Intel Sheds Light on "Penryn" Enhancements

[maroon1]

DailyTech - Intel Sheds Light on "Penryn" Enhancements


This website tells you the improvement that the upcoming 45nm processors will have. And the improvements is not only a die shrink like many AMD fanboys claimed. There will be many tweaks to the micro architecture of penryn to achieve greater performance at the same clock speeds as Conroe processors.

Here are some quotes from the the article in the above website

Penryn enhances Wide Dynamic Execution technology with a fast radix-16 divider and improved Virtualization technology. With a fast radix-16 divider, the processor can process 4-bits per cycle instead of the 2-bits per cycle of Conroe – doubling the divide instruction capabilities. Intel VT technology receives enhancements that reduce virtual machine transition latencies by 25-to-75%

Intel Advanced Smart Cache technology receives additional enhancements, besides the increased L2 cache. Penryn-based quad and dual-core processors will have up to 12MB and 6MB L2 cache, respectively. Intel reduces cache latency in addition to the larger sizes. Penryn features a 24-way associative cache, an upgrade from Conroe's 16-way associative cache.

So this ^^ means that Penryn will not only have more cache, but Intel will aslo reduce cache latency, and Penryn features a 24-way associative cache.

New to the Advanced Digital Media Boost technology is the inclusion of a new Intel SSE4 instruction set. SSE4 introduces 47 new instructions to improve performance of video accelerators, graphics building blocks and streaming load. Intel claims a 2x performance gain in video acceleration tasks. There are 14 new instructions for video accelerator performance enhancement. Intel improves compiler auto-vectorization performance with 32 new instructions.

Intel expects SSE4 optimizations to deliver performance improvements in video authoring, imagine, graphics, video search, off-chip accelerators, gaming and physics applications. Also new to Advanced Digital Media Boost is the Super Shuffle Engine. Intel's Super Shuffle Engine allows for shuffling unpacking, packing, align concatenated sources, wide shifts, insertion and extraction, and setup for horizontal arithmetic functions. Intel claims a “2x faster SSE shuffle instruction execution,” according to briefing documents.

Thats ^^ a good improvement


And thats not all !!

 

[QuaziBee]

Whats that AMD fanboy? No, you may not play Crisis on my pc.

 

[Apokalipse]

Whats that AMD fanboy? No, you may not play Crisis on my pc.

That's okay, I'll play it on my Phenom X4.

This website tells you the improvement that the upcoming 45nm will have. And the improvements is not only a die shrink like many AMD fanboys claimed. There will be many tweaks to the micro architecture of penryn to achieve greater performance at the same clock speeds as Conroe processors.

But not to, for example, the pipeline, the floating point units, the instruction decoders and so on.

 

[alexsabree]

Amd better bounce back in both gpu and cpu market.

 

[b1gapl]

They will, and they have to. Or else Intel will take over.

 

[nu2duo]

They will, and they have to. Or else Intel will take over.

Intel will take over again.

Once those new amds come out, I'll buy one. I used amd from 2003-early 2007.

 

[Apokalipse]

Intel will take over again.

Once those new amds come out, I'll buy one. I used amd from 2003-early 2007.

Once K10 does come out, even if for some reason Intel comes out ahead with Penryn, the difference will be barely anything.

Core 2's have an advantage of having a 4 IPC pipeline, vs K8's 3 IPC pipeline
However K10 is increasing it to 4, as well as making many changes to the architecture - much more significant changes than Penryn is getting, like doubling the FPU's to 128-bit, a much improved caching and prefetching system, better virtualisation, four instruction decoders that can all decode complex [/i]and[/i] simple instructions (vs Core 2's one complex and three simple decoders). Then there's their 12 stage pipeline, Penryn will still be 14-stage (lower is better)

On top of that, all four cores in the Agena models will be on the same die, all sharing 2MB l3 cache, and all of them having direct access to the Hypertransport bus - which is also being increased to 4GHZ I believe.

 

[veg1992]

Once K10 does come out, even if for some reason Intel comes out ahead with Penryn, the difference will be barely anything.

Core 2's have an advantage of having a 4 IPC pipeline, vs K8's 3 IPC pipeline
However K10 is increasing it to 4, as well as making many changes to the architecture - much more significant changes than Penryn is getting, like doubling the FPU's to 128-bit, a much improved caching and prefetching system, better virtualisation, four instruction decoders that can all decode complex [/i]and[/i] simple instructions (vs Core 2's one complex and three simple decoders). Then there's their 12 stage pipeline, Penryn will still be 14-stage (lower is better)

On top of that, all four cores in the Agena models will be on the same die, all sharing 2MB l3 cache, and all of them having direct access to the Hypertransport bus - which is also being increased to 4GHZ I believe.

so ur going Opteron or Phenom?

 

[Apokalipse]

so ur going Opteron or Phenom?

Phenom X4 aka Agena

Agena CPU's will be compatible with normal AM2 boards, so my C51XEM2AA should run one.

 

[veg1992]

oh ok, i've heard they're release price will be near sub-300 or low 300$ range, is this true?