Its bad for sales tactics to make sense. The point is to drive sales in a way that they(the company selling them) want. If people understand the model numbers much clearer, the company wouldn't have the control, the buyers will.
Its bad for sales tactics to make sense. The point is to drive sales in a way that they(the company selling them) want. If people understand the model numbers much clearer, the company wouldn't have the control, the buyers will.
Is Intel targeting the consumer with their naming scheme, or is Intel optimizing their naming scheme for OEMs, in order to extract more money from them through the multitude of bins they sell?
Desktop i5s have all been 4 thread and had turbo, but they haven't all been quad core.Intel's naming system is fine for the desktop, except at the low end when you have atom being rebranded as celeron and pentium. But if you look at mobile, it is a mess. Desktop i5 are all quad cores, but no mobile i5s are. i7 is confusing too since they range from low voltage dual cores to the top end mobile quad. I agree, AMD is no better. The desktop FX lineup is fine (naming wise) but now they had to stick FX on mobile. And whenever i see an a4 or A6 model. i pretty much have to google to see whether it is cat core or big core.
Meanwhile I expect the price of a Windows 7 licensed (upgradeable to Windows 10 for free) Sandy Bridge quad core SFF Pre-built to also drop in price.
I pressed Su for the one thing on AMD's roadmap where she felt the company absolutely had to succeed, either because a misstep might be harmful, or because success might help the company hit its financial targets faster. The answer, in a word, is Zen.
"In the short term, how we do in the PC market is very important," she said, pointing to share losses in the PC market that have knocked AMD around a bit.
AMD is making a big bet, she said, on servers and on high-performance desktops. It is "very deliberate, and we have to execute," she said.
[...]
Su said, about Zen, that it was important to get everything right. That single-threaded performance was important, but so too was overall system architecture, and power performance. She said the two fundamental pillars in the future AMD roadmap included improving IPC across all applications, and performance per watt in graphics.
She also reiterated that "Zen is first in a family of processors," adding that "you can't be a one shot wonder."
In light of Oakridge National Laboratory switching from AMD Opteron + Nvidia to IBM Power + Nvidia, I have to wonder if AMD would add NVLink to their Server CPUs?
CUDA is a quasi standard for academia and research.
First of all, there is only one VM that gets access to the GPU: one VM literally owns the GPU (unlike NVIDIA's GRID technology).
Yeah, I am thinking adding NVLink would not be such a bad idea.
If anything it might help them compete better with Intel rather than cannibalize their own Server APU sales.
Then AMD can improve their Radeon to be ready for whatever Intel/Xeon Phi, IBM/Nvidia, ARM/Nvidia APUs exist in the future.
Since AMD is looking to enter more deeply into the OpenCL compute community - this would be a bad idea. Not to mention the negative effect psychologically with the GPU engineers or externally vis-a-vis the negative PR.
Losing the number 2 Super computer in the World bid to IBM might also have a negative effect on the CPU engineers as well as being negative PR.
AMD didn't have a hope of getting that contract with their current CPU lineup- Zen is unproven, K12 is delayed, and Bulldozer is a walking disaster. They only supercomputers they still had were ones leftover from the K8 days. Adding their direct competitor's proprietary tech wouldn't have changed that.
The TITAN supercomputer uses Bulldozer based Opterons 62xx series.
Also something nobody talks about, number ONE in Green500 is a Xeon with AMD Fire Pro S9150.
The Titan only uses AMD because it is an upgrade of an older supercomputer which was built on AMD. If it were built from scratch it would probably have gone with Intel or IBM.
And your second link just kind of proves the point I was making, about AMD's current CPU lineup. The only time they are competitive is when paired with a Xeon.
I really hope Zen gets them back in the game.
AMD didn't have a hope of getting that contract with their current CPU lineup- Zen is unproven, K12 is delayed, and Bulldozer is a walking disaster. They only supercomputers they still had were ones leftover from the K8 days. Adding their direct competitor's proprietary tech wouldn't have changed that.
The TITAN supercomputer uses Bulldozer based Opterons 62xx series.
There's a flip side to that coin. If the numbers and products are too messy, the consumers will get confused and go with the simple well known alternative instead. If there's a competitor that provides such a simple well known alternative, the company with the confusing product portfolio may end up losing customers to that competitor.
Example: Just get "the" MBA or iPhone in it's latest version and you're done.
Ehm the current Opterons 63xx are almost three years old, they are not able to compete today.
My point is that I would like to see AMD regain that contract.
Just because Oakridge National Laboratory uses CUDA doesn't mean the CPU can't be an AMD one.
So for the future, I would hope the AMD CPUs would have NVLink (if that helps), although I do have to wonder what they would compete with at that time? IBM/Nvidia APUs? ARM/Nvidia APUs? Or maybe still using dCPU and dGPU (with NVLink)?
P.S. Something tells me commodity AMD dCPU server part with NVLInk would probably be much much cheaper than a specialized IBM/Nvidia APU which would be even more expensive than IBM Power + Nvidia card.