Report: Intel Ivy Bridge launch set for April 29, AMD Trinity on May 15

Jos

Posts: 3,073   +97
Staff

More information on Intel and AMD's planned launch for their respective next-generation processors has come to light thanks a couple of reports by CPU World and SWEClockers. According to the former, Intel will kick things off announcing the first desktop and notebook Ivy Bridge processor models in the last week of April, with a bunch of quad-core chips available for sale from April 29.

These will include several new Core i5 and Core i7 chips for the desktop market, and three Core i7 mobile processors -- as detailed below. Dual-core mobile chips are supposedly coming on June 3 along with another handful of quad-core desktop Ivy Bridge processors. Early reports indicate we'll see a 5% to 15% increase in CPU performance and a 20% to 50% boost on the graphics side with these chips compared to their Sandy Bridge equivalents.

Model Cores Threads Frequency Turbo Frequency L3 cache TDP Intel HD Graphics
i5-3570K 4 4 3.4 GHz 3.8 GHz 6MB 77W 4000
i5-3570T 4 4 2.3 GHz 3.3 GHz 6MB 45W 2500
i5-3550 4 4 3.3 GHz 3.7 GHz 6MB 77W 2500
i5-3550S 4 4 3.0 GHz 3.7 GHz 6MB 65W 2500
i5-3450 4 4 3.1 GHz 3.5 GHz 6MB 77W 2500
i7-3770 4 8 3.4 GHz 3.9 GHz 8MB 77W 4000
i7-3770K 4 8 3.5 GHz 3.9 GHz 8MB 77W 4000
i7-3770S 4 8 3.1 GHz 3.9 GHz 8MB 65W 4000
i7-3770T 4 8 2.5 GHz 3.7 GHz 8MB 45W 4000
i7-3720QM 4 8 2.6 GHz 3.6 GHz 6MB 45W 4000
i7-3820QM 4 8 2.7 GHz 3.7 GHz 8MB 45W 4000
i7-3920XM 4 8 2.9 GHz 3.8 GHz 8MB 55W 4000

We should start seeing fast and powerful laptops with new Core i7 processors on the market later this spring. However, the absense of low power mobile SKUs suggests we won't see the next wave of Ultrabooks with Ivy Bridge processors at least until June  -- probably coinciding with the Computex trade show in Taipei -- when Intel rolls out its power sipping Core i5 and i7 chips, including the i5-3427U and i7-3667U.

Meanwhile AMD's Trinity chips, which will succeed the Llano generation of APUs, are rumored to be launching on May 15. No specific models were listed on the report by SWEClockers, only that the launch will include desktop and notebook chips. Some more low-power APUs for "thin-and-light" laptops (essentially, AMD's answer to Intel's Ultrabook efforts) are reportedly set for a June release.

Trinity will feature up to four "Piledriver" cores, which are enhanced variants of the Bulldozer core found on FX Series processors, as well as an AMD Radeon HD 7000-series "Southern Islands" graphics core with DirectX 11-class graphics support, DDR3 memory controller and other improvements.

Permalink to story.

 
Basically a re-release of chips for Intel to dance to the bank with again.
Thanks AMD! :p

The new Intel HD Graphics 4000 + etc is impressive though.
 
When you consider PCI-E 3.0 support, Ivy Bridge is a HUGGGEE improvement over Sandy Bridge. Its not even close.
 
Looks like I'll have plenty of options to choose from in the near future. Ivy Bridge and Kepler, or Trinity. I'm thinking the Trinity APUs will be considerably more affordable though.
 
Since when has Intel had the idea of making triple core CPU's?
 
engineer - Sir ! You need to release a new version of intel processors to gain more profit.

Ceo - **** it, just replace the name sandy with ivy and add 100mhz in HD graphics, we are done !

Engineer - people are so dumb here

Ceo - So much win
 
Guest said:
engineer - Sir ! You need to release a new version of intel processors to gain more profit.

Ceo - **** it, just replace the name sandy with ivy and add 100mhz in HD graphics, we are done !

Engineer - people are so dumb here

Ceo - So much win

Yeah, and add more graphics cores, increase the transistor count by shrinking the transistor to 22nm and produce the first tru-3D transistors for consumers in existence. Do you know anything about semi-conductors, MOSFETs or something else to give a more intelligent opinion? I'm not to buy IB for at least the following year (haven't upgraded my PC in 4 years) so I'm not a fanboy who'll buy when the new products arrive, and while actually is a minor improvement for specifications and performance, in the engineering aspects are quite interesting -10 years of development.
 
Guest said:
When you consider PCI-E 3.0 support, Ivy Bridge is a HUGGGEE improvement over Sandy Bridge. Its not even close.

Really...

With today's graphics cards, is there any proof that PCI-e 3.0 will have any massive advantages?
 
sarcasm said:
Guest said:
When you consider PCI-E 3.0 support, Ivy Bridge is a HUGGGEE improvement over Sandy Bridge. Its not even close.
Really...
With today's graphics cards, is there any proof that PCI-e 3.0 will have any massive advantages?
Maybe not for the mainstream, but for compute functionality the benefits of higher bandwidth are readily apparent. Any gaming GPU performance increase is likely to be marginal at best in most situations simply because most of the computation for frame rendering and post processing is accomplished within the GPU+VRAM. Games that require a higher level of CPU <-> GPU co-operation ( large game maps, intensive AI, CPU physics etc.) would likely start producing tangible differences between PCI-E 2.0 and 3.0 only once top SKU dual cards make an appearance ( HD 7990, GTX 690 ), and the CPU at the other end of the PCI-E bus is capable of living up to its end of the transactions.
 
As usual dividebyzero giving a more objective opinion; as we can see in the link the difference is not huge, but if you have PCIe 3 you may get some frames more with the same hardware which is very welcomed, and if you don't, well... you'll still get what you expected.
 
No 6 Cores...
No 8 Cores....
No 12 Cores...
No 16 Cores....
No 24 Cores...
.
Man they need to get going...
Ya Think..

:)
 
3dcgmodeler said:
No 6 Cores...
No 8 Cores....
No 12 Cores...
No 16 Cores....
No 24 Cores...
.
Man they need to get going...
Ya Think..

:)

Why? Does anything really tax more than 4 threads these days? I'd rather see them offer a CPU with 4 cores and 2x the IPC than an 8 core based on todays Architectures... both would have the same theoretical throughput but one would be massively faster in just about every application.
 
Back