More Intel leaks.. this one is not good though

[pas008]

wasnt this vulnerability found because it targeted intel cpus?

i'm sure we will find many more things like this like always

just like people will always find a way

[AresKrieger]

 All this info is in the original thread, I don't see why we need another one

[Dabombinable]

36 minutes ago, VaneBlade said:

It's bs that such a premium company is having these issues.

Though part of the "premium" is their use of monolithic designs and gigantic iGPU.

[jagdtigger]

50 minutes ago, SilkyDistress said:

Well, screw server buy atm, I am gonna save up for 8-core EPYC then.

Just buy an 1700/1700x...

[SilkyDistress]

4 minutes ago, jagdtigger said:

Just buy an 1700/1700x...

Nah ECC is my priority and more than 4 mem slots but still its a great buy for server.

[jagdtigger]

2 minutes ago, SilkyDistress said:

Nah ECC is my priority and more than 4 mem slots but still its a great buy for server.

I thought that AMD didnt disabled ECC in ryzen CPU's....

[SilkyDistress]

8 minutes ago, jagdtigger said:

I thought that AMD didnt disabled ECC in ryzen CPU's....

Can you link me board with ECC Ram support and I am buying instantly Ryzen with Red Hat for server 

[iRileyx]

53 minutes ago, JTLSound said:

Side note, the graphic for the Spectre is absolutely adorable

Someone gave the snapchat guy a stick. Ooh  

[ravenshrike]

1 hour ago, Delicieuxz said:

So, how will this affect Denuvo games for Intel CPUs? Denuvo makes a lot of kernel calls, right?

The real question is how hard will it be to use MELTDOWN to crack Denuvo and VMProtect.

[mr moose]

5 hours ago, Hellion said:

What difference does this make?

 

Just because an individual doesn't have the credentials to design/manufacture a processor doesn't mean that they aren't entitled to criticize a company for it's short comings. Having reasonable or even excessively high standards isn't an excuse for negligence. Wanting your machine secure isn't unreasonable by any means.

 

There may not be conclusive evidence at this point in time that indicates intel had prior knowledge of this exploit but when the highest ranking employee of the company dumps as much stock as possible prior to the public being made aware of the problem, that's highly suspect behavior that would warrant suspicion as it’s a move that had motive.

 

Or are you one of those blind, brand loyal. consumer tools that will always accept lip service from a corporation regardless of how severely they have wronged you?

 

 

It makes a huge difference, if you are going to accuse someone (or a whole R+D division) of incompetence or intentional neglect) then you should have the product knowledge and industry knowledge to support such claims. 

 

It's got nothing to do with consumer expectations and everything to do with leveling accusations that insinuate the accuser knows what went wrong (pro tip, no one on this forum knows why this eventuated). 

 

As far as the CEO dumping stock, you'd better learn how that all works before jumping on the "insider trading" bandwagon.  It's like accusing some of breaking the speed limit when all you know is they drove fast (not what road they were on or if the road is even governed by a limit),  everything he did was properly disclosed and (this is the fourth time I am saying this) he hasn't brought or made money out of any stock surrounding this, he has only lost value.

 

I don't even know why you would bring brand or loyalty into it,  that doesn't change the facts. 

 

 

[Matu20]

Downplaying to keep their stockholders happy, fun.

[porina]

3 hours ago, AresKrieger said:

Wait so meltdown is only vulnerable to local attack? That is incredibly useless as a hacking tool in the real world as anyone who has direct access to a machine can already hack into the machine at will, if it were remote then it would be a major problem potentially allowing for all kinds of stolen data but without that it is basically meaningless.

Exploits aren't only used in isolation. Other exploits or methods could be used to gain access to your system, and from there these may aid them to gain further access.

[jagdtigger]

29 minutes ago, SilkyDistress said:

Can you link me board with ECC Ram support and I am buying instantly Ryzen with Red Hat for server 

Eh, my bad. I allways forget that epyc and tr have a different socket than ryzen...

[SilkyDistress]

Just now, jagdtigger said:

Eh, my bad. I allways forget that epyc and tr have a different socket than ryzen...

Still is there any mobo with ECC support for Ryzen 7 ?

[cj09beira]

5 minutes ago, SilkyDistress said:

Still is there any mobo with ECC support for Ryzen 7 ?

depends on what you define as ecc support, is there boards that work just fine with ecc ram yes, are they tested to make sure that happens no

[TheCherryKing]

5 minutes ago, SilkyDistress said:

Still is there any mobo with ECC support for Ryzen 7 ?

The Ryzen boards are mainly made for gaming and streaming. You are better off getting on of the entry-level EPYC CPU's instead of Ryzen 7. 

[jagdtigger]

6 minutes ago, SilkyDistress said:

Still is there any mobo with ECC support for Ryzen 7 ?

I only found this:

http://www.overclock.net/t/1629642/ryzen-ecc-motherboards/10

 

 

[SilkyDistress]

3 minutes ago, TheCherryKing said:

The Ryzen boards are mainly made for gaming and streaming. You are better off getting on of the entry-level EPYC CPU's instead of Ryzen 7. 

Any cheapas* motherboard for EPYC ?

[cj09beira]

3 minutes ago, SilkyDistress said:

Any cheapas* motherboard for EPYC ?

maybe supre micro 

https://www.supermicro.nl/Aplus/motherboard/EPYC7000/H11SSL-NC.cfm

[FUR1AN]

So I recently just built a PC so getting manual updates is new to me. I tried checking for updates within windows and it says I am up to date. That being said I do run Anti Virus software from a third party which I have read my effect the visibility of the patch within windows.

 

Does anyone know how I would go about downloading the patch, and or if even should right now considering windows isn't prompting me to?

 

Any help would be great, thanks.

[ItsMitch]

2 minutes ago, wpowell13 said:

So I recently just built a PC so getting manual updates is new to me. I tried checking for updates within windows and it says I am up to date. That being said I do run Anti Virus software from a third party which I have read my effect the visibility of the patch within windows.

 

Does anyone know how I would go about downloading the patch, and or if even should right now considering windows isn't prompting me to?

 

Any help would be great, thanks.

You can get the patch here: https://support.microsoft.com/en-us/help/4056892/windows-10-update-kb4056892 and scroll down to the bottom and it's just a case of waiting for the automatic fixes to be rolled out via Windows Update (if you go that way)

[FUR1AN]

Just now, SC2Mitch said:

You can get the patch here: https://support.microsoft.com/en-us/help/4056892/windows-10-update-kb4056892 and scroll down to the bottom

Is there a reason why it isn't coming up when I check for updates? Is it safe to download as a result?

[anticoralian]

13 minutes ago, wpowell13 said:

Is there a reason why it isn't coming up when I check for updates? Is it safe to download as a result?

The update is blocked if your anti-virus hasn't received an update to make it compatible with the changes, otherwise you may get BSoD's

 

More info on that here: https://www.bleepingcomputer.com/news/microsoft/microsoft-releases-emergency-updates-to-fix-meltdown-and-spectre-cpu-flaws/

[porina]

        before    after    % change
Cinebench R15    CPU    684    685    0.1
                
Prime95 29.4b5    
    64k            
    4c1w    10078.49     9625.22    -4.5
    4c4w    17832.83    17847.94    0.1
    4c1w ht    10437.14     9463.28    -9.3
    4c4w ht    17428.52    17237.07    -1.1
                
    256k            
    4c1w    3455.5    3397.44    -1.7
    4c4w    3738.41    3694.8    -1.2
    4c1w ht    3847.33    3742.44    -2.7
    4c4w ht    3669.07    3655.39    -0.4
                
    2048k            
    4c1w    366.24    367.53    0.4
    4c4w    355.44    355.41    0.0
    4c1w ht    355.02    355.33    0.1
    4c4w ht    338.47    340.27    0.5

 

I'm out of time for tonight, but above are some random Prime95 29.4b5 throughput results (iter/s), and also Cinebench R15 just cos I had it on the system already. System is my gaming laptop: MSI GE62 6QF. This has a i7-6700HQ with 16GB ram (DDR4 2133, dual channel, dual rank). The storage has been replaced, and is now a single Crucial MX300 525GB SATA SSD.

 

On the Prime95 tests, I manually chose 64k to represent tiny work units which easily fit in the CPU cache. 256k is a kinda odd size, and at 4 workers it just fits in this CPU cache, so big cache stress, no significant ram stress. 2048k are big units, lots of ram access. The letters indicate number of cores used, number of workers, and if HT was used. So 4c1w is 4 cores working on the same task (multi-thread). 4c4w is 4 tasks with 1 core each. Similar for the two cases with HT on, where it is not just 1 core but 2 threads used per core.

 

I ran each benchmark 3 times, and the best result is presented. There is typically a couple % variation at the best of times, so I'd consider small differences to be insignificant. The only place there does seem to be a drop in performance is the 64k FFT size with multi-thread mode. You wouldn't run that anyway, but now there is more reason as it seems to be impacted more. Perhaps due to the small size, there is more relative overhead in the threading code, thus any impact there is greater.

 

I also did a single run on the SSD before and after, using CrystalDiskMark. I'm not presenting the results as I want to run more tests in the "after" state to check if there may be some other impact at work. Most results were not significantly changed, but sequential writes saw a big drop. The caution I have here is this contradicts other testing by Hardware Unboxed, suggesting it was actually 4k reads that were impacted. A possible mechanism is that smaller operations have more management overhead. On the plus side, I didn't see a change in my 4k read rate. If my results are repeatable, that may be due to some bottleneck, such that higher write speeds are impacted, yet smaller ones are not. This'll take more looking at.

 

I'll finish the 3D graphical testing tomorrow... but in a very quick look with FFXIV Stormblood benchmark, there is no significant change.

[W-L]

-Related Thread Merged-