A10-5800K undervolting article:
http://www.tomshardware.co.uk/a10-58...iew-32536.html
Printable View
A10-5800K undervolting article:
http://www.tomshardware.co.uk/a10-58...iew-32536.html
Not until someone starts releasing unbuffered 16GB DIMMs though ;) Can't be far off - registered DIMMs are already up to 32GB and 8GB unbuffered DIMMs are common. I guess at the minute there's no driver for larger modules in the consumer space because it's hard to use as much as 32GB anyway...
Of course, 64GB of high speed RAM would mean you could basically cache your entire OS in RAM - how fast would that be... :o
Another IPC comparison between the FX8150 and the A10 5800K:
http://www.technic3d.com/review/cpu-...800k-apu/3.htm
http://www.technic3d.com/review/cpu-...800k-apu/4.htm
There've been all sorts of rumours floating around one way or another. As CAT says, it would be *very* surprising if the AM3+ CPUs didn't use PD. After all, PD is only really BD+ - they've tweaked a few things but there are no major architectural changes (that will come in Steamroller).
Unfortunately, it looks like Vishera will be just as lacking in single thread performance as Bulldozer, which means that x86 reviews will still favour Intel. Desktop power draw hasn't really improved either (although it's disappointing that we don't have any 65W Trinity reviews to do the comparisons v. i3, as that should be a much more even comparison in terms of power draw...).
Surreal review on S/A:
http://semiaccurate.com/2012/10/01/a...ls-ivy-bridge/
Compared with an i7-3770K apparently Trinity is slow, who would have predicted that :D Goes on to say the machines felt exactly the same in use though.
Thing is, I nearly didn't bother reading to the end of the review. Imagine a lot of people really would give up part way through. I also guess the reviewer doesn't game much.
Well the website is appropriately named at least
WTF is going on at AMD?
Silly reviewer hand-tying and RAM disks......within a week of each other?
Next thing they will be doing an Intel and demoing game performance via pre-recorded videos....
TBH,I really think you cannot complain massively about AMD limiting previews. Plenty of websites did company controlled previews of previous AMD and Intel CPUs in the past. Remember the Conroe previews,SB previews,IB previews and Brazos previews done by websites like Anandtech and The Tech Report?? Most were done under the control of Intel and AMD.
Moreover,AMD just gave permission for reviewers to brake NDA,by publishing parts of their previews early. They were not forced to release them and plenty were free to say that AMD put restrictions.
So you have two scenarios here:
1.)Websites doing previews to get more page hits.
2.)Websites whinging about previews,despite doing this loads of times before,to get more page hits.
I think AMD are just trying to stay in the headlines for a bit. BSN says they call it the start of "Speedmonth", so wonder if we will see the same games with the new AM3+ platform next week.
http://www.brightsideofnews.com/news...you-wings.aspx
Far more interesting is that in the worst case it's only providing 43% of the performance. But, doesn't an A10-5800k only cost 40% as much as an i7-3770k? ;)
I think the most pleasing thing is that multi-threaded performance is spot on to compete with the price-equivalent Intel parts. If only they could find some extra single threaded performance from somewhere... *sigh*
In the non-gaming tests,the A10-5800K does do better than I expected:
http://www.tomshardware.co.uk/a10-58...w-32536-5.html
Despite,the criticisms placed at Toms Hardware,they are one of the few websites to use an up to date test suite. Loads of websites use really old software like CS4 which was released 4 years ago.
The main issue appears to be that load power consumption,seems to have not improved over the previous generation,although some reviews say it does,many don't. At least low load power consumption has improved somewhat though.