4 ms·

I've been holding my breath ever since spectre/meltdown for a free cpu upgrade to make up for the slower performance that the mitigations cause. It's intel's b

I've been holding my breath ever since spectre/meltdown for a free cpu upgrade to make up for the slower performance that the mitigations cause. It's intel's bug, they promised a certain processor speed, shouldn't it be their responsibility to replace it since their own security oversight resulted in the hardware not working as advertised? Did you expect the same from intel/amd when those bugs came out? Is it different from this situation?

mirzap 1y ago
They patched the issue, didn't they? Here we are talking about hardware issue that can cause physical harm to people, companies must eat that cost. Same way Samsung did by recalling exploding Galaxy Note a few years ago
aembleton 1y ago
> they promised a certain processor speed How have they advertised that? Was it clock frequency? Their mitigations mean it still runs at that clock frequency.
- tlb 1y ago
  The last desktop processor that actually promised a certain speed was probably the 68020. Everything since then has had inscrutable performance characteristics.
- TheDong 1y ago
  They promised the performance of their processor more than Hyundai promised that the Ioniq 5 was "carjacking-proof, even if the attackers have specialized tools", and this thread is about hyundai replacing parts of cars for free because attackers can hack into cars. With specialized tools. Which of course they can.
bjackman 1y ago
We don't actually know how to build the CPU that would replace your vulnerable one. New CPUs are largely immune to the specific attacks that were published before they were designed. But we aren't gonna get a fast CPU without sidechannels and IMO it's not possible in theory to build a branch predictor that never makes potentially exploitable mispredictions. In a sense this doesn't change your point but I wanted to take the opportunity to point this out. "This CPU is vulnerable to attack X" just means researchers have found an exploit in practice, which we already knew in theory was there. This wasn't the expectation before Spectre/Meltdown but now we live in a world where you need to assume a degradation in your CPU's effective performance as we learn about its vulnerabilities and need to apply software workarounds. I am building "one mitigation to rule them all" called Address Space Isolation but this doesn't fundamentally remove that fact, it just means that when we learn about a CPU's vulns we don't have to build a new mitigation we just have to change the settings on the existing one (and it should be more efficient than the bespoke one would be).
- yencabulator 1y ago
  > IMO it's not possible in theory to build a branch predictor that never makes potentially exploitable mispredictions. I believe the Mill hardware design would be immune by design because the hardware is in-order (relying on other trickery for its performance). Of course, it's still vaporware, but the noises made have been fairly competent. So in some ways, yes, but in other ways, what if you didn't need a branch predictor in the CPU. https://millcomputing.com/blog/wp-content/uploads/2018/01/Spectre.03.pdf https://millcomputing.com/blog/wp-content/uploads/2018/01/Sp... https://www.youtube.com/watch?v=8E4qs2irmpc https://www.youtube.com/watch?v=8E4qs2irmpc
  - snapplebobapple 1y ago
    Thats my favorite vapourware of the decade. I wish some(ne would fund them to finish so we can see how realbor bs the claims are
  - bjackman 1y ago
    > I believe the Mill hardware design would be immune by design because the hardware is in-order Well yes you can dodge this problem by not having a branch predictor but note the way I formulated my claim ;)

mirzap 1y ago

They patched the issue, didn't they? Here we are talking about hardware issue that can cause physical harm to people, companies must eat that cost. Same way Samsung did by recalling exploding Galaxy Note a few years ago

aembleton 1y ago

> they promised a certain processor speed How have they advertised that? Was it clock frequency? Their mitigations mean it still runs at that clock frequency.

tlb 1y ago
The last desktop processor that actually promised a certain speed was probably the 68020. Everything since then has had inscrutable performance characteristics.
TheDong 1y ago
They promised the performance of their processor more than Hyundai promised that the Ioniq 5 was "carjacking-proof, even if the attackers have specialized tools", and this thread is about hyundai replacing parts of cars for free because attackers can hack into cars. With specialized tools. Which of course they can.

bjackman 1y ago

We don't actually know how to build the CPU that would replace your vulnerable one. New CPUs are largely immune to the specific attacks that were published before they were designed. But we aren't gonna get a fast CPU without sidechannels and IMO it's not possible in theory to build a branch predictor that never makes potentially exploitable mispredictions. In a sense this doesn't change your point but I wanted to take the opportunity to point this out. "This CPU is vulnerable to attack X" just means researchers have found an exploit in practice, which we already knew in theory was there. This wasn't the expectation before Spectre/Meltdown but now we live in a world where you need to assume a degradation in your CPU's effective performance as we learn about its vulnerabilities and need to apply software workarounds. I am building "one mitigation to rule them all" called Address Space Isolation but this doesn't fundamentally remove that fact, it just means that when we learn about a CPU's vulns we don't have to build a new mitigation we just have to change the settings on the existing one (and it should be more efficient than the bespoke one would be).

yencabulator 1y ago
> IMO it's not possible in theory to build a branch predictor that never makes potentially exploitable mispredictions. I believe the Mill hardware design would be immune by design because the hardware is in-order (relying on other trickery for its performance). Of course, it's still vaporware, but the noises made have been fairly competent. So in some ways, yes, but in other ways, what if you didn't need a branch predictor in the CPU. https://millcomputing.com/blog/wp-content/uploads/2018/01/Spectre.03.pdf https://millcomputing.com/blog/wp-content/uploads/2018/01/Sp... https://www.youtube.com/watch?v=8E4qs2irmpc https://www.youtube.com/watch?v=8E4qs2irmpc
- snapplebobapple 1y ago
  Thats my favorite vapourware of the decade. I wish some(ne would fund them to finish so we can see how realbor bs the claims are
- bjackman 1y ago
  > I believe the Mill hardware design would be immune by design because the hardware is in-order Well yes you can dodge this problem by not having a branch predictor but note the way I formulated my claim ;)