Tiger Lake CPUs Surface With High Clock Speeds Amid Looming Intel Reveal

By Zhiye Liu
published

Competition is heating up

(Image credit: Shutterstock)

The Intel Core i7-1185G7 and Core i5-1135G7 have shown up once again as the hardware world awaits Intel's big 11th Generation Tiger Lake announcement on September 2 (opens in new tab). With the official launch so close, the specifications that we see today could be the final revisions for the two Tiger Lake-U (opens in new tab) parts.

As we've seen from past submissions, the Core i7-1185G7 reportedly comes equipped with four cores, eight threads and 12MB of L3 cache. The Geekbench 5 submissions (via @TUM_APISAK (opens in new tab)) now reveal the quad-core Tiger Lake-U chip with a 2.99 GHz base clock and a 4.79 GHz boost clock. If we round the clock speeds, they boil down to 3 GHz and 4.8 GHz, respectively.

The Core i7-1185G7's Xe LP graphics (opens in new tab), however, is the protagonist of today's show. The Gen12 unit features up to 96 Execution Units (EUs), which represents a 50% increase in comparison to Gen11's maximum configuration. That's only the appetizer though – Gen12 also comes with higher clock speeds. According to the Geekbench 5 submission (opens in new tab), the Core i7-1185G7's iGPU clocks up to 1.55 GHz, 36.3% faster than the Core i7-1065G7 that it's replacing.

Swipe to scroll horizontally
ProcessorCores / ThreadsBase / Boost Clocks (GHz)L3 Cache (MB)Execution UnitsGraphics Max Frequency (GHz)
Core i7-1185G7*4 / 83.0 / 4.812961.55
Core i7-1065G74 / 81.3 / 3.98641.10
Core i5-1135G7*4 / 82.4 / 4.28801.30
Core i5-1035G74 / 81.2 / 3.76641.05

*Specifications are unconfirmed.

The Core i5-1135G7 (via @davideneco25320 (opens in new tab)) seemingly conforms to the same four-core, eight-thread configuration as the Core i7-1185G7 albeit the smaller L3 cache. There's a difference of 4MB that separates the Core i5 from the Core i7.

Geekbench 5 detected the Core i5-1135G7 with 2.4 and 4.19 GHz base and boost clock, respectively. It's a quantum improvement over the previous Core i5-1035G7 and even surpasses the Core i7-1065G7, which was last generation's flagship Ice Lake processor.

After digging a little bit deeper, we found the Core i5-1135G7 (opens in new tab) with a Xe LP solution that's comprised with 80 EUs at 1.3 GHz. The Tiger Lake-U processor has 25% more EUs than the Core i5-1035G7 at 23.8% higher clocks.

Image 1 of 2
Intel Core i7-1185G7
Intel Core i7-1185G7 (Image credit: Primate Labs Inc.)

Tiger Lake is just a taste of what Intel can do with its 10nm SuperFin process node. There is no doubt that the new 10nm chips represent a tremendous uplift in both processing and graphical performance over Ice Lake. 

It's not like Intel ever lost the mobile device market, but with Tiger Lake, the chipmaker could take back some of AMD's gains that the Red Team made with its Ryzen 4000-series (opens in new tab) (codename Renoir) processors. More importantly, the powerful mixture of Willow Cove cores and Xe LP graphics has the potential to make iGPU gaming on an Intel processor a reality.

Zhiye Liu
Zhiye Liu
RAM Reviewer and News Editor

Zhiye Liu is a Freelance News Writer at Tom’s Hardware US. Although he loves everything that’s hardware, he has a soft spot for CPUs, GPUs, and RAM.

Topics
CPUs
11 Comments Comment from the forums
  • usiname
    Only gold samples are used for i7, just check how low is i5 - 4.8 vs 4.2. Their 10nm + or ++ is still trash and they can't bring more cores, because of that.
    Reply
  • Jimbojan
    How trash is it?
    Reply
  • usiname
    Jimbojan said:
    How trash is it?

    They have few good cores, but most of the production is trash, and that is why i5 is only 4.2, and i7 just 4 cores. Their production is not enough for 8/16 i7 with 4.7+ and i5 with higher clocks. They need another year to refine this process, because is not enough even for their mobile market.
    Reply
  • TerryLaze
    usiname said:
    Only gold samples are used for i7, just check how low is i5 - 4.8 vs 4.2. Their 10nm + or ++ is still trash and they can't bring more cores, because of that.
    They have the same amount of threads though so obviously intel had to make them less desirable somehow and the only options are remove HTT or lower clocks.If the i5 would be running at 4.5 nobody would buy the i7.
    Reply
  • PCWarrior
    usiname said:
    Only gold samples are used for i7, just check how low is i5 - 4.8 vs 4.2. Their 10nm + or ++ is still trash and they can't bring more cores, because of that.
    The only trash I see here is your trash talk about Intel. But let’s follow your own (non-)logic for a moment.
    Would you say the same about TSMC’s 7nm based on AMD’s 3950X release? (Its launch even delayed due to the extreme binning required to make it hit (even intermittently) 4.7GHz on a single core).
    Would you say the same about TSMC’s 7nm based on the scarce availability of the higher-frequency 3000 Ryzen cpus during the first 6 months of their production?
    Would you say the same about TSMC’s 7nm based on the fact that for AMD’s 3000-series cpus the all-core turbo (even with OCing) is only 4.2-4.4GHz?
    Would you say the same about TSMC’s 7nm based on the fact that the 3800X is priced $70 more than the 3700X and that for a mere 100MHz higher single-core frequency (and an intermittent one at that)?
    Would say the same about TSMC’s 7nm based on the fact that the XT lineup is priced $70+ more than the X lineup?
    Would you say the same about TSMC’s 7nm based on the fact that AMD's biggest CPU die is only an octacore, it was an octacore and will remain an octacore?
    Would you say the same about TSMC's 7nm based on the fact that it took AMD a YEAR to release the equivalent CPU die with integrated graphics?
    Would you, by your own logic, call TSMC’s 7nm trash? No, you wouldn’t. Because that doesn’t serve your anti-Intel sentiment. It doesn’t serve your trash talk against the competitor of your beloved AMD.
    Reply
  • Don Frenser
    Maybe this time they actually launch something as opposed to the last 5 "launches".

    Toiletpaper was the only thing it was good for.
    Reply
  • usiname
    PCWarrior said:
    The only trash I see here is your trash talk about Intel. But let’s follow your own (non-)logic for a moment.
    Would you say the same about TSMC’s 7nm based on AMD’s 3950X release? (Its launch even delayed due to the extreme binning required to make it hit (even intermittently) 4.7GHz on a single core).
    Would you say the same about TSMC’s 7nm based on the scarce availability of the higher-frequency 3000 Ryzen cpus during the first 6 months of their production?
    Would you say the same about TSMC’s 7nm based on the fact that for AMD’s 3000-series cpus the all-core turbo (even with OCing) is only 4.2-4.4GHz?
    Would you say the same about TSMC’s 7nm based on the fact that the 3800X is priced $70 more than the 3700X and that for a mere 100MHz higher single-core frequency (and an intermittent one at that)?
    Would say the same about TSMC’s 7nm based on the fact that the XT lineup is priced $70+ more than the X lineup?
    Would you say the same about TSMC’s 7nm based on the fact that AMD's biggest CPU die is only an octacore, it was an octacore and will remain an octacore?
    Would you say the same about TSMC's 7nm based on the fact that it took AMD a YEAR to release the equivalent CPU die with integrated graphics?Would you, by your own logic, call TSMC’s 7nm trash? No, you wouldn’t. Because that doesn’t serve your anti-Intel sentiment. It doesn’t serve your trash talk against the competitor of your beloved AMD.

    Yes TSMC must to improve their process and they did
    ^^^^^
    4.3-4.4 and 4.5 for 3950 is only 200mhz bellow the max frequency, not that bad.
    10850-10900 and many many others like 8700 and 8086
    ^^^^^^^^^^^^^^
    It is 8-core for better price and clocks, that is why intel is 5 years on the same place.
    It is not only TSMC problem, but more AMD delaying it
    TSMC 7nm was trash in the beginning and it is better and better with every month. Sorry, I am not blind fanboy like you. You will stay forever on the same place like intel until you start to think, try it, does not hurt.
    Reply
  • JayNor
    Their linux patches show a Tiger Lake-H will happen, and their recent presentations indicate Tiger Lake's SuperFin process will go out to 65W,

    A few leaks have reported an 8 core Tiger Lake-H.

    I suspect their Tiger Lake U is limited due to Intel's choice to include the fancy GPU with 8K60 capability, av1 hardware decoding, the integrated Thunderbolt 4 and WIFI6 features, the avx512 and AI dlboost per core features. They also added 4 lanes of PCIE4 for the latest SSDs.

    It's just a trade-off ... useful io and simd processing features vs more cores.
    Reply
  • Dsplover
    I’ll take a Tiger Lake quad as long as the cache has increased.
    Don’t need Xe but single core performance on an Intel Core is still King, perfect for my low latency needs.

    My AMD builds can be 8 Cores with acceptable latency as my workhorses.
    Waiting for a custom X570 from ASRockRack I ordered a month ago.
    Vermeers or Cezzanes is what Im waiting for.

    Hope to see some great competition.
    Intel took a beating. You know they have to deliver now.

    cheerz
    Reply
  • urbanman2004
    Looks like competition in the mobile sector is heating up and its anyone's call
    Reply