Is my L3 cache really slower than my DRAM?
I've recently been running a memtest86 on my PC (Threadripper 1920x, Kingston ECC DDR4-2666 CAS19) because I suspected that the stock timings were conservative. And I was right: the memory will run at stock voltages at 3200 speed with JEDEC standard CAS20 timings, which gives a small but noticeable performance increase on benchmarks.
However, when running the memtest86, I noted that it reports my L3 cache as running at 17.36GB/s but my RAM as running at 17.73GB/s, meaning my RAM is actually faster than the processor's L3 cache.
- Is this likely?
- Should I disable on processor L3 cache?
- How can I disable on processor L3 cache it that's appropriate?
memory cpu performance cache
edited Dec 15 '18 at 10:51
JakeGould
31k1093137
asked Dec 15 '18 at 10:31
Christi
7641616
add a comment |
I've recently been running a memtest86 on my PC (Threadripper 1920x, Kingston ECC DDR4-2666 CAS19) because I suspected that the stock timings were conservative. And I was right: the memory will run at stock voltages at 3200 speed with JEDEC standard CAS20 timings, which gives a small but noticeable performance increase on benchmarks.
However, when running the memtest86, I noted that it reports my L3 cache as running at 17.36GB/s but my RAM as running at 17.73GB/s, meaning my RAM is actually faster than the processor's L3 cache.
- Is this likely?
- Should I disable on processor L3 cache?
- How can I disable on processor L3 cache it that's appropriate?
memory cpu performance cache
edited Dec 15 '18 at 10:51
JakeGould
31k1093137
asked Dec 15 '18 at 10:31
Christi
7641616
There are multiple aspects to CPU and cache performance and I understand this can be very complex. They cannot be properly evaluated on the basis of a single benchmark number. CPU manufacturers have done extensive testing and designed the caches to be of benefit under real world conditions, not to do well with a specific benchmark. And I suspect this benchmark number is of little relevance to real world performance.The determination that L3 cache is not beneficial could only be done after extensive and difficult testing, which memtest86 was not designed to do.
– LMiller7
Dec 15 '18 at 14:33
add a comment |
I've recently been running a memtest86 on my PC (Threadripper 1920x, Kingston ECC DDR4-2666 CAS19) because I suspected that the stock timings were conservative. And I was right: the memory will run at stock voltages at 3200 speed with JEDEC standard CAS20 timings, which gives a small but noticeable performance increase on benchmarks.
However, when running the memtest86, I noted that it reports my L3 cache as running at 17.36GB/s but my RAM as running at 17.73GB/s, meaning my RAM is actually faster than the processor's L3 cache.
- Is this likely?
- Should I disable on processor L3 cache?
- How can I disable on processor L3 cache it that's appropriate?
memory cpu performance cache
edited Dec 15 '18 at 10:51
JakeGould
31k1093137
asked Dec 15 '18 at 10:31
Christi
7641616
I've recently been running a memtest86 on my PC (Threadripper 1920x, Kingston ECC DDR4-2666 CAS19) because I suspected that the stock timings were conservative. And I was right: the memory will run at stock voltages at 3200 speed with JEDEC standard CAS20 timings, which gives a small but noticeable performance increase on benchmarks.
However, when running the memtest86, I noted that it reports my L3 cache as running at 17.36GB/s but my RAM as running at 17.73GB/s, meaning my RAM is actually faster than the processor's L3 cache.
- Is this likely?
- Should I disable on processor L3 cache?
- How can I disable on processor L3 cache it that's appropriate?
memory cpu performance cache
memory cpu performance cache
edited Dec 15 '18 at 10:51
JakeGould
31k1093137
asked Dec 15 '18 at 10:31
Christi
7641616
edited Dec 15 '18 at 10:51
JakeGould
31k1093137
asked Dec 15 '18 at 10:31
Christi
7641616
edited Dec 15 '18 at 10:51
JakeGould
31k1093137
edited Dec 15 '18 at 10:51
JakeGould
31k1093137
edited Dec 15 '18 at 10:51
JakeGould
31k1093137
31k1093137
asked Dec 15 '18 at 10:31
Christi
7641616
asked Dec 15 '18 at 10:31
Christi
7641616
asked Dec 15 '18 at 10:31
Christi
7641616
7641616
There are multiple aspects to CPU and cache performance and I understand this can be very complex. They cannot be properly evaluated on the basis of a single benchmark number. CPU manufacturers have done extensive testing and designed the caches to be of benefit under real world conditions, not to do well with a specific benchmark. And I suspect this benchmark number is of little relevance to real world performance.The determination that L3 cache is not beneficial could only be done after extensive and difficult testing, which memtest86 was not designed to do.
– LMiller7
Dec 15 '18 at 14:33
add a comment |
There are multiple aspects to CPU and cache performance and I understand this can be very complex. They cannot be properly evaluated on the basis of a single benchmark number. CPU manufacturers have done extensive testing and designed the caches to be of benefit under real world conditions, not to do well with a specific benchmark. And I suspect this benchmark number is of little relevance to real world performance.The determination that L3 cache is not beneficial could only be done after extensive and difficult testing, which memtest86 was not designed to do.
– LMiller7
Dec 15 '18 at 14:33
There are multiple aspects to CPU and cache performance and I understand this can be very complex. They cannot be properly evaluated on the basis of a single benchmark number. CPU manufacturers have done extensive testing and designed the caches to be of benefit under real world conditions, not to do well with a specific benchmark. And I suspect this benchmark number is of little relevance to real world performance.The determination that L3 cache is not beneficial could only be done after extensive and difficult testing, which memtest86 was not designed to do.
– LMiller7
Dec 15 '18 at 14:33
There are multiple aspects to CPU and cache performance and I understand this can be very complex. They cannot be properly evaluated on the basis of a single benchmark number. CPU manufacturers have done extensive testing and designed the caches to be of benefit under real world conditions, not to do well with a specific benchmark. And I suspect this benchmark number is of little relevance to real world performance.The determination that L3 cache is not beneficial could only be done after extensive and difficult testing, which memtest86 was not designed to do.
– LMiller7
Dec 15 '18 at 14:33
add a comment |
1 Answer
1
active
oldest
votes
The slight difference is perhaps because the RAM is buffered, but the L3 cache
is not (L3 is a kind of buffer by itself).
Doing a write to the RAM will result in the Memory Controller storing it in
a buffer and sending a response back stating that the transaction is complete
(even though it's not). The exception to this is when doing large blocks of writes
and go beyond the capabilities of the Memory Controller buffer.
My guess is that the L3 cache and the RAM are very similar in hardware,
but the Memory Controller gives the RAM a slight advantage, a small one since
memtest86 is writing large block of memory.
So:
Is this likely?
Yes, as described above, although it might only be an illusion.
Should I disable on processor L3 cache?
Absolutely not, even if this is at all possible.
How can I disable on processor L3 cache it that's appropriate?
Not appropriate, as the CPU is hard-wired to use it.
answered Dec 15 '18 at 10:47
harrymc
254k13265565
add a comment |
Your Answer
StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "3"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});
function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});
}
});
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fsuperuser.com%2fquestions%2f1383805%2fis-my-l3-cache-really-slower-than-my-dram%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
1 Answer
1
active
oldest
votes
1 Answer
1
active
oldest
votes
active
oldest
votes
active
oldest
votes
The slight difference is perhaps because the RAM is buffered, but the L3 cache
is not (L3 is a kind of buffer by itself).
Doing a write to the RAM will result in the Memory Controller storing it in
a buffer and sending a response back stating that the transaction is complete
(even though it's not). The exception to this is when doing large blocks of writes
and go beyond the capabilities of the Memory Controller buffer.
My guess is that the L3 cache and the RAM are very similar in hardware,
but the Memory Controller gives the RAM a slight advantage, a small one since
memtest86 is writing large block of memory.
So:
Is this likely?
Yes, as described above, although it might only be an illusion.
Should I disable on processor L3 cache?
Absolutely not, even if this is at all possible.
How can I disable on processor L3 cache it that's appropriate?
Not appropriate, as the CPU is hard-wired to use it.
answered Dec 15 '18 at 10:47
harrymc
254k13265565
add a comment |
The slight difference is perhaps because the RAM is buffered, but the L3 cache
is not (L3 is a kind of buffer by itself).
Doing a write to the RAM will result in the Memory Controller storing it in
a buffer and sending a response back stating that the transaction is complete
(even though it's not). The exception to this is when doing large blocks of writes
and go beyond the capabilities of the Memory Controller buffer.
My guess is that the L3 cache and the RAM are very similar in hardware,
but the Memory Controller gives the RAM a slight advantage, a small one since
memtest86 is writing large block of memory.
So:
Is this likely?
Yes, as described above, although it might only be an illusion.
Should I disable on processor L3 cache?
Absolutely not, even if this is at all possible.
How can I disable on processor L3 cache it that's appropriate?
Not appropriate, as the CPU is hard-wired to use it.
answered Dec 15 '18 at 10:47
harrymc
254k13265565
add a comment |
The slight difference is perhaps because the RAM is buffered, but the L3 cache
is not (L3 is a kind of buffer by itself).
Doing a write to the RAM will result in the Memory Controller storing it in
a buffer and sending a response back stating that the transaction is complete
(even though it's not). The exception to this is when doing large blocks of writes
and go beyond the capabilities of the Memory Controller buffer.
My guess is that the L3 cache and the RAM are very similar in hardware,
but the Memory Controller gives the RAM a slight advantage, a small one since
memtest86 is writing large block of memory.
So:
Is this likely?
Yes, as described above, although it might only be an illusion.
Should I disable on processor L3 cache?
Absolutely not, even if this is at all possible.
How can I disable on processor L3 cache it that's appropriate?
Not appropriate, as the CPU is hard-wired to use it.
answered Dec 15 '18 at 10:47
harrymc
254k13265565
The slight difference is perhaps because the RAM is buffered, but the L3 cache
is not (L3 is a kind of buffer by itself).
Doing a write to the RAM will result in the Memory Controller storing it in
a buffer and sending a response back stating that the transaction is complete
(even though it's not). The exception to this is when doing large blocks of writes
and go beyond the capabilities of the Memory Controller buffer.
My guess is that the L3 cache and the RAM are very similar in hardware,
but the Memory Controller gives the RAM a slight advantage, a small one since
memtest86 is writing large block of memory.
So:
Is this likely?
Yes, as described above, although it might only be an illusion.
Should I disable on processor L3 cache?
Absolutely not, even if this is at all possible.
How can I disable on processor L3 cache it that's appropriate?
Not appropriate, as the CPU is hard-wired to use it.
answered Dec 15 '18 at 10:47
harrymc
254k13265565
answered Dec 15 '18 at 10:47
harrymc
254k13265565
answered Dec 15 '18 at 10:47
harrymc
254k13265565
answered Dec 15 '18 at 10:47
harrymc
254k13265565
254k13265565
add a comment |
add a comment |
Thanks for contributing an answer to Super User!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Some of your past answers have not been well-received, and you're in danger of being blocked from answering.
Please pay close attention to the following guidance:
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fsuperuser.com%2fquestions%2f1383805%2fis-my-l3-cache-really-slower-than-my-dram%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
There are multiple aspects to CPU and cache performance and I understand this can be very complex. They cannot be properly evaluated on the basis of a single benchmark number. CPU manufacturers have done extensive testing and designed the caches to be of benefit under real world conditions, not to do well with a specific benchmark. And I suspect this benchmark number is of little relevance to real world performance.The determination that L3 cache is not beneficial could only be done after extensive and difficult testing, which memtest86 was not designed to do.
– LMiller7
Dec 15 '18 at 14:33