AMD Issues Urgent Warning: New Transient Scheduler Attacks Expose Wide Range of CPUs

What if your computer’s most trusted layer—the silicon brain at its core—could quietly leak your secrets? That unsettling prospect just became a fresh reality for users of AMD processors. In June 2024, AMD publicly disclosed a new class of vulnerabilities called Transient Scheduler Attacks (TSA) that could allow attackers to infer sensitive data across security boundaries on a surprisingly broad range of their CPUs.

If you’re a business owner, IT professional, or even a privacy-minded PC enthusiast, you might be wondering: What are these attacks? How serious is the risk? And most importantly—what should I do about it? Let’s break it down together, in clear language, so you can make informed choices and stay secure.

What Are Transient Scheduler Attacks? A Plain-English Overview

First, let’s demystify the jargon. Transient Scheduler Attacks (TSA) are a newly identified family of speculative execution vulnerabilities. If that phrase rings a bell, it’s because prominent chip flaws like Meltdown and Spectre (discovered in 2018) put speculative execution on the map—showing how modern CPUs’ performance tricks can sometimes go awry.

In a Nutshell: Speculation Gone Wrong

Modern CPUs don’t like to wait. To run as fast as possible, they “speculatively” guess which instructions to process next, even before they’re sure it’s the right choice. Usually, this works wonders for speed. But sometimes, the CPU guesses wrong, and the half-finished computations must be tossed away.

Here’s the twist: Even those discarded operations can leave subtle traces—side effects in memory or timing—that clever attackers can observe. In the case of TSA, the danger comes not from what’s left behind, but from how quickly (or slowly) certain instructions run when the CPU’s scheduler goes off-script.

The Technical Details: How TSA Exploits CPU Scheduling

To understand why TSA is different—and why it matters—let’s step a little deeper behind the scenes.

False Completion and the Scheduler’s Blind Spot

Picture your CPU as a hyper-efficient assembly line. Every instruction is a part being processed, and the scheduler is the foreman, keeping things moving.

False completion happens when the CPU “thinks” a job is finished (say, loading a value from memory), but actually, it isn’t—the real data isn’t ready yet.
Rather than stall the line, the scheduler lets dependent instructions proceed, using placeholder (“invalid”) data.
Later, when the real data arrives, those instructions re-run with the correct values. Normally, this is harmless.

Where’s the issue? The timing of those speculative, invalid operations can subtly change the speed of other instructions—essentially leaving a breadcrumb trail that an attacker with deep system access could sniff out.

Unlike some prior speculative exploits, these false completions don’t flush the CPU’s pipeline or leave traces in cache or memory. They’re stealthier, making TSA a new class of side channel.

TSA Variants: TSA-L1 and TSA-SQ Explained

AMD’s research and the accompanying security advisory highlight two specific types of TSA, each with its own technical nuance:

TSA-L1: The L1 Data Cache Angle

TSA-L1 targets the Level 1 data cache—the CPU’s fastest “scratchpad” for recently used data.
A flaw in how the cache uses “microtags” lets a load instruction mistakenly retrieve invalid data, affecting timing in a way that can leak information.

TSA-SQ: The Store Queue Scenario

TSA-SQ involves the store queue, which manages memory write operations.
Here, a load instruction grabs data from the store queue too soon, before the data is truly available—again, with timing quirks that can be measured.

In both cases, the actual invalid data isn’t directly exposed—but the time it takes to process instructions becomes a leaky signal.

Which AMD Processors Are Affected? The Full List

If you use AMD hardware, this section’s for you. TSA vulnerabilities don’t just hit a handful of obscure chips—they affect a swath of AMD’s recent desktop, server, workstation, and embedded products.

Impacted CPUs (As of June 2024):

3rd and 4th Gen AMD EPYC Processors (data centers, servers)
AMD Instinct MI300A (AI and HPC accelerators)
AMD Ryzen 5000, 6000, 7000, and 8000 Series Desktop & Mobile Processors
AMD Ryzen Threadripper PRO 7000 WX-Series
AMD EPYC Embedded 7003, 8004, 9004, 97X4
AMD Ryzen Embedded 5000, 7000, V3000
…and more (see AMD’s full advisory)

If your PC, workstation, or server runs on a recent Ryzen or EPYC chip—or you rely on embedded AMD silicon—you’re likely affected.

What Data Is at Risk? Scenarios and Severity

Let’s get practical: What can attackers actually do with these flaws? Here’s the real-world impact.

Potential Information Leakage

From OS kernel to user application
From hypervisor (host) to guest virtual machine
Between two user applications

That said, exploiting TSA isn’t trivial. It requires:

Malicious access to the target machine (attackers must run their own code locally)
The ability to repeatedly create specific microarchitectural conditions

Web-based attacks (e.g., via malicious websites) are NOT currently possible with TSA, so drive-by exploit risk is low.

CVEs and Severity Ratings

AMD disclosed four closely related vulnerabilities, each with its own CVE:

CVE-2024-36350 (CVSS 5.6): Data leakage from previous stores (privileged info exposure)
CVE-2024-36357 (CVSS 5.6): L1D cache leaks sensitive info across privilege boundaries
CVE-2024-36348 (CVSS 3.8): User process can infer control registers speculatively, even if UMIP is enabled
CVE-2024-36349 (CVSS 3.8): User process can infer TSC_AUX value, even when read is disabled

Severity ranges from moderate to important, but not critical. Still, any attack that can pierce the veil between user and system, or between virtual machines, deserves attention—especially in shared, multi-tenant, or cloud environments.

How Difficult Is It to Exploit TSA?

Here’s where things get interesting—and perhaps a bit reassuring.

TSA requires very specific conditions and high technical know-how. An attacker typically needs to execute code directly on the target machine (think: malicious app or insider threat, not remote hacker).
The attack relies on timing measurements and repeated triggering of the false completion flaw—usually via a communication path between victim and attacker (such as an app talking to the OS kernel).
You’re not at risk just by browsing the web or opening email attachments. TSA is not “wormable” and cannot be exploited remotely.

Here’s why that matters: For everyday consumers, the immediate risk is low. But for cloud providers, data centers, or anyone running sensitive workloads on shared AMD hardware, the risk is real—especially if untrusted parties share compute resources.

What Is AMD Doing? Patches, Mitigations, and Guidance

AMD has responded quickly, working with Microsoft and ETH Zurich researchers (who found these flaws) to develop:

Microcode updates (CPU firmware) for affected processors – These updates address the root cause at the hardware level. Check here for AMD’s microcode updates.
OS-level mitigations – Microsoft and other OS vendors are integrating workarounds and protections as part of regular security updates.
Guidance for system administrators and users – Instructions on how to check your CPU and apply updates.

Steps You Should Take

Check if your CPU is affected. Use AMD’s security advisory or your system vendor’s documentation.
Update your BIOS/UEFI firmware. Many PC and server makers have already released updates that include AMD’s new microcode.
Install all OS security patches. Stay current with Windows Update, Linux kernel updates, and related security advisories.
For data center or cloud workloads, consider process isolation strategies. Avoid running untrusted code alongside sensitive workloads on the same hardware.

Pro tip: If you’re unsure about BIOS or microcode updates, check your PC manufacturer’s support site or reach out to IT support. These updates are often labeled as “security fixes.”

Why TSA Matters: The Bigger Picture for CPU Security

You might be asking, “So what? Don’t these side-channel attacks come up every year?” It’s a fair question—after all, Spectre and Meltdown shook the industry in 2018, and we’ve seen similar headlines since.

Here’s why TSA is a wake-up call:

It’s a new flavor of speculative side channel. TSA exploits timing, rather than memory or cache residue, making it harder to detect and block.
The affected CPUs are widespread. This isn’t a niche issue—Ryzen and EPYC CPUs power millions of desktops, laptops, and servers worldwide.
Cloud and virtualization environments are especially at risk. If untrusted code can run on the same hardware as sensitive workloads, cross-tenant leakage becomes possible.

A Note on Responsible Disclosure

Kudos to AMD, Microsoft, and ETH Zurich for their transparent, coordinated response. Attacks like TSA are only uncovered—and mitigated—through collaboration between researchers, vendors, and the broader security community.

TSA in Context: How Does It Compare to Spectre, Meltdown, and Friends?

If you’ve followed CPU security over the years, you know speculative execution flaws are an ongoing challenge. Here’s how TSA fits into the landscape:

TSA doesn’t make older methods obsolete, but it broadens the attack surface by highlighting yet another performance tradeoff with security implications.

Practical Advice: How Can You Protect Yourself?

Let’s recap with clear steps:

Home Users: Apply all system and firmware updates. The risk is minimal, but staying current is always best.
Businesses and Enterprises: Assess your AMD hardware inventory, prioritize firmware updates, ensure OS patches are deployed, and review workload placement on shared hardware.
Cloud Providers: Consider risk modeling for multi-tenant hosts. Where possible, use hardware or hypervisor-based isolation for highly sensitive workloads.

No need to panic—but don’t ignore the updates, either. Attackers are always watching for unpatched systems.

Frequently Asked Questions (FAQ) on AMD’s Transient Scheduler Attacks

Q1: What is a Transient Scheduler Attack in simple terms?
A: It’s a way attackers can use the timing of certain CPU operations (after a scheduling hiccup) to infer data they shouldn’t have access to—like secrets from another app or the operating system.

Q2: Am I at risk if I just use my computer for regular web browsing and email?
A: No. TSA attacks require the attacker to run code on your computer—simply browsing the web or opening files won’t put you at risk.

Q3: Which AMD CPUs are affected by TSA?
A: Most modern AMD Ryzen and EPYC processors, including desktop, mobile, server, and embedded chips from the last several generations. Check AMD’s advisory for the full list.

Q4: How do I protect myself or my organization?
A: Install BIOS/microcode updates from your hardware vendor, apply all OS security patches, and avoid running untrusted code on shared systems.

Q5: Are there any known attacks in the wild exploiting TSA?
A: As of now, no widespread exploits have been reported. The attacks are difficult to pull off, but that could change as knowledge spreads.

Q6: Do these vulnerabilities affect AMD graphics cards?
A: No, TSA focuses on CPU behavior. AMD GPUs are not impacted.

Q7: How does TSA differ from Spectre and Meltdown?
A: TSA exploits timing quirks in the CPU’s scheduler, not cache or branch prediction side effects. It’s a distinct, newer class of speculative execution flaw.

Q8: Should I avoid buying AMD CPUs?
A: Not at all. All major CPU makers, including Intel and ARM, have faced similar issues. AMD has released timely fixes—what matters is keeping your systems up to date.

Final Takeaway: Stay Updated, Stay Secure

The discovery of Transient Scheduler Attacks is another reminder that modern CPUs, for all their speed and sophistication, are not invincible. But with transparency, rapid patches, and informed users, we can keep the risks in check.

Bottom line:
– Patch your firmware and operating system promptly. – Understand the real (but limited) risk—especially if you run sensitive or multi-tenant workloads. – Keep following trusted sources for the latest in security news.

Want in-depth security updates and practical tech advice like this delivered straight to your inbox? Subscribe for more, and stay ahead of the curve.

External resources for further reading:

Thanks for reading—and remember, a secure system is an updated system!

Discover more at InnoVirtuoso.com

I would love some feedback on my writing so if you have any, please don’t hesitate to leave a comment around here or in any platforms that is convenient for you.

For more on tech and other topics, explore InnoVirtuoso.com anytime. Subscribe to my newsletter and join our growing community—we’ll create something magical together. I promise, it’ll never be boring!

Stay updated with the latest news—subscribe to our newsletter today!

Thank you all—wishing you an amazing day ahead!

AMD Issues Urgent Warning: New Transient Scheduler Attacks Expose Wide Range of CPUs