinity @ 3.8 / 4.2GHz Turbo
26.039 - Llano @ 2.9GHz
For reference my 24/7 Llano @ 3.55GHz does about 21.1s on 1M. IMO Trinity on desktop needs Hybrid Crossfire tech that will allow it to be paired with any GPU (and not just a low end GPU part like the 76xx).
wonder how these overclock...5ghz possible ;)
hopefully trinity is unlocked, the gpu score is awesome!
But we dont know, how Turbo was worked, dont know RAM setings (it could be only 1333 MHz c9/10), jumping between threads in superpi or downclocks (seems, C1E,C6 was active)
http://www.nordichardware.com/news/6...n-folding.html
pretty good since its compared to the A8-3500m and theres still the A10Quote:
The two cores in a Piledriver module shares one floating point unit which can be split into two smaller, so the boost there is relatively marginal. When it comes to integer performance the boost is nearly 40 percent, this is even thought the new processors are specified to the same TDP. AMD told us we can expect up to 25 percent better processor performance with the same energy consumption over last geenration, which matches the results above. The exact frequency of A8-4500M is uncertain at this time, but should be higher than 1,5 GHz found with A8-3500M.
I believe AMD prefers the term 'Dual Graphics' ... :D ... Hybrid Crossfire is so IGP (and meh).
With alternate-frame sequencing such pairings are unlikely, and the rumahs on the internets have the Radeon HD 6570 Turks Pro discreet being the card of choice in dual graphics with Trinity.
AMD propaganda says "up to 50% greater performance" than Hybrid Cros ...errr.... Dual Graphics on Llano.
With a little more grunt from the Piledriver cores, improved dual graphics performance and better use of faster RAMs, AMD might have a little something going on.
Time to unleash the fury
http://www.sweclockers.com/nyhet/152...pps-den-15-maj
Trinity launch for 15 of may
wonder if blizzard is gonna be a launch partner...
Free Diablo 3 with Trinity based machines :D
I donw know guys...I expected launch one month later...If will be this new true then is good new :)
So trinity in May and full AM3+ piledriver in June? Assuming the sweeds have it right.
Hmm,
I have noticed a lot of manufacturer "refurbished" am3+ boards showing up on various sites that seem to be the complete box and accessories. I also noticed prices starting to drop after many months of relatively little change on some of the more popular am3+ boards.
I wonder if this is due to the rumored new chipset coming soon for AM3+ ? A June releases would be about right to clear inventory and introduce new boards in late May-early June.
I would like to see Vishera core(PD with L3) has 5% more IPC than Bulldozer,let alone Trinity.
Then we will have to see what Trinity can do :)
Of course, if AMD made OC headroom similar to C2 AM2+ X4 to C3 AM3 X4 (200-500+ MHz) we will certainly have an interesting product on our hands. I can see people being happy with 5.0-5.2 stable and benching up around 5.4-5.5 on normal ambient cooling as well as a 5% IPC increase.
...and reasonable power consumption when overclocked.
Trinity will contain Piledriver v1 x86 cores (10h - 1fh). Vishera will have Piledriver v2 cores (Viperfish - 20h - 2fh).
maybe....My tip is, example in old x87 benchmark superpi will be this comparsion:
3.6/4.2 GHz Zambezi = superpi 1M in 20.5s
3.6/4.2 GHz Vishera= superpi 1M in 19.00-19.5s (this is my tip)
You're wrong. Superpi's code was written in the 90'ties. It's fully x87 based which area is totally irrelevant nowadays. Neither Intel or AMD develop this part of the microarchitecture. What you could see as a so called "architectural improvements" is just the speed of the cache or the memory access.
Superpi is useless because you can't even get a realistic comparison between different families from the same company. It's just a cheap toy for boring kids. Nothing else.
Okay, PD and BD are similar cores so if we see an improvement, I'm sure you know that is probably a good thing.
If you think SuperPi is a cheap toy for boring kids, I'm sorry but I disagree. This is XtremeSystems.org, not Super Mario Bros. Win 7 Fan Site...SuperPi is completely irrelevant as far as comparison between different families/brands of CPU but creates great competition for HWBOT or people running similar processors with few variables in the way you can run it.
And the speed of the cache or the memory access it's useless for you.Quote:
What you could see as a so called "architectural improvements" is just the speed of the cache or the memory access.
Superpi is useless because you can't even get a realistic comparison between different families from the same company. It's just a cheap toy for boring kids. Nothing else.
Different families, what you can't compare core i3 with i5 or i7? Scoket 1155 with 2011?
It's a funny thing that Intel Cpu's during last years just improved over and over in Super Pi, with generations and arhitectures, and IPC and overall performance also just over and over improved.
And yet comes BD, with a big hit in Super Pi compared with Phenom.
And what a surprize, single thread is down to earth and multithread almost a fail.
Don't writte of x87, do you think that it's used only in Super Pi.
OLiverda, OK, so diferent comparsion...Cinebench R11.5 multi
Zambezi 3.6GHz/4.2GHZ 6.01p
Vishera 3.6GHZ/4.2 GHz 6.5-6.7p is my estimate