If we go down that road there are a few who should have their right to post here revoked.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Rocket lake leaks
- Thread starter pc-guy
- Start date
If we go down that road there are a few who should have their right to post here revoked.
Associate
- Joined
- 27 Apr 2007
- Posts
- 966
Half of this thread or so seems to consist of AMD 'enthusiasts' gloating, baiting and mocking Intel users and chips.
I find that just as tedious as the posts from the Intel 'enthusiasts'.
At least the Intel posts are possibly more on topic!
Both sides are equally puerile to me.
Ive also seen his veiled insults and actually told him to stop them i thought personal attacks wernt allowed
Associate
- Joined
- 27 Apr 2007
- Posts
- 966
There goes the AVX-512 exclusivity!
https://hothardware.com/news/amd-epyc-7004-zen-4-cpus-avx3-512-bfloat16-instructions
https://hothardware.com/news/amd-epyc-7004-zen-4-cpus-avx3-512-bfloat16-instructions
Associate
- Joined
- 27 Apr 2007
- Posts
- 966
Because they are so fast out of the box there is no need to OC them so the program is futile?Intel has announced it is shutting down its overclocking protection warranty program just before the realise of Rocket Lake - mhm I wonder why
Dave is merely the pertinent example. And yes, there are quite a few people who'd be on the same list.
That is literally not what I said and you know it.
You'd think, but half the moderation these days seems to consist of just deleting the bickering, not actually doing anything about the cause of the problem.
But this is an entirely different subject, let's not derail any further.
Right
"AVX-512, or Advanced Vector Extensions 512, can require quite a bit of power and generate considerable heat in some cases"
This is why they state TDP up to hellish 400 watts!
Permabanned
- Joined
- 28 Sep 2018
- Posts
- 0
Further info: Now 90% sure that memory max is locked at 3733 sync which is pretty terrible for performance. The microcode is there to boost the voltage and bump up vid table to stabilize the platform. It’s not for performance. You won’t be able to clock mem high enough to make up for desync penalty. The issue isn’t the mem, it’s the interconnect being a choke point. No diff to zen in that respect.
unless something dramatic changes (it won’t) I’d get a 10900k. How cml performs on z590 still tbd.
unless something dramatic changes (it won’t) I’d get a 10900k. How cml performs on z590 still tbd.
Last edited:
Right
"AVX-512, or Advanced Vector Extensions 512, can require quite a bit of power and generate considerable heat in some cases"
This is why they state TDP up to hellish 400 watts!
Tbh that isn't really an issue, it is same with any chip that full utilise AVX-512 would be the same and the relevance for anyone whom was on Epyc before would still be seeing same/similar TDP usage anyways. If you change workflow to allow for AVX-512 and all that then you will be updating other systems and tbh a 96 core CPU at 400 watt seems pretty impressive to me tbh. Also isn't it standard 350w up to 400w.
The standard is between 180 and 250 watts.
That's 250 Watt's for a 28 core Xeon, not 96 cores.
AMD's current 64 core EPYC is 200 Watt's.
more numbers for the 11700k
firstly memory latency, we know it's causing issues in games but how bad is it?
The 11700k paired with 3600 CL14 memory with the interconnect in sync produces 57ns latency in AIDA64, much higher than the 44ns the same RAM on 10700k produces and nearly matching the 59ns I get on my Zen3 with the same clocks and sub timings.
Power draw: The 11700k uses 70w more power than the 5800x while gaming
firstly memory latency, we know it's causing issues in games but how bad is it?
The 11700k paired with 3600 CL14 memory with the interconnect in sync produces 57ns latency in AIDA64, much higher than the 44ns the same RAM on 10700k produces and nearly matching the 59ns I get on my Zen3 with the same clocks and sub timings.
Power draw: The 11700k uses 70w more power than the 5800x while gaming
If the 11700K is down on gaming performance that would explain why, as we already know
What have they done to it to bring the interconnect latency up to Zen 3 levels? Zen 3 has an IMC external to the cores and cache, Rocket Lake is monolithic isn't it?
Hang on, Intel's HEDT line suffered similar latency issues compared with mainstream, HEDT had to use a "Mesh" interconnect because the "Ring Bus" wouldn't work with CPU's of that size, the traces are too long.
Soldato
- Joined
- 14 Aug 2018
- Posts
- 3,579
Thanks for the insight.
It's a little troubling that you are finding this 3733Mhz sync limit which seems quite a retrograde move, though I'm curious as to the reason why it is necessary. Are the memory latencies you are seeing already in the low 40ns when you're at 3733Mhz?
The ability to tune memory without sync limits is the main reason that my tuned 9700K can still hang with my tuned 5800X in some lower threaded applications, though as we are a niche clientele then getting the most performance out of the CPU at stock is probably the priority - similar to what AMD have been doing.
Someone on another forum stated the latency might be because the "final" microcode hasn't been released yet. If this is indeed what latency will be like without any fixes that's a letdown.
Memory manufacturers must be kicking themselves if this is the growing trend for RAM being around 3,600. Sure you can bin RAM for 4,600+ but if nobody buys it you have to start putting the way higher chip grades in lesser units to satisfy demand.
Memory manufacturers must be kicking themselves if this is the growing trend for RAM being around 3,600. Sure you can bin RAM for 4,600+ but if nobody buys it you have to start putting the way higher chip grades in lesser units to satisfy demand.
Not sure where the memory latency microcode rumour came from, maybe people just assume it's intel, intel can't have higher latency so something must be wrong. If the latency is reduced to comet lake levels then gaming performance would go up by 10% over what people are getting now. Nothing is gauranteed though, it's not a skylake refresh its brand new architecture and this could just be the normal latency for this design