blightymate wrote:You well illuminate your true character with such words.
Personal computing discussed
Moderators: renee, morphine, SecretSquirrel
blightymate wrote:You well illuminate your true character with such words.
blightymate wrote:You well illuminate your true character with such words.
Savyg wrote:blightymate wrote:You well illuminate your true character with such words.
In the great words of some Earthlings I met a while ago "You are not helping."
blightymate wrote:Not helping what?
Savyg wrote:We do have some early perf claims finally
http://wccftech.com/amd-kaveri-apu-a107 ... r-directx/
Still waiting on the code obviously. I would suspect they want it out before the Kaveri launch, but we'll see if that actually happens.
Airmantharp wrote:blightymate wrote:Not helping what?
We realize you're here to evangelize Mantle, not actually participate in this community. Kyle Bennett was pretty clear about that when he locked your thread on the [H] Forums. Just like there, there's very little reason to discuss Mantle in depth when we have no product to discuss.
chuckula wrote:Having seen those AMD slides... I take back something I had predicted earlier: I had predicted that Kaveri's IGP would beat the Iris Pro parts. Given that AMD went out of its way to avoid comparisons to Iris Pro... even in its own press slides where it is free to fiddle with the knobs to make Kaveri look as good as possible... it's pretty clear that Kaveri isn't beating the Iris Pro by an appreciable margin even when Kaveri has a 95 watt power envelope. When you force Kaveri down to the same power envelopes of the Iris Pro parts, then Kaveri flat-out loses.
JohnC wrote:Airmantharp wrote:blightymate wrote:Not helping what?
We realize you're here to evangelize Mantle, not actually participate in this community. Kyle Bennett was pretty clear about that when he locked your thread on the [H] Forums. Just like there, there's very little reason to discuss Mantle in depth when we have no product to discuss.
Not only locked his thread - this shill/troll was banned at HardOCP, Anandtech and other forums.
Airmantharp wrote:Had to go check and see that he was banned- that never gets old!
sschaem wrote:More on the recent slides.
The demo that show a 3X perf boost with mantle on Kaveri and 2x on an i7-4670k with a GCN discreet GPU will be released on steam this month.
http://forums.elementalgame.com/451041/page/1/#3430480
I think some people are in denial about Mantle, big time.... and are going to be flabbergasted.
sschaem wrote:More on the recent slides.
The demo that show a 3X perf boost with mantle on Kaveri and 2x on an i7-4670k with a GCN discreet GPU will be released on steam this month.
http://forums.elementalgame.com/451041/page/1/#3430480
I think some people are in denial about Mantle, big time.... and are going to be flabbergasted.
chuckula wrote:sschaem wrote:More on the recent slides.
The demo that show a 3X perf boost with mantle on Kaveri and 2x on an i7-4670k with a GCN discreet GPU will be released on steam this month.
http://forums.elementalgame.com/451041/page/1/#3430480
I think some people are in denial about Mantle, big time.... and are going to be flabbergasted.
1. 3X in an isolated benchmark is easy when you go from horrible optimization to "Mantle" optimization. Look at the promises for BF4... well given how poorly the default version of BF4 seems to be doing, is it because Mantle is amazing or because it's easy to look "amazing" when you compare yourself to crap?
2. I've said all along that Mantle is in place to paper over deficiencies in the CPU power of AMD's non-FX series processors and I've seen nothing in AMD's slides that indicates otherwise. AMD came up with Mantle so that a relatively weak Kaveri CPU can drive an R9-290X and likely its successor without choking performance. That's all well and good, but when the CPU isn't a bottleneck to begin with, Mantle isn't going to work any miracles.
sschaem wrote:I think some people are in denial about Mantle, big time.... and are going to be flabbergasted.
maxxcool wrote:show me the TR and Anandtech reviews with a phenonx6, amd8350, Kevari-apu, I7-4670k, i52500k with a amd 7770 and a nvidia 6600 as the graphics cards. *then* i will believe it may work in a limited market with even less consumer support.
*BUT* what will you do when Nivida does the same thing ? and you cannot play your favorite game because it is NOT mantle, but Nvidicized ?
blightymate wrote:AMD says a 45% increase in BF4 Mantle vs. Direct 3D on Kaveri.
Jon1984 wrote:blightymate wrote:AMD says a 45% increase in BF4 Mantle vs. Direct 3D on Kaveri.
Could you post the link to this info?
Savyg wrote:Jon1984 wrote:blightymate wrote:AMD says a 45% increase in BF4 Mantle vs. Direct 3D on Kaveri.
Could you post the link to this info?
I just did.
fhohj wrote:that 45% is a step up from what they were saying before though.
I'm wondering how much of that is Mantle being coded for HSA. I mean, why wouldn't AMD code it for that? So that 45% improvement may only come if you have Kaveri or a similarly HSA cpu, and may not apply if you're just using an AMD GPU with an Intel processor.
fhohj wrote:that 45% is a step up from what they were saying before though.
I'm wondering how much of that is Mantle being coded for HSA. I mean, why wouldn't AMD code it for that? So that 45% improvement may only come if you have Kaveri or a similarly HSA cpu, and may not apply if you're just using an AMD GPU with an Intel processor.
Airmantharp wrote:I'm seeing a whole lot of predictions and numbers without a whole lot of context, and I'll reiterate that this discussion should wait until we have real data to chew on.
fhohj wrote:that 45% is a step up from what they were saying before though.