Search
Search titles only
By:
Search titles only
By:
Log in
Register
Search
Search titles only
By:
Search titles only
By:
Menu
Install the app
Install
Home
Featured
Forums
Trending
Search forums
What's new
New posts
Latest activity
Members
Current visitors
JavaScript is disabled. For a better experience, please enable JavaScript in your browser before proceeding.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an
alternative browser
.
N
naukkis
Senior member
Messages
706
Featured content
0
Profile views
303
Reaction score
578
Points
136
Joined
Jun 5, 2002
Last seen
1 minute ago
Find
Find content
Find all content by naukkis
Find all threads by naukkis
Latest activity
Postings
Featured content
About
Trophies
N
naukkis
replied to the thread
Discussion
Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)
.
No, what they did do with Zen3 is to use marco-ops instead of micro-ops - they reduced PRF usage by letting macro-ops transfer data...
Monday at 5:50 AM
N
naukkis
reacted to
Timorous's post
in the thread
Discussion
Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)
with
Like
.
The correct numbers are clearly the measured ones from the review in question. It may not make much difference or change the conclusion...
Saturday at 6:26 AM
N
naukkis
replied to the thread
Discussion
Apple Silicon SoC thread
.
In hybrid cpu configuration big cores are there for best per thread performance. If they want still to utilize SMT right core to have it...
Apr 4, 2024
N
naukkis
replied to the thread
Discussion
Apple Silicon SoC thread
.
You do know that what you supposed means disabling HT. Splitting each core to two virtual cores = HT on, one thread per core = HT off...
Apr 4, 2024
N
naukkis
replied to the thread
Discussion
Apple Silicon SoC thread
.
This specially was solutions for utilizing wider cores. Vectorization(SIMD) is working only when there's no dependencies between data -...
Mar 30, 2024
N
naukkis
replied to the thread
Discussion
Apple Silicon SoC thread
.
Many part of loops are't vectorizable but can be unrolled. Compilers do unroll loops to extract parallelism but compile time unrolling...
Mar 28, 2024
N
naukkis
replied to the thread
Discussion
Apple Silicon SoC thread
.
There's something that might give good results from very wide cores that aren't yet utilized - like hardware loop unrolling. Complex to...
Mar 27, 2024
N
naukkis
replied to the thread
Discussion
Apple Silicon SoC thread
.
Everything can be buggy or just doesn't work so usually cpu's have ability to switch off almost all performance options. But loading...
Mar 22, 2024
N
naukkis
replied to the thread
Discussion
Apple Silicon SoC thread
.
The most interesting part of that side-channel is that Apple is using pointer finder in their prefetchers and prefetching suspected...
Mar 22, 2024
N
naukkis
replied to the thread
Discussion
Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
.
Whole NPU meaning is to make optimized hardware for very short datatypes with simplified instructions. FPU does very complex...
Mar 20, 2024
N
naukkis
replied to the thread
Discussion
Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
.
Speculation was about replacing FPU with NPU. FPU doesn't usually even support FP16 math, it's single precision is 32 bit and double...
Mar 20, 2024
N
naukkis
replied to the thread
Discussion
Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
.
NPU is exact opposite of FPU. Floating point instructions are varying point so number expression range can be huge, like from 2^-64 to...
Mar 19, 2024
N
naukkis
replied to the thread
Discussion
Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
.
Actually higher leakage in cpu means higher clocking potential. Cpu manufacturers sometimes give golden overclocking samples away that...
Mar 10, 2024
N
naukkis
replied to the thread
Discussion
Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
.
I speculated in risc-v threads that cpu designs should go towards split register file designs. VISC is nothing more than split register...
Mar 9, 2024
N
naukkis
replied to the thread
Discussion
Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
.
Yeah, Intel should probably just made GoldenCove much bigger and run 40 of them @2Ghz with at least 4-way SMT to be competitive with...
Mar 9, 2024
Loading…
Loading…
Loading…
Loading…
RESOURCES
Advertising
Cookies Policies
Privacy
Term & Conditions
FOLLOW
Facebook
Twitter
Top
Bottom
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
Accept
Learn more…