• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Last call before Vishera

lifeblood

Senior member
If rumors are to be believed Vishera gets released tomorrow. Any last minute bets on performance? I am predicting 15% increases (IPC + Clock) in a very few apps, with 5-10% being the average except for games. Because I'm a pessimist I think games are going to see 3-5% improvement on average.

Anyone care to try?
 
If I remember correctly AMD was talking 15% better efficiency - does this mean lower power usage at a given clock speed, or the same power usage but 15% better performance.

I think in the case of Vishera it might be both. The power consumption has dropped considerably judging by Trinity, and the stock clocks have been bumped up ~10%. With single digit IPC increases, the single-threaded workloads should benefit from only IPC bumps while the mutli-threaded workloads should see quite a bit more. All of it, presumably, at lower power consumption but still paling in comparison to Ivy
 
I will wait for benches to come out ... At this point I cannot help to be skeptical. Some bench suggested a 15% increase from bulldozer (trinity vs bulldozer). Some suggested single digit (OBR for the arguments sake). Plus the rumored win8 fix on scheduler optimizations ( techreport suggested 10%-20% improvement on bulldozer)

So if numbers add up are we looking at 25% to 30% improvement over the spectrum? Or still horrible single threaded 5% increase due to increased base clock and 15% multithreaded increase?

I have to say, if this a marketing scheme to create controversy, that certainly got my attention, but not my respect. If vishera is a decent CPU, I will buy one.
 
jeux.jpg


moar benches here...use salt
http://www.cowcotland.com/articles/1199/test-processeur-amd-fx-8350.html
 
Single digit single-threaded increases while multi-threaded should see some considerable gains.

I saw some article, cant remember the source, that was predicting this as well. It would be nice to see any improvement, but the problem is, if this is true, where they trailed the most was in single/lightly threaded workloads. So they got the least improvement in the area where they needed it the most.

But I am waiting for a full test from Anand to make a final judgement. I dont really put any stock in the linux benches or the very preliminary gaming benches posted so far.
 
Back
Top